Heart and skeletal muscle inflammation (HSMI) in Atlantic salmon

Heart and skeletal muscle inflammation (HSMI)
in Atlantic salmon, Salmo salar:
pathology, pathogenesis and experimental infection
Ruth Torill Kongtorp
Thesis for the degree of Philosophiae Doctor
Norwegian School of Veterinary Science
Oslo, 2008

Contents
1. Acknowledgements............................................................................................................... 4
2. List of papers ........................................................................................................................ 5
3. Abbreviations........................................................................................................................ 6
4. Summary............................................................................................................................... 7
5. Norsk sammendrag (summary in Norwegian) .................................................................. 8
6. Introduction .......................................................................................................................... 9
7. Anatomy and physiology of heart and skeletal muscle................................................... 10
7.1. Heart.............................................................................................................................. 10
7.1.1. Organisation ........................................................................................................... 10
7.1.2. Myocardium ........................................................................................................... 10
7.1.3. Epi- and endocardium ............................................................................................ 11
7.1.4. Oxygenation ........................................................................................................... 12
7.2. Skeletal muscle.............................................................................................................. 13
8. Cardiac and skeletal muscle pathology ............................................................................ 13
8.1. Reversible cellular changes........................................................................................... 14
8.1.1. Compensatory hypertrophy.................................................................................... 14
8.1.2. Myofibre atrophy.................................................................................................... 14
8.1.3. Cellular degeneration ............................................................................................. 14
8.2. Irreversible cellular changes.......................................................................................... 14
8.2.1. Apoptosis................................................................................................................ 14
8.2.2. Necrosis.................................................................................................................. 15
8.3. Inflammation ................................................................................................................. 16
8.3.1. Inflammatory cells.................................................................................................. 16
8.3.2. Myositis and myocarditis ....................................................................................... 21
8.3.3. Peri-, epi- and endocarditis..................................................................................... 22
8.4. Neoplasia....................................................................................................................... 22
8.5. Circulatory disturbances................................................................................................ 23
8.6. Regenerative capacity ................................................................................................... 24
9. The study of novel diseases................................................................................................ 25
10. Heart and skeletal muscle inflammation (HSMI) ......................................................... 26
10.1. Occurrence .................................................................................................................. 26
10.2. Description of disease outbreaks................................................................................. 26
10.3. Pathology..................................................................................................................... 27
10.3.1. Clinical signs and macroscopical lesions ............................................................. 27
10.3.2. Histopathology ..................................................................................................... 27
11. Aims of study .................................................................................................................... 28
12. Summary of papers .......................................................................................................... 29
13. Artikkelsammendrag (summary of papers in Norwegian) .......................................... 31
14. Methodological considerations........................................................................................ 33
14.1. Field samplings ........................................................................................................... 33
14.2. Infection experiments.................................................................................................. 34
14.3. Histopathology ............................................................................................................ 35
14.4. Electron microscopy.................................................................................................... 36
14.5. Enzyme histochemistry ............................................................................................... 37
14.6. Immunohistochemistry................................................................................................ 38
14.7. Microbiological examination ...................................................................................... 39
2

14.7.1. Bacteriology ......................................................................................................... 39
14.7.2. Virology ............................................................................................................... 39
15. Main results ...................................................................................................................... 41
16. Discussion of results ......................................................................................................... 42
16.1. Diagnostic criteria for HSMI....................................................................................... 42
16.2. Distinction from other diagnoses ................................................................................ 44
16.2.1. CMS ..................................................................................................................... 44
16.2.2. PD......................................................................................................................... 45
16.3. Pathogenesis................................................................................................................ 49
16.3.1. Pre-clinical phase ................................................................................................. 49
16.3.2. Clinical (outbreak) phase ..................................................................................... 51
16.3.3. Post-clinical phase................................................................................................ 52
16.3.4. Inflammatory cells................................................................................................ 53
16.4. Aetiology..................................................................................................................... 54
16.4.1. The virus hypothesis............................................................................................. 54
16.4.2. Bacteria................................................................................................................. 55
16.4.3. Parasites and fungi ............................................................................................... 56
16.4.4. Toxins................................................................................................................... 56
16.4.5. Autoimmunity ...................................................................................................... 57
16.4.6. Neoplasia.............................................................................................................. 58
16.5. The significance of HSMI for Atlantic salmon aquaculture ....................................... 58
16.5.1. Mortality............................................................................................................... 58
16.5.2. Transmissibility.................................................................................................... 60
17. Main conclusions .............................................................................................................. 61
18. Future work ...................................................................................................................... 62
References ............................................................................................................................... 63
Tables....................................................................................................................................... 75
Table 1: Distribution of HSMI outbreaks 2003-2007 .......................................................... 75
Table 2: Histopathological findings during HSMI outbreaks .............................................. 76
Table 3: Severity of lesions in cardiac and red skeletal muscle during HSMI outbreaks.... 77
Table 4: Occurrence of histopathological changes in several organs .................................. 78
Table 5: Histological stains.................................................................................................. 79
Table 6: Histochemical reactivities...................................................................................... 80
Table 7: Fish cell lines ......................................................................................................... 81
Figures..................................................................................................................................... 82
Figure 1: Anatomy of heart and skeletal muscle.................................................................. 82
Figure 2: Pathology of heart and skeletal muscle ................................................................ 84
Figure 3: HSMI diagnoses in the period 1999-2007............................................................ 86
Papers ...................................................................................................................................... 87
3

1. Acknowledgements
The present work was performed at the National Veterinary Institute (NVI) in the period
between 2003 and 2008. The project was financed by the Research Council of Norway, grant
no. 153000/120.
I wish to express my gratitude to my main supervisor Dr. Torunn Taksdal and the other
supervisors Dr. Knut Falk, Professor Inge Bjerkås and Professor Trygve Poppe for interesting
discussions, guidance, support and co-authorship.
I would also like to thank my co-authors Marianne Halse, Arild Kjerstad, Arthur Lyngøy,
Arne Guttvik, Dr. David Graham, Dr. Erling Olaf Koppang and Professor Hugh Ferguson for
valuable collaboration.
I am grateful for the contributions from technical staff and researchers at the Section for Fish
Health, Section for Pathology, Section for Virology and Serology and Section for
Epidemiology at the NVI, and also to technical staff at the Norwegian School of Veterinary
Science.
Last, but not least, thanks to family and friends for fantastic support!
4

2. List of papers
Paper I: Kongtorp RT, Taksdal T, Lyngøy A (2004)
Pathology of heart and skeletal muscle inflammation (HSMI) in farmed Atlantic salmon
Salmo salar
Diseases of Aquatic Organisms, 59; 217-224.
Paper II: Kongtorp RT, Kjerstad A, Taksdal T, Guttvik A, Falk K (2004)
Heart and skeletal muscle inflammation in Atlantic salmon, Salmo salar L.: a new
infectious disease
Journal of Fish Diseases, 27; 351-358.
Paper III: Kongtorp RT, Halse M, Taksdal T, Falk K (2006)
Longitudinal study of a natural outbreak of heart and skeletal muscle inflammation in
Atlantic salmon, Salmo salar L.
Journal of Fish Diseases, 29; 233-244.
Paper IV: Kongtorp RT, Koppang EO, Bjerkås I, Falk K, Taksdal T
Features of the pathogenesis of heart and skeletal muscle inflammation in Atlantic
salmon, Salmo salar L.
Submitted
Paper V: Kongtorp RT, Taksdal T
Studies with experimental transmission of heart and skeletal muscle inflammation in
Atlantic salmon, Salmo salar L.
Accepted, Journal of Fish Diseases
Paper VI: Ferguson HW, Kongtorp RT, Taksdal T, Graham D, Falk K (2005)
An outbreak of disease resembling heart and skeletal muscle inflammation in Scottish
farmed salmon, Salmo salar L., with observations on myocardial regeneration
Journal of Fish Diseases, 28; 119-123.
5

3. Abbreviations
ASPV: Atlantic salmon paramyxovirus
CMC: cell-mediated cytotoxicity
CMS: cardiomyopathy syndrome
CPE: cytopathic effect
EGC: eosinophilic granule cells
FGF: fibroblast growth factor
HSS: haemorrhagic smolt syndrome
HE: haematoxylin and eosin
HSMI: heart and skeletal muscle inflammation
Ig: immunoglobulin
I.p.: intraperitoneal
IHNV: infectious haematopoietic necrosis virus
ISA: infectious salmon anaemia
ISAV: infectious salmon anaemia virus
IPN: infectious pancreatic necrosis
IPNV: infectious pancreas necrosis virus
MHC: major histocompatibility complex
NVI: National Veterinary Institute
PCNA: proliferating cell nuclear antigen
PCR: polymerase chain reaction
PD: pancreas disease
Pers.com.: personal communication
RNA: ribonucleic acid
SAV: salmonid alphavirus
SPDV: salmon pancreas disease virus
VHS: viral haemorrhagic septicaemia
VHSV: viral haemorrhagic septicaemia virus
6

4. Summary
Heart and skeletal muscle inflammation (HSMI) is a severe disease affecting farmed Atlantic
salmon Salmo salar in Norway. In this thesis, the characteristic features of HSMI is described
from fish sampled during the mid-outbreak phase of field cases Paper I . Lesions most commonly
occur in the atrium and ventricle in the heart, and include inflammation and necrosis in epi-,
endo- and myocardium. Both compact and spongy layers of the ventricle are involved. In
addition, there is usually myositis and necrosis of red skeletal muscle. Fish with severe
cardiac lesions may also have multifocal liver necrosis and signs of circulatory disturbance.
Further, it is shown that HSMI is transmissible from diseased to healthy fish by experimental
infection Paper II . Experimental fish were intraperitoneally injected with cardiac tissue
homogenate from diseased fish. The broad spectered antibiotic gentamycine was added to one
half of the tissue homogenate and injected into a corresponding group of experimental fish in
a separate tank. A group of non-infected cohabitating fish were also held in each of the tanks.
All groups developed histopathological changes consistent with HSMI, but no mortality could
be associated with the lesions. There was a time lag between injected and cohabitant groups
of four weeks, indicating a period of propagation of some causal agent followed by shedding
of infective particles. No virus or bacteria were detected with standard laboratory methods. In
a longitudinal study of a field outbreak of HSMI in Norway Paper III , cardiac lesions appeared
several months before the onset of mortalities. It was also observed that while lesions in red
skeletal muscle and other organs only were present during the clinical outbreak phase, cardiac
lesions persisted for many months following the outbreak. During the clinical outbreak, the
morbidity of fish showing lesions consistent with HSMI was close to 100 %. It was further
attempted to characterise the inflammatory cells by the use of enzyme and
immunohistochemistry Paper IV . The results were non-conclusive, as the reactivity to the
methods used was low. However, the study of the inflammatory reaction also showed that
perivasculitis and endocardial changes are present in tissue sampled prior to a clinical
outbreak. This finding was supported by observations from an experimental study Paper V . The
histopathological lesions in HSMI may resemble lesions observed in cardiomyopathy
syndrome (CMS) and pancreas disease (PD). In typical cases HSMI is histopathologically
distinguishable from these diseases Papers I and III . However, a disease outbreak in Scotland
illustrates the diagnostic challenge associated with differentiating between HSMI and PD in
an area of endemic PD Paper VI .
7

5. Norsk sammendrag (summary in Norwegian)
Norsk tittel: Hjerte- og skjelettmuskelbetennelse (HSMB) hos atlantisk laks (Salmo
salar): patologi, patogenese og eksperimentell infeksjon.
Hjerte- og skjelettmuskelbetennelse (HSMB) er en alvorlig sykdom hos atlantisk laks (Salmo
salar) i Norge. HSMB karakteriseres ved histopatologiske funn slik de forekommer hos fisk
som gjennomgår naturlige sykdomsutbrudd Artikkel I . Hovedfunnene er betennelse og nekrose
(celledød) i epi- endo- og myokard i hjertets for- og hjertekammer. I hjertekammeret er både
kompakt og spongiøst muskellag involvert. Det er vanligvis også betennelse og nekrose i rød
skjelettmuskulatur. Fisk med alvorlige hjerteforandringer kan også ha multifokal
levernekrose, samt tegn til sirkulasjonsforstyrrelse i flere organer. Videre vises det at
sykdommen er eksperimentelt overførbar ved intraperitoneal injeksjon av vevshomogenat
med eller uten tilsetning av det bredspektrede antibiotikumet gentamycin Artikkel II .
Forandringene ble også overført til forsøksfisk (kohabitanter) som ble holdt sammen med
injisert fisk, men med en forsinkelse på fire uker i forhold til de øvrige gruppene. Dette
indikerer at en periode med oppformering av det infeksiøse agens er nødvendig for å smitte
kohabitanter. Det var ikke dødelighet i forsøket som kunne assosieres med HSMB. Det ble
heller ikke isolert bakterier eller virus fra forsøksfisk ved bruk av standard
laboratoriemetoder. I et langtidsstudium av HSMB under naturlige forhold ble det funnet
hjerteforandringer flere måneder før det ble registrert forøket dødelighet i fiskegruppen Artikkel
III
. Det var en gradvis utvikling fra milde til mer alvorlige hjerteforandringer før det kliniske
sykdomsutbruddet. I den kliniske fasen hadde nesten 100 % av den undersøkte fisken
hjerteforandringer. Det var også persistens av hjerteforandringer i månedene etter utbruddet.
Forandringer i andre organer var bare tilstede i utbruddsfasen. Det ble forsøkt å identifisere
betennelsescellene ved hjelp av enzym- og immunhistokjemiske metoder, men resultatene var
inkonklusive Artikkel IV . Det ble imidlertid funnet at betennelse omkring koronarkar og
endokardforandringer oppstår tidlig i sykdomsutviklingen. Disse funnene ble støttet av en
tilsvarende utvikling etter eksperimentell smitte med materiale fra fisk med HSMB Artikkel V .
De histopatologiske forandringene ved HSMB kan ligne forandringene man finner ved
kardiomyopatisyndrom (CMS) og pankreassykdom (PD). I typiske tilfeller er det mulig å
skille de tre sykdommene på grunnlag av histopatologi Artikkel I og III , men i atypiske tilfeller kan
det være diagnostiske utfordringer. Dette illustreres ved beskrivelsen av et sykdomsutbrudd i
et område med endemisk PD i Skottland Artikkel VI .
8

6. Introduction
Atlantic salmon (Salmo salar) aquaculture is one of the most important industries in Norway.
There has been a rapid development from the early beginning in the 1970s to today, regarding
structure of the industry, technology and methods 4,58 . Although this has enabled efficient and
economically feasible management, the margins are small and may be diminished by losses
due to diseases. A longitudinal study of farmed Atlantic salmon in 2006 showed that
cumulative mortality during the first three months of the seawater phase was 2.1 %. Fish were
mostly lost during short time periods associated with disease outbreaks, and 64 % of the
mortality was due to infectious diseases (A Aunsmo, pers.com.). The economic losses due to
disease problems increases with time and money spent on fish growth. Diseases causing the
highest direct costs may therefore occur later in the seawater phase. It has been estimated that
the economic impact due to loss of biomass alone may reach € 1.25 million for a single site
experiencing an outbreak of pancreas disease (PD) 149 . If one also counts additional costs,
including for instance changes in management routines, disposal of dead fish, medical
treatment and biosecurity regimes, disease may have a tremendous economical impact on the
salmon farming industry.
Until the development of efficient oil-adjuvanted vaccines in the early 1990s, bacterial
diseases constituted the dominating health problem in the Norwegian salmon aquaculture 83 .
In the last 20 years, however, viral diseases have become more important 4,149 . In 1999, a new
disease problem was discovered in farmed Atlantic salmon in Norway. During the summer
months, several farms experienced mortalities of unknown cause, and material was submitted
to the National Veterinary Institute (NVI) for examination. Dead and moribund fish from the
outbreak sites showed severe inflammation and necrosis of cardiac tissue and red skeletal
muscle. The histopathological features resembled known diseases, such as cardiomyopathy
syndrome (CMS) 44 and PD 93 , but also had distinct features. No known infectious agents or
other causal factors were identified. Researchers from the NVI thus considered the disease to
represent a new condition, and named it heart and skeletal muscle inflammation (HSMI), due
to its histomorphological appearance. HSMI is the focus of this thesis.
9

7. Anatomy and physiology of heart and skeletal muscle
The main lesions in HSMI are found in the heart and skeletal muscle. Normal morphology
and function of these tissues will therefore be described in the following, with emphasis on
the heart.
7.1. Heart
7.1.1. Organisation
The salmon heart is located in the pericardial cavity, and consists of four compartments: Sinus
venosus, atrium, ventricle and bulbus arteriosus 116 . The sinus venosus is a thin-walled
structure (60-90 m), having a similar volume to the atrium 41 . It is mainly composed of
connective tissue. The cardiac pacemaker tissue is found within the wall of the sinus venosus.
Blood flows into the sinus venosus from the Cuverian ducts, hepatic veins and anterior jugular
veins, and drains into the atrium. The walls of the atrium largely consist of a thin layer of
spongy cardiac muscle 116 . It has a large surface area, and may contract considerably to move
blood into the cardiac ventricle. An atrioventricular valve supported by surrounding
myocardium ensures a unidirectional blood flow 41 . In Atlantic salmon, the ventricle has a
pyramidal shape, which is common in fish species with an active life-style 116 . The outer
muscular layer consists of compact cardiac muscle tissue arranged in a two-layered concentric
pattern lying perpendicularly to each other (Fig. 1A). In the compact layer, endomysial
connective tissue containing a capillary network of coronary vessel branches surrounds the
cardiomyocytes 142 . Towards the centre of the ventricle is a more loosely organised trabecular
spongy muscle tissue (Fig. 1B). Connective tissue is found in the interface between the two
muscle layers 116 . The last compartment is the bulbus arteriosus, a fibroelastic structure
functioning as a pressure regulator and depulsator. Blood from the ventricle flows into this
chamber through semilunar valves 133 .
7.1.2. Myocardium
Both compact and spongy muscle layers in the heart consist of cross-striated cardiac muscle
tissue (Fig. 1C). In fish, cardiomyocytes grow by hyperplasticity, rather than hypertrophy.
Compact myocardial myocytes are generally smaller than myocytes in the spongy layer, but
there is also considerable variation between individual cells. Both cell types are considerably
10

smaller than mammalian cardiomyocytes 41 . The nucleus is centrally placed within the cell,
and takes a pale bluish colour by haematoxylin and eosin (HE) staining 116 . Important
components of myocytes are the contractile myofibrils, which are situated in the cytoplasmic
portion of the cell (sarcoplasm). In fish, there are fewer myofibrils in atrial myocytes,
compared to ventricular myocytes 41 . Myofibrils consist of thin filaments composed of actin,
tropomyosin and troponin, and thick filaments containing myosin. These structures are
organised in a repetitive pattern of bands called sarcomeres, creating a pattern of crossstriation
(Fig. 1D) 67 . The acidophilic nature of myofibrils contributes to a pink colour of the
cytoplasm in HE sections 41 . Myoglobin is the major oxygen-binding protein contained within
the sarcoplasm of myocytes. It has a structure similar to haemoglobin in blood cells, and its
function is associated with oxidative phosphorylation 67 .
In cardiac muscle individual cells are branched, and are interconnected with neighbouring
cells through intercalated discs. The discs consist of a transverse and a lateral part. The
transverse portion serves as an anchoring site for actin filaments, as well as connection point
between the cells through desmosomes. The lateral portion contains gap junctions, enabling
exchange of ions between interconnected cells. A signal to contract may thus pass quickly
from cell to cell. This enables the production of a wave of contraction necessary for pumping
the blood through the heart 67 .
7.1.3. Epi- and endocardium
All compartments of the heart are lined externally by epicardium and internally by
endocardium (Figs. 1A-C). A thin layer of connective tissue covers the myocardium
subepicardially and subendocardially 142 . The epicardium is the visceral layer of the
pericardium. It is mainly composed by a monolayer of epithelial cells, but some areas also
contain haematopoietic cells 116 . The endocardium contains a monolayer of endothelial cells,
which are equivalent to cells lining blood vessels 67 . However, studies with rats have shown
that cardiac endothelium has a different transcription profile compared to endothelium in
coronary vessels and aorta 55 . This probably reflects different functional properties.
Endothelial cells in the atrium of some species of fish, such as platy, swordtail and cod, are
capable of phagocytosis 79,80,152 . These cells may function as scavenger cells, and it has
therefore been suggested that endocardium in these fish is part of the innate immune system
11

147,154 . In salmonids, a scavenger function of the endocardium has not been demontstrated, but
scavenger cells have been found in head kidney 48 .
Development and differentiation of cardiomyocytes depend on the presence of epi- and
endocardial cells 81,105 . Fibroblast growth factors (FGFs) are present in the epi- and
endocardium of several species. The activities of some FGFs appear to be essential for the
development and differentiation of cardiomyocytes in embryonic mice 78 . Expression of FGFs
are also important for the regenerative capacity of the zebrafish heart following
experimentally induced cardiac injury 81 . The epicardium, but not the endocardium, also
secretes erythropoietin and retinoic acid, which enhance survival and proliferation of
cardiomyocytes in chicken 158 . Factors negatively affecting epi- and endocardial function may
therefore potentially have a large impact on the myocardium.
7.1.4. Oxygenation
Venous blood from organs in the body is pumped through the heart and ventral aorta to the
gills, in which gas exchange occurs. Oxygen-rich arterial blood from the second gill arch
artery is carried by the coronary arteries on the ventricular surface. The arteries branch off
into capillaries surrounding cardiomyocytes in the compact layer 35 . The coronary vessel
branches are largely derived from mesenchymal precursor cells in the epicardium. During
cardiac development, these cells migrate into the myocardium to form vessels 30 . The arterial
blood perfusion through the coronary vessel branches supplies the compact myocardium with
necessary oxygen for maintenance and work. This is reflected in the presence of catabolytic
enzymes in compact myocardium that mainly facilitate aerobic metabolism of fatty acids 35 .
Venous blood from the coronary vessel branches within compact myocardium is drained into
the atrium in the atrioventricular region 41 .
The spongy myocardium in the atrium and inner portion of the ventricle contains no blood
vessels, and is oxygenated by venous blood flowing through the cardiac lumen 116 . Due to the
limited access to oxygen, enzymes suited for aerobic and anaerobic metabolism of
carbohydrates are abundant in this tissue. In addition, the myoglobin content is higher,
enabling greater oxygen storage and maintenance of aerobic metabolism at reduced oxygen
pressures 35 .
12

7.2. Skeletal muscle
Skeletal muscle tissue in salmon mainly consists of red and white fibres (Fig. 1E). White
muscle fibres are the most abundant, and make up the bulk of body muscle. They are
anchored to the skeleton and skin by a layer of connective tissue called the membranous
skeleton, arranging the muscle in four units. The tissue is further divided into bundles of
striated skeletal muscle by connective tissue layers called myosepta 163 . White muscle is
explosive in function, enabling rapid movements. This is due to anaerobic metabolism of
glucose, which is stored as glycogen within the muscle tissue.
Red skeletal muscle tissue is located parallel to the body axis along the lateral line of the fish
144
. It consists of smaller bundles of striated skeletal muscle fibres and contains more
myoglobin within the sarcoplasm than white muscle tissue (Fig. 1F) 67 . The red muscle tissue
is highly vascularised 163 . It receives most of its energy from oxidative phosphorylation of
fatty acids, and is involved in the maintenance movements of the fish 67,144 . Between these
muscle types and also partly mixed with white fibres are pink fibres. Functionally, these fibres
are intermediates between red and white muscle 133 .
Skeletal muscle cells are similar to cardiomyocytes, but are organised in multinucleated cell
syncytia instead of the arrangement of individual cells interconnected by intercalated discs
described above. This enables a more regular structure and a uniform, rather than wave-like,
contraction pattern. Myofibrils occupy most of the sarcoplasm, and these are arranged
similarly to cardiac myofibrils. The function is also equivalent 67 .
8. Cardiac and skeletal muscle pathology
Pathological changes in cardiac and skeletal muscle tissue have been observed in a number of
terrestrial and aquatic species, including Atlantic salmon. The cells most often involved in
observable changes are myocytes 90 . They have a limited range of reaction patterns to various
insults 165 . The most important of these will be addressed in the following.
13

8.1. Reversible cellular changes
8.1.1. Compensatory hypertrophy
Hypertrophy of muscle cells may compensate for an increased work load, either due to
physiological stress or pathological processes. This is achieved by an increased number of
sarcomeres, leading to an increase in cell size 124 . Hypertrophic myocytes are histologically
recognised by their enlarged fibres and nuclei (Fig. 2A) 165 . Pathological cardiac hypertrophy
may occur in mammals as a result of valvular stenosis, which is a common congenital
anomaly in dogs and pigs 21 . Also, insufficient valve function due to anomalies or infection,
septal defects and hyperthyroidism may cause cardiac hypertrophy 165 . In Atlantic salmon
with CMS, myocytes with hypertrophic nuclei may be observed within or adjacent to foci of
necrosis and inflammation in the heart 44 . In skeletal muscle, compensatory hypertrophy
usually occurs as a response to atrophy or necrosis of adjacent myofibres 90 .
8.1.2. Myofibre atrophy
Reduction of muscle cell or myofibre diameter (atrophy) is most commonly observed in
skeletal muscle. The causes of atrophy may be denervation, nutritional deficiency or
immobility. Denervation atrophy is a result of neural damage, and results in rapid
disintegration of myofibrils. Generally, myocytic nuclei remain for a longer period, resulting
in the formation of cells only containing a row of nuclei 90 .
8.1.3. Cellular degeneration
Cells undergo degenerative changes if they fail to maintain homeostasis, for instance as a
result of viral infection 21 . Reversible injury to cells in muscle tissues mainly includes
accumulation of lipid droplets in the myocytic sarcoplasm, sarcoplasmic vacuolation and lysis
of myofibrils 165 .
8.2. Irreversible cellular changes
8.2.1. Apoptosis
Irreversible injuries result in cell death. Some infectious agents, such as infectious pancreatic
necrosis virus (IPNV), may induce apoptosis in fish cells 59 . This is a controlled process
14

leading to cell death without leakage of cellular products, and may also occur as a normal
physiological process. Apoptotic cells initially show signs of shrinkage and condensation of
cellular components, including nuclear fragmentation (karyorheksis). This is followed by
break-up of the cell into membrane-enclosed apoptotic bodies, which are subsequently
phagocytosed by macrophages or neighbouring tissue cells 21 .
8.2.2. Necrosis
Cells dying due to pathological processes most often undergo necrosis. This term refers to
post-mortal degradation of cellular components, which is an autolytic process initiated by
release of endogenous enzymes immediately after death of the cell. Necrotic myocytes are
strongly eosinophilic with HE staining, and often show loss of striation on histological
examination (Figs 2B-C). In later stages, vacuolation occurs, and the cells gradually
disintegrate (Fig. 2D). The nuclei may shrink and become basophilic (pyknosis), or undergo
lysis (karyolysis) 21 . Irreversibly damaged cells may be replaced by fibrous or fatty tissue 90 .
Tissue necrosis is usually followed by infiltration of inflammatory cells (Fig. 2C).
Inflammatory processes may be toxic to cells in the host tissue, and may also induce necrosis.
It is therefore not always easy to determine whether primary necrosis has occurred, or if the
necrosis is secondary to inflammation 165 .
Primary necrosis of myocytes in mammals is most often caused by nutritional deficiencies,
toxic substances or trauma, but infectious diseases may also play a role 90,165 . Lesions
occurring at only one site are usually caused by trauma, while systemic diseases usually cause
multifocal or diffuse damage 90 . Inadequate ingestion of vitamin E or selenium causes primary
cardio- and skeletal myopathy in both fish and terrestrial animals 54,90,165 , but the most
important causes of myocyte necrosis in modern fish farming appear to be infectious agents
116,163 .
Several bacterial diseases affect the skeletal muscle of Atlantic salmon. Among the most
important are furunculosis caused by Aeromonas salmonicida subsp. salmonicida, vibriosis
caused by Vibrio anguillarum, cold-water vibriosis (Hitra disease) caused by Vibrio
salmonicida, “winter ulcer” caused by Moritella viscosa and Vibrio viscosus, piscirickettsiosis
caused by Piscirickettsia salmonis, and infection with Tenacibaculum maritimum 131 . These
15

diseases may all cause necrosis of both red and white skeletal muscle. The most significant
disease of certain viral origin causing necrosis in heart and skeletal muscle is PD 116 .
8.3. Inflammation
Inflammation occurs as a response to infectious and non-infectious insults to the tissue.
Inflammatory processes take place in most animal species, including teleosts 135 . There are a
number of chemical mediators that contribute to inflammation. Some of these are present in
the blood plasma, most being complement factors. Others are produced by leukocytes or
macrophages. The main functions of these mediators are vasodilation, increased vascular
permeability, chemotaxis and cytotoxicity 86,98,122 . This results in an increased blood perfusion
and migration of leukocytes into the damaged area 98 . In Atlantic salmon, leukocytes are
recruited from circulating cells in the blood, by activation of cell populations in the kidney
and spleen, and by local activation of cells residing in the tissue 145 .
8.3.1. Inflammatory cells
White blood cells in fish are classified using the nomenclature of human haematology, which
is based on differential staining methods. However, it is not possible to differentiate all fish
cells by this technique, as functionally different cells may have similar staining properties.
This classification system is therefore not optimal for fish cells 38 . Inflammatory cells
involved in acute inflammatory responses of mammals are present in fish, but it appears that
chronic inflammation is a more prominent host response in these species 42 . In the following,
morphology and function of the most important inflammatory cells observed in fish will be
presented.
8.3.1.1. Neutrophils
Characteristic features of mammalian neutrophils are their multilobulated nuclei, the presence
of lysosomal granules in the cytoplasm and their capability of phagocytosis 98 . Mature
mammalian neutrophils generally have few organelles relative to the amount of cytoplasm 7 .
The nucleus morphology of fish neutrophils varies with species, but the nucleus of a salmonid
neutrophil is multilobulated 38 . In addition, the inner nuclear membrane is sometimes
infolded, and pockets containing cytoplasm may be observed. The outer nuclear membrane is
16

covered with ribosomes 84 . Fish neutrophils contain Golgi apparatus, mitochondria,
ribosomes, endoplasmic reticulum, vacuoles, glycogen particles and specific granules 89 . The
granules usually do not stain with standard methods, hence the name neutrophils. Three types
of cytoplasmic granules have been observed: fibrillar, homogenous, and striated or rod like
38,89
. Neutrophils present within inflamed tissue in channel catfish also have phagosomes and
cytoplasmic extensions, called pseudopodia 22 .
In mammalian inflammatory processes, neutrophils are usually the first to arrive at the site of
injury. These cells are continuously circulating, and therefore quickly recruited due to
increased blood flow to the area. Upon recognition of non-self material, neutrophils bind to
and engulf the foreign particles. Within the cell, phagosomes containing engulfed material
fuse with lysosomal granules, which enable digestion 98 . The phagocytic process of
neutrophils is injurious to the host tissue, partly due to leakage of lysosomal contents, and
partly due to active disgorging of enzyme content as an extracellular killing mechanism. This
may result in local tissue necrosis 145 .
Neutrophils may also be recruited in early immune responses of fish, but they appear to play a
less important role 135 . Instead of being primarily phagocytic, it is thought that the main
function of fish neutrophils is extracellular killing of microorganisms by secretion of enzymes
42
. Mammalian neutrophils are capable of expressing major histocompatibility complex
(MHC) class II molecules, and it is suggested that they play a role in antigen presentation 5 .
However, it is uncertain whether fish neutrophils possess this ability.
8.3.1.2. Monocytes and macrophages
Monocytes are mainly found in the blood. In fish they are slightly larger than other circulating
leukocytes, and have phagocytic abilities. The nucleus is oval or kidney shaped, and the
cytoplasm is basophilic with no distinct granules 38 . Macrophages are phagocytic cells present
in tissues. They are generally derived from a monocytic cell line, but some fish macrophages
are related to granulocytes or are derived from cells in connective tissue. Macrophages are
usually agranular, but may sometimes contain cytoplasmic granules 38,122 . Compared to
neutrophils, macrophages are larger. Nuclei of macrophages are folded or kidney shaped and
are centrally located in the cell. The cytoplasm of mammalian macrophages contains
lysosomes and small granules, as well as numerous pseudopodia 7 . Macrophages of channel
17

catfish are similar, containing pseudopodia, vacuoles, lysosomes and rough endoplasmic
reticulum 22 . Melanomacrophages also contain melanin pigment. In lymphomyeloid tissues
they may form nodule-like accumulations in some species 38 , but this is not prominent in
Atlantic salmon. Melanomachrophages are frequently observed in chronic inflammation,
where they possibly play a role in removing free radicals 42 .
Macrophages have many functions in the inflammatory reaction. Most importantly, their
phagocytic capacity is greater than that of neutrophils. In addition to engulfing foreign
material, macrophages play an important role in cleaning up cellular debris in the damaged
area 7 . Macrophages also produce and secrete acid and proteases, complement components,
reactive oxygen species, nitric oxide, eicosanoids and cytokines 98 . These promote
microbicidal activities, act as chemotactic signals for other leukocytes and contribute to
systemic effects 7 . The activity may, however, also cause significant damage to the tissue,
which in turn may prolong the disease process by inducing a more intense immune response.
This self-sustaining inflammation may be the cause of a prolonged duration seen in many
inflammatory diseases 98 . Macrophages are also a part of the adaptive immune system. They
express MHC class II on the surface, and function as antigen presenting cells 162 . Fish
macrophages have similar functions to their mammalian relatives, and are important for
clearing the tissue of infected or necrotic cells 9,86,66 .
8.3.1.3. Lymphocytes
Typical lymphocytes in mammals are described to be smaller than neutrophils, and have a
small rim of cytoplasm surrounding a densely staining nucleus 7 . Fish lymphocytes are also
small cells with a round or oval nucleus. According to Fänge 38 , lymphocytes in fish are
morphologically almost indistinguishable from other types of non-granulated cells.
Circulating stem cells, blast cells, monocytes and cross sections of thrombocytes may
therefore all be described as lymphocyte-like cells. The cytoplasm is scant, basophilic and
contains no or very small granules 38 . Lymphocytes observed within inflamed tissue in
channel catfish may have even less cytoplasm than lymphocytes in peripheral blood 22 .
Ultrastructurally, fish lymphocytes have a chromatin-dense nucleus. Mitochondria, rough and
smooth endoplasmic reticulum, ribosomes and Golgi apparatus are present in the cytoplasm of
circulating lymphocytes, but some structures may be lacking in cells found within tissues
during inflammation 22,38 .
18

Lymphocytes are the main cells involved in adaptive immune responses. They are rectruited
to the site of injury by adhesion molecules and chemokines, and are activated by the
interaction with antigen presenting cells. Although not morphologically different, there are
three main groups of lymphocytes: helper T cells, cytotoxic T cells and B cells 162 . The
presence of these cell types have been demonstrated in Atlantic salmon 74,113,122 . Helper T
cells bind to foreign material presented in association with MHC class II. This is essential for
inducing a specific immune response. B cells specific for the pathogen attach to
corresponding helper T cells bound to MHC class II and become activated. The surface
receptors of activated B cells are able to capture and present circulating antigen to other
immune cells 162 . Unlike the situation in mammals, B lymphocytes in fish are probably also
capable of phagocytosis 82 .
An important part of the defence against viruses and some intracellular bacteria in mammals
is the ability of cell-mediated toxicity (CMC). Cytotoxic responses occur in fish, and both
specific and non-specific inflammatory cells appear to be involved 47 . Specific CMC requires
the expression of T-cell receptors, CD8 molecules and MHC class I compatible with the
antigen. Cytotoxic T cells recognise foreign antigen presented on MHC class I. When binding
to infected cells, cytotoxic T cells induce apoptosis, thereby eliminating the infection 162 . The
components of specific CMC is present in fish from a very early age, and it has been shown
that cytotoxic T cells in fish are involved in the elimination of cells displaying viral antigens
on their MHC class I receptor 47 . Also, non-specific cytotoxic cells have been described in
fish. Morphologically, they are small, agranular lymphocytes 122 . Their function is similar to
mammalian natural killer cells, and their ability to initiate apoptosis of infected cells has been
demonstrated 34 .
8.3.1.4. Plasma cells
Activated B cells eventually differentiate into plasma cells 162 . Fish plasma cells have an
eccentrically positioned nucleus surrounded by a basophilic, non-granulated cytoplasm. They
contain numerous ribosomes on a rough endoplasmic reticulum forming flat or irregular
cisternae 38 . Mammalian plasma cells produce antibodies, consisting of immunoglobulins that
are antigen specific and capable of binding to antigens 162 . This prevents attachment of
pathogens to cell surfaces and entry of viruses into cells. In addition, the antigen-antibody
complex formation promotes phagocytosis and elimination of the pathogens by macrophages
19

162
. Atlantic salmon peripheral blood leukocytes contain receptors for binding to antibodies,
and show a high degree of immune complex receptor binding activity 110 . Salmonids are
highly capable of inducing an antibody response to pathogenic infection, although the serum
concentration of immune globulin (Ig) M in salmonids is low compared to that of for instance
cod 153 .
8.3.1.5. Eosinophilic granule cells
Eosinophilic granule cells (EGC) may be present in several chronic inflammatory conditions
in fish. They do not produce histamine in fish, but may otherwise have a function similar to
histamine-releasing mast cells in other species 128-130 . Mammalian mast cells are mainly
located perivascularly in the tissues. They are morphologically and functionally similar to
circulating basophils. Generally, however, mast cells are mononuclear and basophils are
polymorphnuclear. The cytoplasm of both cell types contain basophilic granules, which
appear blue on HE sections. Instead of being phagocytic, mast cells and basophils degranulate
and release histamine and other inflammatory mediators upon stimulation. In mammals, these
cells thus play a role in acute inflammation and recruitment of other inflammatory cells to the
site of injury 7 .
Fish cells with mast-cell-like morphology may display metachromasia of cytoplasmic
granules when stained with for instance toluidin blue, thionin and Alcian blue, but the
basophilic material in the granules is sensitive to watery fixatives and staining solutions. With
normal fixation, staining properties are mainly acidophilic, and in HE sections the granules
stain red from eosin 129 . Degranulation has been observed in acute bacterial infection and by
injection of a mast cell degranulating agent compound. Also, recruitment of EGCs has been
observed in parasitic and other diseases causing chronic inflammation, such as vaccineinduced
peritonitis 115,127,128,145 .
8.3.1.6. Thrombocytes
Platelets in mammalian species do not contain nuclei, but possess cytoplasmic granules, dense
bodies and a cytocavitary network 7 . Thrombocytes in fish are oval or spindle shaped. They
have a large nucleus and may have an almost invisible rim of surrounding cytoplasm. Usually,
however, the thrombocytes resemble deformed erythrocytes without haemoglobin, or they
20

may be histologically indistinguishable from lymphocytes. Ultrastructurally, the cytoplasm
contains mitochondria, ribosomes, glycogen granules and bundles of microtubules. They also
contain vacuoles that may be arranged linearly and connected to the cell surface 38 . Once
activated and aggregated on damaged endothelial cells, they release mediators related to
coagulation, inflammation and healing processes 7 . Fish thrombocytes are probably also
phagocytic, and may participate in antigen presentation and regulation of the immune
response 42,70 .
8.3.2. Myositis and myocarditis
The most important causes of myositis and myocarditis in mammals are infectious agents or
immune-mediated processes, as well as idiopathic conditions. Bacterial infection of the heart
and skeletal muscle usually trigger an immune response, and suppurative, serohaemorrhagic
or granulomatous inflammation may be observed in mammals. Viruses are not known to
cause primary skeletal muscle inflammation in domestic animals, but in viral diseases
involving the heart, the inflammatory infiltrates are dominated by lymphocytes. In idiopathic
and immune-mediated myositis, lymphocytes, plasma cells and eosinophils are abundant.
Also, parasitic infections of skeletal muscle usually evoke a mononuclear inflammation 90,165 .
Cellular components of the inflammatory reaction in fish are mostly mononuclear, and mainly
consist of macrophages, lymphocytes and plasma cells. Particles that are too large to be
phagocytosed may in some cases not induce an inflammatory response. Instead, encapsulation
may occur 42,145 . Lymphocytes and macrophages also form granulomas, as for instance is
observed in vaccine-induced inflammation 73,115 .
Bacterial infections in fish often cause primary necrosis and secondary inflammation. The
inflammation may be pyogranulomatous, as seen in infections with Renibacterium
salmoninarum. Cells involved in granulomatous inflammation are macrophages, epitheloid
cells, lymphocytes, plasma cells and fibroblasts 42 . Other bacteria, such as Flavobacterium
psychrophilum, may induce a more diffuse necrosis and inflammation in rainbow trout 116,163
23,43
. Fungal infections may occur secondarily to bacterial skin infections, generally causing
necrosis of exposed skeletal muscle, followed by granulomatous inflammation. Systemic
fungal infections caused by Exophialia spp., Phoma spp. and Ichthyophonus hoferi may also
cause granulomatous inflammation or necrosis of myocardium 116,163 . The presence of
parasites may cause focal granulomatous inflammation, as well as infiltration and
21

accumulation of melanin 163,165 . Viral diseases in fish usually elicit a mononuclear
inflammation, with infiltrations mainly consisting of lymphocytes and macrophages. Many
pathogens may be found in the interphase between compact and spongy myocardium in the
heart, triggering inflammation. Viraemia may also cause inflammation. For instance, vessel
damage occurring during generalised infections with viral haemorrhagic septicaemia virus
(VHSV) or infectious haematopoietic necrosis virus (IHNV) may lead to multifocal
myocarditis 116 . Also, primary myocytic necrosis caused by infection with salmonid
alphavirus (SAV) in the heart and skeletal muscle causes secondary inflammation 116,163 .
CMS is a specific inflammatory condition involving the heart of Atlantic salmon. While
skeletal muscle is usually normal, severe mononuclear inflammation and necrosis are
observed in endocardium and spongy myocardium in the atrium and ventricle of the heart
3,44,136
. Pathological lesions appear to be restricted to these tissues, but occasional
inflammation has been observed in epicardium and compact myocardium, especially in
association with coronary vessel branches 136 .
8.3.3. Peri-, epi- and endocarditis
Pericardial inflammation occurring in domestic mammals and birds is caused by various
systemic infections. Cellular components are mostly neutrophils, and there is usually a
considerable amount of fibrin on the pericardial surface. Possible outcomes of severe cases
are fibrosis and adhesion between the pericardial surfaces, which may interfere with cardiac
function 165 . In fish, pericarditis may be induced by bacterial infections 116 . However, mild
epicardial inflammation is often found in apparently normal individuals of farmed Atlantic
salmon in seawater 116 . Endocarditis in mammals is usually associated with systemic bacterial
infection. The most prominent lesions are found on the valves, but this may lead to secondary
mural endocarditis in severe cases 165 . In Atlantic salmon, endocarditis is usually associated
with myocarditis 116 .
8.4. Neoplasia
Neoplasia is uncontrolled division of abnormal cells that may grow locally in a tissue, or also
metastase to other organs. Neoplasia may be caused by physical or chemical agents, such as
ultraviolet radiation and aflatoxins 20 . Liver tumours have been observed in rainbow trout
22

exposed to aflatoxin 88 . Also, several viruses induce neoplasia in mammals, for instance
hepatitis B virus and bovine leukaemia virus 20 . In fish, only retroviruses are known to cause
tumours 88 . Histologically, neoplastic cells in a tumour may all resemble each other, and the
cells in some benign tumours may be almost identical. The degree of pleomorphy generally
increases with malignancy 20 . In fish, neoplastic cells may appear malignant, but not be
particularly invasive 88 .
In several species, tumours may sometimes be observed in cardiac and skeletal muscle, and
usually occur as a result of metastasis from tumours in other organs 165 . However, neoplastic
conditions originating from muscle tissue are rare. In fish these tumours are mainly incidental
findings 88 . Benign muscle tumours are called rhabdomyomas, and malignant tumours are
called rhabdomyosarcomas. Their origin may most easily be identified by the presence of
myoglobin. Rhabdomyomas have been reported from cardiac tissues of pig and sheep, and
laryngeal rhabdomyomas have been observed in dogs. Cells observed in these tumours have
granular eosinophilic cytoplasm and an abundance of mitochondria. Malignant
rhabdomyosarcomas are sometimes seen in dogs, and are highly metastatic, with cell
appearances varying from striated to undifferentiated 90 .
8.5. Circulatory disturbances
Diseases causing increased vascular resistance and decreased blood perfusion through the
heart may result in cardiomorphological changes. This has been observed in fish with
amoebic gill disease 121 . Coronary arteriosclerosis is frequently observed in Pacific salmonids,
and is thought be associated with ageing 39,40 . Affected vessels are histologically characterised
by disruption of the elastic lamina and intimal proliferation of smooth muscle cells 41 .
Coronary arteriosclerosis is also common in farmed and wild Atlantic salmon, but is not
usually associated with mortalities. In farmed fish, it appears that the pathological process
may start already during the first year in seawater 41,146 . A case of sudden mortality in farmed
Atlantic salmon in Norway during transport to the slaughter facility was associated with
coronary arteriosclerosis and subsequent infarction of compact myocardium.
Histopathological observations of acute myocardial infarction included multiple foci of
cardiomyocytes with marked eosinophilia and nuclear pyknosis, as well as multifocal necrosis
of white skeletal muscle 119 .
23

In vivo intravascular coagulation, thrombosis (Fig. 2E), may cause vascular occlusion and
subsequent infarction of muscle tissue 165 . Endothelial damage and subsequent necrosis causes
exposure of subendothelial collagen, which triggers thrombosis 165 . In domestic mammals,
thrombosis may be caused by infection-related vasculitis, for instance due to infection with
heart worm in dogs and bluetongue-virus in sheep 90 . In Atlantic salmon, thrombosis has been
observed in fish with CMS 44,143 .
Immune-mediated vasculitis in horses may cause haemorrhages and infarction in skeletal
muscle 90 . Myocardial haemorrhages may also be associated with nutritional deficiencies, for
instance selenium-vitamin E deficiency in piglets 165 . Haemorrhagic smolt syndrome (HSS), a
disease of unknown aetiology in Atlantic salmon, affects smolts just before transfer to sea. It
is characterised as a haemorrhagic anaemia, causing multifocal haemorrhaging in several
organs, including muscle and cardiac tissue 139 . Multifocal haemorrhages may also be seen
within the muscle tissue of fish affected by generalised bacterial diseases, such as vibriosis
and furunculosis 116 .
8.6. Regenerative capacity
The regenerative capacity of cardiac muscle tissue is limited in mammals, and probably does
not occur in adult animals. Instead, necrotic cells are generally replaced by fibrotic tissue 165 .
Mammalian skeletal muscle has a larger regenerative capacity, as satellite cells are located
between the basal lamina and the sarcolemma. Necrotic segments may therefore regenerate if
the sarcolemmal tube is still intact. The damaged area is cleared of debris by macrophages,
and regeneration occurs through proliferation and migration of satellite cells. Once in
position, satellite cells are able to form myoblasts, which differentiate into mature myocytes.
Regenerating fibres are recognised histologically by basophilia, centrally placed nuclei and a
lack of striation (Fig. 2F) 90 . Myocardial regeneration has been observed in birds and fish
42,165
. Recruitment of subepicardial progenitor cells enables zebrafish to rebuild the ventricular
wall following experimental resection of the apex 120 .
24

9. The study of novel diseases
New diseases continue to arise in Atlantic salmon aquaculture. When they occur,
diagnosticians and researchers must work together to define and characterise the problem. The
most common way of approaching a new disease is to describe its clinical and pathological
appearance, as was done for PD 99 , CMS 3,44 and HSS 139 . Useful knowledge may also be
gained by following the pathological development of the disease for some time 95,102 , or by
epidemiological studies of several disease outbreaks 14,96,164 . An important question is whether
this truly is a new disease. Helpful hints may be found in the literature, or by comparing the
novel pathological pattern with other diseases.
For infectious diseases, a possible association with known or previously unknown pathogens
might be determined with laboratory methods, for instance polymerase chain reaction (PCR)
57,31 23 26,106
, immunohistochemistry or growth in a medium such as blood agar or cell culture .
However, some microbes are not readily retrievable. It is necessary to find out whether a new
disease is infectious, as it may potentially spread to naïve populations and cause significant
problems related to fish welfare and farmer economy. Experimental transmissibility may
strongly indicate infectiousness, and is often the method of choice. Before it was possible to
isolate the aetiological agent for PD, it was shown that material from diseased fish was
sufficient to reproduce typical histopathological changes in experimental fish 91,101,125 .
Ultimately, a causal relationship between the proposed aetiological agent and the disease
should be demonstrated. This is usually done by reproducing the disease in naïve fish by
experimental infection with the isolated pathogen, and re-isolation of the same agent from
challenged fish 26,92 .
25

10. Heart and skeletal muscle inflammation (HSMI)
10.1. Occurrence
As earlier mentioned, HSMI was first discovered in 1999. Records for the occurrence of
HSMI in the period between 1999 and 2002 are lacking. Fish pathologists at the NVI have
estimated that the yearly occurrence of HSMI was 10-15 affected sites from 1999 to 2001,
and in 2002, fish were diagnosed with HSMI on at least 41 sites. A yearly report of the health
status of salmonids in Norway, including data from the diagnostic database at the NVI and
other laboratories 149 , showed that 450 sites had outbreaks of HSMI from 2003 to 2007 (Fig.
3). Of these, 57 were diagnosed in 2003, 54 in 2004, 83 in 2005, 94 in 2006 and 162 in 2007.
The number of diagnosed HSMI outbreaks varied between counties, with Møre og Romsdal,
Sør-Trøndelag and Nord-Trøndelag hosting 60.9 % of the total number of affected sites
(Table 1). Recurrent outbreaks after fallowing and restocking occurred on 86 sites (19.1 %)
experiencing HSMI outbreaks in 2003-2007. Most (69) of these sites had two HSMI
outbreaks in this time period, while 13 sites had three outbreaks, 3 sites had four, and one site
had five outbreaks.
10.2. Description of disease outbreaks
Anamnestic information collected from disease outbreaks in 2003-2005 showed that clinical
signs leading to a HSMI diagnosis appeared on average 7 months after sea transfer, with a
range of 1 to 11 months (records from the NVI). The average weight of the fish was 1 kg, and
it was uncommon that fish weighed more than 1.5 kg at the time of the first HSMI diagnosis.
After an initial HSMI diagnosis, samples were resubmitted one or several times from about 20
% of the sites. This was due to higher than normal mortality over a longer period of time or as
a result of new mortality peaks. On average, the time from the first to the last HSMI diagnosis
on the site in each generation was 2.39 months, with a range from 1 to 7.
26

10.3. Pathology
10.3.1. Clinical signs and macroscopical lesions
Clinical disease is discovered by the farmer due to the presence of moribund fish gathering
along the cage walls, facing the sea current, or increased mortality in a few or several cages
on affected sites. The mortality is variable from almost insignificant to 20 % in affected cages.
The appetite may be reduced, but this is not always observed (A. Lyngøy, pers.com.).
By necropsy, affected fish are usually in good condition with no external lesions. Internally,
the most obvious pathological signs are often haemopericardium and a pale or greyish heart.
The liver may be pale and yellowish, or dark and congested. Some fish have a fibrinous coat
on the liver surface. Signs of circulatory disturbances such as ascitic fluid and swollen spleen
may also be observed. Fatty tissue surrounding pyloric caeca may have red spots, indicating
petechia or congestion of veins.
10.3.2. Histopathology
Due to the recent discovery of HSMI, no description of the associated pathology had been
published scientifically on the commencement of the present study, and diagnostic criteria
have not yet been firmly established within the scientific community. The macroscopical
changes are relatively unspecific, and usually point towards circulatory disturbances. So far,
HSMI has therefore been diagnosed on the basis of histopathology. A histopathological
investigation of various organs in 404 fish sampled during outbreaks of HSMI on 56 sites in
the period from 1999 to 2007 was performed for this thesis, in order to summarise the most
important lesions in fish sampled during the clinical phase of HSMI. The results are presented
in Tables 2, 3 and 4.
27

11. Aims of study
The aims of the present study were:
1) To describe the pathology of HSMI during clinical outbreaks
2) To compare the pathology of HSMI with diseases having similar pathologies
3) To study the development of lesions in HSMI
4) To investigate potential experimental transmissibility of HSMI
5) To examine the possible presence of a causal agent of HSMI
28

12. Summary of papers
Paper I: Heart and skeletal muscle inflammation (HSMI) is a disease affecting farmed
Atlantic salmon Salmo salar in Norway. It was discovered in 1999, and there has since been a
yearly increase in the number of recorded outbreaks. Fish are commonly affected 5 to 9
months after transfer to sea. Clinical signs and pathology of HSMI from three field outbreaks
in Norway were described. Mortality varied from 4 to 20 %, and moribund fish displaying
abnormal swimming behaviour were observed in affected cages. Necropsy findings included
haemopericardium, a pale or greyish heart, yellowish or brown liver, ascites, swollen spleen
and petechiae in the perivisceral fat. By histopathology, severe inflammation and necrosis was
evident in the heart and red skeletal muscle. Pancreatic lesions were not observed, but several
fish had multifocal liver necrosis, oedema-like meshwork in the spleen and signs of
circulatory disturbances in several organs.
Paper II: HSMI was experimentally transmitted to Atlantic salmon post-smolts by
intreaperitoneal injection with tissue homogenate from farmed Atlantic salmon previously
diagnosed with HSMI, and by cohabitation to injected fish. Injected fish developed epi-,
endo- and myocarditis with mononuclear cell infiltrations consistent with HSMI in compact
and spongy layers of the cardiac ventricle after six weeks. Similar lesions were found in
cohabitants after 10 weeks. Experimental fish also developed inflammation and necrosis of
red skeletal muscle. There was no difference in groups challenged with inoculum containing
antibiotics compared to non-treated inoculum, suggesting a viral aetiology for HSMI.
Paper III: A population of Atlantic salmon were studied from three months post sea transfer
until slaughter. Samples from apparently healthy as well as clinically diseased fish were
collected monthly or every two months. Focal myocarditis was observed from five months
post sea transfer, and lesions consistent with HSMI were seen eight months after transfer to
sea. This was followed by a clinical outbreak with increased mortality, lasting two months. In
this period, most sampled fish had cardiac lesions. Many fish also had lesions in red skeletal
muscle, liver and spleen, but these changes disappeared after the clinical phase. Cardiac
lesions also decreased in severity, but multiple foci of cellular infiltration and necrosis
persisted. However, no fish had cardiac lesions in the final sample before slaughter.
29

Paper IV: The development of pathological changes in HSMI was studied more closely.
Sections of heart and skeletal muscle were examined histologically and tested histochemically
for alkaline and acid phosphatase, unspecific esterase, peroxidase and MCH class II. Early
inflammatory changes in the heart were observed in association with coronary vessels or
endocardium. During the clinical phase, lesions were more diffuse. Inflammatory cells were
morphologically similar to lymphocytes and macrophages, but did not show histochemical
characteristics for these. Few cells contained alkaline phophatase, MCH class II was observed
in a moderate number of cells, and some cells were positive to IgM.
Paper V: Two infection experiments with HSMI were performed by intraperitonal injection
of Atlantic salmon smolts. In the first experiment, fish were injected with cardiac tissue, blood
plasma and cell cultured material. In the second experiment, fish were injected with heart,
liver, kidney/spleen and plasma from fish in the mid-outbreak phase. Also, cardiac tissue
samples collected two months before and after the outbreak phase were used. Finally, cardiac
tissue pre-treated with chloroform was tested. All inoculates induced cardiac lesions in
experimental fish. Perivasculitis and endocarditis first appeared after one week, and focal
myocarditis was seen after three weeks. Most fish had widespread lesions after eight weeks.
Paper VI: A disease outbreak in Scottish Atlantic salmon is described; representing the
diagnostic challenge related to distinguishing between pancreas disease (PD) and HSMI in
areas of endemic PD. Histopathological changes were most prominent in spongy
myocardium, and included vacuolation and degeneration of cardiac myocytes, as well as
infiltration by neutrophils and macrophages. Two fish had aggregations of nuclei, possibly
representing attempts of cardiac regeneration. Skeletal muscle necrosis and myophagia were
present in both red and white fibres, but were more pronounced in red muscle. Antibodies to
the causal agent of PD were detected in all sampled fish, suggesting prior exposure.
30

13. Artikkelsammendrag (summary of papers in Norwegian)
HSMB ble først oppdaget ved rutinediagnostikk av syk fisk i 1999, og har siden blitt et
økende problem for norsk oppdrettsnæring. Laksen rammes som regel etter 5 til 9 måneder i
sjøen. I artikkel I beskrives sykdomstegn og patologiske forandringer ved HSMB i tre ulike
sykdomsutbrudd. Dødeligheten varierte fra 4 til 20 % på merdnivå, og det var utbredt
forekomst av tydelig syk og døende fisk (”svimere”). Makroskopiske funn var blodkoagel i
perikardialhule og forkammer, blekt eller gråaktig hjerte, gul- eller brunfarget lever, væske i
bukhulen, svullen milt og blodopphopning i perivisceralt fettvev. De viktigste
histopatologiske forandringene var uttalt betennelse og nekrose i hjertet og rød
skjelettmuskulatur. Det ble ikke observert forandringer i bukspyttkjertelen. Mange fisk hadde
multifokal levernekrose, ødem-lignende nettverk i milten og blodopphopning i flere organer.
I artikkel II ble det vist at HSMB er eksperimentelt overførbar til frisk atlantisk laks ved
intraperitoneal injeksjon av vevshomogenat fra fisk med HSMB, samt ved kohabitasjon med
injisert fisk. Seks uker etter smitte hadde den injiserte fisken utviklet alvorlig epikarditt og
myokarditt i det kompakte og spongiøse muskellaget i hjertet. Lignende forandringer ble også
funnet i kohabitantene etter ti uker. Forsøksfisken hadde også betennelse og nekrose i rød
skjelettmuskulatur. Forsøksgrupper som hadde blitt injisert med inokulat som inneholdt et
bredspektret antibiotikum (gentamycin) utviklet tilsvarende forandringer, noe som indikerer
en mulig virusetiologi.
I artikkel III ble utviklingen av HSMB på et oppdrettsanlegg for atlantisk laks fulgt i fra tre
måneder etter sjøsetting til slakt. Det ble tatt ut prøver fra klinisk frisk fisk, svimere og
dødfisk hver måned frem til etter HSMB-utbruddet, og deretter annenhver måned. Multifokale
betennelsesforandringer i hjertet ble observert fem måneder etter sjøsetting, og økte i
alvorlighetsgrad i påfølgende uttak. En fisk hadde forandringer forenlige med HSMB etter
åtte måneder i sjø. Påfølgende måned startet et klinisk sykdomsutbrudd med forøket
dødelighet, som varte i 2 måneder. Da hadde nær 100 % av de undersøkte fiskene
hjerteforandringer. Mange hadde også betennelse og nekrose i rød skjelettmuskulatur,
multifokal levernekrose, samt ødem og blodkonsentrasjon i kar i gjeller, milt og nyre. I løpet
av de påfølgende to månedene avtok dødeligheten på lokaliteten, og hjerteforandringene ble
mindre alvorlige. Multifokale betennelsesforandringer kunne imidlertid observeres i flere
31

måneder etter sykdomsutbruddet. Dette studiet viste at HSMB er en alvorlig hjertesykdom
som gir økt dødelighet, høy morbiditet og har lang varighet.
I artikkel IV ble uviklingen av hjerteforandringer og karakteristika ved betennelsesinfiltratene
undersøkt nærmere. Vev fra hjerte og skjelettmuskulatur ble undersøkt ved lys- og
elektronmikroskopi, samt testet for alkalisk og sur fosfatase, uspesifikk esterase, peroxidase
ved enzymhistokjemi, samt for MCH II og IgM ved immunhistokjemi. Tidlige
betennelsesforandringer i hjertet ble observert nær koronarkar eller endokard, mens det var
det en mer diffus betennelse under det kliniske utbruddet. Morfologisk lignet
betennelsecellene på lymfocytter og makrofager, men de viste ikke enzym- eller
immunhistokjemiske karakteristika for disse celletypene. Få celler var positive for alkalisk
fosfatase, og ingen var positive for esterase eller sur fosfatase. Et moderat antall celler
uttrykte MHC II, mens et fåtall celler var positive for IgM.
I artikkel V ble HSMB eksperimentelt overført til laksesmolt i ferskvann ved injeksjon i
bukhulen av blodplasma og hjertevev fra fisk med HSMB, samt ved injeksjon av
vevsmateriale som var passert i cellekultur. Det ble også gjort et smitteforsøk i saltvann hvor
materiale fra hjerte, lever, nyre/milt og plasma ga hjerteforandringer hos injisert fisk. Dette
forsøket viste også at hjertevev som var tatt ut to måneder før og to måneder etter klinisk
utbrudd var infektivt, samt at kloroformbehandling av hjertevev ikke hindret utviklingen av
hjerteforandringer hos forsøksfisken. De tidligste hjerteforandringene kom allerede etter en
uke, og bestod av perivaskulitt i kompaktum og endokarditt i spongiosum. Fokal myokarditt
ble også observert hos få fisk etter tre uker. Etter åtte uker var det høy prevalens av
hjerteforandringer i alle smittegruppene.
I artikkel VI er et sykdomsutbrudd hos atlantisk laks i et oppdrettsanlegg i Skottland
beskrevet. Det ble mistenkt at det dreide seg om HSMB, ettersom vevsskadene lignet
beskrivelser fra norske HSMB-tilfeller og det ikke var forandringer i pankreas. Antistoff mot
PD-virus ble funnet hos samtlige (titer 1/40). Dette indikerer at de hadde vært eksponert for
viruset. Studiet representerer derfor en viktig diagnostisk utfordring i forhold til å skille
HSMB og PD.
32

14. Methodological considerations
14.1. Field samplings
Most fish diseases are characterised according to their clinical and pathological presentation
in field outbreaks, and there is an abundance of papers describing disease problems in farmed
Atlantic salmon see for instance 44,95,102,118,121,134,136 . These case studies provide valuable
information, both for diagnosticians attempting to explain the cause of a practical problem,
and for researchers aiming to understand disease processes and give advice on disease
mitigation. When faced with emerging diseases, the pathological features may be all the
information that is available. The histopathology of diseased fish in field outbreaks is
therefore a natural starting point for further investigations.
The description of HSMI in this thesis is mainly based on studies of fish sampled from natural
disease outbreaks Papers I, III and IV . Samplings performed in these studies included dead,
clinically diseased and apparently healthy fish. This was done to study the range of lesions
associated with HSMI, as well as to get an idea of the prevalence of the disease on cage and
farm levels in the outbreak situation.
Longitudinal studies of PD have provided important information about the chronology of
events in the disease development 95,102 . This approach was therefore also used to study the
development of HSMI Papers III and IV . The sampling frequency in the present study was monthly
before the clinical phase, and later every two months. This enabled a crude, but long lasting
study, pointing at key events in the pathological development.
A disadvantage of studying field material is that one does not have control with concurrent
pathologies caused by other diseases 93 . This may partly be overcome by studying disease
outbreaks in several populations and at different time points, as well as gaining knowledge of
other possible causes of the lesions observed. When dealing with infectious diseases,
experimental studies performed in a controlled environment may also be a support for
establishing diagnostic criteria and determining processes involved in disease development.
33

14.2. Infection experiments
The most important purpose of infection experiments is to study the transmissibility of a
disease to naïve fish. While studies of field cases may indicate contagiousness, experimental
transmission of a disease is generally considered as evidence for the involvement of an
infectious agent in the aetiology 97,102,126 . When performing infection experiments,
environmental and other factors are kept as constant as possible. It is therefore possible to
study pathological and other effects of exposure to the infectious agent alone 29,60,148 . It is also
possible to study features of the disease or causal agent that are not easily studied in field
material, such as for instance the infectivity of various tissues and characteristics of the causal
agent by different treatments 28,64 . The sensitivity of experimental fish to pathogens may vary
with physiological stage, seasonal changes and fish strain 94 . This may affect the outcome of
infection experiments. In addition, the design of the challenge study may influence the clinical
presentation of the diseases. For instance, the choice of tissue used as inoculate material,
preparation of inoculates, storage and other factors may be of importance. Thus, a negative
result may be difficult to interpret, and does not necessarily mean that the disease is not
transmissible.
Three infection experiments with HSMI were performed in the present study Papers II and V . The
main variable of the studies was a reproduction of the histopathological changes that had been
observed in fish from natural outbreaks. A standard parallel design was used in all trials, and
fish were mainly challenged by intraperitoneal (i.p.) injection. This approach is commonly
used in challenge experiments with known and unknown infectious agents affecting farmed
fish 11,25,101,155,156 .
Cohabitation of healthy experimental fish with injected fish was also used in two groups in
the first experiment Paper II . Transmission through water is likely to be a more natural route of
infection in real life 151 . The possibility of horizontal transmission of fish pathogens is
therefore often tested by the use of bath challenge or cohabitation experiments 64,161 . Bath
challenge is usually performed in experiments with small fish, while cohabitation is a more
practical approach when working with larger fish 64,156 .
Filtration of the inoculate through 200-220 nm filters is a common strategy for determining
whether a causal agent could have the size of a virus 25,92,101,125 . This was not attempted in the
34

described experiments because some viruses tend to adhere to larger objects, thus not being
able to pass through a fine filter (K Falk, pers.com.). The experiments described in this thesis
were mainly performed to determine the transmissibility of HSMI from diseased to healthy
fish and to study the pathological development, thus this crude approach. In later experiments,
filtered material may also be tested, especially if the causal agent is not successfully isolated.
The first infection experiment was run for 12 weeks, with samplings every two weeks Paper II .
In the second and third experiments, samples were mostly taken weekly in the eight-week
study period Paper V . This appeared to be sufficient to secure the diagnosis and to study the
early pathological development of HSMI. More frequent samplings and a longer experimental
period in future studies may provide additional information about the pathogenesis of the
disease. However, the usefulness of this must be weighed against economical and ethical
considerations associated with infection experiments.
14.3. Histopathology
Currently, histopathological examination of tissue by light microscopy is the only diagnostic
tool for HSMI. This method was therefore of vital importance for the work performed for this
thesis Papers I-VI . Diseased and dead fish often show unspecific macroscopic changes on
necropsy. Histopathology allows the diagnostician to perform a more detailed examination of
microscopic changes in affected tissues. By this method it is therefore possible to differentiate
between diseases with similar disease histories or clinical signs.
Tissue samples used for histopathological examination are usually fixed in phosphate buffered
formalin immediately following resection of the organ. Fixation is achieved by cross-linking
of protein molecules in the tissue. Formalin is a slow fixation liquid, and the histological
structure generally stabilises after one week of fixation. Specimens may be kept in formalin
for a long time, but staining abilities of cell structures may be reduced with time. Following
fixation, water is removed from the tissue with alcohol and xylene, and is replaced by melted
paraffin wax 68 . By this method, the tissue becomes solid, and one may easily cut thin sections
of 1-10 m with a microtome 67 . Sections are mounted on glass slides.
Contrast in tissue sections is produced by the use of dyes. The most common method in
histology is staining with haematoxylin and eosin (HE). Haematoxylin is a product from a tree
35

found in South and Central America (Haematoxylon campechianum). Oxidation changes
haematoxylin to haematein. Aluminium forms a bridge between hematein and nuclear
chromatin. This creates a blue colour. Eosin is derived from fluorescein. It is anionic and
stains most components of the cell red, such as cytoplasm, collagen and keratin 68 . An
overview of histological staining methods used in the present study is presented in Table 5.
Pathological descriptions of disease concentrate on deviations from normal structure. The
histopathological diagnosis is, however, made more complex by individual differences, post
mortem changes and concurrent diseases. Histopathology therefore requires considerable time
and experience, as subtle changes may easily be missed or misinterpreted. This is reflected in
the present study, in which the degree of detail in the observations increased as time went by.
14.4. Electron microscopy
Transmission electron microscopy was used in Paper IV to study the ultrastructural pathology
of HSMI. A common way to fix tissue for electron microscopy is the use of double fixation
with 2-4 % glutaraldehyde and 1-2 % osmium tetroxide in buffers containing cacodylate,
phosphate or tris-maleate. This enables a good fixation of most structures and also provides
some contrast. The fixatives penetrate the tissue at a slow rate, and pieces for electron
microscopy should therefore not exceed 1 mm 3 . Following fixation, the samples are taken
through a dehydration process in ethanol with increasing concentration. Ethanol is thereafter
replaced by propylene oxide, which enables a more efficient infiltration with resin. However,
propylene oxide is carcinogenic, and may also potentially be explosive. In addition, it
dissolves most plastic containers. For preparation of cell culture monolayers in plastic flasks it
is therefore most practical to omit the propylene oxide step. The tissue samples are gradually
infiltrated by the resin, which are epoxy-monomers that will polymerize with time and
become solid. The polymerization process is usually performed at 60 ºC for 1-3 days. Epoxyembedded
specimens are cut to 30-60 nm sections by ultramicrotomy. They are thereafter
placed on a metal grid, and are ready for use.
In the electron microscope, structures are differentiated by their different densities. As most
tissue structures have similar densities, contrast must be added by reacting some of them with
heavy metals. In addition to the contrast provided by fixation with osmium tetroxide, contrast
may be produced en bloc or on each section. These methods are either used separately or,
36

most commonly, in combination. En bloc contrasting may be combined with the fixation step,
such as mixing osmium tetroxide with potassium ferrocyanide. It may also be applied
between the fixation and dehydration steps, as is done with uranyl acetate. Contrast staining
on sections is usually done with lead citrate 12 .
The large range of magnification in transmission electron microscopy allows for examination
of cells in greater detail, as well as search for small pathogens within cells, such as viruses.
Preparation of specimens for electron microscopy is, however, more expensive and timeconsuming
than preparation for histology. In addition, it is possible to spend considerable
work hours on each prepared section. This tool is therefore best used as a supplement to light
microscopy.
14.5. Enzyme histochemistry
As inflammation appears to be an important feature of HSMI, the cells involved in the
reaction were studied in Paper IV. There is a lack of specific phenotypic markers of fish
leukocytes 17 . However, enzyme histochemical methods used to identify mammalian
leukocytes have previously been tested on salmonid cells 6,37,53 . Some of these were therefore
applied on sections from fish with HSMI Paper IV . Reagents tested were alkaline and acid
phosphatase, unspecific esterase and peroxidise. There are no major differences in immune
and enzyme histochemical reactivity in Atlantic salmon compared to other fish species.
However, individual differences in enzyme reactivity may be observed 123 . Described
reactivities for the reagents used are listed in Table 6. The idea of using enzymes in
histochemical studies is that they function as catalysts in chemical reactions between organic
substances (substrate) and other compounds. In enzyme histochemistry, the resulting product
is an insoluble and coloured substance which is fixed on the site where it has been formed.
The colour signal is visible by light microscopy. It is therefore possible to identify cells by
studying their enzymatic activities 68 .
Alkaline and acid phosphatase and esterases are hydrolytic enzymes found in leukocytes.
Their role in the immune process is not clear, but they may take part in the defense against
microbial infection 38 . Functionally, phosphatases are involved in hydrolysis of phosphoric
acid esters. Alkaline phophatase is especially associated with areas where endo- and
pinocytosis take place in the cell, and in mammals they are found in the cytoplasm of
37

neutrophils and basophils. Acid phosphatase is found in lysozymes, and is abundant in
phagocytes 68 . Esterases also catalyse hydrolysis of ester bonds 38 . Several types of esterases
are found in monocytes and macrophages. The esterase most commonly used in fish enzyme
histochemistry acts on -napthyl acetate. In mammals, it is present in monocytes and Tlymphocytes.
Peroxidases are important for oxidation reactions 68 . In the immune response,
peroxidase acts during phagocytosis as a catalyst in a reaction between hydrogen peroxide and
chloride ions, resulting in the production of antibacterial substances 38 .
Although widely used on tissue sections, enzyme histochemical methods are largely
developed by studying chemical reactions in solutions containing purified enzymes. The
availability of enzymes in tissues may be considerably lower than in solutions, due to their
presence within organelles. Also, enzymes are not constantly active, and may fail to bind to
added substrates. In addition, enzymatic activity may be impaired by fixation methods. It is
therefore not always possible to detect enzymes present in the tissue by these methods 68 .
14.6. Immunohistochemistry
The presence of immune cells was explored in Paper IV by the use of immunohistochemistry
for demonstration of MHC class II and Ig Table 6 . Immunohistochemistry is based on the
binding forces between antigens and antibodies. Specific antibodies against surface structures
of microbia or fish cells (primary antibodies) may be produced in laboratory mice or rabbits
and thereafter used for detection of these structures in tissue sections. In
immunohistochemistry, the primary antibody binds to corresponding antigen in the tissue,
forming an insoluble compound. A secondary antibody directed against the primary antibody
and carrying a colour signal binds to the antigen-antibody complex, thus visualising the
presence of antigen 68 . MHC class II and Ig are components of antigen presentation and
clearance in the adaptive immune system 122 . Methods of immunohistochemical detection of
these structures have been developed for Atlantic salmon 37,74,123 . As with enzyme
histochemistry, visualisation of the structures may be reduced due to methodological
problems. For instance, formalin-fixation may mask binding sites for antibodies, thereby
preventing binding of primary antibody. Also, membranes and extracellular matrix may
reduce access to the antigen. This problem may often be overcome by the use of proteolytic
enzymes, surfactants or heating, but long-term fixation may cause permanent masking of
antigen 68 .
38

14.7. Microbiological examination
14.7.1. Bacteriology
Initially, it was thought that HSMI may be caused by pathogenic bacteria (T Taksdal,
pers.com.). This hypothesis was therefore tested in the first infection experiment Paper II . In
Papers I and II, kidney samples were examined for bacteria by cultivation on blood agar with
and without 2 % NaCl. This is a standard method for isolation of the most common
pathogenic bacteria in fish 107 . Also, bacteriological examination was routinely performed by
fish health services in the longitudinal study Paper III and in the investigation of other disease
outbreaks. Not all bacteria grow well on blood agar, and cardiac tissue sampled from two field
outbreaks of HSMI were therefore additionally tested for the presence of 16S r ribonucleic
acid (RNA) by PCR 1,171 in a general investigation of bacterial infection (unpublished). The
method described by Suau et al. 159 was followed, using the primers S20 and A18. The
possible presence of bacteria was also investigated by histopathology, including special stains,
and transmission electron microscopy.
14.7.2. Virology
The standard method for virus isolation is inoculation of material from infected tissue onto
cell culture. This is a very sensitive method, providing that the virus is capable of infecting
the cells used 100 . There is a range of cell lines developed from fish cells 49 . Cell cultures used
in the present study are presented in Table 7 (results from trials with CHH-1, SHK-1 and TO
unpublished). These cell lines have been used to isolate several fish viruses from farmed
Atlantic salmon in Norway, for instance infectious pancreatic necrosis virus (IPNV) 75 ,
infectious salmon anaemia virus (ISAV) 26 , Atlantic salmon paramyxovirus (ASPV) 76 and
salmonid alphavirus (SAV) 18 . In total, heart and/or kidney samples from 63 fish were tested
in the cell lines in the present study.
A standard approach for cell culture work in fish is to incubate inoculated cell cultures at 14-
15 ºC for a length of time, commonly 1-2 weeks 18,27,76 . Part of the supernatant from cell
culture flasks believed to contain viral particles is thereafter passaged to naïve cell cultures in
other flasks and further incubated. By this method, it is possible to propagate the virus in vitro
for further characterisation. Infected cells usually show cytopathic effect (CPE). The most
39

frequent findings are vacuolation, rounding of cells and detachment from the monolayer.
However, some viruses may grow in cell culture without producing CPE 100 .
Experiences from work with other viral diseases causing lesions in cardiac and skeletal
muscle have shown that haematopoietic tissue is infective during viraemia 16,61 . Cell cultures
were therefore mainly inoculated with material from kidney tissue initially. As the
experimental results showed that the causal agent must be present in the heart when typical
histopathology for HSMI is observed Paper II , however, cardiac samples were also included in
some later trials. A standard procedure of three blind passages every 14 days was usually
followed, but in a few trials, blind passages every seven days were also attempted. In two
trials, the first passage was kept parallel with the other passages for six weeks.
Following the successful experimental transmission of HSMI from diseased to healthy fish
described in Paper II, a broad cell culture trial was performed. Material chosen for this trial
was based on the assumption than viraemia occurred at four weeks post challenge. Based on
experience with other viral diseases, it was assumed that the time of viraemia was equal to the
time lag between the appearance of cardiac lesions in injected and cohabitating fish Paper II . In
the experimental transmission of HSMI, this delay was about four weeks. Samples from
injected fish showing lesions at eight weeks post challenge were also included, as tissue from
fish in field outbreaks showing histopathological changes was infective to the experimental
fish Paper II . Material from this trial was successfully transmitted to experimental fish Paper V .
Investigations for determining the possible presence of known viruses in fish with HSMI were
also performed Papers I-III . Standard PCR procedures were used to identify genomic fingerprints
of ISAV, ASPV and SAV, and serological methods were used to investigate prior exposure to
SAV Paper III,31,50,160 .
40

15. Main results
• The clinical phase of HSMI was characterised by epi-, endo- and myocarditis and
inflammation-associated necrosis in the heart, as well as myositis and inflammation-
Papers I, III and IV
associated necrosis of red skeletal muscle
• Lesions were most frequently found in the heart
Papers I-III
• Many fish with moderate or severe cardiac lesions may also had multifocal liver necrosis
and circulatory disturbances in several organs, including an oedema-like meshwork in the
Papers I and III
spleen (“pseudolobulation”)
• The disease was histopathologically distinguishable from PD and CMS, but concurrent
Papers I-III and VI
diseases and atypical cases posed diagnostic challenges
• In a longitudinal field study, focal myocarditis was observed four months before the
clinical phase. Following the clinical outbreak, cardiac lesions persisted for several
Paper III
months
• During clinical outbreaks of HSMI, morbidity appeared to be close to 100 %
Papers I and III
• Skeletal muscle lesions and lesions in other organs appeared to be restricted to the clinical
Paper III
phase
• Early cardiac changes included inflammatory infiltration around vessels in epicardium and
compact layer, and hypertrophy and inflammation of endocardium in the spongy layer of
the ventricle Papers IV-V . A possibly preceding lysis of cells in the compact layer was
Paper IV
observed in field material
• The cell types involved in the inflammation could not be determined by enzyme- and
Paper IV
immunohistochemstry
• HSMI was experimentally transmissible from diseased to healthy Atlantic salmon by
o intraperitoneal injection of material from heart, liver, haematopoietic tissue or
Papers II and V
blood plasma from fish sampled during the mid-outbreak phase
Paper II
o cohabitation to fish injected intraperitoneally with material from heart tissue
o intraperitoneal injection of material from heart sampled 2 months before and 2
Paper V
months after the clinical phase
• The time from injection of infective material to the appearance of inflammation of the
heart and red skeletal muscle was 3-8 weeks in injected fish, with a 4-week delay in
Papers II and V
cohabitating fish
• Transmission of HSMI was not prevented by pre-treatment of the inoculate with
Paper V
chloroform
41

16. Discussion of results
When this project started, almost nothing was known about HSMI. The work in the present
study therefore had to start from scratch. To reach a certain level of understanding of this new
disease problem, there was therefore a need to apply basic methods such as histopathology,
electron microscopy, infection experiments and classical virology and bacteriology. Some
advantages and disadvantages with these methods are discussed in chapter 14. The findings
were compared with knowledge gained from similar diseases in fish and mammals, with
emphasis on other diseases in Atlantic salmon.
16.1. Diagnostic criteria for HSMI
When investigating a suspected HSMI outbreak in the field, the diagnostician is faced with a
lack of specific diagnostic tools. Supportive information may be found in the anamnestic
information and test results showing an absence of known pathogens. Ultimately, however,
the diagnosis must be based on histopathology. This is a common approach for emerging
diseases of unknown cause, such as for instance CMS 15 .
From the very beginning diagnosticians at the NVI observed that HSMI was a severe
inflammatory condition, mainly affecting heart and red skeletal muscle. In the present study,
severe inflammation was seen both in samples from field outbreaks and infection experiments
Papers I-IV
. The changes in these tissues were widely distributed, and were the most consistent
findings. In addition, it was often observed that many severely affected fish had multifocal
liver necrosis and signs of circulatory disturbances in several organs Papers I and III . However,
changes in the heart and skeletal muscle were by far the most frequent findings Papers I and III .
During a clinical outbreak, however, it is likely that at least some fish in a sample of
moribund or dead individuals show all the described lesions Papers I and III . The diagnosis should
thus be based on the general impression from examination of several fish from the same site.
One or several of the histopathological changes observed in HSMI are also observed in other
diseases; see the discussion of important differential diagnoses in 16.2. From this follows that
variations in the pattern of lesions due to different stages of the disease development or other
factors may cause diagnostic problems. Especially in relation to PD, the HSMI diagnosis may
be difficult, as differences in cardiac and red skeletal muscle lesions may be subtle 93 . In
42

atypical cases and concurrency with other diseases therefore, the HSMI diagnosis is far from
clear. In the present study, clinically diseased fish tended to show more severe lesions than
apparently normal individuals, and inflammatory cells were on average more abundant in
clinically diseased than in clinically healthy fish Papers I, III and IV . In mildly affected fish, the
typical pathology for HSMI may be more or less absent, as the heart may be the only organ
showing lesions Paper III . A list of diagnostic criteria for HSMI may therefore have limited
value on apparently normal fish and fish examined in farms with no apparent disease
problems. However, the present study showed that even fish with no clinical signs may have
severe lesions consistent with HSMI Papers I, III and IV . This lack of correspondence between
clinical signs and histopathological changes has also been observed in diseases in other
animals 165 . A possible explanation for this is that a protected life in a sea cage requires
considerably less of the organ capacity to function sufficiently for the fish than a life in the
wild 114 .
Overall, a high number of fish sampled during the mid-outbreak phase of HSMI showed a
Papers I and
characteristic pattern of lesions, and there was not a great deal of individual variation
III
. Also, the pattern of pathological changes did not change significantly with time, as changes
observed in outbreaks occurring in the period 1999-2007 were similar Papers I-V . Thus, in most
HSMI outbreaks it will be possible to give a relatively clear diagnosis based on
histopathological examination of a relatively small number of diseased fish from each case. In
summary, the diagnostic criteria of HSMI may be as in the following:
1) Obligatory findings:
a. moderate or severe myocarditis and inflammation-associated necrosis involving
compact and spongy myocardium of the cardiac ventricle
b. moderate or severe epicarditis
c. moderate or severe endocarditis
d. cellular infiltration is mononuclear
e. inflammation is more dominant than necrosis
2) Supportive findings:
a. myositis and necrosis of red skeletal muscle
b. endo- and myocarditis and associated necrosis involving the cardiac atrium
c. multifocal liver necrosis
43

44
d. “pseudolobulation” of the spleen
e. signs of circulatory disturbances in several organs
f. absence of pancreatic lesions (see 16.2.)
16.2. Distinction from other diagnoses
As the most important differential diagnoses to HSMI appear to be PD and CMS, the
discussion on differential diagnoses will be limited to these in the following. During disease
outbreaks, HSMI is histopathologically distinguishable from these diseases, as discussed in
Papers I, II and III. See also the previous discussion on diagnostic criteria for HSMI in 16.1.
16.2.1. CMS
CMS was first discovered in Norway, and has later been described from Scotland, the Faeroes
and Canada 3,13,44,136 . In Norway, CMS has been observed along most of the coastal line, but
has been most frequently diagnosed from Mid-Norway (Records from the NVI). Fish are
generally affected during the second year in seawater, on average 410 days post transfer 14,15 .
The cause is not known, and the discussion of possible nutritional, environmental and
infectious aetiologies is still going strong 65,71,108,117,150 . However, knowledge on CMS
pathogenesis recently took a leap forward, as lesions consistent with CMS was successfully
transmitted to experimental fish by intraperitoneal injection of cardiac and renal tissue
homogenates from diseased fish (Fritsvold et al., manuscript in prep.). Also, indications of a
viral aetiology have been observed 52,167 , but Koch’s postulates 100 have not yet been fulfilled.
CMS is a severe chronic inflammatory condition affecting endocardium and spongy
myocardium in the atrium and ventricle 3,44,136 . Pathological changes mostly appear to be
restricted to these tissues, but occasional inflammation has been observed in epicardium and
compact myocardium, especially in association with coronary vessel branches. Haemorrhagic
liver necrosis may also be observed, possibly as a result of circulatory failure 136 . Examination
of live fish by the use of diagnostic ultrasound has shown that the atrium is less defined and
the ventricle more compressed in fish with CMS, compared to healthy fish 143 . Skeletal
muscle is usually normal 44 . Common features of CMS and HSMI are their cardiac
involvement and chronic inflammatory nature.

Unlike HSMI, CMS almost exclusively affects spongy myocardium in the ventricle and
especially the atrium 44, Papers I and III . In most cases, the diseases may therefore easily be
differentiated on the basis of the tissues showing pathological changes Paper III . Also, CMS is
most commonly observed in larger fish than those usually affected by HSMI 14 . However,
both CMS and HSMI may affect fish during the entire seawater phase, and one should
therefore not rely on this criterion alone. The geographical distribution of the two diseases are
similar, as most cases of CMS and HSMI are reported from Møre og Romsdal, Sør-Trøndelag
and Nord-Trøndelag counties (records from the NVI). Also, the aetiologies of both diseases
are uncertain. Histopathology is therefore the most important method of differentiation. There
are only a few published descriptions of CMS pathology 3,44,136 , and in the present study
recent cases of CMS were also studied Paper III .
Qualitatively, the histopathology of spongy myocardium in HSMI is similar to CMS. In both
diseases, the changes in this tissue may be characterised as mononuclear inflammation
affecting myo- and endocardium Paper I, 44 . In severe cases, there is a substantial loss of
myocardium in CMS. In some areas, myocardial fibres may be almost completely replaced by
inflammatory cells, lined up like beads on a string under the endocardium 116 . This is not often
observed in HSMI, but may occur. Some researchers have focused on the severity of
endocarditis in CMS, leading to a hypothesis of an initial endocardial insult as the cause of the
subsequent pathology 52,116 . Endocardial lesions are also an important part of the
histopathology of HSMI, but may be less pronounced than in CMS cases Papers I, IV and V .
Common for the two diseases is the observation that even though there may be scattered loss
of endocardial cells, most of the endocardium remains intact or is able to regenerate, even in
severe cases Papers III and IV, 116 . If examining spongy myocardium alone, there are no clear
differences between CMS and HSMI, but a holding point could be that atrial lesions are
almost always more severe in CMS than in HSMI. Thus, in the event of a possible
concurrency of CMS and HSMI in a fish, an extensive atrial inflammation and loss of
myocardium may be of support for the diagnosis.
16.2.2. PD
PD was first discovered in the British Isles, where it caused losses in farmed Atlantic salmon
during the first year in seawater 99 . Reports from recent years, however, show that the
majority of cases occur in larger fish 93 . In the acute phase, fish dying from PD have necrosis
45

of the exocrine pancreatic tissue. Later on, there is significant loss of exocrine pancreas.
These lesions gave rise to the name. In 1986, Ferguson et al 45 described necrosis and
inflammation of compact and spongy myocardium and red skeletal muscle in fish with PD.
The association between myopathy and PD was initially contradicted 96 , but was later
demonstrated in a retrospective study 140 . For almost 20 years, PD was thought to be due to
nutritional deficiencies, but in 1995, Nelson et al 106 demonstrated a viral aetiology. It was
later shown that the disease is caused by an alphavirus 92,106,140,169,170 . This virus was first
named salmon pancreas disease virus (SPDV). After the discovery of closely related viruses
causing similar pathologies, the virus causing PD in Scotland and Ireland are now referred to
as SAV 1, and the Norwegian PD-virus is called SAV 3 56 .
It has now been established that the initial exocrine pancreatic necrosis with subsequent loss
of pancreatic tissue in PD is followed by a primary necrosis of cardiomyocytes and secondary
inflammation involving both compact and spongy myocardium, as well as skeletal myopathy
and myositis mostly in red, but also white skeletal muscle 29,91,95,101 . Due to the lesions in both
heart and red skeletal muscle, PD is the most important differential diagnosis to HSMI. There
has been an ongoing debate in the scientific community for several years about the
justification of calling HSMI a separate disease, instead of a variant of PD 93 .
In Norway, PD was only observed in Hordaland County in Western Norway between 1995
and 2003. When PD outbreaks started appearing in other parts of the country, HSMI had been
reported for some years 160 . During most of the work with the present study, PD had never
been observed in the three counties where the highest number of HSMI cases is recorded.
This enabled a search for SAV in fish with HSMI, in order to determine whether this agent
could be associated with the new disease. It soon became clear that SAV was not retrievable
from HSMI fish by the diagnostic methods routinely used for PD Papers I-III , including
conventional and real-time RT-PCR 56,57,170 , inoculation onto cell cultures susceptible for
SAV 18,106 and serology 50 . As the causal agent of HSMI was present and infective in
moribund fish sampled during the clinical phase Papers II and V , this strongly suggests that HSMI
and PD are caused by separate infectious agents. The causal agent of HSMI was not
successfully isolated Papers I and II , but experimental transmission using a chloroform pre-treated
inoculate Paper V indicated that the causal agent may be non-enveloped. If this result is
repeatable in future experiments, it will confirm that HSMI is not caused by SAV.
46

In addition, the studies of the HSMI pathology showed that there were evident differences
between the two diseases Papers I-III . The lesion locations of PD are similar to HSMI, as they
both affect compact and spongy myocardium, as well as red skeletal muscle. In PD, however,
necrosis and loss of exocrine pancreatic tissue also occurs 93,99,102,160 . In Norwegian cases of
PD, the pancreatic changes are usually observed concurrently with the cardiac and skeletal
muscle lesions 160 . Pancreatic lesions were not observed in fish with HSMI, neither before,
after nor concurrently with the clinical phase when cardiac and skeletal muscle lesions were
present Papers I-III 93,95, Papers I and
. Multifocal liver necrosis as seen in HSMI is not common in PD
III 160
. If present in PD cases, it is rather observed as single cell necrosis or apoptosis . More
widespread necrosis may, however, occur as a result of circulatory failure in several diseases.
This is observed in CMS and infectious salmon anaemia (ISA) 136,155 , and may potentially
occur in severe cases of PD. Red skeletal muscle lesions were similar in PD and HSMI, but
appeared more fibrotic in PD and more inflammatory in HSMI 93,138, Papers I-III . Also, there is
usually an involvement of white skeletal muscle in fish with PD 93,95,160 , while skeletal muscle
lesions in HSMI were mainly restricted to red muscle Papers I and III . In late stages of PD in
Norway, an accumulation of cells containing eosinophilic granules are observed in cells
present along the renal sinusoids 160 . This was not observed in fish with HSMI Papers I-III .
In the heart, an extensive epicarditis was more frequently observed in HSMI Papers I-V , and the
myocardial changes were generally more severe and widespread than in PD cases. There was
also a qualitative difference in the myocardial lesions of the two diseases. A characteristic
early myocardial lesion in fish with PD is necrosis of cardiomyocytes 45,95,160 . Thus, the
myocardial inflammation in PD does not appear to be primary, but may start as a result of the
initial necrosis 95,138 . In PD hearts, therefore, there is a mixture of inflammatory foci and
necrotic cardiomyocytes that are not surrounded by inflammatory cells during a clinical
outbreak 160 . Necrotic cardiomyocytes show marked eosinophilia by HE staining, and are
therefore clearly distinguishable from surrounding normal cells 21 . In clinical field outbreaks
and infection experiments with HSMI, myocardial necrosis was consistently been observed in
association with inflammation Papers I-V , and HSMI mainly appeared to be an inflammatory
condition. By sequential sampling, no period of dominant myocardial necrosis was observed,
and inflammation in the heart appeared at an early stage in the disease process Papers II-V .
Possibly primary ultrastructural lesions observed in pre-outbreak field samples Paper IV indicate
that the inflammation may possibly be set off by an initial tissue necrosis in association with
vessels. However, these changes were not visible by light microscopy, as they are with PD 93 .
47

Also, as the lesions were observed in field material, it is uncertain whether they are at all
associated with HSMI. The longitudinal study Paper III showed that farmed fish may experience
outbreaks of several different diseases during the production cycle. Pathological changes in
fish with no clinical signs may therefore be caused by some other disease process, and their
interpretation is uncertain.
The major differential diagnostic problems arise in areas of endemic PD 93,160 . The diagnostic
tools for SAV will not be of much use in such areas, as there is a high risk of detecting
subclinical infection with SAV as an incidental finding. Thus, the observed histopathology
may not be caused by the SAV infection, but it will be difficult to prove that there is no causal
relationship. In addition, the pathology in endemic areas may possibly be more heterologous
than in previously naïve areas. In Paper VI, such a problem is discussed. A disease outbreak
resembling HSMI occurred in Scotland, and the major question was whether this represented
the introduction of a new disease into the UK, or if this was a new variant of PD. Neutralising
antibodies to SAV were found in all fish, and the exposure to SAV at some stage in the
production was therefore not doubted. The histopathology was different from PD in that
pancreatic lesions were absent, there was evident liver necrosis and the lesions in heart and
skeletal muscle were more severe than usually observed in Scottish PD-outbreaks. The PDvariant
occurring in the British Isles is characterised by a more rapid regeneration of exocrine
pancreas than in Norwegian PD 93 , and this could be the case in this outbreak. The remaining
problem was therefore the marked difference in pathologies of the liver, heart and skeletal
muscle.
The pathology did, however, not mirror that of HSMI Paper I . The inflammatory response in the
heart and skeletal muscle was not as pronounced as observed in Norwegian HSMI outbreaks,
and the inflammatory infiltrate mainly consisted of neutrophils and macrophages. The
presence of neutrophils is unusual both in HSMI and PD 93 . The lesion location was also
somewhat different from HSMI. In the Scottish disease outbreak, spongy myocardium was
more severely affected than compact myocardium. As described above for HSMI, the
opposite is often the case, but cardiac changes have a similar distribution in some HSMI
outbreaks (unpublished). Also, white skeletal muscle was more involved than is common in
HSMI cases. What actually caused the pathology in the Scottish disease outbreak remains
uncertain, but neither PD nor HSMI could be ruled out.
48

16.3. Pathogenesis
The development of lesions in various organs and the cells involved were examined in a
longitudinal field study Papers III and IV , and by experimental infection Papers II and V .
16.3.1. Pre-clinical phase
The first cardiac lesion to be observed in the longitudinal field study Paper III was focal
myocarditis in the spongy layer of the ventricle. This was observed as early as four months
before the clinical outbreak. The association of this lesion to HSMI is highly uncertain, but in
the following samplings, the number of inflammatory foci increased in examined fish. Also,
an initial mild epicarditis commonly observed in farmed fish 116 , increased in severity in the
months prior to the clinical disease outbreak. By experimental infection, the time from initial
observations of myocarditis to development of severe inflammation and necrosis of cardiac
and red skeletal muscle consistent with HSMI was 2-4 weeks Paper II . If the early changes in the
field study Paper III represented an early stage of HSMI, disease development in the field would
therefore be long compared to the experimental situation. This has also been observed in ISA
164
. It is likely that environmental and other factors contribute to a prolonged developmental
process in the field. Cardiac samples taken two months before the clinical phase from fish
Papers III and IV
showing mild multifocal myocarditis, endo-/perivasculitis and/or epicarditis
caused cardiac lesions consistent with HSMI in experimental fish Paper V . This showed that the
causal agent of HSMI was present in the population while the fish were still at the first
seawater site Paper III . Also, the development of HSMI pathology followed similar patterns in
field and experimental studies Papers IV and V . It is therefore likely that the early myocardial
changes observed in the longitudinal study Paper III may indeed have been part of the HSMI
pathogenesis.
Following experimental infection, epicarditis and perivasculitis associated with coronary
vessel branches within the compact layer and endocarditis in the spongy layer were mostly
observed before the onset of myocarditis Paper V . These results indicate that epi- and
endocardium and perivascular tissue may have a role in the early responses to infection with
the causal agent of HSMI. Similar changes were also seen prior to the clinical phase in the
longitudinal study Papers III- IV . It may often be difficult to determine whether the primary
lesions occur in the endocardium, myocardium or epicardium 116 , but results from the
experimental studies Paper V indicated that myocardial changes succeed changes in epi- and
49

endocardium in the development of HSMI. As natural spread of a pathogen in the field occurs
gradually, two fish in a sea cage could be at completely different stages in the disease
development at any given time 160 . It is therefore possible that the myocarditis first observed
in the longitudinal study Paper III , in fact may have been a secondary response to an initial
endocardial response.
Lysis of cells close to coronary vessel branches, possibly preceding inflammatory processes,
was observed by electron microscopy two months before the clinical phase Paper IV . This could
not be found by light microscopy of the same fish. Acute inflammatory responses are likely to
induce vascular and perivascular tissue damage 98 . No inflammatory cells were observed in
the vicinity of the lesions. However, other fish in the same sample showed signs of
inflammation. As this was found in field material, and only in two fish, it is uncertain whether
these changes are a part of the development of HSMI, or if they are caused by some other
disease process. Farmed fish are exposed to several pathogens during the production cycle
4,111
. This was also observed in the longitudinal study, as several diseases were diagnosed on
the studied farm Paper III . As support for this idea of a separate disease process, no early lytic
changes were observed in the infection experiments Papers II and V . However, only a few hearts
from these studies were examined by electron microscopy. There is thus a need for more
thorough examination of experimental material to determine the possible presence or absence
of such ultrastructural changes.
What comes first of myocardial damage and inflammation in HSMI? Even though possibly
early ultrastructural necrosis did not appear to be associated with inflammatory cells Paper IV ,
observations of myocardial necrosis by light microscopy were exclusively associated with
inflammatory cells Papers I-V . This is in contrast to PD, in which myocardial necrosis is evident
in the early stages of myocardial damage 93 . In PD, early myocardial damage is easily
observed by light microscopy by myocardial shrinkage and eosinophilia of single cells 93,160 .
If primary necrosis was part of the HSMI development, it should therefore be possible to
discover this by light microscopy. In HSMI, inflammatory infiltration was the most dominant
finding during the histologically observable period of disease development, and it quickly
became very extensive Papers I-V . This indicates that HSMI is primarily an inflammatory
condition, and that necrosis occurs secondarily as a result of the inflammation.
50

16.3.2. Clinical (outbreak) phase
Following the more subtle cardiac changes in the early samples of the longitudinal field study
Papers III and IV
, the inflammatory nature of HSMI became evident as time drew near the clinical
outbreak. The clinical phase was defined as the period with increased mortality due to HSMI.
In the longitudinal field study Paper III , cardiac lesions became significantly more severe in the
clinical phase than prior to it. Inflammation was the dominating feature, but there was
considerable myocardial necrosis associated with inflammatory foci. A large sample from the
study groups in the longitudinal study Paper III showed that the morbidity was close to 100 %. A
high number of fish showing cardiac lesions were also been observed in other clinical
outbreaks Paper I .
In the clinical phase, lesions were also observed in other organs Papers I and III . Most evidently,
inflammation and necrosis of red skeletal muscle was observed in most fish with moderate or
severe cardiac lesions. In addition, there was multifocal liver necrosis, splenic
“pseudolobulation”, erythrocyte concentration and other signs of circulatory disturbances in
several organs. It is not known whether lesions in other organs are directly caused by the
infection, or if they occur as a result of the changes in cardiac tissue. In mammals, randomly
scattered single cell or multifocal hepatocytic necrosis is generally caused by infectious
agents, while hypoxic conditions, such as cardiac failure, may cause centrilobular or
periacinar necrosis of hepatocytes 85 . The pattern of hepatocyte necrosis in HSMI did not form
an evident zonal pattern, and was only associated with about half of the vessels. The causal
agent was present in liver tissue of fish with HSMI Paper V , and it is possible that infection of
hepatocytes may cause necrosis. However, multifocal liver necrosis observed in other
diseases, such as CMS and ISA, are thought to occur as a result of circulatory failure 136,155 .
Likewise, “pseudolobulation” of the spleen may either be an effect of generalised circulatory
disturbances, or it may be more directly linked to infection with the causal agent. Splenic
“pseudolobulation” has also been observed in fish with CMS, ISA and viral haemorrhagic
septicaemia (VHS) (records from the NVI). The origins and development of lesions in the
liver, spleen and other organs should be further studied.
The experimental studies showed that material from heart, liver and haematopoietic (spleen
and kidney) tissues sampled from diseased fish in the clinical phase were infective to
experimental fish Paper V , showing that the infection with the causal agent of HSMI was
51

generalised. Also, intraperitoneal injection of blood plasma sampled from fish in the midclinical
phase caused cardiac lesions in experimental fish Paper V . Viral particles circulating in
the blood are effectively removed by macrophages. Viraemia may therefore only be
maintained if there is a continuous shedding of virus from infected tissues and into the blood,
or if clearance by macrophages is for some reason reduced 100 . This implies that a viraemic or
septicaemic phase of HSMI probably concurs with the clinical phase and severe cardiac and
skeletal muscle lesions.
16.3.3. Post-clinical phase
Late changes were studied in field material Paper III . When mortality had returned to
background levels at the studied site, the clinical phase was considered to be over. The cardiac
changes were markedly milder and more multifocal, and there were no longer lesions in other
organs, indicating complete regeneration. It is known that regenerative processes occur in fish
hearts, even in those severely affected 120 . It is of course not possible to know whether fish
displaying cardiac lesions after the clinical outbreak ever had lesions in red skeletal muscle, as
lethal sampling is necessary for collecting histological samples. The high morbidity and
frequency of such lesions in apparently normal fish during the outbreak Papers I and III , however,
indicated that skeletal muscle had also been affected at some point in the survivors. Regarding
liver necrosis and changes in the spleen, however, it is not known whether these lesions will
heal, or if fish displaying such changes will eventually die from organ collapse.
Cardiac sections stained with van Gieson for demonstration of collagen, showed that while
some fibrotic material was present during and after the clinical outbreak, this gradually
disappeared as time went by Paper IV . Also, the foci of myocarditis gradually decreased in size
and numbers, and there was an increasing demarcation between pathological and healthy
tissue. Epi- and endocardial changes were also milder. Ten months after the onset of
mortalities due to HSMI, no fish in the sample had visible cardiac lesions. This indicates that
complete recovery is possible following HSMI, even though the healing process may be very
slow. Recovery is also the most likely outcome of survivors from PD-outbreaks 93 .
Regenerative abilities of myocardium in mice are largely dependent on the presence of
endothelial cells 105 . Also, epicardial cells are capable of undergoing transition to
mesenchymal cells following traumatic injury such as removal of the apex of the heart,
enabling revascularisation of the myocardium to occur 81 . Although many cells were lost in
52

the hearts of fish with HSMI, there was no generalised necrosis of epi- or endothelial cells
Paper IV . This may explain why regeneration was possible.
In the months following the clinical phase there was a shift from mainly compact layer
involvement to a greater involvement of the spongy layer in the ventricle and especially the
atrium Paper III . These changes were similar to cardiac changes preceding outbreaks of CMS
(unpublished results). Interestingly, CMS was diagnosed in fish in some other cages on the
site just before slaughter, but this occurred several months later. By the time CMS had given
rise to increased mortality on the farm, the fish group included in the study did not show any
cardiac changes. The reason why the degree of lesions was higher in spongy cardiac tissue
after an outbreak of HSMI is not known. As no reported experimental studies with HSMI
have been extended beyond 12 weeks, the association of these lesions with HSMI is
uncertain. The lesions in spongy myocardium may thus have some other cause. However,
samples from fish with multifocal myocarditis taken two months after the clinical outbreak
induced cardiac lesions similar to HSMI in experimental fish Paper V . Also, multifocal
myocarditis has been observed some time after outbreaks of HSMI on other sites (unpublished
results). A possible association with HSMI may therefore not be excluded. Unlike compact
myocardium, spongy myocardium of Atlantic salmon does not contain coronary vessel
branches 116 , and the metabolism is also different 35 . It is therefore possible that the healing
process in spongy myocardium is somewhat slower than that of compact myocardium.
However, in very late samples Paper III , focal myocarditis was frequently observed also in
compact myocardium.
16.3.4. Inflammatory cells
Morphologically, lymphocyte-like and macrophage-like cells appeared to constitute most of
the cellular infiltrate in HSMI. Some plasma-like cells and EGCs were also observed.
However, the cells did not show enzyme- or immunohistochemical properties as described in
earlier reports, see sections 14.5 and 14.6. Paper IV, Table 6 . It was therefore not possible to
characterise the cells by these methods. The causal agent of HSMI may possibly be able to
down regulate the expression of MHC class II or reduce enzymatic activities 2 . This is known
from some viral diseases in other species 72 . Methodological problems such as masking of
binding sites and destruction of enzymes may also have occurred, even though replication of
the tests was performed to reduce the risk of these problems. In future studies, additional
53

methods may be used to identify the cell populations, for instance in situ hybridization, PCR
or flow cytometry. One may also test cell functions, such as oxidative burst, phagocytosis and
proliferation 17 .
16.4. Aetiology
Upon discovery, HSMI was instantly assumed to be infectious, as the anamnestic information,
its widespread occurrence within the affected farms and apparent pattern of spread between
farms pointed in that direction. A pilot infection experiment performed in 1999 with material
from an outbreak of HSMI gave no results Paper I , but a second pilot study in 2001 indicated
that HSMI was experimentally transmissible (T Taksdal, pers.com.). This was confirmed in
subsequent infection experiments, including intraperitoneal injection of tissue homogenate
and cohabitation to injected fish Papers II and V . It is therefore likely that HSMI is caused by an
infectious agent.
16.4.1. The virus hypothesis
Is HSMI a viral disease? The true answer to this question lies in the future, but some
observations pull in direction of a viral aetiology. Firstly, HSMI appeared to spread rapidly
from fish to fish in field outbreaks Papers I and III , indicating a high degree of contagiousness.
Secondly, no bacteria or parasites were isolated or discovered Papers I-III . So far, no certain
isolation of a virus has been reported, although some candidates have been discovered and
proposed 32,166 . However, it appears that there is considerable morphological variation among
the particles found in these studies. Virus-like particles were also found both in cardiac tissue
from the longitudinal field study Paper IV and from experimental fish (unpublished). However,
viral particles were not discovered by electron microscopic examination of cell culture Paper V .
Incidental findings of subclinical viral infections occur in several species 100 . Some of the
observed particles may therefore be unassociated with HSMI. However, successful
transmission of HSMI by using a filtered inoculate has been reported 166 . This clearly suggests
a viral aetiology. Also, pre-treatment of supernatant from cardiac tissue with chloroform did
not prevent the transmission of HSMI Paper V , indicating that the causal agent may be a nonenveloped
virus. However, other researchers have suggested that the causal virus of HSMI
may be enveloped 32 . These studies have not yet been published in a peer-reviewed journal,
and details regarding methods and results are not known.
54

In the cell culture trials Papers I-III and V , CPE with evident destruction of the cellular layer was
not observed. CPE is common in most viral infections of susceptible cell cultures, and may be
produced by salmonid viruses such as ISAV, IPNV and SAV 27,93,106 . Other research groups
have reported multifocal vacuolation of cells inoculated with material from fish with HSMI
32,166
. This was only occasionally observed in the present study, and the most evident sign of
infection of cell cultures was the formation of plaque-like foci just after inoculation, followed
by an unspecific change to the monolayer in FHM and EPC cultures Paper V . Experimental
infection with some of this material showed that it was infective to experimental fish after two
passages in cell culture Paper V , indicating that some degree of agent propagation probably took
place in the cell cultures.
By testing HSMI material for the presence of known salmonid viruses (ISAV, ASPV and
SAV) Papers I-III , it became clear that there was no firm association between these viruses and
HSMI. The presence of IPNV in some of the samples was also tested by real time RT-PCR, as
the fish group studied in the longitudinal study had an outbreak of IPN early in the seawater
phase Paper III . As IPNV is ubiquitous in Norwegian fish farms and persistent infection may
occur after an outbreak of IPN 75,103 , it was expected that there would be a certain level of
virus present in the fish group. This was also the case. However, the amount of IPNV in the
tissues was not higher than in normal fish or fish with other diseases not normally associated
with this virus (records from the NVI). Also, sampled material was neutralised by rabbit
antisera against the most common serotypes of IPNV 19 Papers I-
before all the cell culture trials
III and V Paper V
. This did not prevent experimental transmission of HSMI , suggesting that IPNV
was not a part of the aetiology of HSMI. However, it is not unlikely that persistent IPNV
infection may enhance the susceptibility of fish to HSMI, even though this does not appear to
be the case for ISA 64,141 . Further studies are needed to determine the cause of HSMI.
16.4.2. Bacteria
Results from the first experimental study Paper II indicated that the causal agent of HSMI was
resistant to gentamycine in the tested concentration (50 g mL -1 ). It does not eliminate a
hypothesis of possibly causal gentamycine resistant bacteria. However, light and transmission
electron microscopy of experimental fish did not reveal any bacteria Papers I-V , bacteria were not
isolated on blood agar Papers I and II and pre-treatment of experimental inoculate with chloroform
did not prevent the transmission of HSMI Paper V . Also, cardiac tissue sampled during two field
55

outbreaks of HSMI was negative for 16S rRNA (unpublished). The 16S rRNA PCR targets a
conserved region within the genome of bacteria, which may be used to discover and identify
bacteria present in a sample 1,171 . Based on these results is therefore not likely that HSMI has
a bacteriological aetiology.
16.4.3. Parasites and fungi
Some parasites and pathogenic fungi are found in muscle tissue, such as Ichthyophonus hoferi
116 . However, parasitic and fungal infections are usually easily discovered, as they are caused
by uni- or multicellular organisms detectable by light microscopy. Incidental findings of
parasites and fungi have been observed in fish with HSMI (records from the NVI), but the
frequency of these concurrent infections do not appear to be higher than average for the
population of farmed Atlantic salmon in Norway.
HSMI concurred with parvicapsulosis in the longitudinal study Paper III , which is an infection of
the microsporidial parasite Parvicapsula pseudobranchicola 109,157 . This parasite has also been
observed in several other HSMI outbreaks, especially in Mid-Norway (records from the NVI)
109,157 Papers I and III
, but was not consistently observed during field outbreaks of HSMI . The main
areas for parvicapsulosis and HSMI are not equal, as parvicapsulosis mostly occurs in
Northern and Mid-Norway, and is rarely seen in Western Norway 111 . In addition, HSMI was
experimentally transmissible in the absence of this parasite Papers II and V . While infection with
Parvicapsula pseudobranchicola may be a potential stress factor for the immune system and
possibly make fish more susceptible to HSMI, it is therefore not likely that it is the
aetiological agent.
16.4.4. Toxins
One plausible hypothesis is that HSMI may be due to a toxic agent in the water or produced
by an infectious agent, and that injection of the toxin may have produced the tissue lesions
that were observed in experimental fish Papers II and V . However, the observed 4-week delay in
appearance of tissue lesions in cohabitant groups Paper II indicate that the causal agent was
propagated in i.p.-injected fish before transmission occurred. This observation was similar in
both tanks, and thus appeared to be a true effect of cohabitation. One explanation could be
56

that the clearance of a possible toxic agent was somehow delayed by the intraperitoneal
injection, but this is only speculative.
16.4.5. Autoimmunity
In human medicine, myocarditis often occurs secondarily to viral infection 62,69 . This may
cause sudden mortalities, or possibly result in dilated cardiomyopathy 33 . Several studies show
that viruses from different families, viruses such as coxsackie (CB3) and murine
cytomegalovirus, may induce an autoimmune condition leading to widespread myocarditis.
One proposed mechanism is epitope mimicry, in which the virus mimics human myosin
receptors and may cause a faulty production of self-antibodies. Another mechanism is called
the “adjuvant effect”, in which viruses act as immunostimulatory agents that increase the risk
of a development towards autoimmunity 36 . A third possible cause of autoimmunity may be a
massive myocardial destruction and release of myosin, which may in turn lead to production
of antibodies against myosin 62 . It is highly uncertain if these events may occur in fish, but
possible autoimmunity has been associated with reaction to vaccine adjuvant in Atlantic
salmon 73 . However, a hypothesis of autoimmunity was contradicted by the observation that
most fish appeared to regain control of the inflammation after a relatively short period of
time, and that the pathological changes eventually disappeared Paper III .
The severity of the myocardial inflammation and prolonged disease process in HSMI in
relation to the difficulty in finding the causal agent within infective tissue may indicate that
the immune reaction is somewhat more intense than what could be expected. This is
supported by the observation that dead fish generally had a more severe inflammation than
moribund and apparently normal fish Paper III . Perhaps mortalities were associated with an
immune response blown out of proportions? It is well established that the inflammatory
process itself may cause extensive tissue destruction, and induce a prolonged inflammatory
condition due to release of cytokines, lysosymal contents and other tissue damaging factors 98 .
Lesions in other organs may thus be secondary effects of a massive myocardial selfdestruction.
It is not unlikely that the immune response may be an important factor in the
progression towards clinical disease. The disease process may thus have been started by an
infectious agent, which is thereafter to a certain extent self-sustained due to the immune
response.
57

16.4.6. Neoplasia
Instead of actually being inflammatory cells, the cellular infiltrate observed in HSMI may
have consisted of transformed or neoplastic cells. Neoplastic cells may be metastatic and
establish infiltrative tumours in several organs 132 . Cancer is not considered to be contagious,
and will therefore not spread horizontally through water. As HSMI was transmitted by
cohabitation to injected fish Paper II , it is likely that the primary cause is infectious. Some virus
families may induce neoplasia in mammals, for instance adenoviridae, herpesviridae and
retroviridae 100 . Retroviridially induced neoplasia has also been observed in fish 88 . A
preliminary investigation of the presence of cells in active mitosis in hearts of HSMI fish by
proliferating cell nuclear antigen (PCNA) 112 showed that the number of PCNA + cells were
relatively small (EO Koppang, pers.com.). This does not support the idea of an infiltrative
neoplasia, but future studies may further illuminate this question.
16.5. The significance of HSMI for Atlantic salmon aquaculture
16.5.1. Mortality
Although outbreaks of HSMI have caused high mortality in some farms, the general
impression is that losses due to HSMI are small compared to outbreaks of other diseases, such
as CMS and PD. The lower mortality during HSMI outbreaks may explain why the disease
has received less attention from Norwegian fish farmers and researchers in recent years than
PD. However, the high number of HSMI outbreaks may still result in significant cumulative
losses. Also, fish groups developing tissue lesions due to HSMI may potentially be less
resistant to other diseases or environmental challenges. So far, no studies have specifically
looked at losses directly or indirectly associated with HSMI. It is therefore not yet possible
give firm conclusions on the economical impact of this disease, but the yearly increase in the
number of outbreak sites Figure 3 and the apparent spread of the disease Table 1 is certainly
worrying.
The high number of fish showing severe lesions in cardiac and skeletal muscle without
corresponding clinical signs during the outbreak phase of HSMI in the present study
III
, indicated that HSMI did not always have a lethal outcome. Also, results from a recent
longitudinal study of fish in the seawater phase showed lesions consistent with HSMI in a
high number of fish on two farms, even though no clinical signs were detected in the
58
Papers I and

populations (unpublished). This implies that subclinical cases of HSMI probably occur, and it
is therefore highly possible that some cases of HSMI may go unnoticed. A sampling
performed in the middle of the clinical outbreak of the longitudinal study Paper III , showed that
close to 100 % of 112 sampled fish in three cages had cardiac lesions. There is therefore little
doubt that the tissue lesions consistent with HSMI were widespread in the sea cages included
in the study during the clinical phase Paper III , indicating a very high morbidity of HSMI. In
addition, there was elevated mortality in all sea cages in this period. Experiences from other
field cases also indicate a high morbidity Paper I , but some fish health services report that they
sometimes observe elevated mortality in only a few cages on sites with HSMI outbreaks (A
Østvik, pers.com.). The prevalence of cardiac lesions across affected farms should be further
investigated. This would be a help in determining potential variation in morbidity and
mortality at the cage and farm levels.
Stress experiences during the clinical phase appear to contribute to an increase in mortality
due to HSMI (A Lyngøy, pers.com.). An explanation for this could be that fish with severe
tissue damage in vital organs such as the heart, red skeletal muscle and liver, probably do not
have the capacity to handle the increased level of activity or metabolism that stress situations
might require. It is known that stress factors such as crowding, handling and transportation
may have a negative effect on the immune system of fish and possibly reduce their resistance
to disease 4,63,122 . The change of sites two months before the clinical outbreak of HSMI in the
longitudinal study Paper III could therefore possibly have hastened the pathological process,
thereby inducing mortalities.
What are possible biological mechanisms leading to mortalities in an outbreak of HSMI?
Observations from the field showed that moribund and especially the dead fish had a more
widespread inflammatory response than apparently normal fish Paper III . A possible explanation
for this is that the immune reaction itself may be most damaging factor in the disease process.
This is a common side-effect of diseases causing chronic inflammation 98,115 . The highest
mortality in PD outbreaks occurs during the chronic phase when muscle lesions appear
93,137,138
. It has therefore been hypothesised that necrosis of muscle fibres largely contributes
to mortalities in that disease 93 . For HSMI, lesions in red skeletal muscle were mainly present
during the clinical phase of the disease process Paper III . A resulting malfunction of this tissue
could therefore increase the risk of dying from HSMI. On the other hand, a large number of
59

apparently normal fish also had severe lesions in the red skeletal muscle Paper III . The cause of
mortalities due to HSMI therefore remains uncertain.
16.5.2. Transmissibility
Results from the experimental studies Papers II and V showed that HSMI was experimentally
transmissible, indicating that it is a contagious disease. Particularly, the successful
transmission of the disease to cohabitating fish Paper II strongly suggested that injected fish
were able to shed infective material and that horizontal transmission occurred. The time-lag in
the appearance of lesions between injected and cohabitating fish Paper II indicated that a stage
of pathogen propagation was required before this horizontal spread could occur 100 . Taken
together with the observations of high morbidity during field outbreaks Papers I and III and the
dramatic increase in diagnosed outbreaks Figure 3 , HSMI appears to be an infectious disease
with a potential of rapid horizontal dispersal.
Testing of different inoculates showed that the causal agent was simultaneously present in
cardiac, hepatic, renal and splenic tissue and in plasma during the clinical phase Paper V .
Shedding of infective particles to the environment may therefore possibly occur during this
phase. It is not known how early infected fish may start shedding infective particles, but the
possible transmission of the causal agent to cohabitating fish before the onset of
histopathological changes in i.p. injected fish Paper II indicated that this may occur at an early
stage in disease development. The presence of infective particles in renal tissue at four weeks
post challenge was confirmed by the induction of cardiac changes in experimental fish with
this material Paper V . Experimental infection with HSMI material from the longitudinal study
Paper III
also showed that infective particles were present in cardiac tissue as early as two
months before and as late as two months after the clinical phase Paper V . It is possible that
shedding of infective particles do not occur as early or as late as this, but there is a potential
that particles may be released from dead and decaying fish. In the longitudinal study, there
was an increase in mortality due to proliferative gill inflammation at the time of sampling for
the infection experiment Papers III and V . There may therefore potentially have been release of the
causal agent of HSMI from the population to the environment during this period.
Interestingly, experimental infection showed that infective particles were present in the fish
group at all three sites included in the longitudinal study Papers III and V . This indicated that
60

dispersal of the causal agent may potentially have occurred at all these sites, even though fish
did not show clinical signs at all times. Movement of fish during the seawater phase may
therefore have contributed to the spread of HSMI in Norway, possibly due shedding of virus
at several sites from the same population, or also due to contamination of well boats 104 .
In addition, this study showed that smolts in freshwater are susceptible to the causal agent of
HSMI by experimental infection Paper V . Taken together, these observations may have
implications for future disease mitigation plans and management practices.
17. Main conclusions
• HSMI is a severe cardiac disease in farmed Atlantic salmon characterised by
inflammation and inflammation-associated tissue necrosis.
• Lesions in other organs were restricted to the period of the clinical disease outbreak with
increased mortalities. These changes included inflammation and necrosis of red skeletal
muscle, liver necrosis and signs of circulatory disturbances.
• HSMI was histologically distinguishable from PD and CMS
• The disease process may start as a subclinical infecton several months before a clinical
outbreak and have a prolonged duration after the outbreak. This was supported by the
infectivity of cardiac tissue from pre- and post-outbreak phases to experimental fish.
• Early cardiac lesions in experimental and field material included perivasculitis associated
with coronary vessel branches, endocarditis and focal myocarditis.
• HSMI is probably an infectious disease with a potential of rapid horizontal spread.
• The viraemic or septicaemic phase of the causal agent appeared to be concurrent with the
clinical phase.
• The aetiology was not established, but a viral hypothesis is the most plausible.
61

18. Future work
To facilitate future control of the disease, it is necessary that the aetiology of HSMI is firmly
established. This should be performed by an increased effort to isolate the causal agent by in
vitro methods, and by further attempts to fulfil Koch’s postulates. In future infection
experiments, filtration of inoculates should be tested. Also, chloroform pre-treatment of
infective material before inoculation onto cell cultures or injection into fish should be retested
using both positive and negative controls. Recently developed molecular methods designed to
search for unknown microbes may also be applied, in order to target the search for a causal
agent for HSMI.
More accurate diagnostic tools are needed to separate HSMI from other diseases for use in
diagnostics and research. Further, an epidemiological investigation of disease occurrence,
spread and risk factors for HSMI would be a useful basis for mitigation plans. There is also a
need to increase the knowledge of possible mechanisms for autoimmunity in fish, and the role
of the immune system in the development of HSMI. This may possibly be studied by the use
of molecular methods, such as microarray or targeted PCR. Also, more traditional methods
such as haematology and clinical chemistry may contribute to a more profound understanding
of the mechanisms leading to HSMI.
62

References
1. Amann, R. I., W. Ludwig, and K. H. Schleifer. 1995. Phylogenetic identification and in situ
detection of individual microbial cells without cultivation. Microbiological Reviews 59:143-
169.
2. Amati L, Passer ME, Lippolis D, Lio C, Caruso E, Jirillo E, Covelli V. 2006. Taking
advantage of viral immune evasion: Virus-derived proteins represent novel
biopharmaceuticals. Current Medicinal Chemistry 13:325-333.
3. Amin AB, Trasti J. 1988. Endomyocarditis in Atlantic salmon in Norwegian seafarms.
Bulletin of the European Association of Fish Pathologists 8:70-71.
4. Ashley PJ. 2007. Fish welfare: Current issues in aquaculture. Applied Animal Behaviour
Science 104:199-235.
5. Ashtekar AR, Saha B. 2003. Poly's plea: membership to the club of APCs. Trends in
Immunology 24:485-490.
6. Bakke-McKellep AM, Press CM, Bæverfjord G, Krogdahl Å, Landsverk T. 2000. Changes in
immune and enzyme histochemical phenotypes of cells in the intestinal mucosa of Atlantic
salmon, Salmo salar L., with soybean meal-induced enteritis. Journal of Fish Diseases 23:115-
127.
7. Bochsler PN, Slauson DO. 2002. Inflammation and repair of tissue, p. 140-245. In Slauson
DO and Cooper BJ (eds.), Mechanisms of disease. A textbook of comparative general
pathology. Mosby, St. Louis.
8. Bols NC, Barlian A, Chirino-Trejo M, Caldwell SJ, Goegan P, Lee LEJ. 1994. Development
of a cell line from primary cultures of rainbow trout, Oncorhynchus mykiss (Walbaum), gills.
Journal of Fish Diseases 17:601-611.
9. Boshra H, Li J, Sunyer JO. 2006. Recent advances on the complement system of teleost fish.
Fish & Shellfish Immunology 20:239-262.
10. Boucher P, Castric J, Baudin Laurencin F. 1994. Observation of virus-like particles in rainbow
trout Oncorhynchus mykiss infected with sleeping-disease virulent material. Bulletin of the
European Association of Fish Pathologists 14:215-216.
11. Boucher P, Raynard RS, Houghton G, Baudin Laurencin F. 1995. Comparative experimental
transmission of pancreas disease in Atlantic salmon, rainbow trout and brown trout. Diseases
of Aquatic Organisms 22:19-24.
12. Bozzola JJ, Russell LD. 1999. Electron microscopy. Principles and techniques for biologists.
Jones and Bartlett Publishers, Sudbury.
13. Brocklebank J, Raverty S. 2002. Sudden mortality caused by cardiac deformities following
seining of preharvest farmed Atlantic salmon (Salmo salar) and by cardiomyopathy of
postintraperitoneally vaccinated Atlantic salmon parr in British Columbia. Canadian
Veterinary Journal 43:129-130.
63

64
14. Brun E, Poppe TT, Skrudland A, Jarp J. 2003. Cardiomyopathy syndrome in farmed Atlantic
salmon Salmo salar: occurrence and direct financial losses for Norwegian aquaculture.
Diseases of Aquatic Organisms 56:241-247.
15. Bruno DW, Poppe TT. 1996. Diseases of uncertain aetiology. Cardiac myopathy syndrome, p.
140-141. A colour atlas of salmonid diseases. Academic press, London.
16. Castric J, Baudin Laurencin F, Bremont M, Jeffroy J, Le Ven A, Bearzotti M. 1997. Isolation
of the virus responsible for sleeping disease in experimentally infected rainbow trout
(Oncorhynchus mykiss). Bulletin of the European Association of Fish Pathologists 17:27-30.
17. Chilmonczyk S, Monge D. 1999. Flow cytometry as a tool for assessment of the fish cellular
immune response to pathogens. Fish & Shellfish Immunology 9:319-333.
18. Christie KE, Fyrand K, Holtet L, Rowley HM. 1998. Isolation of pancreas disease virus from
farmed Atlantic salmon, Salmo salar L., in Norway. Journal of Fish Diseases 21:391-394.
19. Christie KE, Ness S, Djupvik HO. 1990. Infectious pancreatic necrosis in Norway: partial
serotyping by monoclonal antibodies. Journal of Fish Diseases 13:323-327.
20. Cockerell GL, Cooper BJ. 2002. Disorders of cell growth and cancer biology, p. 298-377. In
Slauson DO and Cooper BJ (eds.), Mechanisms of disease. A textbook of comparative general
pathology. Mosby, St. Louis.
21. Cooper BJ. 2002. Disease at the cellular level, p. 16-75. In Slauson DO and Cooper BJ (eds.),
Mechanisms of disease. A textbook of comparative general pathology. Mosby, St. Louis.
22. Croll NB, Grizzle JM. 1998. Ultrastructure of leukocytes in acutely inflamed dermis and
muscle of channel catfish at different temperatures. Journal of Aquatic Animal Health 10:381-
389.
23. Crumlish M, Diab AM, George S, Ferguson HW. 2007. Detection of the bacterium
Flavobacterium psychrophilum from a natural infection in rainbow trout, Oncorhyncus mykiss
(Walbaum), using formalin-fixed, wax-embedded fish tissues. Journal of Fish Diseases 30:37-
41.
24. Dannevig BH, Brudeseth BE, Gjøen T, Rode M, Wergeland HI, Evensen Ø, Press CM. 1997.
Characterisation of a long-term cell line (SHK-1) developed from the head kidney of Atlantic
salmon (Salmo salar L.). Fish & Shellfish Immunology 7:213-226.
25. Dannevig BH, Falk K, Krogsrud J. 1993. Leucocytes from Atlantic salmon, Salmo salar L.,
experimentally infected with infectious salmon anaemia (ISA) exhibit an impaired response to
mitogens. Journal of Fish Diseases 16:351-359.
26. Dannevig BH, Falk K, Namork E. 1995. Isolation of the causal virus of infectious salmon
anaemia (ISA) in a long-term cell line from Atlantic salmon head kidney. Journal of General
Virology 76:1353-1359.
27. Dannevig BH, Falk K, Press CM. 1995. Propagation of infectious salmon anemia (ISA) virus
in cell culture. Veterinary Research 26:438-442.
28. Dannevig BH, Falk K, Skjerve E. 1994. Infectivity of internal tissues of Atlantic salmon,
Salmo salar L., experimentally infected with the aetiological agent of infectious salmon
anaemia (ISA). Journal of Fish Diseases 17:613-622.

29. Desvignes L, Quentel C, Lamour F, Le Ven A. 2002. Pathogenesis and immune response in
Atlantic salmon (Salmo salar L.) parr experimentally infected with salmon pancreas disease
virus (SPDV). Fish & Shellfish Immunology 12:77-95.
30. Dettman RW, Pae SH, Morabito C, Bristow J. 2003. Inhibition of alpha 4-integrin stimulates
epicardial-mesenchymal transformation and alters migration and cell fate of epicardially
derived mesenchyme. Developmental Biology 257:315-328.
31. Devold M, Krossøy B, Aspehaug V, Nylund A. 2000. Use of RT-PCR for diagnosis of
infectious salmon anaemia virus (ISAV) in carrier sea trout Salmo trutta after experimental
infection. Diseases of Aquatic Organisms 40:9-18.
32. Eliassen TM, Solbakk IT, Evensen Ø, Gravningen K. 2004. Isolation of heart and skeletal
muscle inflammation virus (HSMIV) from salmon. Poster at the 6th International Symposium
on Viruses of Lower Vertebrates, Hokkaido, Japan, p 56.
33. Eriksson U, Penninger JM. 2005. Autoimmune heart failure: new understandings of
pathogenesis. International Journal of Biochemistry & Cell Biology 37:27-32.
34. Evans DL, Leary III JH, Jaso-Friedmann L. 2001. Nonspecific cytotoxic cells and innate
immunity: regulation by programmed cell death. Developmental and Comparative
Immunology 25:791-805.
35. Ewart HS, Driedzic WR. 1987. Enzymes of energy-metabolism in salmonid hearts - spongy
versus cortical myocardia. Canadian Journal of Zoology 65:623-627.
36. Fairweather D, Kaya Z, Shellam GR, Lawson CM, Rose NR. 2001. From infection to
autoimmunity. Journal of Autoimmunity 16:175-186.
37. Falk K, Press CM, Landsverk T, Dannevig BH. 1995. Spleen and kidney of Atlantic salmon
(Salmo salar L.) show histochemical changes in the course of experimentally induced
infectious salmon anaemia (ISA). Veterinary Immunology and Immunopathology 49:115-126.
38. Fänge R. 1992. Fish blood cells, p. 1-54. In Hoar WS, Randall DJ, and Farrell AP (eds.),
Cardiovascular system, part B, vol. 12. Academic Press Inc., San Diego.
39. Farrell AP. 2002. Coronary arteriosclerosis in salmon: growing old or growing fast?
Comparative Biochemistry and Physiology A-Molecular & Integrative Physiology 132:723-
735.
40. Farrell AP, Johansen JA, Saunders RL. 1990. Coronary lesions in Pacific salmonids. Journal
of Fish Diseases 13:97-100.
41. Farrell AP, Jones DR. 1992. The heart, p. 1-88. In Hoar WS, Randall DJ, and Farrell AP
(eds.), The cardiovascular system, part A, vol. 12. Academic Press Inc., San Diego.
42. Ferguson HW. 2006. Introduction, p. 10-22. In Ferguson HW (ed.), Systemic pathology of
fish. A text and atlas of normal tissue in teleosts and their responses in disease. Scotian Press,
London.
43. Ferguson HW, Girons A, Rizgalla G, LaPatra S, Branson E, McKenzie K, Davis MK, Diab A,
Crumlish M. 2006. Strawberry disease in cage-reared rainbow trout (Oncorhynchus mykiss) in
Scotland: pathological description and association with Flavobacterium psychrophilum. The
Veterinary Record 158:630-631.
65

66
44. Ferguson HW, Poppe T, Speare DJ. 1990. Cardiomyopathy in farmed Norwegian salmon.
Diseases of Aquatic Organisms 8:225-231.
45. Ferguson HW, Roberts RJ, Richards RH, Collins RO, Rice DA. 1986. Severe degenerative
cardiomyopathy associated with pancreas disease in Atlantic salmon, Salmo salar L. Journal
of Fish Diseases 9:95-98.
46. Fijan N, Sulimanovic D, Bearzotti M, Muzinic D, Zwillenberg LO, Chilmonczyk S, Vautherot
JF, de Kinkelin P. 1983. Some properties of the Epithelioma papulosum cyprini (EPC) cell
line from carp Cyprinus carpis. Annals of Virology (Inst Pasteur) 134:207-220.
47. Fischer U, Utke K, Somamoto T, Köllner B, Ototake M, Nakanishi T. 2006. Cytotoxic
activities of fish leucocytes. Fish & Shellfish Immunology 20:209-226.
48. Frøystad MK, Rode M, Berg T, Gjøen T. 1998. A role for scavenger receptors of proteincoated
particles in rainbow trout head kidney macrophages. Developmental and Comparative
Immunology 22:533-549.
49. Fryer JL, Lannan CN. 1994. Three decades of fish cell culture: A current listing of cell lines
derived from fishes. Journal of Tissue Culture Methods 16:87-94.
50. Graham DA, Jewhurst VA, Rowley HM, McLoughlin MF, Todd D. 2003. A rapid
immunoperoxidase-based virus neutralization assay for salmonid alphavirus used for a
serological survey in Northern Ireland. Journal of Fish Diseases 26:407-413.
51. Gravell M, Malsberg RG. 1965. A permanent cell line from fathead minnow (Pimephales
promelas). Annals of the New York Academy of Sciences 126:555-565.
52. Grotmol S, Totland GK, Kryvi H. 1997. Detection of a nodavirus-like agent in heart tissue
from reared Atlantic salmon Salmo salar suffering from cardiac myopathy syndrome (CMS).
Diseases of Aquatic Organisms 29:79-84.
53. Grove S, Johansen R, Reitan LJ, Press CM. 2006. Immune- and enzyme histochemical
characterisation of leukocyte populations within lymphoid and mucosal tissues of Atlantic
halibut (Hippoglossus hippoglossus). Fish & Shellfish Immunology 20:693-708.
54. Hardy RW. 2001. The nutritional pathology of teleosts, p. 347-366. In Roberts RJ (ed.), Fish
Pathology. W.B. Saunders, Edinburgh.
55. Hendrickx J, Doggen K, Weinberg EO, Van Tongelen P, Fransen P, De Keulenaer DW. 2004.
Molecular diversity of cardiac endothelial cells in vitro and in vivo. Physiological Genomics
19:198-206.
56. Hodneland K, Bratland A, Christie KE, Endresen C, Nylund A. 2005. New subtype of
salmonid alphavirus (SAV), Togaviridae, from Atlantic salmon Salmo salar and rainbow trout
Oncorhynchus mykiss in Norway. Diseases of Aquatic Organisms 66:113-120.
57. Hodneland K, Endresen C. 2006. Sensitive and specific detection of salmonid alphavirus
using real-time PCR (TaqMan). Journal of Virological Methods 131:184-192.
58. Holm JC. 1999. Aktuelle fiskearter i oppdrett (Species in current fish farming), p. 49-60. In
Poppe TT (ed.), Fiskehelse og fiskesykdommer (Fish health and fish diseases; in Norwegian).
Universitetsforlaget, Oslo.
59. Hong JR, Lin TL, Hsu YL, Wu JL. 1998. Apoptosis precedes necrosis of fish cell line with
infectious pancreatic necrosis virus infection. Virology 250:76-84.

60. Houghton G. 1994. Aquired protection in Atlantic salmon Salmo salar parr and post-smolts
against pancreas disease. Diseases of Aquatic Organisms 18:109-118.
61. Houghton G. 1995. Kinetics of infection of plasma, blood leucocytes and lymphoid tissue
from Atlantic salmon Salmo salar experimentally infected with pancreas disease. Diseases of
Aquatic Organisms 22:193-198.
62. Huber SA. 1997. Autoimmunity in myocarditis: Relevance of animal models. Clinical
Immunology and Immunopathology 83:93-102.
63. Iversen M, Finstad B, McKinley RS, Eliassen RA, Carlsen KT, Evjen T. 2005. Stress response
in Atlantic salmon (Salmo salar L.) smolts during commercial well boat transports, and effects
on survival after transport to sea. Aquaculture 243:373-382.
64. Johansen LH, Sommer AI. 2001. Infectious pancreatic necrosis virus infection in Atlantic
salmon Salmo salar post-smolts affects the outcome of secondary infections with infectious
salmon anemia virus or Vibrio salmonicida. Diseases of Aquatic Organisms 47:109-117.
65. Johansen R. 2004. Mulige årsaker til Cardiomyopatisyndrom (CMS) (Possible causes of
cardiomyopathy syndrome), p. 56-57. In Tørud B and Hillestad M (eds.), Hjerterapporten
2004. Rapport om hjertelidelser hos laks og regnbueørret (The heart report 2004, in
Norweigan).
66. Jørgensen JB, Robertsen B. 1995. Yeast beta-glucan stimulates respiratory burst activity of
Atlantic salmon (Salmo salar L.) macrophages. Developmental and Comparative Immunology
19:43-57.
67. Junqueira LC, Carneiro J, Kelley RO. 1995. Basic histology. Prentice-Hall International, Inc.,
London.
68. Kiernan JA. 1999. Histological and histochemical methods. Theory and practice. Butterworth-
Heinemann, Oxford.
69. Kishimoto C, Hiraoka Y, Takamatsu N, Takada H, Kamiya H, Ochiai H. 2003. An in vivo
model of autoimmune post-coxsackievirus B3 myocarditis in severe combined
immunodeficiency mouse. Cardiovascular Research 60:397-403.
70. Köllner B, Fischer U, Rombout JHWM, Taverne-Thiele JJ, and Hansen JD. 2004. Potential
involvement of rainbow trout thrombocytes in immune functions: a study using a panel of
monoclonal antibodies and RT-PCR. Developmental and Comparative Immunology 28:1049-
1062.
71. Kongtorp RT, Taksdal T, Lillehaug A. 2005. Cardiomyopathy syndrome (CMS): a literature
review. Report from the National Veterinary Institute, Oslo, Norway.
72. Koppang EO, Dannevig BH, Lie Ø, Rønningen K, Press CM. 1999. Expression of MHC class
I and II mRNA in a macrophage-like cell line (SHK-1) derived from Atlantic salmon, Salmo
salar L., head kidney. Fish & Shellfish Immunology 9:473-489.
73. Koppang EO, Haugarvoll E, Hordvik I, Aune L, Poppe TT. 2005. Vaccine-associated
granulomatous inflammation and melanin accumulation in Atlantic salmon, Salmo salar L.,
white muscle. Journal of Fish Diseases 28:13-22.
74. Koppang EO, Hordvik I, Bjerkås I, Torvund J, Aune L, Thevarajan J, Endresen C. 2003.
Production of rabbit antisera against recombinant MHC class II B chain and identification of
67

68
immunoreactive cells in Atlantic salmon (Salmo salar). Fish & Shellfish Immunology 14:115-
132.
75. Krogsrud J, Håstein T, Rønningen K. 1989. Infectious pancreatic necrosis virus in Norwegian
fish farms, p. 284-291. In Ahne W and Kurstak E (eds.), Viruses of lower vertebrates.
Springer-Verlag, Berlin.
76. Kvellestad A, Dannevig BH, Falk K. 2003. Isolation and partial characterization of a novel
paramyxovirus from the gills of diseased seawater-reared Atlantic salmon (Salmo salar L.).
Journal of General Virology 84:2179-2189.
77. Lannan CN, Winton JR, Fryer JL. 1984. Fish cell lines: Establishment and characterization of
nine cell lines from salmonids. In Vitro 20:671-676.
78. Lavine KJ, Yu K, White AC, Zhang X, Smith C, Partanen J, Ornitz JM. 2005. Endocardial
and epicardial derived FGF signals regulate myocardial proliferation and differentiation in
vivo. Developmental Cell 8:85-95.
79. Leknes IL. 2001. Endocytosis of ferritin and hemoglobin by the trabecular endocardium in
swordtail, Xiphophorus helleri L. and platy, Xiphophorus maculatus L. (Poecilidae :
Teleostei). Annals of Anatomy 183:251-254.
80. Leknes IL. 2004. Two types of endothelial cells in the heart of platyfish (Poeciliidae :
Teleostei). Tissue & Cell 36:369-371.
81. Lepilina A, Coon AN, Kikuchi K, Holdway JE, Roberts RW, Burns CG, Poss KD. 2006. A
dynamic epicardial injury response supports progenitor cell activity during zebrafish heart
regeneration. Cell 127:607-619.
82. Li J, Barreda DR, Zhang YA, Boshra H, Gelman AE, LaPatra S, Tort L. 2006. B lymphocytes
from early vertebrates have potent phagocytic and microbicidal abilities. Nature Immunology
7:1116-1124.
83. Lillehaug A, Lunestad BT, Grave K. 2003. Epidemiology of bacterial diseases in Norwegian
aquaculture - a description based on antibiotic prescription data for the ten-year period 1991 to
2000. Diseases of Aquatic Organisms 53:115-125.
84. Lovy J, Wright GM, Speare DJ. 2007. Ultrastructural examination of the host inflammatory
response within gills of netpen reared chinook salmon (Oncorhyncus tshawytscha) with
microsporidial gill disease. Fish & Shellfish Immunology 22:131-149.
85. Maclachlan, NJ, Cullen JM. 1995. Liver, biliary system, and exocrine pancreas, p. 81-115. In
Carlton WW and McGavin MD (eds.), Thomson's special veterinary pathology. Mosby, St.
Louis.
86. Magnadóttir B. 2006. Innate immunity of fish (overview). Fish & Shellfish Immunology
20:137-151.
87. Manning MJ, Nakanishi T. 1996. The specific immune system: cellular defences, p. 159-205.
In Iwama G and Nakanishi T (eds.), The fish immune system. Organism, pathogen, and
environment, vol. 15. Academic Press, San Diego.
88. Martineau D, Ferguson HW. 2006. Neoplasia, p. 313-334. In Ferguson HW (ed.), Systemic
pathology of fish. A text and atlas of normal tissues in teleosts and their responses in disease.
Scotian Press, London.

89. Martínez JL, López-Dóriga MV. 1995. Neutrophil granulocytes in the epidermis of the brown
trout, Salmo trutta L.: ultrastructural characterization. Journal of Submicroscopic Cytology
and Pathology 27:459-465.
90. McGavin MD. 1995. Muscle, p. 393-422. In Carlton WW and McGavin MD (eds.),
Thomson's special veterinary pathology. Mosby, St. Louis.
91. McLoughlin M, Nelson RT, McCormick JI, Rowley H. 1995. Pathology of experimental
pancreas disease in freshwater Atlantic salmon parr. Journal of Aquatic Animal Health 7:104-
110.
92. McLoughlin M, Nelson RT, Rowley HM, Cox DI, Grant AN. 1996. Experimental pancreas
disease in Atlantic salmon Salmo salar post-smolts induced by salmon pancreas disease virus
(SPDV). Diseases of Aquatic Organisms 26:117-124.
93. McLoughlin MF, Graham DA. 2007. Alphavirus infections in salmonids - a review. Journal of
Fish Diseases 30:511-531.
94. McLoughlin MF, Graham DA, Norris A, Matthews D, Foyle L, Rowley HM, Jewhurst H,
MacPhee J, Todd D. 2006. Virological, serological and histopathological evaluation of fish
strain susceptibility to experimental infection with salmonid alphavirus. Diseases of Aquatic
Organisms 72:125-133.
95. McLoughlin MF, Nelson RN, McCormick JI, Rowley HM, Bryson DB. 2002. Clinical and
histopathological features of naturally occurring pancreas disease in farmed Atlantic salmon,
Salmo salar L. Journal of Fish Diseases 25:33-43.
96. McVicar AH. 1987. Pancrease disease of farmed Atlantic salmon, Salmo salar, in Scotland -
epidemiology and early pathology. Aquaculture 67:71-78.
97. McVicar AH. 1990. Infection as a primary cause of pancreas disease in farmed Atlantic
salmon. Bulletin of the European Association of Fish Pathologists 10:84-87.
98. Mitchell RN, Cotran RS. 1997. Acute and chronic inflammation, p. 25-46. In Kumar V,
Cotran S, and Robbins RS (eds.), Basic Pathology. W.B. Saunders Company, Philadelphia.
99. Munro ALS, Ellis AE, McVicar AH, McLay HA, Needham EA. 1984. An exocrine pancreas
disease of farmed Atlantic salmon in Scotland. Helgolander Meereuntersuchungen 37:571-
586.
100. Murphy FA, Gibbs EPJ, Horiznek MC, Studdert MJ. 1999. Veterinary virology. Academic
Press, San Diego.
101. Murphy TM, Drinan EM, Gannon F. 1995. Studies with an experimental model for pancreas
disease of Atlantic salmon Salmo salar L. Aquaculture Research 26:861-874.
102. Murphy TM, Rodger HD, Drinan EM, Gannon F, Kruse P, Korting W. 1992. The sequential
pathology of pancreas disease in Atlantic salmon farms in Ireland. Journal of Fish Diseases
15:401-408.
103. Murray AG. 2006. Persistence of infectious pancreatic necrosis virus (IPNV) in Scottish
salmon (Salmo salar L.) farms. Preventive Veterinary Medicine 76:97-108.
104. Murray AG, Smith RJ, Stagg RM. 2002. Shipping and the spread of infectious salmon
anaemia in Scottish aquaculture. Emerging Infectious Diseases 8:1-5.
69

70
105. Narmoneva DA, Vukmirovic R, Davis ME, Kamm RD, Lee RT. 2004. Endothelial cells
promote cardiac myocyte survival and spatial reorganization. Implications for cardiac
regeneration. Circulation 110:962-968.
106. Nelson RT, McLoughlin MF, Rowley HM, Platten MA, McCormick JI. 1995. Isolation of a
toga-like virus from farmed Atlantic salmon Salmo salar with pancreas disease. Diseases of
Aquatic Organisms 22:25-32.
107. Nygaard S. 1999. Diagnostikk (Diagnostics), p. 371-379. In T. Poppe (ed.), Fiskehelse og
fiskesykdommer (Fish health and fish diseases; in Norwegian). Universitetsforlaget, Oslo.
108. Nylund A. 2001. Hjertesprekk, kardiomyopatisyndrom (CMS) (Cardiomyopathy syndrome).
Fiskehelse (Fish Health; in Norwegian) 1:29-37.
109. Nylund A, Karlsbakk E, Sæther PA, Koren C, Larsen T, Nielsen BD, Brøderud AE, Høstlund
C, Fjellsøy KR, Lervik K, Rosnes L. 2005. Parvicapsula pseudobranchicola (Myxosporea) in
farmed Atlantic salmon Salmo salar: tissue distribution, diagnosis and phylogeny. Diseases of
Aquatic Organisms 63:197-204.
110. O'Dowd AM, Ellis AE, Secombes CJ. 1998. Binding of immune complexes to Atlantic
salmon peripheral blood leucocytes. Developmental and Comparative Immunology 22:439-
448.
111. Olsen AB, Bornø G, Colquhoun D, Flesjå K, Haldorsen R, Mo TA, Nilsen H, Skjelstad HR,
Hjeltnes B. 2007. Helsesituasjonen hos laksefisk i 2006 (Diseases in farmed salmonids in
2006; in Norwegian). Report from the National Veterinary Institute, Norway.
112. Ortego LS, Hawkins WE, Walker WW, Krol RM, Benson WH. 1995. Immunohistochemical
detection of proliferating cell nuclear antigen (PCNA) in tissues of aquatic animals utilized in
toxicity bioassays. Marine Environmental Research 39:271-273.
113. Partula S. 1999. Surface markers of fish T-cells. Fish & Shellfish Immunology 9:241-257.
114. Poppe T. 1999. Systematisk patology og patofysiology (Systematic pathology and
pathophysiology), p. 11-36. In Poppe T (ed.), Fiskehelse og fiskesykdommer (Fish health and
fish diseases; in Norwegian). Universitetsforlaget, Oslo.
115. Poppe TT, Breck O. 1997. Pathology of Atlantic salmon Salmo salar intraperitoneally
immunized with oil-adjuvanted vaccine. A case report. Diseases of Aquatic Organisms
29:219-226.
116. Poppe TT, Ferguson HW. 2006. Cardiovascular system, p. 141-167. In Ferguson HW (ed.),
Systemic pathology of fish. A text and atlas of normal tissues in teleosts and their responses in
disease. Scotian Press, London.
117. Poppe TT, Seierstad SL. 2003. First description of cardiomyopathy syndrome (CMS)-related
lesions in wild Atlantic salmon Salmo salar in Norway. Diseases of Aquatic Organisms 56:87-
88.
118. Poppe TT, Taksdal T. 2000. Ventricular hypoplasia in farmed Atlantic salmon Salmo salar.
Diseases of Aquatic Organisms 42:35-40.
119. Poppe TT, Taksdal T, Bergtun PH. 2007. Suspected myocardial necrosis in farmed Atlantic
salmon, Salmo salar L.: a field case. Journal of Fish Diseases 30:615-620.
120. Poss K, Wilson L, Keating M. 2002. Heart regeneration in zebra fish. Science 298:2188-2190.

121. Powell MD, Nowak BF, Adams MB. 2002. Cardiac morphology in relation to amoebic gill
disease history in Atlantic salmon, Salmo salar L. Journal of Fish Diseases 25:209-215.
122. Press CM. 1998. Immunology of fishes, p. 3-62. In Pastoret PP, Griebel P, Bazin H, and
Govaerts A (eds.), Handbook of vertebrate immunology. Academic Press, San Diego.
123. Press CM, Dannevig BH, Landsverk T. 1994. Immune and enzyme histochemical phenotypes
of lypmoid and nonlymphoid cells within the spleen and head kidney of Atlantic salmon
(Salmo salar L.). Fish & Shellfish Immunology 4:79-93.
124. Radostits OM, Gay CC, Blood DC, Hindcliff KW. 2000. Veterinary medicine. A textbook of
the diseases of cattle, sheep, pigs, goats and horses. WB Saunders, London.
125. Raynard RS, Houghton GH. 1993. Development towards an experimental protocol for the
transmission of pancreas disease of Atlantic salmon Salmo salar. Diseases of Aquatic
Organisms 15:123-128.
126. Raynard RS, Snow M, Bruno DW. 2001. Experimental infection models and susceptibility of
Atlantic salmon Salmo salar to a Scottish isolate of infectious salmon anaemia virus. Diseases
of Aquatic Organisms 47:169-174.
127. Reite OB. 2005. The rodlet cells of teleostean fish: their potential role in host defence in
relation to the role of mast cells/eosinophilic granule cells. Fish & Shellfish Immunology
19:253-267.
128. Reite OB. 1997. Mast cells/ eosinophilic granule cells of salmonids: staining properties and
responses to noxious agents. Fish & Shellfish Immunology 7:567-584.
129. Reite OB. 1998. Mast cells/ eosinophilic granule cells of teleostean fish: A review focusing on
staining properties and functional responses. Fish & Shellfish Immunology 8:489-513.
130. Reite OB, Evensen Ø. 2006. Inflammatory cells in teleostean fish: A review focusing on mast
cells/eosinophilic granule cells and rodlet cells. Fish & Shellfish Immunology 20:192-208.
131. Roberts, RJ. 2001. The bacteriology of teleosts, p. 297-331. In Roberts RJ (ed.), Fish
Pathology. W.B. Saunders, Edinburgh.
132. Roberts RJ. 2001. Neoplasia of teleosts, p. 151-168. In Roberts RJ (ed.), Fish Pathology. W.B.
Saunders, Edinburgh.
133. Roberts RJ, Ellis AE. 2001. The anatomy and physiology of teleosts, p. 12-54. In Roberts RJ
(ed.), Fish Pathology. W.B. Saunders, Edinburgh.
134. Roberts RJ, Pearson MD. 2005. Infectious pancreatic necrosis in Atlantic salmon, Salmo salar
L. Journal of Fish Diseases 28:383-390.
135. Roberts RJ, Rodger HD. 2001. The pathophysiology and systematic pathology of teleosts, p.
55-132. In Roberts RJ (ed.), Fish Pathology. W.B. Saunders, Edinburgh.
136. Rodger H, Turnbull T. 2000. Cardiomyopathy syndrome in farmed Scottish salmon.
Veterinary Record 146:500-501.
137. Rodger HD. 1991. Summer lesion syndrome in salmon - a retrospective study. Veterinary
Record 129:237-239.
71

72
138. Rodger HD, Murphy TM, Drinan EM, Rice DA. 1991. Acute skeletal myopathy in farmed
Atlantic salmon Salmo salar. Diseases of Aquatic Organisms 12:17-23.
139. Rodger HD, Richards RH. 1998. Haemorrhagic smolt syndrome: a severe anaemic condition
in farmed salmon in Scotland. Veterinary Record 142:538-541.
140. Rodger HD, Turnbull T, Richards RH. 1994. Myopathy and pancreas disease in salmon - a
retrospective study in Scotland. Veterinary Record 135:234-235.
141. Sadasiv EC. 1995. Immunological and pathological responses of salmonids to infectious
pancreatic necrosis virus (IPNV). Annual Review of Fish Diseases 5:209-223.
142. Sanchez-Quintana D, Garcia-Martinez V, Climents V, Hurle JM. 1995. Morphological
analysis of the fish heart ventricle: Myocardial and connective tissue architecture in teleost
species. Annals of Anatomy 177:267-274.
143. Sande RD, Poppe TT. 1995. Diagnostic ultrasound examination and echocardiography in
Atlantic salmon (Salmo salar). Veterinary Radiology & Ultrasound 36:551-558.
144. Sänger AM, Stoiber W. 2001. Muscle fiber diversity and plasticity, p. 187-250. In Johnston IA
(ed.), Muscle development and growth, vol. 18. Academic Press, San Diego.
145. Secombes CJ. 1996. The nonspecific immune system: Cellular defences, p. 63-103. In Iwama
G and Nakanishi T (eds.), The fish immune system. Organism, pathogen, and environment,
vol. 15. Academic Press, San Diego.
146. Seierstad SL, Poppe TT, Koppang EO, Svindland A, Rosenlund G, Frøyland L, Larsen S.
2005. Influence of dietary lipid composition on cardiac pathology in farmed Atlantic salmon,
Salmo salar L. Journal of Fish Diseases 28:677-690.
147. Seternes T, Sørensen K, Smedsrød B. 2002. Scavenger endothelial cells of vertebrates: A
nonperipheral leukocyte system for high-capacity elimination of waste macromolecules.
Proceedings of the National Academy of Sciences of the United States of America 99:7594-
7597.
148. Simko E, Brown LL, MacKinnon AM, Byrne PJ, Ostland VE, Ferguson HW. 2000.
Experimental infection of Atlantic salmon, Salmo salar L., with infectious salmon anaemia
virus: a histopathological study. Journal of Fish Diseases 23:27-32.
149. Skjelstad HR, Bornø G, Flesjå K, Hansen H, Nilsen H, Wasmuth MA, Hjeltnes B. 2008.
Helsesituasjonen hos laksefisk 2007 (Diseases in farmed salmonids 2007; in Norwegian).
Report from the National Veterinary Institute, Norway.
150. Skrudland A, Poppe TT, Jarp J, Brun E. 2002. Hjertesprekk hos norsk oppdrettslaks. En
feltundersøkelse (Cardiomyopathy syndrome in farmed salmon in Norway. A field study; in
Norwegian).
151. Smail DA, Munro ALS. 2001. The virology of teleosts, p. 169-253. In Roberts RJ (ed.), Fish
Pathology. W.B. Saunders, Edinburgh.
152. Smedsrød B, Olsen R, Sveinbjornsson B. 1995. Circulating collagen is catabolized by
endocytosis mainly by endothelial cells of endocardium in cod (Gadus morhua). Cell and
Tissue Research 280:39-48.

153. Solem ST, Stenvik J. 2006. Antibody repertoire development in teleosts - a review with
emphasis on salmonids and Gadus morhua L. Developmental and Comparative Immunology
30:57-76.
154. Sørensen KK, Melkko J, Smedsrød B. 1998. Scavenger-receptor-mediated endocytosis in
endocardial endothelial cells of Atlantic cod Gadus morhua. Journal of Experimental Biology
201:1707-1718.
155. Speilberg L, Evensen Ø, Dannevig BH. 1995. A sequential study of the light and electron
microscopic liver lesions of infectious anaemia in Atlantic salmon (Salmo salar L.).
Veterinary Pathology 32:466-478.
156. Stangeland K, Høie S, Taksdal T. 1996. Experimental induction of infectious pancreatic
necrosis in Atlantic salmon, Salmo salar L., post-smolts. Journal of Fish Diseases 19:323-327.
157. Sterud E, Simolin P, Kvellestad A. 2003. Infection by Parvicapsula sp. (Myxozoa) is
associated with mortality in sea-caged Atlantic salmon Salmo salar in Northern Norway.
Diseases of Aquatic Organisms 54:259-263.
158. Stuckmann I, Evans S, Lassar AB. 2003. Erythropoietin and retinoic acid, secreted from the
epicardium, are required for cardiac myocyte proliferation. Developmental Biology 255:334-
349.
159. Suau A., Bonnet R, Sutren M, Godon JJ, Gibson GR, Collins MD, Doré J. 1999. Direct
analysis of genes encoding 16S rRNA from complex communities reveals many novel
molecular species within the human gut. Applied and Environmental Microbiology 65:4799-
4807.
160. Taksdal T, Olsen AB, Bjerkås I, Hjortaas MJ, Dannevig BH, Graham DA, McLoughlin MF.
2007. Pancreas disease in farmed Atlantic salmon, Salmo salar L., and rainbow trout,
Oncorhynchus mykiss (Walbaum), in Norway. Journal of Fish Diseases 30:545-558.
161. Taksdal T, Stangeland K, Dannevig BH. 1997. Induction of infectious pancreas necrosis (IPN)
in Atlantic salmon Salmo salar and brook trout Salvelinus fontinalis by bath challenge of fry
with infectious pancreatic necrosis virus (IPNV) serotype Sp. Diseases of Aquatic Organisms
28:39-44.
162. Tizard IR. 1996. Veterinary immunology. An introduction. W.B.Saunders Company,
Philadelphia.
163. Turnbull J. 2006. Musculoskeletal system, p. 288-311. In Ferguson HW (ed.), Systemic
pathology of fish. A text and atlas of normal tissues in teleosts and their responses in disease.
Scotian Press, London.
164. Vågsholm I, Djupvik HO, Willumsen FV, Tveit AM, Tangen K. 1994. Infectious salmon
anemia (ISA) epidemiology in Norway. Preventive Veterinary Medicine 19:277-290.
165. Van Vleet JF, Ferrans VJ. 1995. Pathology of the cardiovascular system, p. 175-208. In
Carlton WW and McGavin MD (eds.), Thomson's special veterinary pathology. Mosby, St.
Louis.
166. Watanabe K, Karlsen M, Devold M, Isdal E, Litlabø A, Nylund A. 2006. Virus-like particles
associated with heart and skeletal muscle inflammation (HSMI). Diseases of Aquatic
Organisms 70:183-192.
73

74
167. Watanabe K, Økland S, Holand T, Midttun B, Nylund A. 1995. Hjertesprekk (CMS):
Forårsaket av virus? (Cardiomyopathy syndrome: caused by virus?). Norsk Fiskeoppdrett
(Norwegian Fish Farming; in Norwegian) 21:28-29.
168. Wergeland HI, Jakobsen HA. 2001. A salmonid cell line (TO) for production of infectious
salmon anaemia virus (ISAV). Diseases of Aquatic Organisms 44:183-190.
169. Weston J, Villoing S, Bremont M, Castric J, Pfeffer M, Jewhurst V, McLoughlin M, Rodseth
O, Christie KE, Koumans J, Todd D. 2002. Comparison of two aquatic alphaviruses, salmon
pancreas disease virus and sleeping disease virus, by using genome sequence analysis,
monoclonal reactivity, and cross-infection. Journal of Virology 76:6155-6163.
170. Weston JH, Welsh MD, McLoughlin MF, Todd D. 1999. Salmon pancreas disease virus, an
alphavirus infecting farmed Atlantic salmon, Salmo salar L. Virology 256:188-195.
171. Woese CR. 1987. Bacterial evolution. Microbiological Reviews 51:221-271.
172. Wolf K, Quimby MC. 1966. Lymphocystis virus: isolation and propagation in centrarchid fish
cell lines. Science 151:1004-1005.
173. Zapata AG, Chiba A, Varas A. 1996. Cells and tissues of the immune system of fish, p. 1-62.
In Iwama G and Nakanishi T (eds.), The fish immune system. Organism, pathogen and
environment, vol. 15. Academic Press, San Diego.

Tables
Table 1: Distribution of HSMI outbreaks 2003-2007
Table 2: Histopathological findings during HSMI outbreaks
Table 3: Severity of lesions in cardiac and red skeletal muscle during HSMI outbreaks
Table 4: Occurrence of histopathological changes in several organs
Table 5: Histological stains
Table 6: Histochemical reactivities
Table 7: Fish cell lines

Table 1: Distribution of HSMI outbreaks 2003-2007
Percentage distribution by county for sites with disease outbreaks diagnosed as heart and
skeletal muscle inflammation (HSMI) by the National Veterinary Institute in the period
between 2003 and 2007. Counties are listed by their geographical location from north to
south.
County 2003 2004 2005 2006 2007 Total
Finnmark 0 0 0 2.4 3.9 1.8
Troms 0 1.9 1.2 8.2 5.8 4.2
Nordland 0 3.8 3.6 8.2 16.0 8.5
Nord-Trøndelag 24.6 17.0 21.8 22.4 17.3 20.1
Sør-Trøndelag 14.0 11.3 12.1 20.0 17.3 15.7
Møre og Romsdal 35.1 39.6 32.5 16.5 17.3 25.1
Sogn og Fjordane 12.3 9.4 10.8 3.5 4.5 7.1
Hordaland 0 5.7 4.8 3.5 5.8 4.4
Rogaland 10.5 5.7 9.6 12.9 10.3 10.1
Vest-Agder 3.5 5.6 3.6 2.4 1.8 3.0
Total number of sites 57 54 83 94 162 450
75

Table 2: Histopathological findings during HSMI outbreaks
Histopathological findings in 404 fish sampled during 56 field outbreaks of heart and skeletal
muscle inflammation (HSMI) in the period 1999-2007. The investigation was performed for
this thesis. Fish are categorised by their clinical status at sampling.
Total = exact number of fish examined. % = percentage of examined fish having this lesion.
Organ
Compartment Lesion
Normal Moribund Dead
Total % Total % Total %
Myocarditis 127 80 115 85 147 93
Atrium
Assoc. necrosis
Endocarditis
127
127
80
81
115
115
85
86
147
147
93
93
Thrombosis 127 6 115 9 147 20
Ventricle epicardium Epicarditis 134 99 117 100 153 100
Perivasculitis 134 93 117 95 153 82
Leukocytosis 134 5 117 3 153 3
Heart
Ventricle compactum
Myocarditis
Assoc. necrosis
134
134
97
97
117
117
98
98
153
153
99
99
Only necrosis 134 0 117 0 153 0
Thrombosis 134 0 117 0 153 0
Myocarditis 134 95 117 96 153 99
Assoc. necrosis 134 95 117 96 153 99
Ventricle spongiosum Only necrosis 134 0 117 0 153 0
Nuclear aggregations 134 66 117 81 153 71
Thrombosis 134 2 117 3 153 12
Ventricle endocardium Endocarditis 134 95 117 96 153 99
Red muscle
Skeletal muscle
White muscle
Myositis
Degeneration/necrosis
Myositis
Degeneration/necrosis
82
82
82
82
85
91
2
36
91 87
91 86
92 2
92 17
136
136
138
138
88
89
0
22
Multifocal necrosis 69 26 66 50 113 43
Liver
Inflammation 69 1 66 8 113 4
Perivasculitis 69 23 66 23 113 26
Pancreas Exocrine
Necrosis
Partly loss of cells
79
79
1
0
71
71
0
0
121
121
3
2
Peritoneum Peritonitis 79 68 71 80 121 62
Spleen
Blood concentration
Pseudolobulation
61
61
79
43
47
47
85
53
98
98
88
48
Kidney
Blood concentration
Oedema
63
63
22
19
47
47
30
21
102
102
40
33
Gills
Lamellae
Haemmorhage
Inflammation
62
62
32
6
54
54
22
33
86
86
21
17
Central venous sinusoid Blood concentration 62 19 54 20 86 16
Total number of fish sampled 134 117 153
76

Table 3: Severity of lesions in cardiac and red skeletal muscle during HSMI outbreaks
Percentage distribution of lesion severity in cardiac and red skeletal muscle in apparently normal, moribund and dead fish sampled in the mid-
outbreak phase of heart and skeletal muscle inflammation (HSMI) on 56 sites in the period between 1999 and 2007. The investigation was
performed for this thesis. Fish are categorised by their clinical status at sampling.
Total = exact number of fish examined (100 %). 0 no lesions, + mild and focal to multifocal lesions, ++ moderate multifocal lesions, +++ severe
diffuse lesions.
Atrium Ventricle Red skeletal muscle
Clinical
Epicardium Compactum Spongiosum
status
Total 0 + ++ +++ Total 0 + ++ +++ 0 + ++ +++ 0 + ++ +++ Total 0 + ++ +++
Normal 127 20 24 28 28 134 1 6 32 61 3 15 45 37 5 17 40 38 82 15 38 22 25
Moribund 113 13 21 34 32 117 0 10 23 67 2 8 38 52 4 7 36 53 91 13 43 24 20
Dead 147 8 33 31 28 153 0 3 20 77 1 12 41 46 1 7 28 64 136 12 46 28 14
77

Table 4: Occurrence of histopathological changes in several organs
Combination of histopathological changes in the heart, red skeletal muscle, liver and spleen in
191 Atlantic salmon (Salmo salar) sampled during 47 clinical outbreaks of heart and skeletal
muscle inflammation (HSMI). The investigation was performed for this thesis.
+ lesion found, - lesion not found
Myocarditis
heart
78
Myositis
red skeletal
muscle
Necrosis
liver
Pseudolobulation
spleen
Number of
fish
Percentage
distribution
+ + + + 37 19.4
+ + + - 26 13.6
+ + - + 44 23.0
+ + - - 54 28.3
+ - + + 3 1.6
+ - + - 2 1.0
+ - - + 7 3.7
+ - - - 16 8.4
- + + + 0 0
- + + - 0 0
- + - + 0 0
- + - - 1 0.5
- - + + 0 0
- - + - 0 0
- - - + 0 0
- - - - 1 0.5
Total 191 100

Table 5: Histological stains
Staining properties of dyes commonly used in histology 68 .
HE = haematoxylin and eosin; PAS = Periodic acid-Schiff; ZN = Ziehl Nielsen; vG = van
Gieson; MSB = Martius Scarlet Blue.
Method Main components Structures stained Colour
HE
PAS
Gram
Giemsa
ZN
vG
MSB
Haematoxylin
Aluminium
Eosin
Periodic acid
Schiff’s reagent
Nuclear chromatin Blue
Cytoplasm, erythrocytes, collagen,
keratin
Hexose- and sialic acid-containing
mucosubstances
Red
Pink to red
Nuclear chromatin Pink to purple
Hucker’s crystal violet
Gram’s iodine
Gram-positive bacteria Blue
Ethanol Gram-negative bacteria Red
Methylene blue eosinate
Azure A and B eosinates
Methylene blue
Nuclei, bacteria Blue to purple
Erythrocytes, collagen, keratin Pink
Carbolfuchsine
Nuclei, cytoplasm Blue
Methylene blue Lipofuscine Red
Weigert’s iron-haematoxylin Nuclei Black
Acid fuchsine
Collagen Red
Piric acid Cytoplasm, keratin, erythrocytes Yellow
Celestine blue
Haematoxylin
Martius yellow
Phosphotungstic acid
Brilliant crystal scarlet
Aniline blue
Nuclei Blue to black
Collagen Blue
Fibrin Red
Erythrocytes Yellow
79

Table 6: Histochemical reactivities
Enzyme- and immunohistochemical reactivity of fish leukocytes, epithelial and endothelial
cells as reviewed in the literature.
ACP = acid phosphatase; ALP = alkaline phosphatase; NSE = non-specific esterase; Px =
peroxidase; MHC II = major histocompatibility complex class II; Ig = immunoglobulin.
– no reactivity, + positive reaction.
80
Cells ACP ALP NSE Px MHC II
Neutrophils + 38,173 + 123 + 38,122,173
Basophils - 122,173
EGC + 145 + 145 + 38,122
Macrophages + 6,122,123
+ 6,122,123,145
+ 122
- 145
Melanomacrophages + 123 +/- 53 + 123 + 74
+ 74,122,145
T lymphocytes + 74 - 87,122
B lymphocytes + 74 + 87,122
Plasma cells - 74
Epithelial cells + 6 + 6 + 74
Endothelial cells + 123
Ig

Table 7: Fish cell lines
Cell lines used for inoculation of material from fish with heart and skeletal muscle
inflammation (HSMI) in the present study.
Derived from species Tissue Abbreviation Reference
Common carp Cyprinus carpio Epithelioma EPC
Bluegill sunfish Lepomis machrochirus Fry (muscle) BF-2
Fathead minnow Pimephales promelas Epithelium FHM
Chinook salmon Oncorhynchus tschawytscha Embryo CHSE-214
Chum salmon Oncorhynchus keta Heart CHH-1
Rainbow trout Oncorhynchus mykiss Gill RT-Gill
Atlantic salmon Salmo salar Head kidney ASK II
Atlantic salmon Salmo salar Head kidney SHK-1
Atlantic salmon Salmo salar Head kidney TO
46
172
51
77
77
8
31
24,26
168
81

Figures
Figure 1: Anatomy of heart and skeletal muscle
Figure 2: Pathology of heart and skeletal muscle
Figure 3: HSMI diagnoses in the period 1999-2007

Figure 1: Anatomy of heart and skeletal muscle
Normal morphology of heart and skeletal muscle in Atlantic salmon, Salmo salar.
1A: Heart, ventricle. The heart is lined externally by epicardium ().Underneath the
epicardium are two concentric layers of compact myocardium lying perpendicularly to
each other (a and b). It is oxygenated through coronary vessel branches (). Spongy
myocardium (c) is found towards the centre of the ventricle. It is lined by endocardium
(). HE. Bar = 50 m
1B: Heart, ventricle. Compact myocardium is separated from spongy myocardium by an
intermediate layer () mainly consisting of connective tissue. The spongy layer consists
of loosely organised myocardium (*) and endocardium (). HE. Bar = 50 m
1C: Heart, ventricle. The nucleus () of a cardiomyocyte is placed centrally in the cell. The
sarcoplasm (*) mainly consists of myofibrils. Endocardium () separating myocardium
from the ventricular lumen consists of endothelial cells and connective tissue. HE. Bar =
20 m
1D: Heart, ventricle. Myofibrils consist of thin and thick filaments organised in a repetetive
pattern of sarcomeres, running from one Z-line () to the next. The dark and light bands
are visible by light microscopy, creating the cross-striation of cardiac and skeletal muscle
tissue. Transmission electron microscopy. Bar = 500 nm.
1E: Skeletal muscle. White muscle enables rapid movements. It consists of muscle fibres with
a large diameter, and is the main component of the filet. Red muscle is necessary for
maintenance movements, and is found along the lateral line. It contains smaller muscle
fibres, but more myoglobin than white muscle. HE. Bar = 50 m
1F: Red skeletal muscle. Multiple nuclei () are situated excentrally to the muscle fibre (*).
The muscle tissue is oxygenated through a capillary network (). HE. Bar = 50 m
82

Figure 1
83

Figure 2: Pathology of heart and skeletal muscle
Pathological changes in the heart and skeletal muscle of Atlantic salmon, Salmo salar. HE.
Samples are taken from fish with heart and skeletal muscle inflammation.
2A: Heart, ventricle. Hypertrophic nuclei of cardiomyocytes () and endothelial cells () are
signs of an increased work load. Bar = 20 m
2B: Red skeletal muscle. Necrotic muscle cells (*) show strong eosinophilia, centration of
nuclei and loss of cross-striation. Bar = 20 m
2C: Heart, ventricle. Necrotic cardiomyocytes () shrink, become eosinophilic and lose cross
striation. Inflammation is characterised by an abundance of inflammatory cells within and
around muscle fibres (*). Bar = 20 m
2D: Heart, ventricle. Necrotic cardiomyocytes finally disintegrate and homogenous
eosinophilic debris may be observed (). Bar = 20 m
2E: Heart, atrium. Cellular thrombi () are a result of intravascular coagulation due to
circulatory disturbances and/or damage to the endothelium. Bar = 50 m
2F: Red skeletal muscle. Signs of regeneration in muscle tissue may be observed as basophilic
areas (). Bar = 20 m
84

Figure 2
85

Figure 3: HSMI diagnoses in the period 1999-2007
Diagnosed cases of heart and skeletal muscle inflammation (HSMI) in Norway in the period
between 1999 and 2007. Numbers for 1999-2002 are uncertain.
86
Number of cases
180
160
140
120
100
80
60
40
20
0
HSMI diagnoses 1999-2007
1999 2000 2001 2002 2003 2004 2005 2006 2007
Year

Papers
87

Paper I
Pathology of heart and skeletal muscle inflammation (HSMI) in
farmed Atlantic salmon Salmo salar

DISEASES OF AQUATIC ORGANISMS
Vol. 59: 217–224, 2004 Published June 11
Dis Aquat Org
Pathology of heart and skeletal muscle inflammation
(HSMI) in farmed Atlantic salmon Salmo salar
INTRODUCTION
Heart and skeletal muscle inflammation (HSMI) is a
novel disease in farmed Norwegian Atlantic salmon
Salmo salar. HSMI was first diagnosed in 1999, and has
since become an increasing problem in Norwegian
salmon farms. Salmon are commonly affected 5 to 9 mo
after transfer to sea, but outbreaks have been recorded
as early as 14 d following seawater transfer. Affected
fish are anorexic and display abnormal swimming behaviour.
Autopsy findings typically include a pale
heart, yellow liver, ascites, swollen spleen and petechiae
in the perivisceral fat. Diagnosis of HSMI is
presently based on histological examination. HSMI is
characterised by extensive panmyocarditis and myositis,
particularly involving red skeletal muscle. Morbidity
may be very high, while mortalities are variable and
may reach 20% in affected cages. Environmental stress
during an outbreak seems to increase mortalities (A.
*Email: ruth-torill.kongtorp@vetinst.no
R. T. Kongtorp 1, *, T. Taksdal 1 , A. Lyngøy 2
1 National Veterinary Institute, Oslo, Norway
2 Nordvest fiskehelse A/S (North West Fish Health Ltd), Gurskøy, Norway
ABSTRACT: This is the first description of heart and skeletal muscle inflammation (HSMI), a novel
disease affecting farmed Atlantic salmon Salmo salar in Norway. HSMI was first diagnosed in 1999,
and there has since been a yearly increase in the number of recorded outbreaks. Atlantic salmon are
commonly affected 5 to 9 mo after transfer to sea, but outbreaks have been recorded as early as 14 d
following seawater transfer. Affected fish are anorexic and display abnormal swimming behaviour.
Autopsy findings typically include a pale heart, yellow liver, ascites, swollen spleen and petechiae in
the perivisceral fat. While mortality is variable (up to 20%), morbidity may be very high in affected
cages. Until more accurate tests are available, HSMI is diagnosed on the basis of histopathology. The
major pathological changes occur in the myocardium and red skeletal muscle, where extensive
inflammation and multifocal necrosis of myocytes are evident. HSMI is transmissible and, although
most likely caused by a virus, the causal agent has not yet been isolated. This paper describes clinical
signs and pathology of HSMI from 3 field outbreaks in Norway. Microscopic lesions are compared
and discussed in relation to published descriptions of pancreas disease (PD) and cardiomyopathy syndrome
(CMS). It is concluded that HSMI is histopathologically distinguishable from PD and CMS.
KEY WORDS: Heart and skeletal muscle inflammation · HSMI · Atlantic salmon · Myocarditis ·
Myositis
Resale or republication not permitted without written consent of the publisher
Lyngøy unpubl. data). HSMI is experimentally transmissible
to healthy Atlantic salmon by intraperitoneal
injection with affected cardiac tissue and by cohabitation
with injected fish (R. T. Kongtorp et al. 2004).
Myocarditis and myopathy have previously been
described in pancreas disease (PD) (Ferguson et al.
1986, Rodger et al. 1991, 1994, Murphy et al. 1992,
McLoughlin et al. 1995, 1996, 1997, 2002), and in cardiomyopathy
syndrome (CMS) (Ferguson et al. 1990,
Rodger & Turnbull 2000).
Outbreaks of PD have been reported from Scotland
(Munro et al. 1984), USA (Kent & Elston 1987), Norway
(Poppe et al. 1989), Ireland (Murphy et al. 1992),
France and Spain (Raynard et al. 1992). PD is caused
by salmon pancreas disease virus (SPDV) (Nelson et al.
1995), characterised as an alphavirus (Weston et al.
1999). SPDV is closely related to sleeping disease virus
(SDV) of farmed rainbow trout Oncorhynchus mykiss
(Castric et al. 1997, Villoing et al. 2000a,b).
© Inter-Research 2004 · www.int-res.com

218
CMS was first described in Norway (Ferguson et al.
1990), and subsequently in the Faeroe Islands (Bruno &
Poppe 1996) and Scotland (Rodger & Turnbull 2000).
The causality of CMS is unclear, but Grotmol et al.
(1997) described nodavirus-like particles in association
with heart lesions consistent with CMS in Norway.
As the causal agent has not yet been found, it is
important to establish a protocol for distinguishing
HSMI histologically from other muscular disorders of
Atlantic salmon. In this paper we describe the clinical
signs, gross pathology and major histopathological
findings from 3 field outbreaks of HSMI in Atlantic
salmon during 1999 and 2003, and compare and discuss
these features in relation to previously published
descriptions of PD and CMS.
MATERIALS AND METHODS
Field samples. Atlantic salmon were sampled during
disease outbreaks in a seawater farm on the Norwegian
west coast in July 1999 and on 2 different
locations in western Norway in May and June 2003.
Clinically diseased and dead fish were sampled nonrandomly,
whereas asymptomatic fish were collected
randomly from affected cages. Fish sampled from the
1999 outbreak were killed by a blow to the head before
blood samples were collected from the caudal vein,
and macroscopic changes were recorded before tissues
were sampled for histology. Fish sampled from
the 2003 outbreaks were euthanised at the fish farm
and sent on ice to the National Veterinary Institute
(NVI), Oslo, where autopsy and sampling for histological
examination took place 1 d after euthanasia.
Histology. Tissue samples were obtained from gills,
heart, skeletal muscle, mid-kidney, spleen and pyloric
caeca with pancreatic tissue, and fixed in 10% neutral
phosphate buffered formalin. Samples from skeletal
muscle consisted of longitudinal and transverse sections
of the area ventral to the dorsal fin and adjacent
to the lateral line, including red and white muscle
tissue. Formalin fixed samples were prepared for histological
examination by standard paraffin wax techniques
and stained with haematoxylin and eosin (H&E).
Microbiological examination. Kidney swabs were
streaked onto blood agar supplemented with 2% NaCl
and incubated at 15°C for 7 d.Kidney swabs from the
2003 outbreak were also streaked onto blood agar and
incubated at 20°C for 7 d. Kidney samples were also
tested for infectious salmon anaemia virus (ISAV) by
an indirect fluorescent antibody technique (IFAT)
(Falk & Dannevig 1995). Kidney and myocardial tissues
from the 1999 outbreak were inoculated onto
Chinook salmon embryo (CHSE-214) (Lannan et al.
1984), Blue-gill fry (BF-2) (Wolf & Quimby 1966) and
Dis Aquat Org 59: 217–224, 2004
epithelioma papillosum cyprini (EPC) (Fijan et al.
1983) cell cultures according to standard procedures at
the NVI. Mid-kidney material from the 2003 outbreak
was inoculated onto BF-2 and EPC cell cultures. The
material was passaged after 1 wk and incubated for an
additional week.
Challenge study. Homogenates of mid-kidney from
the 1999 outbreak were injected intraperitoneally into
20 Atlantic salmon. Injected salmon were placed in a
tank of dechlorinated freshwater together with 20 noninjected
fish (cohabitants). On commencement of the
experiment, the fish had a mean weight of 25 g. The
water flow was 90 l h –1 , and the average temperature
was 12°C. Challenged fish were observed for 7 wk. At
the end of the experiment, heart samples were prepared
for histology as described above.
RESULTS
Clinical signs
The most prominent signs of disease in the sea cages
were anorexia and abnormal swimming behaviour.
During the 1999 disease outbreak, there was an initial
decrease in appetite and mortalities varying from 4 to
15.5% prior to sampling. Peracute mortalities similar to
those observed in outbreaks of CMS were also seen.
During the 2003 disease outbreaks, anorexia was not as
evident. In the first of these outbreaks, mortality had
been recorded as negligible when HSMI was first diagnosed
3 mo prior to sampling, but slowly increasing in
the mo before fish were collected for the present study.
The second 2003 outbreak had negligible cage mortality
when HSMI was diagnosed 2 wk before the study
samples were taken. One wk prior to sampling the mortality
suddenly rose to 20% in cages with affected fish.
Fish with abnormal swimming behaviour were dark
and lethargic, typically positioned near the cage wall
and facing the sea current. The fish retained normal
balance, but did not respond to feeding or other stimuli.
Gross pathology
Five clinically diseased and 15 asymptomatic fish
were sampled from the 1999 outbreak. In the 2003 outbreaks,
5 dead and 10 asymptomatic fish were examined.
There were no evident differences between the 3
groups of clinically diseased, dead or asymptomatic
fish. All were in good condition but with no gut contents.
There were no obvious external lesions. Of the
35 fish sampled, 21 had a pale or greyish heart
(Fig. 1A,B). The pericardial cavities of 15 of these fish
contained small amounts of haemorrhagic fluid or

Kongtorp et al.: Heart and skeletal muscle inflammation in Atlantic salmon
Fig. 1. Salmo salar. Macroscopical findings in heart and skeletal muscle inflammation (HSMI). (A) In this salmon, the heart is pale
and there is haemopericardium. The spleen appears swollen. (B) Closer picture of a pale heart. (C) A few fish had a fibrinous
layer on the liver. a: haemopericardium, b: spleen, c: heart, d: fibrinous layer
coagulum. Small amounts of ascitic fluid were seen in
8 fish. Petechiae were observed in the perivisceral fat
of 10 fish. The liver was pale or yellowish in 20 fish and
stained brown in 6 fish. A fibrinous coat was present on
the surface of the livers of 5 fish (Fig. 1C). In 15 fish the
spleen was swollen. Gross pathology of the skeletal
muscle was sparse, but in 3 fish the muscular layer
appeared loose. Of asymptomatic fish sampled from
the 1999 outbreak, 3 had no macroscopic lesions.
Histopathology
Heart
HSMI was recognised by an extensive panmyocarditis
that was a consistent finding in the 10 clinically
diseased and dead fish. Twelve of 15 asymptomatic
219
fish sampled in 1999 and all 10 sampled in 2003 displayed
panmyocarditis.
A severe epicarditis, characterised by a massive
infiltration of mononuclear cells, was present in 34 fish
(Fig. 2A). Deposits of fibrinous material in the epicardium
were observed in 4 fish.
A range of histological changes was present in the
hearts of affected fish. The most common lesions were
ventricular, in which both the spongy and compact
layers were infiltrated by mononuclear cells, comprising
macrophages, lymphocyte- and plasma-like cells
(Fig. 2B). The inflammatory cells were localised within
and around myocytes in a diffuse or focal pattern, most
evident in the compact layer. In 3 mildly affected fish,
cellular infiltrates were seen only in association with
vessels in the compact myocardium, and in 2 fish
lesions were restricted to the interface between the
spongy and compact layers. Nine moderately affected

220
a
b
b
hearts were focally hypercellular (Fig. 2C). In 20
severe cases, cellular infiltration was diffuse and
extensive, and a large number of necrotic myocytes
were observed (Fig. 2B).
Dis Aquat Org 59: 217–224, 2004
A
C
E
Fig. 2. Salmo salar. Micrographs of sections from affected fish (H&E). (A) Ventricle showing severe epicarditis and myocarditis in
the underlying compact myocardium. (B) Focal cellular infiltration in the compact myocardium. (C) Severe, diffuse myocarditis in
the compact myocardium. (D) Cross section of red skeletal muscle, featuring severe myositis. (E) Vacuolisation and degeneration
of red skeletal muscle fibres. Infiltration of mononuclear cells. (F) Longitudinal section of red skeletal muscle. The muscle fibres
show signs of degeneration. a: epicardium, b: compact myocardium, c: focal cellular infiltration. Scale bars = 20 μm
B
D
F
Affected myocytes exhibited signs of degeneration
with condensation, eosinophilia, loss of muscle striation,
vacuolation and central nuclei. Several cells were
also undergoing pyknosis or karyolysis. Necrotic cells
c

Kongtorp et al.: Heart and skeletal muscle inflammation in Atlantic salmon
appeared in association with inflammatory infiltrates.
Hypertrophic nuclei could be observed in a few
myocytes within the spongy muscle layer. Endocardial
cells were also multifocally hypertrophic, appearing
most often adjacent to myocyte inflammation.
Skeletal muscle
Three fish with mild cardiac lesions did not have any
lesions in the skeletal muscle, while 31 fish with moderate
or severe heart lesions also had lesions in the skeletal
muscle. In the latter individuals the red skeletal
muscle was hypercellular, with mononuclear cells within
and between muscle fibres (Fig. 2D). Loss of striation,
eosinophilia and vacuolation of myocytes were present,
and single cells were necrotic (Fig. 2E,F). White muscle
was mildly affected in 13 fish. Single white fibres were
degenerative, with no visible inflammation. In 5 cases,
the inflammation extended to white fibres adjacent to affected
red fibres. There were no lesions in scales or skin.
20 μm
a
100 μm
a
a
A B
C D
Liver
Multifocal necrosis of hepatocytes was found in 15
fish displaying severe panmyocarditis. There was no
evident pattern to the necrotic foci, and they varied in
size and shape (Fig. 3A). Affected cells were vacuolated
and pyknotic or karyolytic (Fig. 3B). Phagocytosis
of normal hepatocytes was associated with necrotic
foci. There was no prominent cellular response to the
hepatocytic lesions.
Other organs
221
Pancreatic lesions were not observed (Fig. 3C). Focal
haemorrhage and accumulation of erythrocytes was
observed in the kidney, spleen and gills of 7 fish. The
splenic parenchyma of 16 fish also appeared pseudolobulated
(Fig. 3D). Moderate to severe peritonitis was
present in all 35 fish. Macrophage-like cells with a
hyaline-like eosinophilic content were observed in
Fig. 3. Salmo salar. Micrographs of sections from affected fish (H&E). (A) Multifocal hepatocyte necrosis in fish with severe
myocardial lesions. (B) Necrotic focus in the liver. (C) Intact acinar pancreatic tissue in HSMI diseased fish. (D) Pale-coloured cells
create pseudolobulation of the spleen. a: necrotic foci, b: splenic pseudolobulation
20 μm
20 μm
a
b

222
these lesions. Similar cells have been associated with
vaccine-induced peritonitis (Poppe & Breck 1997). Epitheliocystis-like
epithelial inclusions and Ichtyobodo
necator (Costia) were observed in the gills of 20 fish.
Microbiological examination
Pericardial samples from 1 of 4 asymptomatic fish
with pale hearts showed sparse growth on blood agar
with 2% NaCl. The cultures were identified as a Vibrio
sp. Four kidney samples from clinically diseased and 8
from asymptomatic fish did not reveal any bacteriological
growth on blood agar with or without 2% NaCl.
There was no evident cytopathic effect in any of the
cell cultures incubated with tissue material from 10
asymptomatic fish with myocarditis. IFAT tests on kidney
imprints from 17 fish were negative for ISAV.
Challenge study
None of the 20 fish showed clinical signs of disease
during the experiment. On histological examination,
1 injected fish displayed epicarditis. There were no
cardiac lesions in the remaining fish.
DISCUSSION
This study describes the clinical and histopathological
features of HSMI for the first time. The pathological
findings presented correlate with typical observations
during other field outbreaks of HSMI (R. T. Kongtorp &
A. Lyngøy pers. obs). The morbidity was remarkably
high in asymptomatic fish. At autopsy, the most common
finding was a pale heart, possibly due to a severe
epicarditis. Histopathologically, heart and red skeletal
muscle appeared to be most severely affected, showing
a severe inflammation of myocytes that was consistent
in the myocardium and common in skeletal muscle.
Indications of myocyte damage, including loss of
striation, eosinophilia and necrosis, were also present.
While lesions were restricted to heart and skeletal
muscle in mild and moderate HSMI, they were more
widespread in severe cases. Multifocal liver necrosis
was present in under half of the fish diagnosed with
HSMI, and accumulation of erythrocytes was observed
in the gills, spleen and kidney of approximately 20% of
the fish under study.
The panmyocarditis found in Atlantic salmon with
HSMI was remarkably extensive. We suggest, therefore,
that a putative agent causing HSMI must be
present in the myocardium at some stage in the pathogenesis,
thus inducing an immune response. The con-
Dis Aquat Org 59: 217–224, 2004
sistency of cardiac involvement further supports this
hypothesis. When present, the severity of myositis in
red muscle appears to correlate with the degree of
cardiac pathology. It is not clear whether the inflammation
or myocyte degeneration is the primary
response in outbreaks of HSMI. When diagnosed in
the field, inflammation is the most evident component
of the muscle histopathology. It is likely, however, that
myocytes degenerate prior to the infiltration of inflammatory
cells. A causal agent could possibly induce
degeneration primarily, while the massive cellular
infiltration may occur secondarily in an attempt to
remove damaged cells from the affected area. Alternatively
or simultaneously, the severe inflammation itself
may induce myocyte degeneration and necrosis in
unaffected myocytes. A sequential field study may provide
more understanding of the pathogenesis of HSMI.
The role of hepatocytes in this disease is uncertain,
but the magnitude of hepatocyte damage is likely to
affect homeostasis or vice versa. Hepatocyte necrosis
is, however, likely to be a secondary effect of the cardiac
pathology, as no inflammatory response can be
detected in association with the hepatic lesions. General
tissue involvement is probably also an effect of
circulatory disturbance. The splenic pathology in
HSMI may result from a reaction to the massive release
of leucocytes into the bloodstream, but the cause of
this pattern is only speculative.
The investigations conducted in this study indicate
that the morbidity of an outbreak is much higher than
its clinical appearance. Random sampling of salmon
from affected sea cages revealed a high frequency of
asymptomatic individuals with moderate to severe
histopathological changes. This has also been described
for cardiac diseases of several mammalian species,
where lesion severity is not always correlated to
clinical signs (Carlton & McGavin 1995). The magnitude
of damage required to cause mortality in HSMI is
unclear. Stress factors that are known to reduce disease
resistance may play a critical role in the outcome
of the disease. There have been some reports of particularly
high mortality of HSMI in association with management
stresses such as delousing and grading (A.
Lyngøy pers. obs.). Sudden deaths may occur due to
heart failure in such stressful situations because of
extensive myocardial damage.
Results from diagnostic examinations and transmission
experiments (R. T. Kongtorp et al. 2004), suggest
that HSMI is an infectious disease. These studies indicate
that HSMI is likely to be caused by a virus. No
known viruses have been isolated using standard diagnostic
methods at NVI, and we suggest therefore that a
novel pathogenic virus causes HSMI. Although the incubation
period of cell cultures used in the present
study was probably too short to rule out the presence of

Kongtorp et al.: Heart and skeletal muscle inflammation in Atlantic salmon
Table 1. Histopathological lesions appearing in heart and
skeletal muscle inflammation (HSMI), pancreas disease (PD)
and cardiomyopathy syndrome (CMS) in Atlantic salmon
Lesions HSMI PD CMS
Epicarditis + + +
Myocarditis and degeneration of
compact myocardium
+ + –
Myocarditis and degeneration of
spongy myocardium
+ + +
Skeletal muscle inflammation and
degeneration
+ + –
Multifocal necrosis of hepatocytes + – +/–
Necrosis of exocrine pancreas – + –
SPDV, histological examination does not support a PD
diagnosis. The negative results from the challenge experiment
in the present study could be explained by the
short experimental period. Other explanations could be
that the inoculating material was prepared from fish
that were past the infectious stage of pathogenesis, that
a failure occurred in the preparation method or that the
study was conducted in freshwater. In the field, HSMI
has only been reported during the seawater stage.
However, pancreatic necrosis has been successfully
transmitted to salmon in freshwater as early as 4 d postinjection
with PD homogenate (Boucher et al. 1995,
Murphy et al. 1995), and fish presented pancreatic
changes up to 13 wk after challenge (Houghton 1994).
Whether HSMI is a new disease in its own right or a
variant expression of a known agent is not clear at this
stage. The lesions associated with HSMI are partly
similar to those exhibited in PD and CMS (Table 1).
Thus, it may be difficult to distinguish between the 3
diseases. Both PD and HSMI occur most commonly
within the first year in seawater (McLoughlin et al.
2002). CMS often affects older salmon, usually 14 to
18 mo after sea transfer (Ferguson et al. 1990). A characteristic
feature of CMS is sudden death with little
evidence of clinical disease. This was also seen in the
1999 HSMI outbreak, but is a rather uncommon occurrence
in field outbreaks (A. Lyngøy pers. obs.). Mortality
is highly variable in both PD (Crockford et al. 1999)
and HSMI. Following an outbreak of PD, however, up
to15% of the survivors fail to grow (Munro et al. 1984).
This has not been observed in association with HSMI.
Histopathological changes in the compact ventricular
myocardium are consistent in both PD and HSMI.
In PD the compact myocardial necrosis is described as
a focal degenerative myopathy, and appears in the
early stages of disease (Ferguson et al. 1986, Murphy
et al. 1992, McLoughlin et al. 1995). Massive, usually
diffuse, mononuclear cell infiltration is clearly evident
in HSMI cases. Lesions attributed to CMS are typically
restricted to the spongy myocardium (Ferguson et al.
1990). In the spongy portion of the ventricle and the
atrium, the pathologic changes in HSMI, PD and CMS
are similar, and are not easily distinguished. In the
event of simultaneous occurrence of CMS and HSMI, it
may be difficult to differentiate between the 2 diseases.
Because of the similarity of lesions in the spongy
myocardium, HSMI is likely to mask CMS. Incidences
of sudden death in HSMI outbreaks may be the result
of such co-occurrence, but it may be just as likely that
severe damage to the spongy myocardium will cause a
similar mortality pattern regardless of cause.
Skeletal muscle lesions are present in both HSMI
and PD (Ferguson et al. 1986, Murphy et al. 1992). In
HSMI, as well as in PD (Murphy et al. 1992), the red
muscle tissue is most severely affected. Only limited
lesions in skeletal muscle have been described from
fish with CMS (Ferguson et al. 1990).
Widespread multifocal hepatocytic necrosis, as seen
in HSMI, has not been reported in PD. In CMS, degeneration
and necrosis of hepatocytes has been observed
distal to hepatic vessels (Ferguson et al. 1990). In
HSMI, liver necroses are observed both proximal and
distal to hepatic vessels.
One of the major histopathological findings in PD is
an acute necrosis of acinar pancreatic cells, followed
by severe or complete loss of exocrine pancreatic tissue
(Munro et al. 1984). The inflammatory response is
minimal (Munro et al. 1984, Ferguson et al. 1986,
McLoughlin et al. 1995, Murphy et al. 1995). Investigations
of PD in Norwegian salmon farms (A. B. Olsen
pers. comm.) suggest that the process of pancreatic
damage is prolonged and that the pancreatic tissue
regenerates at a later stage than described by Desvignes
et al. (2002). In PD outbreaks in Norway, one
would therefore expect pancreatic lesions to occur
concurrently with heart and muscle lesions. Fish sampled
from outbreaks of HSMI have intact pancreatic
tissue.
In conclusion, the present work shows that HSMI is
histopathologically distinguishable from PD and CMS,
although there are similarities that challenge correct
diagnosis. To increase the accuracy of the diagnosis,
more work has to be done in order to establish the
exact aetiology of HSMI.
Acknowledgements. Sincere thanks to A. B. Olsen at the
National Veterinary Institute in Bergen, Norway, for valuable
participation in sampling and diagnostics during the 1999
outbreak.
LITERATURE CITED
223
Boucher P, Raynard RS, Houghton G, Baudin Laurencin F
(1995) Comparative experimental transmission of pancreas
disease in Atlantic salmon, rainbow trout and brown
trout. Dis Aquat Org 22:19–24

224
Bruno DW, Poppe TT (1996) A colour atlas of salmonid diseases.
Academic Press, London, p 140–141
Carlton WW, McGavin MD (1995) Pathology of the cardiovascular
system. In: Duncan LL (ed) Thomson’s special veterinary
pathology. Mosby, St. Louis, p 175–208
Castric J, Baudin Laurencin F, Bremont M, Jeffroy J, Le Ven
A, Bearzotti M (1997) Isolation of the virus responsible for
sleeping disease in experimentally infected rainbow trout
(Oncorhynchus mykiss). Bull Eur Assoc Fish Pathol 17:
27–30
Crockford T, Menzies FD, McLoughlin MF, Wheatley SB,
Goodall EA (1999) Aspects of the epizootiology of pancreas
disease in farmed Atlantic salmon Salmo salar in
Ireland. Dis Aquat Org 36:113–119
Desvignes L, Quentel C, Lamour F, Le Ven A (2002) Pathogenesis
and immune response in Atlantic salmon (Salmo salar
L.) parr experimentally infected with salmon pancreas disease
virus (SPDV). Fish Shellfish Immunol 12:77–95
Falk K, Dannevig BH (1995) Demonstration of infectious
salmon anaemia (ISA) viral antigens in cell cultures and
tissue sections. Vet Res 26:499–504
Ferguson HW, Roberts RJ, Richards RH, Collins RO, Rice DA
(1986) Severe degenerative cardiomyopathy associated
with pancreas disease in Atlantic salmon, Salmo salar L.
J Fish Dis 9:95–98
Ferguson HW, Poppe T, Speare DJ (1990) Cardiomyopathy
in farmed Norwegian salmon. Dis Aquat Org 8:225–231
Fijan N, Sulimanovic D, Bearzotti M, Muzinic D, Zwillenberg
LO, Chilmonczyk S, Vautherot JF, de Kinkelin P (1983)
Some properties of the Epithelioma Papulosum Cyprini
(EPC) cell line from carp Cyprinus carpis. Ann Virol 134:
207–220
Grotmol S, Totland GK, Kryvi H (1997) Detection of a
nodavirus-like agent in heart tissue from reared Atlantic
salmon Salmo salar suffering from cardiac myopathy
syndrome (CMS). Dis Aquat Org 29:79–84
Houghton G (1994) Aquired protection in Atlantic salmon
Salmo salar parr and post-smolts against pancreas disease.
Dis Aquat Org 18:109–118
Kent ML, Elston RA (1987) Pancreas disease in pen-reared
Atlantic salmon in North America. Bull Eur Assoc Fish
Pathol 7:29–31
Kongtorp RT, Kjerstad A, Taksdal T, Guttvik A, Falk K (2004)
Heart and skeletal muscle inflammation in Atlantic
salmon Salmo salar L.: a new infectious disease. J Fish Dis
(in press)
Lannan CN, Winton JR, Fryer JL (1984) Fish cell lines: establishment
and characterization of nine cell lines from
salmonids. In Vitro 20:671–676
McLoughlin MF, Nelson RT, McCormick JI, Rowley H (1995)
Pathology of experimental pancreas disease in freshwater
Atlantic salmon parr. J Aquat Anim Health 7:104–110
McLoughlin MF, Nelson RT, Rowley HM, Cox DI, Grant AN
(1996) Experimental pancreas disease in Atlantic salmon
Salmo salar post-smolts induced by salmon pancreas
disease virus (SPDV). Dis Aquat Org 26:117–124
Editorial responsibility: David Bruno,
Aberdeen, UK
Dis Aquat Org 59: 217–224, 2004
McLoughlin MF, Nelson RT, Rowley HM, Cox DI, Grant AN
(1997) The development of an experimental model of pancreas
disease in Atlantic salmon (Salmo salar) post-smolts.
Fish Vaccinol 90:467–467
McLoughlin MF, Nelson RN, McCormick JI, Rowley HM,
Bryson DB (2002) Clinical and histopathological features
of naturally occurring pancreas disease in farmed Atlantic
salmon, Salmo salar L. J Fish Dis 25:33–43
Munro ALS, Ellis AE, McVicar AH, McLay HA, Needham EA
(1984) An exocrine pancreas disease of farmed Atlantic
salmon in Scotland. Helgol Meeresunters 37:571–586
Murphy TM, Rodger HD, Drinan EM, Gannon F, Kruse P,
Korting W (1992) The sequential pathology of pancreas
disease in Atlantic salmon farms in Ireland. J Fish Dis
15:401–408
Murphy TM, Drinan EM, Gannon F (1995) Studies with an
experimental model for pancreas disease of Atlantic
salmon Salmo salar L. Aquac Res 26:861–874
Nelson RT, McLoughlin MF, Rowley HM, Platten MA,
McCormick JI (1995) Isolation of a toga-like virus from
farmed Atlantic salmon Salmo salar with pancreas disease.
Dis Aquat Org 22:25–32
Poppe T, Rimstad E, Hyllseth B (1989) Pancreas disease in
Atlantic salmon (Salmo salar) postsmolts infected with
infectious pancreatic necrosis virus (IPNV). Bull Eur Assoc
Fish Pathol 9:83–85
Poppe TT, Breck O (1997) Pathology of Atlantic salmon Salmo
salar intraperitoneally immunized with oil-adjuvanted
vaccine. A case report. Dis Aquat Org 29:219–226
Raynard RS, Houghton G, Munro ALS (1992) Pancreas
disease of Atlantic salmon: proceedings of a European
Commission Workshop. Scottish Office Agriculture
Report, No. 1. Scottish Office Agriculture and Fisheries
Department, Aberdeen, p 2–4
Rodger HD, Turnbull T (2000) Cardiomyopathy syndrome in
farmed Scottish salmon. Vet Rec 146:500–501
Rodger HD, Murphy TM, Drinan EM, Rice DA (1991) Acute
skeletal myopathy in farmed Atlantic salmon Salmo salar.
Dis Aquat Org 12:17–23
Rodger HD, Turnbull T, Richards RH (1994) Myopathy and
pancreas disease in salmon—a retrospective study in
Scotland. Vet Rec 135:234–235
Villoing S, Bearzotti M, Chilmonczyk S, Castric J, Bremont M
(2000a) Rainbow trout sleeping disease virus is an atypical
alphavirus. J Virol 74:173–183
Villoing S, Castric J, Jeffroy J, Le Ven A, Thiery R, Bremont M
(2000b) An RT-PCR-based method for the diagnosis of the
sleeping disease virus in experimentally and naturally
infected salmonids. Dis Aquat Org 40:19–27
Weston JH, Welsh MD, McLoughlin MF, Todd D (1999)
Salmon pancreas disease virus, an alphavirus infecting
farmed Atlantic salmon, Salmo salar L. Virology 256:
188–195
Wolf K, Quimby MC (1966) Lymphocystis virus: isolation and
propagation in centrarchid fish cell lines. Science 151:
1004–1005
Submitted: September 29, 2003; Accepted: March 1, 2004
Proofs received from author(s): May 3, 2004

Paper II
Heart and skeletal muscle inflammation in Atlantic salmon, Salmo
salar L.: a new infectious disease

Ó 2004
Blackwell Publishing Ltd
Journal of Fish Diseases 2004, 27, 351–358
Heart and skeletal muscle inflammation in Atlantic salmon,
Salmo salar L.: a new infectious disease
R T Kongtorp 1 , A Kjerstad 2 , T Taksdal 1 , A Guttvik 3 and K Falk 1
1 National Veterinary Institute, Oslo, Norway
2 Havbrukstjenesten A/S (Fish Health Service Ltd), Sistranda, Norway
3 VESO Vikan, Namsos, Norway
Abstract
Heart and skeletal muscle inflammation (HSMI) is
a disease syndrome of unknown aetiology first
observed in farmed Atlantic salmon, Salmo salar, in
1999. In the present study we have demonstrated
for the first time that HSMI is an infectious disease.
It was induced in Atlantic salmon post-smolts after
injection with tissue homogenate from farmed
Atlantic salmon previously diagnosed with HSMI.
The lesions were also induced in cohabitating salmon
given a corresponding injection without tissue
homogenate. Six weeks post-challenge the fish that
had been injected with tissue homogenate developed
a serious epicarditis and myocarditis with
mononuclear cell infiltrations in compact and
spongy layers of the heart. Similar lesions were
found in cohabitants after 10 weeks. The lesions
were consistent with samples from field outbreaks
of HSMI. No lesions were found in control fish. A
viral aetiology is strongly suggested, as no difference
in disease induction between an inoculum containing
antibiotics and a non-treated inoculum was
found. Further investigations are required in order
to make conclusions regarding the cause and pathogenesis
of HSMI.
Keywords: Atlantic salmon, heart and skeletal muscle
inflammation, pathology, transmission.
Correspondence R T Kongtorp, Section for Fish Health,
National Veterinary Institute, PO Box 8156 Dep, 0033 Oslo,
Norway
(e-mail: ruth-torill.kongtorp@vetinst.no)
351
Introduction
During 1999, outbreaks of a previously undescribed
disease in sea caged Atlantic salmon, Salmo salar L.,
were reported along the Norwegian West coast. The
disease was given the name Ôheart and skeletal
muscle inflammation (HSMI)Õ because of the
characteristic histopathological lesions that occurred
in affected fish. The number of reported outbreaks
is increasing [as recorded by the National Veterinary
Institute in Norway (NVI), 2003]. In 2002, fish
from 41 salmon farms along the Norwegian coast
were diagnosed with HSMI at the NVI laboratories.
Disease occurrence has been reported the whole
year round, but seems to be most common during
spring and early summer, 5–9 months after sea
transfer. Morbidity is high, as most fish in affected
sea cages show histopathological lesions in heart and
skeletal muscle. Mortality varies from almost insignificant
up to 20%. Management stress seems to
increase the mortality rate and prolong the healing
process during outbreaks.
Macroscopic changes include a pale heart with
loose texture, pericardial haemorrhage, ascites and a
pale or stained liver. Haematocrit levels are usually
normal.
The most significant histopathological lesions of
HSMI are found in heart and red skeletal muscle.
In the heart, necrosis of muscle fibres and a massive
inflammatory response involves both compact and
spongy layers of the ventricle. Infiltrations are
mainly mononuclear. An extensive epicarditis is
usually found in association with the myocarditis.
Red muscle inflammation follows the same pattern
as seen in the heart, but is not a consistent finding.
Other lesions include focal liver necrosis and

Ó 2004
Blackwell Publishing Ltd
Journal of Fish Diseases 2004, 27, 351–358 R T Kongtorp et al. Transmission of heart and skeletal muscle inflammation
circulatory disturbances such as oedema and erythrocyte
accumulation in several organs (Kongtorp,
Taksdal & Lyngøy 2004).
Investigations regarding the cause of HSMI have
previously only been speculative. Field observations
indicate that the disease has a contagious nature
(Kongtorp, Taksdal & Lyngøy 2004).The pathological
changes resemble those found in pancreas
disease (PD) (Ferguson, Roberts, Richards, Collins
& Rice 1986; Rodger, Murphy, Drinan & Rice
1991; Murphy, Rodger, Drinan, Gannon, Kruse &
Korting 1992; Rodger, Turnbull & Richards 1994;
McLoughlin, Nelson, McCormick, Rowley &
Bryson 2002) and cardiomyopathy syndrome
(CMS) (Ferguson, Poppe & Speare 1990), but do
not share all features of the two diseases (Kongtorp
et al. 2004). Results from a pilot study conducted
in 2001 (A. Kjerstad, A. Guttvik, H. Skjelstad,
T. Taksdal & K. Falk, unpublished results),
indicate that HSMI is infectious.
This paper describes a challenge study conducted
in 2002. In an experimental challenge, we demonstrated
that HSMI is transmissible both by injection
and cohabitation.
Materials and methods
Preparation of tissue homogenate
Organ homogenates were made from Atlantic
salmon displaying lesions consistent with HSMI
(Fig. 1A, E) from a field outbreak. Atrial and
ventricular myocardium from seven fish (50% wv)
were homogenized in Leibovitz L-15 cell culture
medium (L-15), and centrifuged at 2500 g for
7 min. The resulting supernatant was further
diluted in L-15 media with and without gentamycin
(final concentration 50 lg mL )1 ). L-15 media
with and without gentamycin in the same concentration
was used as control inocula.
Transmission trial
Experimental fish originated from a fish farm from
which no signs of disease had been reported. They
had hatched in January 2001 and were transported
to the research station VESO Vikan (Namsos,
Norway) in mid February 2002. The fish were
transferred to sea water in late April 2002. On
commencement of the challenge study in late June
2002, they had an average weight of 90 g. A total of
500 salmon were placed in three tanks, measuring
352
1 · 1 · 0.5 m. Water flow through the tanks was
0.8 L kg fish )1 min )1 . The temperature ranged
from 10 to 12 °C during the study and the salinity
in the tanks was 33& on average. The inoculate
(0.2 mL) was injected intraperitoneally.
In tank 1, experimental fish were injected with
inocula supplemented with gentamycin. One hundred
fish were injected with tissue homogenate. An
additional 100 fish were injected with a control
inoculum and placed in the tank as cohabitants.
Similarly, 200 fish were injected with antibiotic-free
inocula in tank 2. One hundred fish were challenged
with tissue homogenate and 100 fish with
L-15 media. A control group of 100 fish was given a
corresponding injection with L-15 media and kept
in a separate tank. After inoculation of the fish, data
relating to mortality, feeding, water temperature
and salinity were recorded daily.
Blood and tissue sampling
Samples were taken 1, 2, 4, 6, 8 and 10 weeks postchallenge.
At each sampling, five fish from each of
the five groups were killed by a blow to the head
and bleeding. Blood samples were collected from
the caudal vein for haematocrit measurements.
Tissue samples from heart, skeletal muscle, mid
kidney and pyloric caeca with pancreas were collected
and fixed in 10% neutral phosphate buffered formalin.
Fixed samples were prepared for histological
examination by standard paraffin wax techniques
and stained with haematoxylin and eosin. Sampled
fish were classified histologically as diseased or nondiseased
based on the presence of myocarditis with
mononuclear cell infiltration and myocardial necrosis
consistent with HSMI (Kongtorp et al. 2004).
The presence of an associated serious epicarditis was
used to confirm the diagnosis.
One, 2 and 4 weeks after challenge, samples
from mid kidney were cultivated on blood agar
plates with 2% NaCl at 10 °C for 4–6 days. In the
event of mortalities during the study, the same
procedure was performed on the dead fish.
Tissue from mid kidney was used to inoculate
cell cultures. Samples from 10 fish that had been
challenged by inoculation and sampled 4 and
8 weeks post-infection (p.i.) were diluted in L-15
media, homogenized and centrifuged at 1000 g for
10 min. As infectious pancreatic necrosis virus
(IPNV) is ubiquitous in Norwegian salmon farms
(Melby, Krogsrud, Ha˚stein & Stenwig 1991),
rabbit antisera against the WB, Ja and N1 serotypes

Ó 2004
Blackwell Publishing Ltd
Journal of Fish Diseases 2004, 27, 351–358 R T Kongtorp et al. Transmission of heart and skeletal muscle inflammation
Figure 1 Salmo salar. Micrographs of sections from experimental fish (haematoxylin and eosin). (A–D) Heart. (A) Compact
myocardium from fish used as inoculate material. There is a severe diffuse myocarditis and epicarditis. (B) Compact myocardium and
epicardium of control fish 10 weeks after study commencement. (C) Compact myocardium of the ventricle in cohabitant fish 10 weeks
after study commencement. Myocardial fibres are degenerative and there is a massive inflammation in the surrounding tissue. (D)
Ventricle of a cohabitant fish 10 weeks after challenge. There is a severe epicarditis and a diffuse hypercellularity in the underlying
compact myocardium. (E–H) Red skeletal muscle. (E) Red skeletal muscle from fish used as inoculate material. There is extensive
degeneration and cellular infiltration. (F) Red skeletal muscle in control fish 10 weeks after challenge. (G) Red skeletal muscle in fish
injected with tissue homogenate from diseased fish 12 weeks earlier. Focal inflammation and myocyte degeneration. (H) Red skeletal
muscle in cohabitant fish 10 weeks after challenge. A degenerative myocyte shows evidence of nuclear lysis. There is an infiltration of
mononuclear cells in the myocyte fibre and surrounding interstitium. a, epicardium; b, compact myocardium; c, myocyte necrosis.
353

Ó 2004
Blackwell Publishing Ltd
Journal of Fish Diseases 2004, 27, 351–358 R T Kongtorp et al. Transmission of heart and skeletal muscle inflammation
of IPNV [mAb N1-H8 (Christie, Ness & Djupvik
1990), kindly provided by K.E. Christie, Intervet
Norbio AS] were added to the tissue suspension in
appropriate dilutions to inhibit replication of this
virus in the cell cultures. The inoculate was used to
infect the following cell cultures: rainbow trout gill
(RT-Gill), chinook salmon embryo (CHSE-214),
Atlantic salmon head kidney (ASK-2) (Devold,
Krossøy, Aspehaug & Nylund 2000), epithelioma
papulosum cyprini and fat head minnow. The cells
had been grown at 20 °C in L-15 supplemented
with 10% foetal calf serum, l-glutamine (4 mm)
and gentamycin (50 lg mL )1 ). After inoculation
with tissue homogenates, cells were incubated at
15 °C. All cultures were blind passaged twice at
14 day intervals.
Results
The challenged fish collected sequentially by random
sampling developed lesions consistent with
HSMI (Table 1). The lesions appeared 6 weeks p.i.
in fish injected with diseased material and 10 weeks
p.i. in cohabitant groups. A severe epicarditis could
be seen in all fish that had developed myocardial
lesions. Mild to moderate epicarditis was present in
fish showing signs of early myocardial damage. Fish
with no myocardial lesions did not have epicarditis.
No lesions were detected in fish from the control
tank (Fig. 1B, F).
In fish injected with tissue homogenate, the heart
lesions detected at 6 weeks p.i. were mild to
moderate. Only two fish had lesions that could
clearly be distinguished as HSMI, but all inoculated
fish in the sample showed mild to moderate
eosinophilia and nuclear pyknosis of myocytes in
the compact and spongy layers of the ventricle,
indicating early myocardial degeneration. These
myocardial changes were not found in the control
fish. At 8 weeks p.i. the heart lesions in most
homogenate-injected fish were extensive and
inflammatory in character. The severity of the
lesions had increased at 10 weeks p.i. At 12 weeks
p.i. the myocardial lesions were evident, but more
moderate in character. The cohabitant group
showed the same histopathological pattern with a
4-week delay, and the first fish to be diagnosed with
HSMI were sampled 10 weeks p.i. (Fig. 1C, D).
The preceding degenerative signs in myocardial
cells were present at 8 weeks p.i., similar to the
findings described for the inoculated group.
Mild to moderate inflammation of red skeletal
muscle was found in 10 fish inoculated with tissue
homogenate and 15 cohabitants sampled 10 and
12 weeks p.i. (Fig. 1G, H). There were only a small
number of cells with degenerative changes in these
samples. No pancreatic or renal lesions were found
in any of the sampled fish.
In tank 1, mortalities occurred only during the
first days of the challenge study (Fig. 2A). These
fish had all been injected with an inoculum
containing antibiotics. In tank 2, where the inoculum
had not been treated with antibiotics, the
mortality was more evenly distributed during the
course of the study (Fig. 2B). The largest number
of mortalities in tank 2 was in the cohabitant group,
1–2 months after study commencement. In the
control tank, one fish died at the start of the study.
There was no growth on blood agar from any of
the sequential bacteriology samples.
Investigations of the fish that had died during
the study showed mainly a mixed flora of bacteria,
including Moritella viscosa (Table 2). The
cell culture trial gave no conclusive results. The
Table 1 Challenge study of heart and skeletal muscle inflammation (HSMI). Number of salmon showing histopathologic changes in the
heart consistent with HSMI at each sample date. The sample size was five fish in each group at each sampling
Weeks after
challenge
Tank 1, inoculum treated with
gentamycin at challenge
Injected with
tissue homogenate
Cohabitants
(injected with L-15)
Tank 2, inoculum not treated
at challenge
Injected with
tissue homogenate
Cohabitants
(injected with L-15)
1 week 0 0 0 0 0
2 weeks 0 0 0 0 0
4 weeks 0 0 0 0 0
6 weeks 1 0 1 0 0
8 weeks 4 0 3 0 0
10 weeks 4 5 5 2 0
12 weeks 5 4 5 4 0
In total 14 9 14 6 0
354
Tank 3, inoculum not treated
at challenge
Control group
(injected with L-15)

Ó 2004
Blackwell Publishing Ltd
Journal of Fish Diseases 2004, 27, 351–358 R T Kongtorp et al. Transmission of heart and skeletal muscle inflammation
Percentage dead
30
20
10
0
Table 2 Challenge study of heart and skeletal muscle inflammation. Bacterial growth on blood agar from fish that died during the study
Isolated bacteria
haematocrit values of the experimental fish did not
differ significantly from the control fish.
Discussion
Cumulative mortality in tank 1
1 11 21 31 41 51 61 71
Days after inoculation
Tank 1, inoculum treated with
gentamycin at challenge
Injected with
tissue homogenate Cohabitants
In this study we have demonstrated the transmissibility
of a newly discovered disease in farmed
Atlantic salmon recognized by extensive inflammation
in the heart and skeletal muscle. We have
shown that HSMI can be induced in naïve Atlantic
salmon by intraperitoneal injection of homogenized
material from diseased fish and by cohabitation.
The verification of transmission was based on
histopathological findings of heart lesions that were
consistent with field cases. The control group had
been given similar treatment and environmental
conditions as the experimental groups, but did not
show any signs of disease during the study.
From this we conclude that HSMI is an infectious
disease.
The first heart lesions appeared at 6 weeks p.i. in
the group that had been inoculated with tissue
homogenate from diseased fish. Corresponding
lesions were found in the cohabitants at 10 weeks
A
Tank 2, inoculum not treated
at challenge
Injected with
tissue homogenate Cohabitants
Moritella viscosa 0 1 1 1 0
Mixed flora including
Moritella viscosa
4 3 7 15 1
No bacterial growth 0 1 5 13 0
Percentage dead
30
20
10
0
Cumulative mortality in tank 2
1 11 21 31 41 51 61 71
Days after inoculation
Figure 2 Challenge study of heart and skeletal muscle inflammation. Cumulative mortality in challenged groups. In tank 1, the inocula
had been treated with gentamycin prior to injection into the experimental fish. In tank 2, the inocula were injected without gentamycin
additives. j Inoculated with tissue homogenate, cohabitants.
355
Tank 3, inoculum not
treated at challenge
Control
group
p.i., showing that HSMI can be transmitted via
water. Treatment of the inoculum with a broadspectrum
antibiotic prior to inoculation did not
affect the transmission of HSMI from sick to
healthy fish, as there was no significant difference in
the appearance of lesions between the two challenged
groups. Bacteriological investigations were
negative in all the sequentially sampled fish. This is
consistent with investigations from field outbreaks
of HSMI, from which no specific bacteria have been
isolated (personal observations). These results
strongly suggest a viral aetiology of HSMI. Inoculation
of cell cultures in the present study did not
reveal any causal agent, but were conducted on a
small scale. Further, the time delay in the disease
outbreak of the cohabitating fish compared with the
fish that had been injected with tissue homogenate
indicates a peak in possible virus shedding occurring
at 4 weeks p.i. From studies of infectious salmon
anaemia it has been observed that viraemia occurs
concurrently with a peak in virus shedding
(Dannevig, Falk & Skjerve 1994; Rimstad, Falk,
Mikalsen & Teig 1999; Devold et al. 2000;
Raynard, Snow & Bruno 2001). The conclusion
B

Ó 2004
Blackwell Publishing Ltd
Journal of Fish Diseases 2004, 27, 351–358 R T Kongtorp et al. Transmission of heart and skeletal muscle inflammation
that HSMI is transmissible under experimental
conditions without causing mortality is supported
by the results from the pilot study conducted in
2001 (A. Kjerstad, A. Guttvik, H. Skjelstad,
T. Taksdal & K. Falk, unpublished results). In this
study, the heart lesions were transmitted both by
injection of homogenized tissue from diseased fish
and by cohabitation.
In the present study, mortalities were observed
among the challenged fish. The cause of death was,
however, not likely to be HSMI, as the morbidity
because of HSMI was equal in the two challenged
groups. After 2 weeks p.i., no fish died in tank 1,
while the number of deaths in tank 2 was high
throughout the study (Fig. 2). Overall, the number
of deaths was considerably higher in the group of
fish that had been injected with an inoculum
without antibiotics, compared with the antibiotics
group. Moritella viscosa, a Gram-negative bacterium
that has been associated with winter ulcer in
Atlantic salmon (Benediktsdóttir, Verdonck,
Spröer, Helgason & Swings 2000; Lunder, Sørum,
Holstad, Steigerwalt, Mowinckel & Brenner 2000),
was isolated from several of the dead fish, but not
from any of the sequentially sampled and killed fish
in the study. Winter ulcer most often occurs during
the winter months, when the sea temperature is low
(Salte, Rørvik, Reed & Norberg 1994; Bruno &
Poppe 1996). In a challenge study conducted on
winter ulcer in Norway, M. viscosa was shown to
infect healthy fish at a water temperature of 10 °C
(Lunder, Evensen, Holstad & Ha˚stein 1995).
Under experimental conditions, M. viscosa is most
readily isolated from fish that have been kept at low
water temperatures (A. Guttvik, unpublished
results). The present study of HSMI was conducted
at water temperatures ranging from 10 to 12 °C.
Taking these findings together, there is likely to be
an association between the mortality and the
isolation of M. viscosa.
In the present study, the HSMI diagnosis was
based on the findings of heart lesions, because of the
non-consistency of muscular lesions in field outbreaks.
The lesions were milder than those reported
from field cases, but were consistent with naturally
occurring HSMI, presenting focal or diffuse cellular
infiltration in all layers, accompanied by myocyte
necrosis. In the skeletal muscle, there were only mild
to moderate lesions, never reaching the severity that
can be seen in field cases. Similar experiences have
been made from challenge studies of PD (McLoughlin,
Nelson, McCormick & Rowley 1995; Desvig-
356
nes, Quentel, Lamour & Le Ven 2002). This may
indicate that lesions in the skeletal muscle appear at a
later stage in the pathogenesis, or that agent
concentration in a field outbreak could be higher
than that of this experiment, causing a potentially
more aggressive disease. It also underlines the fact
that a HSMI diagnosis must not be based solely upon
lesions in the skeletal muscle.
The morbidity of HSMI seems to be high. In
field outbreaks, it has been observed that most fish
in an affected sea cage have lesions in the heart and
skeletal muscle. Field observations have also shown
that surviving fish in affected sea cages will recover
(T. Taksdal, unpublished data). Histopathological
lesions have been found in fish for a long time after
symptomatic recovery. Examination of the samples
from the challenge study supports the field observations
of high morbidity. During the last weeks of
the study, diseased fish were frequently detected in
the samples by random sampling of fish from the
experimental tanks.
Pancreas disease and CMS are the most relevant
differential diagnoses to HSMI at the histopathological
level. Cardiac lesions associated with PD
(Rodger et al. 1991; Murphy et al. 1992;
McLoughlin, Nelson, Rowley, Cox & Grant
1996; McLoughlin et al. 2002) show great similarity
to the lesions observed in HSMI outbreaks.
Pancreatic lesions are, however, important hallmarks
of PD, as indicated by the name (Munro,
Ellis, McVicar, McLay & Needham 1984). Also,
experimental trials on PD have consistently resulted
in pancreatic lesions (Boucher, Raynard, Houghton
& Baudin Laurencin 1995; Murphy, Drinan &
Gannon 1995; McLoughlin et al. 1996; Desvignes
et al. 2002) As pancreas was normal in all samples
in the present study of HSMI, the hypothesis of a
possible connection between the two diseases is not
supported. This result is consistent with investigations
from field outbreaks (Kongtorp et al. 2004).
To further distinguish between HSMI and PD,
RNA was extracted from frozen kidney samples of
15 fish that had been sampled 4 and 8 weeks after
challenge and tested by reverse transcriptase
polymerase chain reaction for salmon pancreas
disease virus (SPDV) according to standard procedures
at the NVI. SPDV was not detected in any of
the samples.
When HSMI was first recognized as a unique
disease in 1999, only a few locations in a limited
area of the Norwegian coast seemed to be affected.
Since then, HSMI has been diagnosed at an

Ó 2004
Blackwell Publishing Ltd
Journal of Fish Diseases 2004, 27, 351–358 R T Kongtorp et al. Transmission of heart and skeletal muscle inflammation
increasing number of locations and in a large
geographical area, indicating an infectious nature.
In 2002, most parts of the coastline had outbreaks
of disease that were verified as HSMI. The disease
seems to be having an increasing impact on the
health status of Norwegian salmon farms and will
consequently be of economic significance for the
industry. Some reports indicate that HSMI has a
tendency to reappear annually at the same locations
(A. Kjerstad, unpublished data). It is not known,
however, whether this is because of a carrier status
in some fish, persistence in the farm cages, or if it
represents a reintroduction of the infection. The
present study confirms the infectious nature indicated
by the field observations. HSMI is therefore a
potential threat to salmon farms both within and
outside Norway, following water-born spread of the
causal agent. Further investigations are required in
order to make conclusions regarding the aetiology
and pathogenesis of HSMI.
Acknowledgements
This project was partly funded by the Norwegian
Research Council. The technical assistance provided
by H. Welde at the NVI was of great value in this
study. We would also like to thank B. Dannevig
and M.J. Hjortaas at the NVI Section for Virology
and Serology for help with the PCR.
References
Benediktsdóttir E., Verdonck L., Spröer C., Helgason S. &
Swings J. (2000) Characterization of Vibrio viscosus and Vibrio
wodanis isolated at different geographical locations: a proposal
for reclassification of Vibrio viscosus as Moritella viscosa comb.
nov. International Journal of Systematic and Evolutionary
Microbiology 50, 479–488.
Boucher P., Raynard R.S., Houghton G. & Baudin Laurencin F.
(1995) Comparative experimental transmission of pancreas
disease in Atlantic salmon, rainbow trout and brown trout.
Diseases of Aquatic Organisms 22, 19–24.
Bruno D.W. & Poppe T.T. (1996) A Colour Atlas of Salmonid
Diseases, pp. 144–145. Academic Press, London.
Christie K.E., Ness S. & Djupvik H.O. (1990) Infectious
pancreatic necrosis virus in Norway: partial serotyping by
monoclonal antibodies. Journal of Fish Diseases 13, 323–327.
Dannevig B.H., Falk K. & Skjerve E. (1994) Infectivity of
internal tissues of Atlantic salmon, Salmo salar L., experimentally
infected with the aetiological agent of infectious
salmon anaemia (ISA). Journal of Fish Diseases 17, 613–622.
Desvignes L., Quentel C., Lamour F. & Le Ven A. (2002)
Pathogenesis and immune response in Atlantic salmon (Salmo
salar L.) parr experimentally infected with salmon pancreas
357
disease virus (SPDV). Fish and Shellfish Immunology 12,
77–95.
Devold M., Krossøy B., Aspehaug V. & Nylund A. (2000) Use
of RT-PCR for diagnosis of infectious salmon anaemia virus
(ISAV) in carrier sea trout Salmo trutta after experimental
infection. Diseases of Aquatic Organisms 40, 9–18.
Ferguson H.W., Poppe T. & Speare D.J. (1990) Cardiomyopathy
in farmed Norwegian salmon. Diseases of Aquatic
Organisms 8, 225–231.
Ferguson H.W., Roberts R.J., Richards R.H., Collins R.O. &
Rice D.A. (1986) Severe degenerative cardiomyopathy associated
with pancreas disease in Atlantic salmon, Salmo salar L.
Journal of Fish Diseases 9, 95–98.
Kongtorp R.T., Taksdal T. & Lyngøy A. (2004) Pathology of
heart and skeletal muscle inflammation (HSMI) in farmed
Atlantic salmon Salmo salar. Diseases of Aquatic Organisms (in
press).
Lunder T., Evensen Ø., Holstad G. & Ha˚stein T. (1995)
ÔÔWinter ulcerÕÕ in Atlantic salmon (Salmo salar). Pathologic
changes, bacteriological investigations, and transmission
experiments. Diseases of Aquatic Organisms 23, 39–49.
Lunder T., Sørum H., Holstad G., Steigerwalt A.G., Mowinckel
P. & Brenner D.J. (2000) Phenotypic and genotypic characterization
of Vibrio viscosus sp. nov. and Vobrio wodanis sp.
nov. isolated from Atlantic salmon (Salmo salar) with ÔÔwinter
ulcerÕÕ. International Journal of Systematic and Evolutionary
Microbiology 50, 427–450.
McLoughlin M., Nelson R.T., McCormick J.I. & Rowley H.
(1995) Pathology of experimental pancreas disease in freshwater
Atlantic salmon parr. Journal of Aquatic Animal Health
7, 104–110.
McLoughlin M., Nelson R.T., Rowley H.M., Cox D.I. &
Grant A.N. (1996) Experimental pancreas disease in Atlantic
salmon Salmo salar post-smolts induced by salmon pancreas
disease virus (SPDV). Diseases of Aquatic Organisms 26,
117–124.
McLoughlin M.F., Nelson R.N., McCormick J.I., Rowley H.M.
& Bryson D.B. (2002) Clinical and histopathological features
of naturally occurring pancreas disease in farmed Atlantic
salmon, Salmo salar L. Journal of Fish Diseases 25, 33–43.
Melby H.P., Krogsrud J., Ha˚stein T. & Stenwig H. (1991) All
commercial Atlantic salmon seawater farms in Norway
harbour carriers of infectious pancreatic necrosis virus (IPNV).
In: Proceedings from the 2nd International Symposium of Viruses
of Lower Vertebrates (ed. by J.L. Fyrer), pp. 211–217. Oregon
State University, Corvallis, OR, USA.
Munro A.L.S., Ellis A.E., McVicar A.H., McLay H.A. &
Needham E.A. (1984) An exocrine pancreas disease of farmed
Atlantic salmon in Scotland. Helgolander Meeresuntersuchungen
37, 571–586.
Murphy T.M., Rodger H.D., Drinan E.M., Gannon F., Kruse P.
& Korting W. (1992) The sequential pathology of pancreas
disease in Atlantic salmon farms in Ireland. Journal of Fish
Diseases 15, 401–408.
Murphy T.M., Drinan E.M. & Gannon F. (1995) Studies with
an experimental model for pancreas disease of Atlantic salmon
Salmo salar L. Aquaculture Research 26, 861–874.

Ó 2004
Blackwell Publishing Ltd
Journal of Fish Diseases 2004, 27, 351–358 R T Kongtorp et al. Transmission of heart and skeletal muscle inflammation
Raynard R.S., Snow M. & Bruno D.W. (2001) Experimental
infection models and susceptibility of Atlantic salmon Salmo
salar to a Scottish isolate of infectious salmon anaemia virus.
Diseases of Aquatic Organisms 47, 169–174.
Rimstad E., Falk K., Mikalsen A.B. & Teig A. (1999) Time
course distribution of infectious salmon anaemia virus in
experimentally infected Atlantic salmon Salmo salar. Diseases of
Aquatic Organisms 36, 107–112.
Rodger H.D., Murphy T.M., Drinan E.M. & Rice D.A. (1991)
Acute skeletal myopathy in farmed Atlantic salmon Salmo
salar. Diseases of Aquatic Organisms 12, 17–23.
358
Rodger H.D., Turnbull T. & Richards R.H. (1994) Myopathy
and pancreas disease in salmon – a retrospective study in
Scotland. Veterinary Record 135, 234–235.
Salte R., Rørvik K.-A., Reed E. & Norberg K. (1994) Winter
ulcers of the skin in Atlantic salmon, Salmo salar L.: pathogenesis
and possible aetiology. Journal of Fish Diseases 17,
661–665.
Received: 10 November 2003
Revision received: 22 March 2004
Accepted: 25 March 2004

Paper III
Longitudinal study of a natural outbreak of heart and skeletal
muscle inflammation in Atlantic salmon, Salmo salar L.

Ó 2006
Blackwell Publishing Ltd
Journal of Fish Diseases 2006, 29, 233–244
Longitudinal study of a natural outbreak of heart
and skeletal muscle inflammation in Atlantic salmon,
Salmo salar L.
R T Kongtorp 1 , M Halse 2 , T Taksdal 1 and K Falk 1
1 National Veterinary Institute, Oslo, Norway
2 Havbrukstjenesten A/S (Fish Health Service Ltd), Sistranda, Norway
Abstract
Heart and skeletal muscle inflammation (HSMI) is a
transmissible disease of farmed Atlantic salmon,
Salmo salar L. It is characterized by significant epi-,
endo- and myocarditis, as well as myositis, particularly
involving red skeletal muscle. The aetiology of
HSMI is currently unresolved, though a viral cause is
suspected. Since its discovery in 1999, HSMI has
become an increasing problem for the Norwegian
farming industry, with some farms experiencing
yearly outbreaks and subsequent economic losses. In
the present study an Atlantic salmon farm was studied
from December 2003 to April 2005. Samples
from apparently healthy as well as clinically diseased
fish were collected monthly and examined histopathologically.
The first fish to be diagnosed with
HSMI was sampled in May, 8 months after transfer
to sea. A clinical outbreak of HSMI followed in June,
when all fish in the sample had lesions consistent with
HSMI. Subsequent samples revealed that cardiac
lesions decreased in severity 2 months after the start
of the outbreak, but that multiple foci of cellular
infiltration and necrosis persisted throughout the
year. There appeared to be a shift in lesion location
from being most severe in the compact myocardium
in early stages of disease to a greater involvement of
the atrium and spongy layer of the ventricle in later
samples. Late samples also showed increased fibrosis
of cardiac tissue. In conclusion, HSMI appears to be a
severe disease with elevated mortality, morbidity
close to 100% and prolonged duration.
Correspondence R T Kongtorp, Section for Fish Health,
National Veterinary Institute, PO Box 8156 Dep, N-0033 Oslo,
Norway
(e-mail: ruth-torill.kongtorp@vetinst.no)
233
Keywords: Atlantic salmon, epidemiology, heart
and skeletal muscle inflammation, histopathology,
longitudinal study, pathology.
Introduction
Heart and skeletal muscle inflammation (HSMI) is
a disease syndrome of uncertain aetiology affecting
farmed Atlantic salmon, Salmo salar L., during the
seawater stage (Kongtorp, Taksdal & Lyngøy
2004b). Farms experiencing HSMI report variable
mortality, ranging from almost negligible to 20% in
affected cages. Many fish show clinical signs
including anorexia and abnormal swimming behaviour.
These individuals are lethargic and usually
positioned near the cage wall and facing the sea
current. On autopsy most fish have a pale heart,
yellowish liver, ascites, swollen spleen and petechiae
in the perivisceral fat. HSMI is diagnosed histologically
based on findings of significant epi-, endoand
myocarditis, as well as myositis, particularly
involving red skeletal muscle. The disease is
experimentally transmissible to healthy Atlantic
salmon by intraperitoneal injection of cardiac tissue
from diseased fish and by cohabitation with injected
fish (Kongtorp, Kjerstad, Guttvik, Taksdal & Falk
2004a). It is believed that HSMI is of viral origin,
but the isolation of the causal agent has not yet been
published.
Since its discovery in 1999, HSMI has become a
major challenge for the Norwegian salmon farming
industry. Records from the National Veterinary
Institute (NVI) show a yearly increase in the
number of diagnosed cases, indicating that the
disease is spreading along the coastline. In 2003
HSMI was considered to be the most important

Ó 2006
Blackwell Publishing Ltd
Journal of Fish Diseases 2006, 29, 233–244 R T Kongtorp et al. Longitudinal study of HSMI in salmon
disease problem in farmed Atlantic salmon in mid-
Norway (A. Lyngøy, personal communication).
Some farms have reported yearly outbreaks with
subsequent economic losses. A case of suspected
HSMI has also been reported from Scotland
(Ferguson, Kongtorp, Taksdal, Graham & Falk
2005), indicating that the disease may become a
widespread problem.
The present study was conducted in order to
further investigate the disease pattern and epidemiology
of naturally occurring HSMI. This was
performed by a longitudinal study through the
seawater production cycle of a farm experiencing
recurrent outbreaks of HSMI.
Materials and methods
Fish and management
An Atlantic salmon sea farm with a history of
recurrent outbreaks of HSMI was studied for
16 months, from December 2003 to April 2005.
Fish originated from two smolt producers, and were
transferred to the farm in late September to early
October 2003, weighing 71–80 g (Fig. 1). The
number of fish transferred was 1.13 million, all of
which were vaccinated in a standard vaccination
regime.
The studied farm consisted of a steel rig
construction with automatic feeding. From October
to the end of April fish were reared in an 18-cage
Farm 1
Location 2
Smolt
producer 1
Farm 1
Location 1
Farm 1
Location 2
Farm 1
Location 3
Smolt
producer 2
Farm 2
Figure 1 Longitudinal study of an outbreak of heart and skeletal
muscle inflammation. Flow chart of fish movement during the
period of study. Grey boxes indicate studied farm and locations.
234
rig, keeping the two smolt groups apart. On this
first location, the average sea current was
2.2 cm s )1 . In May all fish were moved to an
8-cage rig at another site, where the separation of
the original smolt groups was maintained. The
average sea current on the second location was
5.0 cm s )1 . In September the fish were sorted and
split into two groups, of which one was moved to a
third location, with a sea current of 3.3 cm s )1 on
average. All fish at location 2 were harvested in
April and May 2005. Seawater temperatures were
recorded daily throughout the production cycle,
and the average monthly temperatures during the
study period are shown in Fig. 2.
Fish from two cages, one from each smolt group
(hereafter denoted cages 2 and 5 because of their
position at location 2) were chosen for the study. As
the cages each contained smolts from only one of
the original groups, a comparison related to origin
was possible. After the fish were moved, the study
continued in the cages holding the same fish that
had been studied previously. Sampled fish were
anaesthetized and killed by a blow to the head. Fish
showing abnormal swimming behaviour as well as
randomly sampled healthy individuals were included.
Blood samples were collected from the caudal
vein. Macroscopic pathological changes were recorded
before tissues were sampled for histopathology.
Descriptive epidemiology
Regular health monitoring was performed by the
local fish health service in collaboration with the
farmer throughout the period of study. Mortalities
were collected daily from a landing net at the
bottom of each cage. Records of mortality, growth
and feeding were routinely fed into a computerized
database on the farm. Samples from suspected
disease outbreaks were sent to NVI for further
diagnostic examination. An investigation of biosecurity
routines was also performed by interviewing
the farmer. Contact fish were traced through
computerized records of fish movement from
hatching to slaughter.
Histology
Tissue samples were obtained from gills, pseudobranchs,
heart, skeletal muscle, liver, mid-kidney,
spleen and pyloric caeca with pancreatic tissue.
They were fixed in 10% neutral phosphate-buffered
formalin, prepared for histology and stained with

Ó 2006
Blackwell Publishing Ltd
Journal of Fish Diseases 2006, 29, 233–244 R T Kongtorp et al. Longitudinal study of HSMI in salmon
Figure 2 Longitudinal study of an outbreak
of heart and skeletal muscle inflammation.
Temperature fluctuation on farm 1 during
the study period. Records were taken on
location 1 from December to the end of
April, and then the fish were moved to
location 2. Temperatures from May and
onward are from location 2.
16.0
14.0
12.0
10.0
haematoxylin and eosin, using standard techniques.
The sampling dates and number of fish collected are
detailed in Table 1.
Fish were diagnosed with HSMI based on the
presence of epi-, endo- and myocarditis with
mononuclear cell infiltration, myocardial degener-
Temperature (˚C)
8.0
6.0
4.0
2.0
0.0
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Month
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
ation and necrosis, consistent with the findings
described by Kongtorp et al. (2004b). Myositis and
myocyte necrosis in red skeletal muscle were also
given weight in the diagnosis. Cardiac tissue
showing few and scattered foci of lesions were
denoted mild. Multifocal or diffuse lesions in
Table 1 Longitudinal study of an outbreak of heart and skeletal muscle inflammation (HSMI). Summary of lesions found in fish on
each sampling date. Cardiac tissue showing few and scattered foci of lesions were scored mild. Multifocal or diffuse lesions in cardiac
tissue that were clearly separated by areas of healthy tissue were scored moderate. Lesions were scored severe when most tissue had diffuse
cellular infiltrates and necrosis, showing no or only small areas of healthy tissue. Number of fish with each severity score is given in
brackets. Results for normal fish and fish showing clinical signs are shown together, as there were no obvious differences in lesion severity
between these groups
Sampling date No. of fish sampled Lesions found by histology
Year Month Date Normal Symptoms Epicardium Compactum Spongiosum Atrium
Skeletal
muscle Liver
2003 Dec 17 5 5 + (10) – – – – –
2004 Jan 6 6 4 + (10) + (1) – – – –
2004 Feb 12 5 5 + (10) – + (3) – – + (1)
2004 Mar 10 8 2 + (5)
++ (5)
+ (3) + (3) – – + (1)
2004 Mar 31 6 4 ++ (10) + (3) + (3) – – –
++ (1) ++ (1)
2004 Apr 20 0 10 ++ (4)
+++ (2)
+ (2) ++ (2) ++ (2) – + (2)
2004 May 13 7 3 ++ (9) + (1) + (1) – ++ (1) + (2)
+++ (1) +++ (1) +++ (1)
2004 Jun 10 0 3 +++ (3) +++ (3) +++ (3) – +++ (3)
2004 Jun 16 6 4 +++ (10) +++ (10) +++ (10) – + (3)
++ (4)
+++ (1)
++ (4)
2004 Jul 6 63 49 + (12)
++ (54)
+++ (45)
+ (12)
++ (54)
+++ (45)
+ (12)
++ (54)
+++ (45)
+ (37)
++ (14)
+++ (19)
+ (45)
++ (26)
+++ (30)
2004 Aug 31 0 5 +++ (5) ++ (5) ++ (5) ++ (4) – + (2)
2004 Sep 27 5 5 ++ (10) ++ (4) ++ (9) ++ (5) – –
2004 Nov 5 6 0 ++ (6) + (6) ++ (6) – – –
2004 Dec 14 3 2 ++ (5) + (5) ++ (5) – – + (1)
2005 Feb 17 6 0 + (6) + (6) + (6) – – –
2005 Apr 25 5 0 – – – – – –
+, mild lesions; ++, moderate lesions; +++, severe lesions.
235
++ (49)

Ó 2006
Blackwell Publishing Ltd
Journal of Fish Diseases 2006, 29, 233–244 R T Kongtorp et al. Longitudinal study of HSMI in salmon
cardiac tissue that were clearly separated by areas of
healthy tissue were scored as moderate. Lesions were
categorized as severe when most tissue had diffuse
cellular infiltrates and necrosis, showing no or only
small areas of healthy tissue.
Comparative histology
Atlantic salmon with no record of HSMI were
examined by histology. Sections from these fish
were compared with the histological findings in the
longitudinal study. The fish had been sampled as
negative controls during challenge studies or as a
part of routine diagnostics of diseased fish from
several commercial farms in Norway. Fish examined
in the comparative study were of various
origins, in order to reflect some of the natural
variation among farmed Atlantic salmon in Norway.
The samples included 53 healthy Atlantic
salmon parr and smolts in fresh water, 37 healthy
smolts and growing fish in sea water, as well as 61
fish suffering from diseases other than HSMI. Of
these, 10 fish were suffering from salmon pancreas
disease (PD) and 15 fish were suffering from
cardiomyopathy syndrome (CMS).
Virology and serology
As a possible causal virus was expected to be present in
mid kidney during the viraemic stage of the disease,
samples were taken from this organ for virological
investigations. Five mid kidney tissue samples from
each of the April, May and June sampling points were
used to inoculate cell cultures from Chinook salmon
embryo (CHSE-214), Atlantic salmon head kidney
(ASK-2) (Devold, Krossøy, Aspehaug & Nylund
2000), epithelioma papulosum cyprini and fat head
minnow. Kidney samples were prepared, inoculated
and passaged according to the procedure described by
Kongtorp et al. (2004a). The samples originated
from both apparently healthy and moribund fish, and
kidneys were not pooled.
A total of 55 kidney samples were examined for
the presence of infectious salmon anaemia virus
(ISAV), Atlantic salmon paramyxovirus (ASPV)
and salmon pancreas disease virus (SPDV) by
standard reverse transcriptase-polymerase chain
reaction (RT-PCR) procedures at the NVI. Samples
were collected in April (five samples), May (10
samples), June (10 samples), July (25 samples) and
August (five samples). The samples included both
apparently healthy and moribund fish. RNA was
236
extracted from the samples using TRIZOL reagent
(Gibco BRL, Gaithersburg, ML, USA) according to
the manufacturer’s instructions. The presence of
ISAV was investigated using the procedure described
by Devold et al. (2000). Primers and
probes used were FA-3 (5¢-GAAGAGTCAGGAT
GCCAAGACG-3¢), RA-3 (5¢-GAAGTCGAT
GAACTGCAGCGA-3¢), ILAV FL (5¢-GCTGTG
TAGCATTGTCTTCAGGTCCTTC X) and
ILAV LC (Red640-ACATCGTCTTCTCCT
CCGCCATGTC p). ASPV was investigated by a
real-time one-step RT-PCR, using a LighCycler Ò
RNA Master HydProbe Kit (Roche, Mannheim,
Germany). Primers and probes used were Matrix S
(GTGCCTCAATCAATATCGTT S), Matrix A
(CGCAGAGTAGACCTTCTTC A), Matrix FL
(GCAGCAGACTGTATCTTTCCTAGTCCFL S)
and Matrix LC (640-CATCGGACATTTCG
TACGGGAC p S). For SPDV, RT-PCR was
performed using a Qiagen Ò one-step RT-PCR kit
(Qiagen, Hilden, Germany) and the primers PDF1
(5¢-CCCCGTTCGATCGCAAAGTA-3¢)andPDR2
(5¢-CGTGTAAGCCACGTGCACAT-3¢).
Thirty-eight plasma samples were collected during
the clinical outbreak in July (23 samples), as
well as at the end of the outbreak in August (5
samples) and September (10 samples). These were
investigated for neutralizing antibodies against
SPDV at the NVI and at the Veterinary Sciences
Division, Department of Agriculture and Rural
Development for Northern Ireland in Belfast by
D.A. Graham, according to the method described
by Graham, Jewhurst, Rowley, McLoughlin &
Todd (2003). The samples collected during the
outbreak originated from moribund fish sampled
8 weeks after the initial HSMI diagnosis. The other
samples were collected after mortalities were back at
a normal level. August samples were from moribund
fish, but July and September samples were
from both moribund and apparently normal fish.
Results
Descriptive epidemiology
The studied farm practised an all-in-all-out procedure
to increase biosecurity. Both locations used by
the farm during the period of study had been fallowed
for 7 months before introduction of fish. Cages and
equipment had been washed and disinfected before
restocking. After initial sea transfer of fish in October
2003, no additional fish were placed on the farm.

Ó 2006
Blackwell Publishing Ltd
Journal of Fish Diseases 2006, 29, 233–244 R T Kongtorp et al. Longitudinal study of HSMI in salmon
All fish that had been transferred to the farm
originated from two smolt producers. At the time of
sea transfer, both smolt producers also delivered fish
to one other farm (Fig. 1). Records of health
monitoring on the other farm showed that outbreaks
of parvicapsulosis (caused by Parvicapsula
pseudobranchicola), proliferative gill inflammation
and CMS occurred during the period from transfer
to slaughter. There was, however, no outbreak of
HSMI.
A timeline of events and diagnoses on the farm
under study is shown in Fig. 3. In mid December
there was an outbreak of infectious pancreatic
necrosis (IPN). This led to an increase in mortality,
peaking in January with 10% cumulative mortality
in cage 2. Cage 5 was the least severely affected cage
on the farm, with cumulative mortality at 1.7% in
the same period. By February mortality had
dropped to normal levels throughout the farm,
and stayed low until May, when HSMI was first
detected. In May there was a marked elevation of
mortality, especially in cage 8, but it was not until
June that aberrant swimmers started to appear.
During the first 2 weeks of June average mortality
continued to rise, and by mid June the clinical
outbreak was overt. The cumulative mortality
during the first 5 weeks of the outbreak is described
in Fig. 4. Clinical signs included abnormal swimming
behaviour, anorexia and mortalities. The
HSMI
Fish
relocated
Clinical signs
and
mortalities
occurrence of moribund fish originated in cage 1,
and appeared thereafter in a linear fashion, apparently
spreading from cage to cage. Mortality records
showed a more scattered pattern than the observations
of aberrant swimmers (Fig. 5). On a weekly
basis, cages 1 and 2 generally had the highest
number of mortalities during the outbreak period.
By late August, the number of aberrant swimmers
and mortalities had decreased to normal levels
and the outbreak was considered to be over. There
was, however, a slight increase in mortality in
September, presumably because of proliferative gill
inflammation which was diagnosed on the farm.
The gill problems continued until November, when
the health status appeared to improve. After the
HSMI outbreak had ceased, fish were sorted and
split into two groups. One group remained at the
studied location (location 2), while the other group
was transferred to another location (location 3). No
disease outbreaks occurred at location 3 after the
move. Early in March fish from two cages were
diagnosed with CMS. The farmer then started
slaughtering fish to reduce economic losses and by
May the farm had been emptied.
Development of cardiac lesions
The results from the histopathological investigation
are summarized in Table 1. Epicardial infiltrates
O N D J F M A M J J A S O N D J F M A
Transfer
to sea
IPN PC PGI
Fish group
split
CMS
Slaughter
Figure 3 Longitudinal study of an outbreak of heart and skeletal muscle inflammation (HSMI). Timeline of events and diagnoses on the
farm during the period of study. Sampling dates are indicated by vertical arrows. Months are indicated by their first letter. The first
cardiac changes were observed in February, and there was an escalation in lesion severity until HSMI was first diagnosed histologically in
May. The last cardiac lesions were seen in February the following year. The observed period of clinical disease and mortality was from
June to August. IPN, infectious pancreatic necrosis; PC, parvicapsulosis; PGI, proliferative gill inflammation; CMS, cardiomyopathy
syndrome.
237

Ó 2006
Blackwell Publishing Ltd
Journal of Fish Diseases 2006, 29, 233–244 R T Kongtorp et al. Longitudinal study of HSMI in salmon
Number of dead
10 000
9000
8000
7000
6000
5000
4000
3000
2000
1000
0
Number of dead
31.05.–06.06.
700
600
500
400
300
200
100
0
07.06.–13.06.
14.06.–20.06.
Time period
21.06.–27.06.
could be seen from December. Signs of myocardial
change first appeared in February, when mild
degeneration of cardiomyocytes and inflammatory
infiltrates were observed in a few fish (Fig. 6a).
Multifocal myocarditis was found with increasing
severity in March and April. In May one fish
displaying abnormal swimming behaviour met the
preset criteria for an HSMI diagnosis (Fig. 6b). The
atrium was not affected, but there were severe
inflammatory changes in all layers of the ventricle.
Samples taken in June confirmed the HSMI
28.06.–04.07.
Time period
Figure 4 Longitudinal study of an outbreak
of heart and skeletal muscle inflammation.
Cumulative weekly mortality during the
early outbreak period in June 2004.
31.05.–06.06. 07.06.–13.06. 14.06.–20.06. 21.06.–27.06. 28.06.–04.07.
Cage 1 Cage 2 Cage 3 Cage 4 Cage 5 Cage 6 Cage 7 Cage 8
Figure 5 Longitudinal study of an outbreak of heart and skeletal muscle inflammation (HSMI). Number of mortalities recorded weekly
during the early outbreak period. The numbers of fish were approximately equal in each cage. First HSMI case suspected by the fish
health service appeared in the second week of June. Mortality peaked in cages 1, 6 and 7 during week 2, and decreased thereafter. Cage 2
followed with a similar peak in week 3, whereas cages 3, 4 and 5 had their highest recorded mortality in week 4. In cage 8 mortality was
decreasing throughout June, apparently having peaked already in May.
238
diagnosis. There was no evident difference in lesion
severity in fish showing clinical signs compared with
those behaving normally. In fact, all 10 fish sampled
showed evidence of severe pancarditis and myocardial
degeneration. Atrial lesions varied from mild to
severe.
During the first week of July a large sample was
collected from cages 2 and 5, including 30 fish
showing aberrant swimming behaviour, 62 apparently
healthy and 10 dead fish. The majority of fish,
including apparently healthy fish, had moderate or

Ó 2006
Blackwell Publishing Ltd
Journal of Fish Diseases 2006, 29, 233–244 R T Kongtorp et al. Longitudinal study of HSMI in salmon
Figure 6 Longitudinal study of an outbreak of heart and skeletal muscle inflammation (HSMI). Micrographs of cardiac lesions. (a) The
first cardiac changes appeared in the February sample, 5 months after transfer to sea. Lesions included scattered foci of myocardial
inflammation, mainly in the spongy layer of the ventricle. (b) One fish met the preset criteria for the HSMI diagnosis in May. There
were diffuse inflammatory and degenerative changes in all layers of the ventricle. A section of compact myocardium is shown here
(bar ¼ 50 lm). (c) During the clinical disease outbreak in June–August most fish had severe inflammatory infiltrates in the epicardium
(asterisk) and underlying compact muscle. (d) Lesions in spongy myocardium were likewise mainly inflammatory in character. (e) In
September–November there was a tendency towards more severe lesions in the atrium than during the clinical outbreak. Several fish had
a moderate multifocal inflammation in this tissue, even though ventricular lesions could be sparse. (f) After the clinical outbreak, lesions
persisted in the heart as multifocal inflammation for many months, such as in this fish sampled in December. (g) Inflammation and
degeneration of myocytes in the red skeletal muscle were commonly observed in fish showing moderate or severe cardiac lesions in the
period between May and August. (h) Several fish showing severe cardiac inflammation also had multifocal to anastomosing liver necrosis
(asterisk) (bar ¼ 50 lm).
239

Ó 2006
Blackwell Publishing Ltd
Journal of Fish Diseases 2006, 29, 233–244 R T Kongtorp et al. Longitudinal study of HSMI in salmon
severe pancarditis; however, atrial lesions were
usually milder than those observed in the ventricle.
Lesions in the cardiac ventricle included multifocal
to diffuse inflammation and degeneration of myocardial
fibres in the compact and spongy layers, as
well as a severe epicardial cell infiltration
(Fig. 6c,d). Only one fish had no visible cardiac
lesions. In this sample a real difference between
lesion severities in the two cages could be seen for
the first time, in that lesions in samples taken from
cage 2 were generally more severe than samples
from cage 5. At this point in the outbreak, cage 8
contained the only fish group not having a higher
than normal mortality. To compare lesions between
this low mortality group with that of the other
groups, nine clinically diseased and one apparently
healthy fish were sampled from this cage. All 10 fish
had cardiac lesions consistent with HSMI, thus
mirroring the impression from the larger sample in
cages 2 and 5.
In late August, 2 months after the onset of
clinical signs, the number of mortalities had
gradually decreased. Cage 5 was chosen for further
sampling, because of a higher number of aberrant
swimmers and mortalities compared with cage 2.
Histopathology showed that epicardial infiltrates
were severe, while myocardial changes were moderate
and multifocal in all samples. A noticeable
change from fish sampled during the outbreak was
an increased involvement of lesions in the atrium
(Fig. 6e). There was obvious myocardial inflammation
and necrosis in the atrium of all but one
sampled fish. Involvement of the atrium was also
present in the September sample, but by November
this had ceased. Myocardial involvement in this and
later samples formed a multifocal pattern of cellular
infiltration, enlarged myocardial nuclei and fibrosis
in the ventricle, especially in the spongy layer
(Fig. 6f). CMS was diagnosed in two other cages in
March, but no evidence of CMS was observed in
the cages included in the study. Thus, in mid April,
no cardiac changes were seen in sampled fish.
Development of lesions in other organs
Skeletal muscle involvement first appeared in the
fish that was diagnosed with HSMI in May. There
was moderate inflammation and necrosis of myocytes
in red skeletal muscle, but no changes in white
muscle fibres. A month later eight fish showed
involvement of red skeletal muscle. There was also
moderate to severe myositis and necrosis in most
240
red muscle samples collected in July (Fig. 6g), but
no muscle lesions were seen in later samples. White
muscle did not show pathological changes.
Multifocal to anastomosing liver necrosis was
found in several fish showing moderate to severe
cardiac affection, regardless of sampling date
(Fig. 6h). Additionally, from early March and
onward, cellular infiltrates were usually seen in
connective tissue surrounding bile ducts. The
presence of these lesions was not restricted to fish
showing cardiac lesions, however, and its association
with the disease is therefore uncertain.
In the February sample, one fish had islets of
necrosis in the pancreatic tissue. This individual was
examined for IPNV and SPDV, using standard
immunohistochemical techniques. The necrotic
areas showed a strongly positive staining for IPNV,
but were negative for SPDV. There were no
pancreatic lesions in the other samples obtained
during the course of the study.
In September to November the farm was affected
by proliferative gill inflammation. Fish showing
abnormal behaviour during this period were considered
to be suffering from gill problems rather
than HSMI. This was supported by histopathology,
as only clinically diseased fish showed evidence of
proliferative changes in the gills. There were also no
differences in cardiac lesion severity between apparently
healthy fish and clinically diseased fish in the
samples during this period.
Signs of circulatory disturbances were not common
at any time during the study. There was a
slight increase in the number of fish displaying
erythrocyte concentration in the spleen, kidney,
liver and gills when cardiac lesions were moderate
or severe, but this pattern was not consistent.
Comparative histology
Examination of fish sampled for the comparative
histological study showed that 33 had a mild to
moderate epicarditis and five had a mild endocarditis.
One fish diagnosed with IPN had a small
focus of myocarditis in the spongy layer of the
ventricle. All 10 fish suffering from PD showed
severe loss of exocrine pancreatic tissue. Cardiac
lesions included mild to moderate epicarditis,
moderate to severe degeneration of cardiomyocytes
and mild to moderate infiltration of inflammatory
cells. Skeletal muscle lesions included severe degeneration
of red and white fibres and only mild
inflammation. All 15 fish diagnosed with CMS

Ó 2006
Blackwell Publishing Ltd
Journal of Fish Diseases 2006, 29, 233–244 R T Kongtorp et al. Longitudinal study of HSMI in salmon
displayed severe degeneration and inflammation of
the atrium and spongy layer of the cardiac ventricle.
No lesions were seen in the compact layer and
epicardium of the heart, or in any other organs.
Virology and serology
No cell cultures showed evidence of cytopathic
effect after inoculation. RT-PCR did not indicate
presence of ISAV, ASPV or SPDV in the samples.
Plasma samples collected during and after the
disease outbreak were negative for SPDV antibodies.
Discussion
In the present study, we have examined the
complete duration of a natural HSMI outbreak
through a succession of histopathological investigations.
The studied farm had been experiencing
repeated problems with HSMI for several years, in
spite of good biosecurity routines. During the study
year the farm was again struck by HSMI, giving rise
to clinical signs and mortalities in the population
during the summer months. Recurrence of outbreaks
on the farm is consistent with observations of
some other natural outbreaks of HSMI (A. Kjerstad,
personal communication), but the reason for
this pattern is yet to be determined.
On commencement of the study, the fish had no
obvious lesions or clinical signs indicative of
HSMI, suggesting that the disease was not already
ongoing in the population. In this case, therefore,
there is little reason to believe that the fish were
affected by HSMI in the freshwater or early
seawater stages. This is further supported by the
absence of a HSMI diagnosis in contact fish
transferred to another seawater farm. Healthy
smolts sampled for histological comparison also
showed no cardiac lesions. The first signs of cardiac
involvement appeared as early as February,
5 months after transfer to sea. HSMI can at
present be diagnosed only by histopathology, and
there is therefore great uncertainty over the significance
of these early cardiac lesions. There was,
however, an increasing number of fish with cardiac
lesions, as well as an escalation of lesion severity,
from February to June. This gives support to the
idea that the causal agent may have a seawater
reservoir, and that HSMI may have been developing
on the farm for many months before clinical
signs of a disease outbreak became evident.
241
Challenge studies of HSMI have shown a time
lag from initial infection to obvious cardiac lesions
of approximately 6–8 weeks, at a temperature
ranging from 10 to 12 °C (Kongtorp et al.
2004a). If all fish on the farm were infected at
once, this would have been likely to occur in late
March to early April, producing the severe cardiac
lesions that were recorded in May and June. In a
natural environment fish may be exposed to the
infectious agent unequally, in respect of time and
dose. Water temperature may also affect the
production and shedding of virus. A study of
infectious salmon anaemia (ISA) in Norway showed
that the time from infection to clinical outbreak
could be several months, even though transmission
trials have resulted in mortalities within 3 weeks
(Va˚gsholm, Djupvik, Willumsen, Tveit & Tangen
1994). It is therefore not unlikely that some fish
could have developed cardiac lesions at an earlier
date than the majority of the population on the
studied farm. Such individuals may perhaps be
more susceptible to cardiac disease, either because
of prior or concurrent diseases, or from some
natural weakness. No macroscopic cardiac lesions
were noted in fish showing early lesions.
When the clinical outbreak first started on the
farm, morbidity was remarkably high. Of a total of
112 fish sampled from the farm 3 weeks after the
onset of clinical disease, as many as 111 had lesions
in the heart and skeletal muscle that were consistent
with HSMI. Furthermore, there was no evident
histopathological difference between apparently
healthy fish and fish showing clinical signs. In fact,
moderately and severely affected fish made up
almost 90% of the sample, even when random
selection was applied. HSMI appeared to spread in
a cage-to-cage pattern, as observed by the number
of fish showing clinical signs such as abnormal
swimming behaviour. However, mortality data
from the farm indicates a different pattern. It
appears that the low mortality group in cage 8 had
experienced peak mortality already in May with
more than 5000 dead, and that mortality in this
cage was decreasing thereafter. All the other cages
had the highest recorded mortality in June, but with
a much less obvious chronology than the observed
onset of clinical signs. Neither increased mortality
nor clinical signs seemed to be good indicators for
the extent of the outbreak. Taken together, HSMI
may have morbidity near 100%, and infection with
HSMI is likely to affect all cages at the farm
location.

Ó 2006
Blackwell Publishing Ltd
Journal of Fish Diseases 2006, 29, 233–244 R T Kongtorp et al. Longitudinal study of HSMI in salmon
What determines the onset of clinical signs?
Studies of other field outbreaks may provide
possible answers to this question. After a clinical
outbreak of HSMI 14 days after transfer to sea,
samples were taken from the smolt producer from
which the fish originated. On histological examination,
cardiac and skeletal muscle lesions were
equally severe in moribund fish from the outbreak
and apparently healthy fish at the smolt production
site (R.T. Kongtorp, unpublished data). Moving of
fish itself, therefore, seemed to have triggered onset
of clinical signs. The causal agent of HSMI also
appeared to have been introduced to smolts and not
growing fish. The fish had been exposed to sea
water prior to transfer, which may in part explain
the early outbreak. Similar observations of clinical
outbreaks just after major stressful episodes have
also been reported from seawater farms containing
growing fish (A. Lyngøy, personal communication).
In the present study, lesions were scarce and
relatively mild in fish sampled before relocation
from location 1 to location 2. Onset of clinical signs
was also delayed compared with earlier observations
of stress-related outbreaks. The first fish showing
lesions consistent with HSMI was, however, sampled
only a fortnight after the move between sites.
Stress associated with the move may have escalated
the progress of an ongoing infection in the
population by negatively affecting the immune
system of the fish (Press 1998). Another explanation
may be that infection was not introduced into the
population until after the move. If so, the incubation
period for some fish must have been somewhat
shorter than that observed experimentally by
Kongtorp et al. (2004a).
It appears that the heart is the primary target
organ in HSMI. Lesions were more common and
more severe in cardiac tissue than in any other
organ. Cardiac lesions were also the first to be
observed and persisted for many months after other
tissues had returned to an apparently normal
condition. The severity of cardiac lesions was,
however, gradually reduced, suggesting that the fish
were able to control the infection and reduce tissue
damage. Fish are able to regenerate myocardial cells
(Poss, Wilson & Keating 2002). Compensatory
cellular hypertrophy may therefore be just one of
the mechanisms helping the affected fish to survive
serious myocardial insults (Ferguson, Poppe &
Speare 1990). In the present study cellular infiltration
concentrated largely at one end of the
myocardial cells in the spongy layer of some late
242
samples. It is possible that these cell aggregations
indicate an ongoing elongation of myocardial cells
and may thus represent attempts at myocardial
regeneration. They were seldom as striking as the
nuclear nests described by Ferguson et al. (2005),
but an association between these features is still
highly possible.
Foci of fibrosis were commonly seen in late
cardiac samples. These were observed in the
intermediate layer between compact and spongy
tissues of the ventricle and were not usually
associated with the presence of inflammatory infiltrates.
Myocardial scarring is a frequent outcome of
cardiac disease in terrestrial species, as the regenerative
capacity is limited (Junqueira, Carneiro &
Kelley 1995). In severely affected fish, therefore,
tissue damage may have been too extensive to allow
for full recovery, thereby substituting muscles with
connective tissue. Whether the cardiac tissue
completely regenerates at a later stage remains to
be determined.
Consistent with the observations of Kongtorp
et al. (2004b), red skeletal muscle was severely
affected in fish with severe cardiac changes. The
presence of lesions in this organ was not consistent,
however, and limited in time to the clinical
outbreak itself. Once the cardiac lesions ceased in
severity, skeletal muscle lesions were no longer
found. From this study, therefore, there is no reason
to believe that red skeletal muscle lesions persist in
fish after an outbreak of HSMI, or that lesions
appear in white muscle at a later stage. A larger
study must however be undertaken if one is to
determine whether HSMI can cause reduced fillet
quality. As the presence of severe inflammatory
changes in both heart and red skeletal muscle is
quite common in moribund fish during the clinical
stage of the disease (Kongtorp et al. 2004b), the
name Ôheart and skeletal muscle inflammationÕ can
still be considered appropriate.
Lesions in the liver were relatively common in
fish with severe circulatory disturbances caused by
myocardial inflammation and damage. The majority
of these observations were made during the
clinical outbreak of HSMI. Multifocal liver necrosis
has been observed in other diseases causing circulatory
disturbances, for instance CMS and ISA
(Speilberg, Evensen & Dannevig 1995; Rodger &
Turnbull 2000). Liver necroses are also observed in
outbreaks of IPN, but these occur concurrently with
pancreatic lesions, and in general about 3 months
post-transfer to sea water (Roberts & Pearson

Ó 2006
Blackwell Publishing Ltd
Journal of Fish Diseases 2006, 29, 233–244 R T Kongtorp et al. Longitudinal study of HSMI in salmon
2005). It is therefore likely that the liver lesions
observed in the present study are secondary to the
cardiac changes.
Pancreas disease and CMS are the most obvious
differential diagnoses to HSMI (Kongtorp et al.
2004b). The present study was conducted in an area
where PD has never been diagnosed, but where CMS
is known to be a common problem (A. Kjerstad,
personal observations). Cardiac lesions in CMS and
HSMI may be distinguished histopathologically by
the marked compact layer involvement in HSMI,
which is rarely observed in CMS. Atrial lesions are
also not usually as severe in HSMI as in CMS
outbreaks in which severity may be such as to lead to
atrial rupture (Ferguson et al. 1990; Kongtorp et al.
2004b). An interesting observation in the present
study was the marked shift from initial involvement
of the ventricle, to more serious lesions in the atrium
after the clinical outbreak. In the ventricle, there was
also a time lag in the recovery of spongy tissue
compared with that of the compact tissue. These late
lesions were not as extensive as one would expect
from a case of CMS, but could still be confused with
an early stage of this disease. In the present study, fish
from cages 2 and 5 did not show lesions consistent
with CMS (Ferguson et al. 1990) during the later
outbreak of CMS on the farm. A possible association
between HSMI and CMS is therefore unlikely, but
cannot be entirely excluded.
Phagocytic activity occurs in cardiac endothelium
of teleosts, especially in the atrium, although this is
not so avid in salmonids (Ferguson 1989). The late
involvement of spongy tissue may therefore indicate
that the endocardium is active in clearance of
antigen and debris from the heart. Another possible
explanation may be an autoimmune reaction of
spongy myocardium, triggered by antigen activity.
Post-viral myocarditis is known from studies of
terrestrial species (Huber 1997; Kishimoto, Hiraoka,
Takamatsu, Takada, Kamiya & Ochiai
2003). It has been suggested that some viruses are
capable of mimicking epitopes within host proteins,
thus inducing autoimmune myocarditis (Davies
1997). The different properties of spongy and
compact myocardium (Ewart & Driedzic 1986) or
proximity to phagocytic endocardium may account
for the location specificity of these lesions.
Cardiac and red skeletal lesions in PD and HSMI
may appear similar, as inflammation and degeneration
in these organs are characteristic features of
both diseases (Ferguson, Roberts, Richards, Collins
& Rice 1986; Murphy, Rodger, Drinan, Gannon,
243
Kruse & Korting 1992; Kongtorp et al. 2004b). In
PD there is, however, a tendency towards a more
degenerative picture, as opposed to the inflammatory
impression of HSMI (Kongtorp et al. 2004b).
Additionally, lesions are also seen in the white
skeletal muscle and in the exocrine pancreatic tissue
in fish suffering from PD. Over the duration of the
study, there were no signs of pancreatic involvement,
except from lesions associated with IPN in
one fish early in the study. Exocrine pancreatic
necrosis and loss are described to be major features
of the pathological picture of PD (Munro, Ellis,
McVicar, McLay & Needham 1984). Although
regeneration of pancreatic tissue occurs at a
relatively early stage of PD (Desvignes, Quentel,
Lamour & Le Ven 2002), one would have expected
to observe pancreatic necrosis and loss of acinar
cells, if present, within the time frame of the present
study. PD is caused by SPDV (Nelson, McLoughlin,
Rowley, Platten & McCormick 1995;
McLoughlin, Nelson, Rowley, Cox & Grant
1996). Aside from different pathology, to exclude
involvement of PD from the diagnosis, sera from
the study were examined for SPDV antibodies.
These were all negative. If SPDV was indeed
responsible for the lesions observed in the present
study, a negative serological result would be highly
unlikely (Graham, Jewhurst, Rowley, McLoughlin,
Rodger & Todd 2005). Kidneys from the April,
May, June, July and August samples were also
investigated for the presence of SPDV by RT-PCR.
These samples all tested negative, while controls
from Norwegian PD outbreaks were positive.
In conclusion, HSMI appears to be a severe
disease with high morbidity and prolonged duration.
A gradual increase in lesion severity in the
months prior to onset of clinical signs indicates that
infection may be present subclinically in the farm
for a very long time. From a histopathological
viewpoint, HSMI is primarily a cardiac disease.
Lesions were also observed in other organs, but were
not consistent. Most fish on the farm under study
recovered from HSMI, although recovering fish
retained foci of degenerated myocytes and inflammatory
infiltration for several months after the
clinical outbreak. Further investigation into the
aetiology and pathogenesis of HSMI is required.
Acknowledgements
This project was partly funded by the Norwegian
Research Council. Sincere thanks to A. Kjerstad and

Ó 2006
Blackwell Publishing Ltd
Journal of Fish Diseases 2006, 29, 233–244 R T Kongtorp et al. Longitudinal study of HSMI in salmon
D. Vollstad at Havbrukstjenesten A/S for their
valuable contribution in sampling material for this
study. Thanks to D. Graham, I. Modahl and
M. Heum for serological testing and technical
assistance with PCR. We are also very grateful for
helpful comments given by H.W. Ferguson.
References
Davies J.M. (1997) Molecular mimicry: can epitope mimicry
induce autoimmune disease? Immunology and Cell Biology 75,
113–126.
Desvignes L., Quentel C., Lamour F. & Le Ven A. (2002)
Pathogenesis and immune response in Atlantic salmon (Salmo
salar L.) parr experimentally infected with salmon pancreas
disease virus (SPDV). Fish and Shellfish Immunology 12,
77–95.
Devold M., Krossøy B., Aspehaug V. & Nylund A. (2000) Use
of RT-PCR for diagnosis of infectious salmon anaemia virus
(ISAV) in carrier sea trout Salmo trutta after experimental
infection. Diseases of Aquatic Organisms 40, 9–18.
Ewart H.S. & Driedzic W.R. (1986) Enzymes of energy metabolism
in salmonid hearts: spongy vs. cortical myocardia.
Canadian Journal of Zoology 65, 623–627.
Ferguson H.W. (1989) Systemic Pathology of Fish. Iowa State
University Press, Ames, IA.
Ferguson H.W., Roberts R.J., Richards R.H., Collins R.O. &
Rice D.A. (1986) Severe degenerative cardiomyopathy associated
with pancreas disease in Atlantic salmon, Salmo salar L.
Journal of Fish Diseases 9, 95–98.
Ferguson H.W., Poppe T. & Speare D.J. (1990) Cardiomyopathy
in farmed Norwegian salmon. Diseases of Aquatic
Organisms 8, 225–231.
Ferguson H.W., Kongtorp R.T., Taksdal T., Graham D.A. &
Falk K. (2005) An outbreak of disease resembling heart and
skeletal muscle inflammation in Scottish farmed salmon,
Salmo salar L., with observations on myocardial regeneration.
Journal of Fish Diseases 28, 119–123.
Graham D.A., Jewhurst V.A., Rowley H.M., McLoughlin M.F.
& Todd D. (2003) A rapid immunoperoxidase-based virus
neutralization assay for salmonid alphavirus used for a serological
survey in Northern Ireland. Journal of Fish Diseases
26, 407–413.
Graham D.A., Jewhurst V.A., Rowley H.M., McLoughlin
M.F., Rodger H. & Todd D. (2005) Longitudinal serological
surveys of Atlantic salmon, Salmo salar L.,
using a rapid immunoperoxidase-based neutralization assay
for salmonid alphavirus. Journal of Fish Diseases 28,
373–379.
Huber S.A. (1997) Autoimmunity in myocarditis: relevance of
animal models. Clinical Immunology and Immunopathology 83,
93–102.
244
Junqueira L.C., Carneiro J. & Kelley R.O. (1995) Basic Histology.
Prentice-Hall International, Inc., London.
Kishimoto C., Hiraoka Y., Takamatsu N., Takada H., Kamiya
H. & Ochiai H. (2003) An in vivo model of autoimmune
post-coxsackievirus B3 myocarditis in severe combined
immunodeficiency mouse. Cardiovascular Research 60,
397–403.
Kongtorp R.T., Kjerstad A., Guttvik A., Taksdal T. & Falk K.
(2004a) Heart and skeletal muscle inflammation in Atlantic
salmon, Salmo salar L.: a new infectious disease. Journal of Fish
Diseases 27, 351–358.
Kongtorp R.T., Taksdal T. & Lyngøy A. (2004b) Pathology of
heart and skeletal muscle inflammation (HSMI) in farmed
Atlantic salmon Salmo salar. Diseases of Aquatic Organisms 59,
217–224.
McLoughlin M., Nelson R.T., Rowley H.M., Cox D.I. & Grant
A.N. (1996) Experimental pancreas disease in Atlantic salmon
Salmo salar post-smolts induced by salmon pancreas disease
virus (SPDV). Diseases of Aquatic Organisms 26, 117–124.
Munro A.L.S., Ellis A.E., McVicar A.H., McLay H.A. &
Needham E.A. (1984) An exocrine pancreas disease of farmed
Atlantic salmon in Scotland. Helgolander Meeresuntersuchungen
37, 571–586.
Murphy T.M., Rodger H.D., Drinan E.M., Gannon F., Kruse P.
& Korting W. (1992) The sequential pathology of pancreas
disease in Atlantic salmon farms in Ireland. Journal of Fish
Diseases 15, 401–408.
Nelson R.T., McLoughlin M.F., Rowley H.M., Platten M.A. &
McCormick J.I. (1995) Isolation of a toga-like virus from
farmed Atlantic salmon Salmo salar with pancreas disease.
Diseases of Aquatic Organisms 22, 25–32.
Poss K., Wilson L. & Keating M. (2002) Heart regeneration in
zebra fish. Science 298, 2188–2190.
Press C.McL. (1998) Immunology of fishes. In: Handbook of
Vertebrate Immunology (ed. by P.P. Pastoret, P. Griebel,
H. Bazin & A. Govaerts), pp. 3–62. Academic Press,
San Diego, CA.
Roberts R.J. & Pearson M.D. (2005) Infectious pancreatic
necrosis in Atlantic salmon, Salmo salar L. Journal of Fish
Diseases 28, 383–390.
Rodger H. & Turnbull T. (2000) Cardiomyopathy syndrome in
farmed Scottish salmon. The Veterinary Record 146, 500–501.
Speilberg L., Evensen Ø. & Dannevig B.H. (1995) A sequential
study of the light and electron microscopic liver lesions of
infectious anaemia in Atlantic salmon (Salmo salar L.).
Veterinary Pathology 32, 466–478.
Va˚gsholm I., Djupvik H.O., Willumsen F.V., Tveit A.M. &
Tangen K. (1994) Infectious salmon anemia (ISA) epidemiology
in Norway. Preventive Veterinary Medicine 19, 277–290.
Received: 12 October 2005
Revision received: 8 February 2006
Accepted: 9 February 2006

Paper IV
Features of the pathogenesis of heart and skeletal muscle
inflammation in Atlantic salmon, Salmo salar L.

Features of the pathogenesis of heart and skeletal muscle inflammation in Atlantic
salmon, Salmo salar L.
R T Kongtorp 1 , E O Koppang 2 , I Bjerkås 2 , K Falk 1 and T Taksdal 1
1 National Veterinary Institute, Oslo, Norway
2 Norwegian School of Veterinary Science, Oslo, Norway
Correspondence: R T Kongtorp, Section for Epidemiology, National Veterinary Institute, Pb
750 Sentrum, N-0106 Oslo, Norway (e-mail: ruth-torill.kongtorp@vetinst.no)
Keywords: Atlantic salmon, HSMI, inflammation, myocardium, pathogenesis
Running title: Pathogenesis of HSMI
1

Abstract
Heart and skeletal muscle inflammation (HSMI) is a transmissible disease affecting farmed
Atlantic salmon. It is histologically recognised by inflammation and necrosis in the heart and
red skeletal muscle. In order to obtain better understanding of the pathogenesis, this study
aimed at characterising key features of the pre- and mid-clinical phases. Material from field
outbreaks were investigated by light and transmission electron microscopy. Sections of heart
and skeletal muscle were also examined for alkaline and acid phosphatase, unspecific
esterase, peroxidase, MCH class II, IgM and fibronectin. In the pre-clinical phase,
perivascular inflammation of coronary vessels in epi- and myocardium was abundant, but the
vascular wall was usually intact. In the spongy layer, vacuolation, inflammation and necrosis
of the endocardium was evident. In the clinical outbreak phase, myocardial changes became
more severe, and inflammation and necrosis were more diffusely distributed in the heart.
Collagen was present in moderate amounts in the most damaged areas in this phase, but this
gradually became less prominent. The disease process of HSMI may thus start with
perivascular cell infiltrations and changes in endocardium while myocardial lesions appear to
occur somewhat later. The morphology of the cells that constituted the early infiltrations
around coronary vessels as well as the severe cell infiltrations during the clinical outbreak
clearly suggested that they were inflammatory cells. However, only a few of them reacted as
anticipated to the histochemical reagents applied in the present study.
2

Introduction
Heart and skeletal muscle inflammation (HSMI), a disease affecting farmed Atlantic salmon
Salmo salar, is an increasing problem in the Norwegian salmon farming industry. From its
first discovery in 1999 with an initial 10-15 registered disease outbreaks per year, the number
of diagnosed outbreaks registered by the National Veterinary Institute (NVI) has increased
dramatically. In 2006, fish on 94 sites were diagnosed with HSMI, and in 2007, the number of
affected sites was as high as 162 (Skjelstad, Bornø, Flesjå, Hansen, Nilsen, Wasmuth &
Hjeltnes 2008). A longitudinal study of fish in the seawater phase has recently revealed that
farms with no apparent disease problems may contain fish with severe lesions consistent with
HSMI (RT Kongtorp, unpublished results). Also, the number of fish with severe lesions in the
heart and red skeletal muscle during clinical outbreaks may be close to 100 % (Kongtorp,
Halse, Taksdal & Falk 2006). However, the mortality due to HSMI may sometimes be very
low, and significant mortality is often associated with additional stress episodes in the clinical
phase of the disease. It is therefore likely that some outbreaks may go unnoticed.
HSMI is experimentally transmissible by intraperitoneal (i.p.) injection of tissue material
from diseased fish and by cohabitation with injected fish (Kongtorp, Kjerstad, Guttvik,
Taksdal & Falk 2004a). A viral aetiology has long been suspected, and the characterisation of
a possibly causal virus is ongoing (Eliassen, Solbakk, Evensen & Gravningen 2004). Electron
microscopy has revealed several different viral-like particles in cardiac tissue, but a
conclusive association of either of these to HSMI has not been established (Watanabe,
Karlsen, Devold, Isdal, Litlabø & Nylund 2006). Due to the lack of agent-specific diagnostic
tools, HSMI is diagnosed on the basis of clinical signs, necropsy findings and histopathology.
Fish suffering from HSMI may be anorectic or moribund. On necropsy, affected fish may
have haemopericardium, pale cardiac ventricle, yellowish liver, ascites, swollen spleen and
hyperemia in the perivisceral fat. HSMI is histologically recognised by moderate to severe
3

inflammation of epi-, endo- and myocardium involving the atrium and all layers of the cardiac
ventricle, and myositis in red skeletal muscle. Tissue necrosis is associated with inflammation
in both heart and red skeletal muscle. Fish with HSMI may also have multifocal to confluent
liver necrosis and an oedema-like meshwork (“pseudolobulation”) in the spleen (Kongtorp,
Taksdal & Lyngøy 2004b).
4
Lesions observed in HSMI may resemble other diseases, especially pancreas disease (PD)
(McLoughlin, Nelson, McCormick, Rowley & Bryson 2002; McLoughlin & Graham 2007)
and cardiomyopathy syndrome (CMS) (Ferguson, Poppe & Speare 1990). In PD, an initial
necrosis and subsequent atrophy of exocrine pancreas is followed by myocardial necrosis and
inflammation of compact and spongy layers of the heart, as well as necrosis and myositis of
skeletal muscle (McLoughlin et al. 2007; Taksdal, Olsen, Bjerkås, Hjortaas, Dannevig,
Graham & McLoughlin 2007). CMS mainly affects spongy myocardium and adjacent
endocardium, in which there are severe inflammatory changes (Ferguson et al. 1990). Typical
cases of HSMI, PD and CMS are clearly distinguishable (Kongtorp et al. 2006). Problems
may arise, however, in atypical cases and when multiple diseases concur. A recent
longitudinal study of the disease development indicated that HSMI may start as a subclincal
infection, followed by an acute clinical phase in the course of a few weeks, and thereafter a
recovery phase lasting several months (Kongtorp et al. 2006). Fish may therefore be in
various stages of HSMI during outbreaks of other diseases in the seawater phase. The
pathogenesis of HSMI is largely unknown. This study therefore aimed at characterising some
key pathological features of the pre-, and mid-clinical phases of HSMI, with emphasis on
cardiac changes. As inflammation appears to be a dominant feature of HSMI, inflammatory
cell markers were used in histochemical and immunohistochemical examinations in order to
further characterise the cells.

Materials and methods
Histology
The material used was part of a longitudinal study of an outbreak of HSMI conducted from
2003 to 2005. In this study, material from five diseased and five apparently healthy fish had
been collected monthly for six months before a clinical outbreak of HSMI with increased
mortality commenced on the farm. Clinical presentation, necropsy results and
histopathological features have been described previously (Kongtorp et al. 2006). In addition,
40 fish from four other farms experiencing clinical outbreaks of HSMI were sampled. Fish
with clinical disease and histopathological lesions consistent with HSMI as described by
Kongtorp et al. (2004b) were included. The HSMI diagnosis was based on the presence of
mononuclear inflammation in the epi-, endo- and myocardium of the heart and myositis in red
skeletal muscle, as well as necrosis associated with these changes. In addition to diseased and
dead fish, samples of fish with no clinical signs were randomly collected from the same net
pens. Samples were taken from pseudobranchs, gills, heart, liver, pyloric caeca with pancreas,
spleen, kidney and skeletal muscle. Tissue samples were fixed in 10 % neutral phosphate
buffered formalin, prepared for histology by standard paraffin wax techniques and stained
with hematoxylin and eosin (HE). Selected samples from heart and skeletal muscle were
stained with Giemsa, Ziehl Nielsen (ZN), Periodic-Acid Shiff (PAS), Gram, van Gieson (vG)
and Martius Scarlet Blue (MSB) (Bancroft & Stevens 1990).
Electron microscopy
In order to examine the development of cardiac lesions towards the clinical outbreak phase,
fish sampled monthly from two months prior to the outbreak to the mid-outbreak period were
also studied by transmission electron microscopy (TEM). Samples were taken from compact
and spongy myocardium and fixed in either 10 % neutral phosphate buffered formalin or 3%
5

glutaraldehyde in 0.1 M sodium cacodylate buffer (CB). Fixed samples were washed three
times in CB and covered with 2% osmium tetroxyde (OsO4) and ferrocyanide in CB for 2
hours. After washing, samples were en bloc contrasted in uranyl acetate. Samples were then
again washed, before they were taken through a dehydration process with graded ethanols.
Samples were thereafter transferred to 1,2-propylene oxide. Finally, samples were infiltrated
by a 1:1 solution of Lx-112 (Epon-equivalent epoxy monomers) and 1,2-propylene oxide, and
a 3:1 solution, the latter being left over night. Pure Lx-112 was added, and the cell material
was incubated at 37 ºC for 24 hours and 60 ºC for four days. Specimens were prepared for
electron microscopy by ultra thin slicing, stained with lead citrate and examined in a Phillips
EM 208 S TEM.
Histochemistry
For enzyme- and immunohistochemistry, samples from heart, skeletal muscle and pyloric
caeca with pancreas from three of the field outbreaks were coated with Tissue-Tek Optimum
Cutting Temperature Compund (OCT, Sakura Finetek) and quickly frozen in isopentane
chilled in liquid nitrogen. Frozen samples were wrapped in aluminium foil and stored at -80
ºC. Cryosections of 4-6 m thickness were allowed to dry for 1 h on Polysine microscope
slides (Menzel) before they were frozen at -80 ºC. Frozen sections were thawed and dried in
room temperature for 1 h before they were used for enzyme- or immunohistochemistry. In
addition to the frozen sections, formalin fixed and paraffin embedded sections from heart and
skeletal musle were tested for alkaline phosphatase, peroxidase, MHC class II, IgM and
fibronectin. These were deparaffinated by heating at 60˚C for 20 min. Sections were thereafter
placed in 0.1 M citrate buffer and demasked in a microwave oven for 2 x 5 min and washed in
0.1 M Tris buffered saline (TBS). Demasked sections were rehydrated through xylene and
graded ethanol baths.
6

Alkaline phosphatase: Frozen sections from heart and skeletal muscle of 12 fish were fixed
in formolcalsium for 5 min at 4 ˚C. Formalin fixed and paraffin embedded sections from heart
and skeletal muscle of ten fish were deparaffinated, demasked and and rehydrated. All
sections were washed in distilled water and incubated for 1 h at room temperature with 10 mg
Fast Red TR salt (Sigma) dissolved in 2 mg Napthol AS-MX phosphate (Sigma), 0.2 mL
N,N dimethylformamide and 9.8 mL TBS at pH 9.2. Duplicate sections were used as negative
controls by preheating at 80 ˚C in TBS for 10 min before incubation. The sections were
counterstained with haematoxylin and mounted in Aquamount (BHD Ltd.).
Acid phosphatase: Frozen sections from heart and skeletal muscle of 13 fish were incubated
in a solution consisting of 4 mL hexazonium pararosaniline in 48 mL sodium acetate, 500 L
N,N dimethylformamide and 20 mg Napthol-AS-TR phosphate (Sigma), pH 5.5, for 1 h at
room temperature. Duplicate sections were used as negative controls by adding 10 mM
sodium fluoride in the incubation solution for inhibition of the reaction. Sections were
counterstained in haematoxylin and mounted with Neo-Mount (EMD Chemicals).
Peroxidase: Frozen sections from heart and skeletal muscle of three fish, as well as paraffin
embedded sections from nine fish were incubated with 6 mg diaminobenzidine in 10 mL
imidazole buffer and 3 % H2O2 in methanol for 10 minutes in room temperature. Duplicate
sections were used as negative controls by preheating at 80 ˚C for 5 min. Sections were
counterstained in haematoxylin and mounted with Neo-Mount.
Unspecific esterase: Frozen sections from heart and skeletal muscle of 12 fish were
incubated for 15 minutes at room temperature with an incubation solution consisting of 2.8 %
di-sodium hydrogen phosphate in distilled water, 0.06 % hexazonium pararosaniline, and 1 %
-napthyl acetate in acetone, pH 7.2. Duplicate sections were used as negative controls by
preheating at 80 ˚C for 5 min. Sections were fixed in 10 % phosphate buffered formalin for 2
h before counterstaining with hematoxyllin and mounting with Neo-Mount.
7

8
MHC class II immunohistochemistry: Frozen sections from heart and skeletal muscle of
three fish and paraffin embedded sections from 20 fish were examined for the expression of
MHC class II by immunohistochemistry based on polyclonal antibodies from Atlantic salmon
(Koppang, Hordvik, Bjerkås, Torvund, Aune, Thevarajan & Endresen 2003). Frozen sections
were fixed in 100 % acetone for 10 min. Paraffin embedded sections were deparaffinated,
demasked and rehydrated. Endogenous peroxidase activity was inhibited by by adding 10 %
H2O2 in methanol. Sections were thereafter washed in phosphate buffered saline (PBS). To
prevent non-specific binding, sections were treated with caprine serum diluted 1:50 in 5 %
bovine serum albumin (BSA) in TBS. Frozen sections were incubated for 1 h with the
primary MHC class II antibody diluted 1:1000 in 1 % BSA in TBS. For deparaffinated
sections, a 1:700 dilution was used. After incubation, the sections were washed in PBS and
incubated for 30 min in avidin-biotin complex/ horse radish peroxidase (Labelled polymer,
HRP anti-rabbit, DAKO Cytomation) antibody. After washing in PBS, bound antibody was
detected with 3-amino-9-ethyl-carbazole (AEC substrate chromogen, DAKO Cytomation) for
15 min. Sections were counterstained with haematoxylin and mounted with polyvinyl alcohol
mounting media, pH 8.2.
Immunoglobulin (IgM) immunohistochemistry: Frozen sections from heart and skeletal
muscle of three fish and paraffin embedded sections from 20 fish were examined for the
presence of immunoglobulin by immunohistochemistry. The signal was based on a polyclonal
rabbit-anti salmon Ig from Atlantic salmon (Falk, Press, Landsverk & Dannevig 1995),
diluted 1:400 in 1 % BSA in TBS. The same procedure as described above for MHC class II
was followed.
Fibronectin immunohistochemistry: Paraffin embedded sections from heart and skeletal
muscle of six fish were investigated for fibronectin expression (Monoclonal anti-human

fibronectin, Clone FN-15, Sigma, St. Louis, MO, USA) in dilutions 1:200 to 1:600 following
the procedure described above.
Results
Pre-clinical phase
The first phase was defined as the period from sea transfer to the clinical outbreak phase
commenced with an increase in mortalities and moribund fish due to HSMI. In the present
study we focused on a period of two months prior to this event.
Initially, the cardiac changes that could be observed by both light and electron microscopy
were mainly associated with the epicardium, coronary vessel branches and endocardium. In
some samples, there was a marked increase of cells within the lumen of coronary vessel
branches (Figs. 1-2), which was not observed in vessels in other tissues. Epicarditis was
evident, especially around coronary vessels. Several coronary vessel branches in the compact
layer were surrounded by inflammatory cells (Fig. 3). However, the vessel wall was not
visibly affected, as the inflammatory cells were mainly located in a space between the
adventitia of the vessel and the surrounding myocardium. Some fish also had perivasculitis in
the liver and red skeletal muscle, but this was not consistent. Inflammatory cells within the
myocardium were usually closely related to the perivascular changes, forming a multifocal
pattern of inflammation in the compact layer. By light microscopy, myocardial necrosis was
only found within such foci. Several myocardial fibres had loss of striation and were
condensed. A few fibres were hypercontracted. The budding activity within the myocardium
was high, especially across membranes bordering necrotic areas. Towards the clinical
outbreak phase, inflammatory cells were more abundant and diffusely distributed.
Ultrastructural changes in the compact layer could be observed in two fish from the early
samples. Endothelium of some coronary vessel branches were either hypertrophic and
9

irregular or almost completely lysed. Myocardial cells were also necrotic, although no
inflammation could be observed in the vicinity of the lesions. Gaps in the myocardium could
be observed, mainly containing remnants of membranes and spherical structures of various
sizes (Fig. 4). A condensed or partially lysed nucleus was usually present within the damaged
foci.
10
In the spongy layer, endothelial cells were hypertrophic in most sampled fish during the pre-
clinical phase (Figs. 5-6). Large vacuoles were present within the endothelial cells, the
underlying myocardium, or in an intercellular space between the basal membrane and the
myocardial cell border (Figs. 7-8). Inflammatory cells were usually associated with these
changes (Figs. 9-10). Some cells within the myocardium or associated with endocardium also
had diffuse cell borders and highly lobulated nuclei by TEM. The membranes of several
mitochondria were swollen or lysed. Myocardial fibres were condensed in a few
cardiomyocytes. Just as in the compact layer, damaged areas also had evident budding across
cell membranes.
The inflammatory infiltrate in the pre-clinical phase mainly consisted of two populations of
cells. The most abundant cell type had dark, symmetrical or spherical nuclei surrounded by a
thin sheet of cytoplasm with pseudopods. The other cell type had larger, more elliptical and
less electron dense nuclei. The cytoplasm contained heteromorphous vacuoles, some
containing myofibril-like structures. Pseudopodia could also be observed in this cell type
within the cardiac lumen, but usually not in cells embedded in the tissue.
Clinical outbreak phase
This phase was defined as the period of increased mortality due to HSMI. In this period, the
number of moribund and dead fish was significantly higher than normal at all the sites where
the samples were collected. Most sampled fish had severe inflammatory changes in epi- endo-

and myocardium in both spongy and compact layers of the ventricle, although clinically
diseased fish tended to have more severe lesions than clinically normal individuals. Lesions in
the atrium were usually more moderate, but a few fish had more severe lesions in the atrium
than in spongy myocardium of the ventricle. Several fish also had additional changes in other
organs, especially red skeletal muscle, liver and spleen.
Both compact and spongy myocardium had widespread inflammation and associated
necrosis of myocytes in the histological sections (Figs. 11-12). Both cell populations observed
in the pre-clinical phase were present, but the cells with elliptical nuclei were now more
common. Also, several cells had large, asymmetrical or bean-shaped nuclei and amorphous
vacuoles in the cytoplasm. A small number of cells with spherical nuclei had their cytoplasm
densely packed with endoplasmic reticulum-like membranous structures. Within and close to
the epicardium, a few cells contained large electron dense cytoplasmic vacuoles.
Necrosis of myocardial cells was restricted to areas of inflammation. There was a high
number of lytic myocytes and budding activity across bordering membranes. Several
myocardial fibres lacked striation and were condensed, but even in damaged areas, some
myocardial fibres remained striated. A few fibres were hypercontracted. In these fibres,
myofibrils were also condensed or lysed. Some cells with these changes had viral-like
particles of unknown origin within the cytoplasm. The particles were 60-80 nm in diameter
and had an electron dense core surrounded by a less electron dense crust (Fig. 13).
Epicardial inflammation was severe, and was most abundant in the vicinity of coronary
vessels. The border between the epi- and myocardial layers was obscure, as inflammatory
cells were embedded in the myocardium just beneath the epicardium. Most of the fish
sampled during the clinical outbreak phase had a diffuse pattern of inflammation and
myocardial necrosis in the heart, and a possible association of myocardial lesions with
coronary vessels was more unclear than in the pre-clinical phase. However, when myocardial
11

changes were mild or moderate and multifocally distributed, perivascular inflammation was
still evident (Fig. 14). Inflammatory infiltrates of fish with moderate to severe cardiac lesions
contained a moderate number of PAS positive cells, predominantly in the epicardium. In
addition, a few necrotic cardiomyocytes stained slightly positive by PAS. However, no acid
resistant material could be observed by ZN staining, and no bacteria were observed by Gram
or Giemsa staining.
12
The spongy layer had marked endothelial hypertrophy, vacuolation and necrosis. These
changes were clearly associated with the presence of inflammatory cells. The most severely
affected fish had extensive lysis of endocardium, resulting in a direct contact between
myocytes and inflammatory cells in the cardiac lumen. Such myocytes generally had a rough
or obscure cell border, lysis of cytoplasmic contents and pyknosis. However, even though this
endocardial damage was frequently observed during the outbreak phase, thrombosis was not a
common finding in any of the sampled fish.
MSB staining revealed moderate fibrin deposits in the early outbreak phase, especially at
the border between the compact and spongy myocardium (Fig. 15). The few cellular thrombi
in the atrial and ventricular lumen also had moderate amounts of fibrin. Scattered foci of vG
positive extracellular matrix were present in areas with multifocal or diffuse inflammation in
the compact myocardium already at the start of the clinical outbreak. Bundles of collagen
fibres were present in these areas (Fig. 16). Equivalently damaged areas of the spongy
myocardium were vG negative, and did not contain visible collagen. During the course of the
outbreak, the number of vG positive areas in cardiac tissue increased, but was never
dominating in the sections. In the late outbreak phase, some distinct foci in both compact and
spongy myocardium were fibrotic, especially close to the border between the compact and
spongy layers of the ventricle. Following this, foci containting vG positive material became
more condensed, and the number of observable vG positive areas gradually decreased.

Characterisation of cells
Many of the cells described as inflammatory in this and previous papers (Kongtorp et al.
2004a; Kongtorp et al. 2004b; Kongtorp et al. 2006; Watanabe et al. 2006) did not react
positive to the histochemical markers used. When positive, the cells were scattered in the
tissues, and there were no observable differences between cells in perivascular regions,
subendocardially or within the myocardium of the heart. Only a few cells in the infiltrates
were postive to alkaline phosphatase during the mid-outbreak phase (Fig. 17). No cells were
positive to acid phosphatase or peroxidase. Myocardial cells and red skeletal myocytes were
strongly positive to unspecific esterase, but white skeletal muscle was negative. However,
presumed inflammatory cells were negative to unspecific esterase. A moderate number of the
cells within cardiac tissue were positive to MCH class II antisera (Fig. 18). MHC class II +
cells were mainly present in the most damaged areas, escpecially where cells could be
observed in distinct foci. A small to moderate number of cells were positive to IgM, but no
particular distribution pattern was evident. The cells had no reactivity for fibronectin.
Discussion
The present study has shown that myocarditis and associated necrosis characteristic of HSMI
may be preceeded by an initial stage of perivascular inflammation around coronary vessel
branches in epicardium and compactum as well as endothelial changes in spongiosum. This is
consistent with corresponding findings from recent infection experiments (Kongtorp &
Taksdal, submitted manuscript). Previous field studies have not been able to identify a starting
point of the pathogenesis, and the observations from this study might contribute to a further
understanding of the development of HSMI. Further, the results indicate that fibrosis does not
play a major role in the disease process, even in severely affected fish. This supports a
13

previous assumption that complete recovery is a possible outcome of HSMI (Kongtorp et al.
2004b).
14
In material from the longitudinal study, inflammatory cells were present around coronary
vessel branches in epicardium and in the compact layer of the cardiac ventricle before
myocardial damage could be observed by light microscopy. Some fish also had an abundance
of cells in the lumen of coronary vessel branches, but this may represent an unspecific
response in the fish. However, myocardial changes in the compact layer were first observed
adjacent to coronary vessels, suggesting that the initial insult to this tissue may occur in
association with vessels. Diffuse inflammatory infiltration of the epicardium was also
observed from an early stage. Epicarditis is a relatively common finding in farmed Atlantic
salmon and especially in fish with diseases affecting the heart. However, the mechanisms
leading to this condition are unclear (Poppe & Ferguson 2006). In the clinical outbreak phase
of HSMI, a more severe epicarditis is typical (Kongtorp et al. 2004b). The areas surrounding
coronary vessels in the epicardium were particularly affected in the present study, both in the
pre- and mid-clinical phases. It is therefore possible that epicardial and perivascular changes
are closely associated. The possible association between these changes and later occurrence of
HSMI is supported by 1) the presence of experimentally transmissible material from fish
sampled two months before the clinical outbreak, 2) the similar pattern in cardiac lesion
development in infection experiments (Kongtorp & Taksdal, submitted manuscript), and 3)
the increasing severity of these lesions towards the clinical outbreak phase (Kongtorp et al.
2006).
Although strongly suspected for the early myocardial changes, an association with
perivascular infiltrates was not as evident in fish with more severe lesions. The cause of this is
not known, but once started, it is possible that the pathological process in the myocardium is
self-sustaining. This could be due to virus propagation within myocardial cells, to leakage of

toxic substances from necrotic or inflammatory cells, or due to autoimmunity (Mitchell &
Cotran 1997; Huber 1997). Perivascular cell infiltrations were also seen in liver and red
skeletal muscle of a few fish, but there were no indications of generalised vessel-associated
changes. It therefore appears that the perivascular changes were heart-specific.
While not observed by light microscopy, early endothelial necrosis in coronary vessel
branches were observed by TEM in two fish, indicating that vascular damage may indeed be
part of the early stages of the pathogenesis. These fish also had degenerative changes in
cardiomyocytes that could not be linked to a presence of infitrating cells. It is uncertain
whether these changes were actually associated with HSMI, as they were not observed in all
sampled fish. However, it is known from other species that inflammatory cells, such as
macrophages, may induce vascular damage (Sundy & Haynes 1999). The observed lesions
may therefore be the result of an immune response, instead of preceeding it. This may be
explained by a possible presence of inflammatory cells above or below the actual sections. It
may also represent a part of the disease process which is limited in time or subtle enough to
be missed on histological examination. On the other hand, it may represent changes not
caused by HSMI at all. As reported previously, the longitudinal study of HSMI showed that
farmed fish are exposed to many threats during the production cycle (Kongtorp et al. 2006).
Lesions in apparently healthy fish sampled in the field are therefore not necessarily part of the
HSMI development. In diseased fish during the mid-clinical phase in the present study,
myocardial necrosis consistently occurred in association with inflammation.
Lesions in the spongy layer were also observed from initial stages, but as there are no
coronary vessels in this tissue (Farrell 2002), it is not very likely that the process in this part
of the heart may have been associated with epicarditis. It was evident that the endocardium
was involved from an early stage, as hypertrophy, vacuolation and separation from the
underlying myocardium were frequently observed. It is possible that some of the early
15

endocardial changes were a result of post mortal autolytic processes. However, the lesions
were mainly associated with the presence of inflammatory cells, suggesting that they were
part of the pathological development. Also, inflammation in spongy myocardium in the pre-
clinical phase was closely associated with endocardial changes. In the mid-clinical phase, the
inflammation and tissue lesions were more diffuse, and few normal areas were left for
comparison. Future studies may clarify the association of the observed changes with HSMI.
16
A high degree of fibrosis was not seen, although myocardial inflammation-associated
necrosis was a common observation during the mid-clinical phase. The collagen content in
damaged areas increased during the mid-clinical phase. However, in fish studied at the end of
the clinical phase, only few and scattered foci contained visible collagen. Additional
examination of material from the longitudinal study indicated a decreasing frequency of vG +
areas in the hearts of fish collected from the post-clinical phase. Also, the final samples prior
to slaughter did not contain visible vG + areas (RT Kongtorp, unpublished results). This
suggests that formation of scar tissue is not a dominant part of the HSMI-related changes.
Also, these findings support the idea that the fish heart may regenerate following an outbreak
of HSMI, and that they possibly clear fibrotic tissue in the healing process. It is known that
the zebrafish heart is capable of regeneration following experimental resection (Poss, Wilson
& Keating 2002). Similar processes may occur following insults leading to other lesions.
Many of the cells described as inflammatory cells in this and previous papers (Kongtorp et
al. 2004a; Kongtorp et al. 2004b; Kongtorp et al. 2006; Watanabe et al. 2006) were
surprisingly non-responsive and did not react as anticipated in the histochemical and
immunohistochemical examinations applied in the present study (Koppang, Fisher, Satoh &
Jirillo 2007). Observations of reactivity for alkaline phosphatase and MHC class II indicated
that some cells may have been macrophages, but this was obscured by the negative results for
acid phosphatase or unspecific esterase (Press, Dannevig & Landsverk 1994). Cells being

positive to MCH class II, but not to IgM may also have represented T lymphocytes and
immature B lymphocytes (Koppang et al. 2003). The presence of some IgM + cells indicated
that a fraction of the cell infiltrate may have consisted of plasma cells.
It is possible that the cells accumulating in the heart and red skeletal muscle of fish with
HSMI are not inflammatory cells. However, the morphological and histochemical
observations in the present study did not provide evidence for a widespread occurrence of
other known cell populations. The cells were negative for fibronectin, suggesting that
fibroblasts were not present in large numbers. Also, the observed cells were morphologically
similar to that reported from other salmon diseases described as inflammatory conditions, for
instance CMS and PD (Ferguson et al. 1990; Taksdal et al. 2007). Different pathogens, and
especially virus, can interfere with the phenotype of leukocytes. For instance, some viruses
are known to downregulate the expression of MHC molecules, which also has been suggested
as a mechanism in fish (Koppang, Dannevig, Lie, Rønningen & Press 1999). Further, viruses
may use a cocktail of other strategies to downregulate immune responses (Amati, Passer,
Lippolis, Lio, Caruso, Jirillo & Covelli 2006), all of which are largely unexplored in piscine
medicine. It is therefore not unlikely that the causal agent of HSMI may possess
immunoregulatory properties.
Cells observed in HSMI may alternatively represent proliferating and abnormal cells of
unknown origin, for instance endothelial cells or myocytes with abnormal phenotype caused
by the disease. The heterogenous appearance of the infiltrates and the marked nuclear
lobulation in several cells may provide some support for this hypothesis. Retroviruses,
herpesviruses and adenoviruses may have the ability to induce tumors in mammals (Murphy,
Gibbs, Horiznek & Studdert 1999), but only retroviruses have so far been associated with
epizootic tumors in fish (Martineau & Ferguson 2006). However, degenerative changes in
myo- and endothelial cells in the present study was highly associated with the presence of cell
17

accumulations, especially during the mid-clinical phase. The abnormal cellular changes may
therefore also reflect normal processes occuring during necrosis or apoptosis.
18
As inflammatory cells of Atlantic salmon are poorly characterised as compared to that of
mammals, broader investigations are needed to elucidate questions regarding the origin and
classification of the cell accumulations in HSMI and other fish diseases affecting Atlantic
salmon. At present we therefore find it appropriate not to change the established nomenclature
currently used to describe the inflammatory conditions in Atlantic salmon.
Acknowledgements
This study was funded by the Norwegian Research Council, grant no. 153000/120. We would
like to thank H B Nilsen and E Engeland for technical assistance. Thanks to L Aune for
excellent guidance in the immunohistochemical techniques. Many thanks also to A Kvellestad
and M Gjessing for valuable discussions.
References
Amati, L., Passer, M. E., Lippolis, D., Lio, C., Caruso, E., Jirillo, E., & Covelli, V. (2006)
Taking advantage of viral immune evasion: Virus-derived proteins represent novel
biopharmaceuticals. Current Medicinal Chemistry 13, 325-333.
Bancroft, J. D. & Stevens, A. (1990) Theory and practice of histological techniques. Churchill
Livingstone, London.
Eliassen, T. M., Solbakk, I. T., Evensen, Ø., & Gravningen, K. (2004) Isolation of heart and
skeletal muscle inflammation virus (HSMIV) from salmon. The 6th International
Symposium on Viruses of Lower Vertebrates, Hokkaido, Japan, p 56.
Falk, K., Press, C. M., Landsverk, T., & Dannevig, B. H. (1995) Spleen and kidney of
Atlantic salmon (Salmo salar L.) show histochemical changes in the course of
experimentally induced infectious salmon anaemia (ISA). Veterinary Immunology and
Immunopathology 49, 115-126.
Farrell, A. P. (2002) Coronary arteriosclerosis in salmon: growing old or growing fast?
Comparative Biochemistry and Physiology A-Molecular & Integrative Physiology
132, 723-735.

Ferguson, H. W., Poppe, T., & Speare, D. J. (1990) Cardiomyopathy in farmed Norwegian
salmon. Diseases of Aquatic Organisms 8, 225-231.
Huber, S. A. (1997) Autoimmunity in myocarditis: Relevance of animal models. Clinical
Immunology and Immunopathology 83, 93-102.
Kongtorp, R. T., Halse, M., Taksdal, T., & Falk, K. (2006) Longitudinal study of a natural
outbreak of heart and skeletal muscle inflammation in Atlantic salmon Salmo salar L.
Journal of Fish Diseases 29, 233-244.
Kongtorp, R. T., Kjerstad, A., Guttvik, A., Taksdal, T., & Falk, K. (2004a) Heart and skeletal
muscle inflammation in Atlantic salmon, Salmo salar L.: a new infectious disease.
Journal of Fish Diseases 27, 351-358.
Kongtorp, R. T., Taksdal, T., & Lyngøy, A. (2004b) Pathology of heart and skeletal muscle
inflammation (HSMI) in farmed Atlantic salmon Salmo salar. Diseases of Aquatic
Organisms 59, 217-224.
Koppang, E. O., Dannevig, B. H., Lie, Ø., Rønningen, K., & Press, C. M. (1999) Expression
of MHC class I and II mRNA in a macrophage-like cell line (SHK-1) derived from
Atlantic salmon, Salmo salar L., head kidney. Fish & Shellfish Immunology 9, 473-
489.
Koppang, E. O., Fisher, U., Satoh, M., & Jirillo, E. (2007) Inflammation in fish as seen from a
morphological point of view with special reference to the vascular compartment.
Current Pharmaceutical Design 13, 3649-3655.
Koppang, E. O., Hordvik, I., Bjerkås, I., Torvund, J., Aune, L., Thevarajan, J., & Endresen, C.
(2003) Production of rabbit antisera against recombinant MHC class II B chain and
identification of immunoreactive cells in Atlantic salmon (Salmo salar). Fish &
Shellfish Immunology 14, 115-132.
Martineau, D., & Ferguson, H.W. (2006) Neoplasia. In:Systemic pathology of fish. A text and
atlas of normal tissues in teleosts and their responses in disease (Ed. by H. W.
Ferguson), pp. 313-334. Scotian Press, London.
McLoughlin, M. F. & Graham, D. A. (2007) Alphavirus infections in salmonids - a review.
Journal of Fish Diseases 30, 511-531.
McLoughlin, M. F., Nelson, R. N., McCormick, J. I., Rowley, H. M., & Bryson, D. B. (2002)
Clinical and histopathological features of naturally occurring pancreas disease in
farmed Atlantic salmon, Salmo salar L. Journal of Fish Diseases 25, 33-43.
Mitchell, R.N., & Cotran, R.S. (1997) Acute and chronic inflammation. In:Basic Pathology
(Ed. by V. Kumar, S. Cotran & R. S. Robbins), pp. 25-46. W.B. Saunders Company,
Philadelphia.
Murphy, F. A., Gibbs, E. P. J., Horiznek, M. C. & Studdert, M. J. (1999) Veterinary virology.
629 pp. Academic Press, San Diego.
19

Poppe,T.T., & Ferguson, H.W. (2006) Cardiovascular system. In:Systemic pathology of fish.
A text and atlas of normal tissues in teleosts and their responses in disease (Ed. by H.
W. Ferguson), pp. 141-167. Scotian Press, London.
Poss, K., Wilson, L., & Keating, M. (2002) Heart regeneration in zebra fish. Science 298,
2188-2190.
Press, C. M., Dannevig, B. H., & Landsverk, T. (1994) Immune and enzyme histochemical
phenotypes of lypmoid and nonlymphoid cells within the spleen and head kidney of
Atlantic salmon (Salmo salar L.). Fish & Shellfish Immunology 4, 79-93.
Skjelstad, H. R., Bornø, G., Flesjå, K., Hansen, H., Nilsen, H., Wasmuth, M. A., & Hjeltnes,
B. (2008) Helsesituasjonen hos laksefisk 2007 (Diseases in farmed salmonids 2007)
(in Norwegian).
Sundy, J.S., & Haynes, B.F. (1999) Vasculitis: Pathogenic mechanisms of vessel damage.
In:Inflammation. Basic principles and clinical correlates (Ed. by J. I. Gallin & R.
Snyderman), pp. 995-1016. Lippincott Williams & Wilkins, Philadelphia.
Taksdal, T., Olsen, A. B., Bjerkås, I., Hjortaas, M. J., Dannevig, B. H., Graham, D. A., &
McLoughlin, M. F. (2007) Pancreas disease in farmed Atlantic salmon, Salmo salar
L., and rainbow trout, Oncorhynchus mykiss (Walbaum), in Norway. Journal of Fish
Diseases 30, 545-558.
Watanabe, K., Karlsen, M., Devold, M., Isdal, E., Litlabø, A., & Nylund, A. (2006) Virus-like
particles associated with heart and skeletal muscle inflammation (HSMI). Diseases of
Aquatic Organisms 70, 183-192.
20

Legends to figures:
Plate 1:
Fig. 1. Heart, Atlantic salmon, Salmo salar. Pre-clinical phase of heart and skeletal muscle
inflammation. Concentration of inflammatory cells in the lumen of a coronary vessel branch
within compact myocardium (arrow). HE. Bar = 20 m.
Fig. 2. Heart, Atlantic salmon, Salmo salar. Pre-clinical phase of heart and skeletal muscle
inflammation. Transmission electron micrograph of inflammatory cells residing in the lumen
of a coronary vessel branch (arrow). Bar = 5 m.
Fig. 3. Heart, Atlantic salmon, Salmo salar. Pre-clinical phase of heart and skeletal muscle
inflammation. Inflammatory cell infiltration (arrow) surrounding a coronary vessel branch
(asterisk) within compact myocardium. HE. Bar = 20 m.
Fig. 4. Heart, Atlantic salmon, Salmo salar. Pre-clinical phase of heart and skeletal muscle
inflammation. Early myocardial lesions in association with a coronary vessel branch in
compact myocardium, with cellular lysis and accumulation of membranous structures (arrow).
Bar = 5 m.
Fig. 5. Heart, Atlantic salmon, Salmo salar. Pre-clinical phase of heart and skeletal muscle
inflammation. Hypertrophic nuclei in the endocardium lining spongy myocardium (arrows).
HE. Bar = 20 m.
Fig. 6. Heart, Atlantic salmon, Salmo salar. Pre-clinical phase of heart and skeletal muscle
inflammation. Ultrastructure of hypertrophic endothelial cells in the endocardium. Bar = 2
m.
21

Plate 2:
Fig. 7. Heart, Atlantic salmon, Salmo salar. Pre-clinical phase of heart and skeletal muscle
inflammation. Vacolation of endothelial cells in the endocardium (asterisk). Basal lamina is
indicated by an arrow. Bar = 2 m.
Fig. 8. Heart, Atlantic salmon, Salmo salar. Pre-clinical phase of heart and skeletal muscle
inflammation. Separation of endocardium from spongy myocardium, creating an extracellular
space beneath endocardial cells (asterisk). Degenerative changes in endothelial cell in the
endocardium (arrow) and underlying myocardium. Bar = 5 m.
Fig. 9. Heart, Atlantic salmon, Salmo salar. Pre-clinical phase of heart and skeletal muscle
inflammation. Inflammatory cell infiltration in the space between endocardium and spongy
myocardium (arrow). HE. Bar = 20 m.
Fig. 10. Heart, Atlantic salmon, Salmo salar. Pre-clinical phase of heart and skeletal muscle
inflammation. Ultrastructural changes associated with endocardium. Various cells (arrows)
reside in a sub-endocardial space. In other areas, necrotic endothelial cells (asterisk) are
observed. Bar = 10 m.
Fig. 11. Heart, Atlantic salmon, Salmo salar. Inflammation (arrows) and vacuolation
(asterisk) in compact myocardium in fish with heart and skeletal muscle inflammation. HE.
Bar = 20 m.
Fig. 12. Heart, Atlantic salmon, Salmo salar. Inflammation (arrows) and myocardial necrosis
(asterisk) in spongy myocardium in fish with heart and skeletal muscle inflammation. HE. Bar
= 20 m.
22

Plate 3:
Fig. 13. Heart, Atlantic salmon, Salmo salar. Virus-like particles observed in necrotic
cardiomyocytes of fish with heart and skeletal muscle inflammation in this study were 60-80
nm in diameter and had an electron dense core surrounded by a less electron dense crust. Bar
= 200 nm.
Fig. 14. Heart, Atlantic salmon, Salmo salar. Perivasculitis (arrow) of a coronary vessel
branch (asterisk) in fish with heart and skeletal muscle inflammation. Inflammation may also
be observed in surrounding myocardium. HE. Bar = 50 m.
Fig. 15. Heart, Atlantic salmon, Salmo salar. Fibrin deposits (red) and collagen (blue) may be
observed at the border between compact and spongy myocardium in the ventricle in fish with
heart and skeletal muscle inflammation. Martius Scarlet Blue. Bar = 50 m.
Fig. 16. Heart, Atlantic salmon, Salmo salar. Collagen fibres (arrow) near condensed
myofibrils in compact myocardium in fish with heart and skeletal muscle inflammation. Bar =
500 nm.
Fig. 17. Heart, Atlantic salmon, Salmo salar. Only a few scattered cells were alkaline
phosphatase + in the cell accumulations of myocardium in fish with heart and skeletal muscle
inflammation (arrow). Counterstained with hematoxyllin. Bar = 50 m.
Fig. 18. Heart, Atlantic salmon, Salmo salar. A moderate number of cells in the myocardium
of fish with heart and skeletal muscle inflammation were MHC class II + (arrows).
Counterstained with hematoxyllin. Bar = 50 m. Bar of inserted micrograph = 20 m.
23

Plate 1
24

Plate 2
25

Plate 3
26

Paper V
Studies with experimental transmission of heart and skeletal
muscle inflammation in Atlantic salmon, Salmo salar L.

Studies with experimental transmission of heart and skeletal muscle
inflammation in Atlantic salmon, Salmo salar L.
R T Kongtorp and T Taksdal
National Veterinary Institute, Oslo, Norway
Correspondence R T Kongtorp, Section for Epidemiology, National Veterinary Institute, Pb.
750 Sentrum, 0106 Oslo, Norway (e-mail: ruth-torill.kongtorp@vetinst.no)
Running title
Experimental studies with heart and skeletal muscle inflammation
1

Abstract
Heart and skeletal muscle inflammation (HSMI) is a transmissible disease causing mortality
in farmed Atlantic salmon, Salmo salar L. It is characterised by epi-, endo- and myocarditis
and myocardial necrosis, as well as myositis and necrosis of red skeletal muscle. The present
paper describes two infection experiments, with the aim of further exploring the infectivity
and pathogenesis of HSMI. In both experiments, Atlantic salmon were intraperitonally
injected with tentatively infective material. The first experiment was carried out in freshwater,
using cardiac tissue, blood plasma and cell cultured material as inoculates. In the second
experiment, various tissues sampled from fish in the mid-outbreak phase were used to
inoculate experimental fish in seawater. Also, cardiac tissue sampled before and after the
outbreak phase was used. Finally, cardiac tissue pre-treated with chloroform was tested. In
both experiments, all inoculates resulted in cardiac inflammation during the study period of
eight weeks. Early cardiac changes included perivasculitis and endocarditis, which were
observed from 1-3 weeks post challenge (p.c.). Focal myocarditis first appeared three weeks
p.c., and the number of fish showing myocardial changes at eight weeks p.c. was high in all
groups. A possible mechanism for the development of HSMI is discussed.
Keywords
Atlantic salmon, heart and skeletal muscle inflammation, transmission, plasma, cell culture,
chloroform
2

Introduction
Heart and skeletal muscle inflammation (HSMI) is a severe disease in farmed Atlantic
salmon, Salmo salar L. (Kongtorp, Taksdal & Lyngøy 2004b). From its discovery in diseased
fish in Norway in 1999, HSMI has become an increasing problem for the Atlantic salmon
farming industry in that country. A suspected case has also been observed in Scotland, but the
association of this disease outbreak with pancreas disease (PD) is yet to be clarified
(Ferguson, Kongtorp, Taksdal, Graham & Falk 2005). HSMI is characterised by epi-, endo-
and myocarditis and myocardial necrosis, as well as myositis and necrosis of red skeletal
muscle (Kongtorp et al. 2004b). In natural cases of HSMI the morbidity of lesions may be
very high, while the mortality varies considerably between outbreaks (Kongtorp et al. 2004b;
Kongtorp, Halse, Taksdal & Falk 2006). The histopathological lesions observed in HSMI may
resemble those found in fish with cardiomyopathy syndrome (CMS) (Ferguson, Poppe &
Speare 1990; Bruno & Poppe 1996) and PD (Ferguson, Roberts, Richards, Collins & Rice
1986; McLoughlin, Nelson, McCormick, Rowley & Bryson 2002; McLoughlin & Graham
2007). In typical cases, however, HSMI is clearly distinguishable from these diseases
(Kongtorp et al. 2006).
The cause of HSMI is still unclear, but the tissue lesions have been successfully transmitted
to experimental fish in seawater by intraperitoneal (i.p.) injection of cardiac tissue
homogenate from diseased fish and cohabitation to injected fish (Kongtorp, Kjerstad, Guttvik,
Taksdal & Falk 2004a). No pathogens have been firmly associated with HSMI, although there
have been several attempts to isolate and characterise a causal agent (Eliassen, Solbakk,
Evensen & Gravningen 2004; Kongtorp et al. 2004a; Kongtorp et al. 2006; Watanabe,
Karlsen, Devold, Isdal, Litlabø & Nylund 2006).
Little is known about introduction of HSMI in the field, the pathogenesis and risk factors
for clinical disease outbreaks. A longitudinal study of HSMI showed that cardiac changes
3

may be observed several months before and after a clinical outbreak, suggesting a slow and
prolonged disease development (Kongtorp et al. 2006). Further investigation of material
sampled in that study indicated that myocarditis may possibly be preceeded by perivasculitis
of coronary vessel branches, as well as endocardial changes (Kongtorp et al., unpublished
results).
4
The present paper describes two new infection experiments, with the aim of further
exploring the infectivity and pathogenesis of HSMI. Infectivity was examined by inoculation
of tissue material from several organs and blood plasma collected during an outbreak of
HSMI, as well as heart tissues from apparently healthy fish collected prior to and after an
outbreak of HSMI. Also, inoculates made from cell cultured material and cardiac tissue pre-
treated with chloroform were examined. Based on these experiments the chronological
development of cardiac lesions under experimental conditions was studied.
Materials and methods
Preparation of inoculates
Fish used for preparation of inoculates were examined by necropsy and histopathology. To
exclude possible bacterial infections, material from kidney tissue was streaked onto blood
agar with and without the presence of 2 % NaCl. The presence of salmonid alphavirus (SAV)
was also excluded by examination of selected cardiac tissue samples according to a standard
real-time RT-PCR protocol used for diagnostics at the National Veterinary Institute (NVI).
Tissue homogenates: Material from heart, liver and pooled kidney/spleen were homogenised
and diluted 1:10 in Leibowitz’ L-15 cell culture medium, centrifuged at 2500 g for 10 min.
The supernatant was pipetted off and frozen at -80 ºC until used. Before commencement of
the experiments, the supernatant was defrosted and further diluted 1:2 in L-15 containing
gentamycin (final concentration 50 g mL -1 ).

Blood plasma: Blood from moribund Atlantic salmon sampled during ongoing field outbreaks
of HSMI was collected from the caudal vein into vacutainer glass tubes containing heparin.
The HSMI diagnosis of each sampled fish was confirmed by histology. Blood samples were
centrifuged at 5000 g for 10 minutes, and plasma was pipetted off and frozen at -80 ºC until
used. For preparation of the experimental inoculate, blood plasma was defrosted and diluted
1:2 in L-15.
Chloroform pre-treatment: Cardiac tissue was homogenised, centrifuged an diluted as
described above. One unit of chloroform was thereafter added to two units of supernatant. The
mixture was vortexed for 10 min at room temperature and thereafter centrifuged at 2000 x g
for 10 min (Falk, Namork, Rimstad, Mjaaland & Dannevig 1997; Kvellestad, Dannevig &
Falk 2003).
Material from cell culture: Cell cultures from the lines of epithelioma papulosum cyprini
(EPC) and fat head minnow Pimephales promelas (FHM) were grown at 20 ºC in L-15
supplemented with 10% foetal calf serum (FCS), L-glutamine (4 mM) and gentamycine (50
μg ml -1 ). Mid kidney from a transmission trial in seawater (Kongtorp et al. 2004a) was used
to inoculate cells, in order to propagate material for inoculation of experimental fish. Samples
originated from fish that had been i.p. injected with material from cardiac tissue and sampled
four and eight weeks post challenge (p.c.). Tissue material from these fish was diluted 1:10 in
L-15, homogenized and centrifuged at 1000 g for 10 min. Neutralising antiserum against N1
serotype of IPNV (mAb N1-H8 (Christie, Ness & Djupvik 1990), kindly provided by KE
Christie, Intervet Norbio AS) was diluted 1:100 and added to the tissue suspension. This was
done to inhibit replication of infectious pancreatic necrosis virus (IPNV) in the cell cultures,
as this organism is ubiquitous in Norwegian salmon farms (Krogsrud, Håstein & Rønningen
1989). Cells were inoculated with 1 ml of the resulting supernatant for 2 h at 15 ºC, and
thereafter removed, but not washed off. Following this, 5 ml L-15 supplemented with 2%
5

FCS, L-glutamine (4 mM) and gentamycine (50 μg ml -1 ) were added to the flasks, and the
cells were incubated at 15 ºC. Inoculated cell cultures were blind passaged after 14 days and
incubated at 15 ºC for another 14 days. After inoculation or passage they were inspected daily
for four days and thereafter every three days. In cell cultures showing plaque-like foci and
loss of cells (Fig 1 A-B), supernatant from the second passage was sampled, diluted 1:2 and
inoculated into experimental fish.
Negative control inoculates: In experiment 1, L-15 supplemented with 10% foetal calf serum
(FCS), L-glutamine (4 mM) and gentamycine (50 μg ml -1 ) was injected into control fish. In
experiment 2, two negative control inoculates were used. Cardiac tissue from healthy Atlantic
salmon was diluted, homogenised and centrifuged as described above. Half of the resultant
supernatant was directly injected into control fish, and half of the supernatant was pre-treated
with chloroform by the method described above.
Experiment 1
The first experiment was performed at the joint experimental research facility of the NVI and
School of Veterinary Science, Oslo. Healthy Atlantic salmon from a commercial smolt
producer with no history of prior disease were used as experimental fish. The smolt producer
was located in an area from which no outbreaks of HSMI nor any important differential
diagnoses such as PD or CMS have been reported. The fish had been subjected to a mixture of
freshwater and seawater at the smolt producing site, and were undergoing smoltification, most
being silver grey without parr markings. On commencement of the infection experiment the
fish had an average weight of 40.0 g. They had been vaccinated with a commercial
multivalent vaccine. The experiment was performed in dechlorinated freshwater at a
temperature of 12 ºC (fluctuation ± 1 ºC). A total of 390 fish were placed in four tanks, each
measuring 60 x 60 x 100 cm. During the experiment, the fish were handfed twice daily with a
6

commercial pelleted feed. Experimental fish were allocated randomly to five groups,
anaesthetised in chlorine buthanol (300 mg L -1 ) and i.p. injected with an inoculum of 0.15
mL. The groups are listed in table 1. To separate the two inoculation regimes in tank 3, the
adipose fin was clipped in one of the groups.
Experiment 2
The second experiment was performed at VESO Vikan, Namsos. Non-vaccinated healthy
Atlantic salmon from the smolt production site at VESO Vikan were used as experimental
fish. Smolts were produced in an area where PD has not been reported. The fish were
transferred to seawater one week before commencement of the study, and had an average
weight of 49.6 g. The salinity was 32.1 ‰ on average (range 29.9-33.8 ‰), and the
temperature was 12 ºC (fluctuation ± 0.1 ºC). A total of 540 fish were placed in seven tanks,
each measuring 100 x 100 x 50 cm. During the experiment, the fish were automatically fed
with a commercial pelleted feed.
Experimental fish were allocated randomly to nine groups, each containing 60 fish,
anaesthetised in chlorine buthanol (300 mg L -1 ) and i.p. injected with an inoculum of 0.2 mL.
The groups are listed in table 1. To separate the inoculation regimes in tanks 2 and 7, the
adipose fin was clipped in one of the groups.
Sampling
Samples for histology and microbiology were collected from five fish before commencement
of the studies and regularly from 5-7 fish thereafter. The sampling intervals and number of
fish sampled are listed in table 1. At the sampling dates, experimental fish from each group
were randomly collected, anaesthetised in chlorine buthanol (300 mg L -1 ) and killed by
7

leeding. Samples were taken according to the procedures described by Kongtorp et al
(Kongtorp et al. 2004a).
8
Tissue samples from several organs were collected and prepared for histology and
transmission electron microscopy. For the purpose of this study, cardiac samples were
examined histologically, and fish were classified as diseased or non-diseased based on the
presence or absence of cardiac lesions consistent with HSMI (Kongtorp et al. 2004b). The
criteria followed were observations of epi-, endo- and myocarditis and inflammation-
associated necrosis in compact and spongy layers of the cardiac ventricle.
In experiment 1, material from mid kidney was cultivated on blood agar for bacteriological
exmination, from all sampled fish. In experiment 2, bacteriological samples were only taken
from fish dying during the study. For virological examination, heart and kidney tissue were
frozen at -80ºC. Selected tissue samples were later defrosted and inoculated onto EPC and
FHM cell cultures according to the procedures described above. Cell cultures were passaged
at least 3 times.
Transmission electron microscopy
Cell monolayers showing changes following inoculation with tissue from experimental fish in
experiment 1 were fixed in 3% glutaraldehyde in 0.1 M sodium cacodylate buffer (CB) for 3
days, washed 3 times in CB and covered with 2% osmium tetroxyde (OsO4) in CB for 2
hours. After washing 3 times in CB, the cells were taken through a dehydration process with
graded ethanols. Cells were then infiltrated by a 1:3 solution of Lx-112 (Epon-equivalent
epoxy monomers) and 100 % ethanol, a 1:1 solution and a 3:1 solution, the latter being left
over night. Pure Lx-112 was added, and the cell material was incubated at 37 ºC for 24 hours
and 60 ºC for 4 days. Cardiac tissue from experimental fish showing histopathological lesions
was fixed and prepared by a similar method substituting ethanol with 1, 2-propylene oxide in

the infiltration stages. Specimens were prepared for electron microscopy by ultra thin slicing
and examined in a Phillips EM 208 S transmission electron microscope (TEM).
Results
Fish from all challenged groups developed epi-, endo- and myocarditis and inflammation-
associated necrosis consistent with HSMI during the experiment (Tables 2 and 3). Inoculate
material sampled from heart, liver, kidney/spleen and blood plasma of fish in the mid-
outbreak phase were all highly infective. In addition, cardiac tissue from clinically normal
fish sampled two months before and two months after the outbreak-phase also caused cardiac
lesions in experimental fish. The chloroform-treated inoculate was infective, indicating that
the infectious agent does not have an outer lipid layer sensitive to chloroform. Cardiac lesions
were also observed in fish inoculated with cell cultured material. The frequency of lesions at
eight weeks p.c. was 80-90 % in most groups, but somewhat lower in fish injected with cell
cultured material from EPC cells. Cardiac lesions were milder in fish sampled in experiment 1
than observed in experiment 2. Fish in control tanks injected either with L-15 or with
homogenates of heart tissue from healthy fish did not have cardiac lesions similar to HSMI.
Early signs of cardiac lesions were infiltration of inflammatory cells around coronary
vessels in the epicardium and compact myocardium (Fig 2), as well as hypertrophy of spongy
endocardium and mild endocarditis. These lesions were observed from three weeks p.c in
experiment 1, and as early as one week p.c. in some groups in experiment 2. Epicarditis was
observed throughout the experiments in all groups, including the negative controls, but was
more severe and frequent in the challenged groups towards the end of the experiments. Focal
myocarditis first appeared in one fish inoculated with material cultured on FHM cells at three
weeks p.c. in experiment 1. In experiment 2, one fish from each of the groups injected with
cardiac tissue sampled from the mid-outbreak phase, from the pre-outbreak phase and
9

chloroform pre-treated cardiac tissue, also showed myocarditis at three weeks p.c. Myocardial
necrosis was consistently associated with inflammatory foci. The number of fish showing
cardiac inflammation and necrosis increased over time in both experiments, and was most
severe at eight weeks p.c. (Fig 3) . However, the frequency of lesions observed in fish
sampled before the eight-week time point was higher in experiment 1 than in experiment 2.
Negative control fish did not develop myocarditis, but in experiment 1, one negative control
fish showed mild endocarditis, and one had cell infiltration around a coronary vessel in
compact myocardium.
10
The mortality was low in both experiments. No fish died during experiment 1. Seven fish
died during experiment 2. The dead fish were found 1, 2, 4 and 7 weeks p.c. and were evenly
distributed across the groups. Due to the low number of dead fish these were not examined
histologically, but Vibrio sp. were isolated from kidney tissue of three fish.
Cell cultures inoculated with material from experimental fish showing cardiac changes
showed a pattern of cellular changes similar to the cell cultures used to inject fish in
experiment 1. One day after inoculation, EPC and FHM cells started forming small plaque-
like foci, and several cells showed evidence of swelling (Fig 1 A). On the second day of
incubation, degenerative plaques were usually larger and more distinct, and by the third day
the foci were more diffuse. From the fourth day and onward, the cells appeared indistinctively
changed and displayed a rough surface. Vacuolation was also observed in a few cells, but this
was not observed in all cell cultures. The cellular changes were observed for 2-4 passages, but
were rarely observed in later passages. Cells inoculated with material from negative control
fish did not show such changes and were similar to non-inoculated control cells (Fig 1 B).
Cell culture monolayers and cardiac tissue from experimental fish showing cardiac changes
were also examined by TEM, but no viral particles were identified.

Discussion
This study confirms and expands knowledge from earlier infection experiments and field
studies of HSMI. Until now, only heart tissue sampled during the mid-outbreak phase of
HSMI has been tested for infectivity (Kongtorp et al. 2004a). In the present study, blood
plasma and several tissues from clinically diseased fish were infective to naïve experimental
fish. These results show that the infection is not restricted to organs with pathological changes
during the disease outbreak. Moreover, the infectivity of hearts sampled 2 months before and
2 months after the the clinical outbreak phase suggests that the causal agent is also present in
fish before and after clinical outbreaks. The results also indicate that the causal agent may be
a nonenveloped virus.
Fish with HSMI show severe epi- myo and endocarditis during clinical outbreaks (Kongtorp
et al. 2004b). In the present study, early responses to injection of HSMI material were
infiltration of inflammatory cells around coronary vessel branches in epicardium and compact
myocardium, and mild endocarditis in spongiosum of the cardiac ventricle. These findings
may give important information for further studies of HSMI aetiology, development of
diagnostic tools for agent detection, experimental models for HSMI and biosecurity
management in farming of Atlantic salmon.
The early appearance of cardiac perivasculitis and endocarditis in the present study is
consistent with similar findings from a longitudinal study of a natural outbreak of HSMI
(Kongtorp et al., unpublished results), indicating that these lesions are part of the HSMI
development. A revisit to the sections from a previous infection experiment in seawater
(Kongtorp et al. 2004a), has later shown that these changes were also present from 2 weeks
p.c. in that study (RT Kongtorp, unpublished results). These findings may point towards a
previoulsy unknown mechanism in the pathogenesis of HSMI: The causal agent may possibly
not primarily infect cardiomyocytes, but rather endothelial or other cells associated with the
11

coronary vessel branches and the endocardium. Infective particles may thus be transported by
blood components and possibly infect these cells initially, thereafter inducing myocarditis.
Virus-like particles have previously been observed within affected myocardial cells
(Kongtorp et al., unpublished results; Watanabe et al. 2006), but there have been no reports of
such particles in endothelial cells. Also, vascular lesions have not been consistenly observed
in other organs in fish with HSMI (Kongtorp et al. 2004b; Kongtorp et al. 2006).
Perivasculitis and endocarditis may thus possibly be the result of an initial passage of
infective particles entering from the blood on its way to the nearby myocardium. This may be
clarified in future studies.
12
In the present study, the minimum time from challenge to observable myocardial
inflammation was three weeks. This is earlier than in the transmission trial from which the
material for the cell culture challenge was derived (Kongtorp et al. 2004a). In that study,
cardiac changes first appeared six weeks post challenge in the injected groups, and ten weeks
post challenge in the cohabitant groups. Some of the early changes may have been due to
other causes, for instance infectious bacteria circulating among the experimental fish. If that
was the case, however, the cardiac changes should be evenly distributed among the groups,
and not concentrated in the challenged groups. In experiment 1, bacteriological samples from
all fish were examined by standard methods, with negative results. In experiment 2, bacteria
were only isolated in three of the dead fish. Also, no bacteria were discovered by
histopathological examination of sampled fish. In addition, the similarity between findings in
the two experiments and with results from a field study (Kongtorp et al., unpublished results)
support the idea that the changes were associated with HSMI.
In the present study, the frequency of cardiac lesions in experimental fish was low until
seven or eight weeks p.c. This may possibly be due to propagation of the causal agent,
gradually causing an increased concentration of infective particles. Another explanation may

e a possible development of a self-sustaining immune response which contributes to further
tissue damage. Myocardial necrosis was consistently associated with the presence of
inflammatory cells, indicating that necrosis may be a secondary effect of inflammation. It is,
however, possible that chemotactic signals may be released from infected, but not visibly
damaged cells, and that degenerative and inflammatory changes therefore appear to concur.
These mechanisms should be studied further.
The successful transmission of HSMI by injection of blood plasma suggests that the causal
agent may be present in peripheral blood during the clinical outbreak. Given the most likely
routes of natural infection and shedding through the gill epithelium, intestine or epidermis, a
certain number of infective particles is expected to be present within peripheral blood during
infection. However, foreign material in the blood is quickly cleared by circulating leukocytes.
A high concentration of infective particles in blood plasma therefore indicates that particles
are constantly released from infected cells in the host tissue (Murphy, Gibbs, Horiznek &
Studdert 1999). In HSMI, viraemia therefore appears to concur with the period when infected
fish have the most severe lesions. In the future, a combination of histopathological
examination and specific antigen detection methods should therefore enable a precise
diagnosis of HSMI. The infectivity of blood plasma opens a possibility for development of
tools for non-lethal agent detection in peripheral blood. This may be useful in diagnostics and
epidemiological studies. Also, tissues from several organs and blood plasma may be used to
inoculate experimental fish in future challenge trials with HSMI.
In a population of fish, HSMI may potentially be transmissible to naïve fish months before
the disease is discovered. As experiment 2 showed, cardiac tissue was infective as early as
two months before the onset of mortalitites in a field outbreak, during the mid-clinical
outbreak phase, and as late as two months after this phase. Fish sampled for the preparation of
inoculates testing early and late infectivity did not show clinical signs of disease. It is
13

uncertain whether live fish are actually capable of shedding infective particles for many
months, but the causal agent appears to be present in cardiac tissue and infective for a very
long time. This may have implications for virus dispersal, at least from dead and decaying
fish.
14
The cardiac lesions observed in experiment 1 were significantly milder than what is
observed in natural outbreaks (Kongtorp et al. 2006), in a previous experimental trial in
seawater (Kongtorp et al. 2004a) and in the present experiment 2 in seawater. The number of
fish showing cardiac lesions consistent with HSMI at eight weeks p.c. in experiment 1 was,
however, similar to observations from earlier studies, and actually higher than in experiment
2. This indicates that Atlantic salmon is susceptible to HSMI both in the freshwater and
seawater stages, and that HSMI is a potential threat for the smolt production. Although HSMI
has been diagnosed at smolt production sites, natural cases of HSMI have almost exclusively
been associated with exposure to seawater (records from the NVI). The cause of this is not
known. There may perhaps be a higher degree of exposure to the causal agent when fish are in
the seawater phase, increasing the risk of infection. Another explanation is that infection in
the freshwater phase may possibly cause milder tissue lesions and lower mortality compared
to the seawater phase, thus going unnoticed.
Experiment 1 showed that whatever agent is causing HSMI, it was still infective after two
passages in cell culture. However, later TEM examination of these cell cultures and cardiac
tissue did not reveal any viral particles, and changes in inoculated cell cultures disappeared
after just a few passages. It is possible that only a few cells in the cultures were infected, and
that the inoculate consisted of a diluted, yet viable, version of the initial viral load.
Propagation of the agent in the present study may therefore mainly have taken place in the
natural host, possibly explaining why the number of fish being affected in the EPC group was
somewhat less than in the other groups. Earlier attempts to grow the causal agent of HSMI in

cell culture have been performed in cell cultures from Chinook salmon embryo (CHSE-214),
Blue-gill fry (BF-2), Rainbow trout gill (RT-gill) and Atlantic salmon head kidney (ASK-2)
(Kongtorp et al. 2004a; Kongtorp et al. 2004b; Watanabe et al. 2006). Neither of these cell
lines have proven to be more susceptible to the causal agent of HSMI nor sustain the infection
longer than the cell lines used in the present study. Until an appropriate cell line is found,
therefore, it may be difficult to identify the causal agent of HSMI.
In an attempt to characterise the causal agent to a certain extent, pre-treatment of the
inoculate with chloroform was attempted in experiment 2. This did not prevent transmission
of cardiac lesions, indicating that the causal agent is nonenveloped. The results from this part
of the experiment are in contrast to findings of apparently enveloped virus-like particles
previously suggested to be associated with HSMI (Eliassen et al. 2004; Watanabe et al.
2006). Further studies are therefore needed to test the repeatibility of these results. If the
causal agent of HSMI repeatedly shows resistance to chloroform, this will further confirm that
HSMI and PD are separate diseases.
In conclusion, results from this study show that fish farmers moving apparently healthy
Atlantic salmon during the seawater phase may take a significant risk for spreading HSMI to
new locations. Movement of fish of unknown infective status should therefore be avoided,
until further knowledge about the risks for agent introduction and dispersal is obtained.
Acknowledgements
This study was financed by the Norwegian Research Council (grant no 153000/120). Many
thanks to H Welde and M Heum for skilful technical assistance, and to K Falk for guidance in
the virology studies. Many thanks also to A Ramstad and her staff at VESO Vikan for
technical running of experiment 2.
15

References
Bruno D.W. & Poppe T.T. (1996) A colour atlas of salmonid diseases., pp. 141-142
Academic press, London.
Christie K.E., Ness S. & Djupvik H.O. (1990) Infectious pancreatic necrosis virus in Norway:
partial serotyping by monoclonal antibodies. Journal of Fish Diseases 13, 323-327.
Eliassen T.M., Solbakk I.T., Evensen Ø. & Gravningen K. (2004) Isolation of heart and
skeletal muscle inflammation virus (HSMIV) from salmon. Poster at the 6th
International Symposium on Viruses of Lower Vertebrates, Hokkaido, Japan, p 56.
Falk K., Namork E., Rimstad E., Mjaaland S. & Dannevig B.H. (1997) Characterization of
infectious salmon anaemia virus, an orthomyxo-like virus isolated from Atlantic
salmon, (Salmo salar L.). Journal of Virology 71, 9016-9023.
Ferguson H.W., Kongtorp R.T., Taksdal T., Graham D. & Falk K. (2005) An outbreak of
disease resembling heart and skeletal muscle inflammation in Scottish farmed salmon,
Salmo salar L., with observations on myocardial regeneration. Journal of Fish
Diseases 28, 119-123.
Ferguson H.W., Poppe T. & Speare D.J. (1990) Cardiomyopathy in farmed Norwegian
salmon. Diseases of Aquatic Organisms 8, 225-231.
Ferguson H.W., Roberts R.J., Richards R.H., Collins R.O. & Rice D.A. (1986) Severe
degenerative cardiomyopathy associated with pancreas disease in Atlantic salmon,
Salmo salar L. Journal of Fish Diseases 9, 95-98.
Kongtorp R.T., Halse M., Taksdal T., & Falk K. (2006) Longitudinal study of a natural
outbreak of heart and skeletal muscle inflammation in Atlantic salmon Salmo salar L.
Journal of Fish Diseases 29, 233-244.
Kongtorp R.T., Kjerstad A., Guttvik A., Taksdal T. & Falk K. (2004a) Heart and skeletal
muscle inflammation in Atlantic salmon, Salmo salar L.: a new infectious disease.
Journal of Fish Diseases 27, 351-358.
Kongtorp R.T., Taksdal T. & Lyngøy A. (2004b) Pathology of heart and skeletal muscle
inflammation (HSMI) in farmed Atlantic salmon Salmo salar. Diseases of Aquatic
Organisms 59, 217-224.
Krogsrud J., Håstein T. & Rønningen K. (1989) Infectious pancreatic necrosis virus in
norwegian fish farms. In:Viruses of lower vertebrates (Ed. by W. Ahne & E. Kurstak),
pp. 284-291. Springer-Verlag, Berlin.
Kvellestad A., Dannevig B.H. & Falk K. (2003) Isolation and partial characterization of a
novel paramyxovirus from the gills of diseased seawater-reared Atlantic salmon
(Salmo salar L.). Journal of General Virology 84, 2179-2189.
McLoughlin M.F. & Graham D.A. (2007) Alphavirus infections in salmonids - a review.
Journal of Fish Diseases 30, 511-531.
16

McLoughlin M.F., Nelson R.N., McCormick J.I., Rowley H.M. & Bryson D.B. (2002)
Clinical and histopathological features of naturally occurring pancreas disease in
farmed Atlantic salmon, Salmo salar L. Journal of Fish Diseases 25, 33-43.
Murphy F.A., Gibbs E.P.J., Horiznek M.C. & Studdert, M.J. (1999) Veterinary virology. 629
pp. Academic Press, San Diego.
Watanabe K., Karlsen M., Devold M., Isdal E., Litlabø A. & Nylund A. (2006) Virus-like
particles associated with heart and skeletal muscle inflammation (HSMI). Diseases of
Aquatic Organisms 70, 183-192.
17

Table 1: Infection experiments with heart and skeletal muscle inflammation (HSMI) in Atlantic salmon Salmo salar L. The table shows details
of the study design, inoculates and sampling of fish during the studies. * Material sampled from an infection experiment, 4 and 8 weeks post
challenge. FF = fat fin clipped. W= week post infection. L-15 = Leibovitz’ L-15 cell culture medium.
Experiment Habitat Inoculate Inoculate clinical stage Tank Fish Sampling No. fish
injected weeks sampled
W 1-4 W 5-8
1 Freshwater Cardiac tissue Mid-outbreak 1 90 1,3,4,5,6,7,8 5 7
Blood plasma Mid-outbreak 2 100 1,3,4,5,6,7,8 5 7
FHM cell culture Pre-outbreak* 3 FF 50 3,4,5,6,7,8 5 7
EPC cell culture Pre-outbreak* 3 50 3,4,5,6,7,8 5 7
L-15 - 4 100 1,3,4,5,6,7,8 5 7
2 Seawater Cardiac tissue Mid-outbreak 1 60 1,2,3,4,5,6,7,8 5 5
Cardiac tissue Healthy fish 2 60 1,2,3,4,6,7,8 5 5
Cardiac tissue, chloroform Healthy fish 2 FF 60 1,2,3,4,6,7,8 5 5
Cardiac tissue, chloroform Mid-outbreak 3 60 1,2,3,4,6,7,8 5 5
Cardiac tissue 2 months pre-outbreak 4 60 1,2,3,4,6,7,8 5 5
Cardiac tissue 2 months post outbreak 5 60 1,2,3,4,6,7,8 5 5
Blood plasma Mid-outbreak 6 60 1,2,3,4,6,7,8 5 5
Kidney/spleen pooled Mid-outbreak 7 60 1,2,3,4,6,7,8 5 5
Liver Mid-outbreak 7 FF 60 1,2,3,4,6,7,8 5 5
18

Table 2: Infection experiment no 1 with heart and skeletal muscle inflammation (HSMI) in Atlantic salmon, Salmo salar L. smolts in freshwater.
The table shows the number of sampled fish with various cardiac lesions. Five fish in each group were sampled at each of the sampling points in
weeks 1-4, while seven fish in each group were sampled at each sampling point in weeks 5-8. FHM = cell culture from fat head minnow
Pimephales promelas, EPC = cell culture from epithelioma papilosum cyprini, L-15 = Leibovitz’ L-15 cell culture medium (negative control
inoculate).
Epicarditis Endocarditis Perivasculitis Myocarditis compact Myocarditis spongy
myocardium
myocardium
1 3 4 5 6 7 8 1 3 4 5 6 7 8 1 3 4 5 6 7 8 1 3 4 5 6 7 8 1 3 4 5 6 7 8
Inoculate/
weeks
Heart 0 3 3 6 6 7 7 0 0 2 4 5 7 7 0 3 3 5 6 7 7 0 0 1 4 4 7 7 0 0 2 4 5 7 7
Plasma 2 2 2 4 4 6 7 0 1 2 1 2 2 7 0 3 2 3 4 5 7 0 0 0 1 2 0 7 0 0 2 1 2 2 7
FHM - 4 2 2 7 5 6 - 2 1 3 2 4 7 - 2 2 2 4 5 3 - 0 0 2 2 5 6 - 1 1 3 2 4 7
EPC - 1 2 5 6 5 4 - 0 0 4 0 4 4 - 3 2 5 4 3 5 - 0 0 0 0 3 5 - 0 0 4 0 4 3
L-15 1 3 1 2 0 2 3 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
19

Table 3: Infection experiment 2 with heart and skeletal muscle inflammation (HSMI) in Atlantic salmon, Salmo salar L. smolts in seawater. The
table shows the number of sampled fish with various cardiac lesions. Five fish were sampled from each group at every sampling point. Heart + =
cardiac lesions consistent with HSMI in inoculate fish. Heart - = negative control fish with normal hearts. Chl. = chloroform treated. Heart early
= hearts sampled 2 months before outbreak phase of HSMI. Heart late = hearts sampled 2 months after outbreak phase of HSMI. - = no sampling
Epicarditis Endocarditis Perivasculitis Myocarditis compact Myocarditis spongy
myocardium
myocardium
1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8
Inoculate/
weeks
Heart + 0 0 0 3 2 2 4 5 0 0 0 0 0 3 3 5 1 2 0 1 2 2 4 4 0 0 1 0 1 1 3 5 0 1 0 0 0 3 3 5
Heart - 0 0 0 1 - 1 1 3 0 0 0 0 - 0 0 0 0 0 0 0 - 0 0 0 0 0 0 0 - 0 0 0 0 0 0 0 - 0 0 0
Heart - chl. 0 0 0 0 - 1 1 1 0 0 0 0 - 0 0 0 0 0 0 0 - 0 0 0 0 0 0 0 - 0 0 0 0 0 0 0 - 0 0 0
Heart + chl. 2 0 0 2 - 1 5 4 0 0 0 0 - 1 3 3 1 0 1 2 - 1 4 4 0 0 1 1 - 2 4 4 0 0 0 0 - 0 2 2
Heart early 1 0 2 1 - 4 3 4 0 0 0 0 - 0 2 4 0 1 2 0 - 1 4 5 0 0 1 0 - 1 1 5 0 0 0 0 - 0 2 4
Heart late 0 0 1 2 - 3 2 4 0 1 0 0 - 2 2 4 0 0 2 1 - 3 2 5 0 0 0 1 - 2 2 5 0 0 0 0 - 2 2 4
Plasma 0 0 0 0 - 4 3 4 0 0 0 0 - 0 0 2 1 1 3 1 - 4 1 4 0 0 0 0 - 0 1 4 0 0 0 0 - 0 0 2
Kidney/spleen 0 0 1 0 - 3 3 5 1 0 1 1 - 3 4 4 0 0 1 0 - 5 4 5 0 0 0 0 - 4 3 5 0 0 0 0 - 3 3 4
Liver 1 0 0 0 - 3 5 5 0 0 1 0 - 1 5 4 0 0 1 1 - 4 5 4 0 0 0 0 - 1 4 3 0 0 0 0 - 0 5 4
20

Legends to figures:
Figure 1
(A) FHM cell culture, 4 th passage of tissue from fish with heart and skeletal muscle
inflammation (HSMI) in Atlantic salmon, Salmo salar L., one day after inoculation. A distinct
focus of rounded cells (arrow). (B) FHM cell culture, negative control.
Figure 2
Experimental infection of heart and skeletal muscle inflammation (HSMI) in Atlantic salmon,
Salmo salar L. Heart of fish injected with hepatic tissue from fish with HSMI, four weeks
post challenge (p.c.) Perivasculitis (arrow) of a coronary vessel branch (asterisk) in the
compact layer. H&E
Figure 3
Experimental infection of heart and skeletal muscle inflammation (HSMI) in Atlantic salmon,
Salmo salar L. Heart of fish injected with chloroform pre-treated supernatant from cardiac
tissue of fish with HSMI, eight weeks p.c. Severe epicarditis (asterisk) and myocarditis
(examples pointed out with arrows) in compact and spongy layers of the cardiac ventricle.
H&E
21

Figures
22

Paper VI
An outbreak of disease resembling heart and skeletal muscle
inflammation in Scottish farmed salmon, Salmo salar L., with
observations on myocardial regeneration

Ó 2005
Blackwell Publishing Ltd
Journal of Fish Diseases 2005, 28, 119–123
Short communication
An outbreak of disease resembling heart and skeletal
muscle inflammation in Scottish farmed salmon, Salmo
salar L., with observations on myocardial regeneration
H W Ferguson 1 , R T Kongtorp 2 , T Taksdal 2 , D Graham 3 and K Falk 2
1 Institute of Aquaculture, University of Stirling, Stirling, UK
2 National Veterinary Institute, Oslo, Norway
3 Veterinary Sciences Division, Stormont, Belfast, UK
Keywords: Anitschkow-like nuclei, cardiac regeneration,
heart, myositis, salmon, Scotland.
Heart disease in farmed Atlantic salmon, Salmo salar
L., is becoming a frequently encountered diagnostic
problem. Examples include cardiomyopathy syndrome
(CMS), a condition of unknown aetiology in
which spongy myocardium of both ventricle and
atrium is targeted, and which may become so severely
affected that the atrium ruptures (Ferguson, Poppe &
Speare 1990). Myocardium is also affected in the
viral condition salmon pancreas disease (SPD),
although in that case both compact and spongy
layers are involved, and lesions are not as severe as in
CMS. In addition, striated muscle elsewhere is
affected, including axial skeletal muscle and oesophagus,
and, as may be inferred from the name of the
disease, exocrine pancreas is also involved (Ferguson,
Roberts, Richards, Collins & Rice 1986).
Heart and skeletal muscle inflammation (HSMI)
is an infectious condition of undetermined cause
that has recently been described in Norwegian
farmed salmon (Kongtorp, Kjerstad, Taksdal, Guttvik
& Falk 2004). Once again, major lesions are
encountered in myocardium, in both spongy and
compact layers, but cardiac lesions are more severe
than those typically seen in SPD and in addition
there is often an epicardial infiltrate. Skeletal muscle
Correspondence Professor H W Ferguson, Institute of
Aquaculture, University of Stirling, Stirling FK9 4LA, UK
(e-mail: hwf1@stir.ac.uk)
119
is the other major location for lesions, red fibres
being mostly affected.
The present report documents an outbreak of
disease in sea-caged salmon in Scotland in which
the lesions mirror those described for HSMI. If it is
indeed the same as HSMI, this outbreak would
represent the first time the condition has been
described outside Norway.
A marine salmon farm on the west of Scotland
experienced significant mortality of second year
grower fish from the third week in June 2004 until
the site was cleared in September. Cumulative
mortality for this period approached 9% and at the
height of the outbreak was escalating sharply,
precipitating the decision to slaughter out. Prior
to this, health status had been good and the only
significant health problem had been clinical infectious
pancreatic necrosis (IPN) as post-smolts
during summer 2003, resulting in approximately
7% mortalities.
Some cages had been held at a relatively high
stocking density, and as a result had been split using
a well boat in June 2004. Mortality started in these
cages approximately 2 weeks after the handling but
a rising trend subsequently developed in all cages
within 5 weeks of the early losses. Mortalities were
initially mostly large fish in good condition and
with no gross lesions. As the problem developed,
however, an increasing number of dead fish showed
poor condition, scale loss, skin and fin erosion and
abraded or ruptured eyes. Some cages showed a
reduction in feed intake prior to mortalities

Ó 2005
Blackwell Publishing Ltd
Journal of Fish Diseases 2005, 28, 119–123 H W Ferguson et al. Heart disease in Atlantic salmon
increasing, but this was not consistently recorded
throughout the site.
Apparently lethargic fish were seen stacking up
and holding station in the tidal current, but these
were not easily caught for examination. As losses
increased, divers brought in to help clear mortalities
reported numbers of pale lethargic fish accumulating
on the net floor, sometimes lying on their sides.
These appeared to be dying and could be caught by
hand, but would sometimes swim off actively when
handled. Moribund fish caught around the sides of
the cages showed pale flanks, a slightly swollen
abdomen, raised scales along the lower flanks with
dermal oedema, and sometimes slight exophthalmos.
Fish with these gross lesions were also found
among the mortalities, but not in large numbers.
Internally, these fish showed clear or cloudy strawcoloured
ascites and pericardial fluid, enlarged pale
mottled liver, sometimes with a thin gelatinous
membrane over the capsule, slightly swollen kidney
and a soft, flabby heart (Fig. 1). Internally many of
the early mortalities showed few gross lesions and
some had fresh feed pellets within the stomach.
Routine bacteriological swabs were taken from the
kidneys and plated onto tryptone soya agar (TSA;
Oxoid, Basingstoke, UK) and TSA + 1.5% salt
(NaCl). For histopathology, tissues from 10 fish with
typical signs were fixed in 10% phosphate-buffered
formalin for at least 24 h, and then dehydrated in a
graded alcohol series before routine processing and
embedding in paraffin wax. Sections were stained
with haematoxylin and eosin (H&E). Blood was
taken from 16 fish, and serum tested for antibodies to
SPD virus (SPDV) using an immunoperoxidasebased
microtitre assay and for the presence of
120
viraemia (Graham, Jewhurst, Rowley, McLoughlin
& Todd 2003; Jewhurst, Todd, Rowley, Walker,
Weston, McLoughlin & Graham 2004).
The most significant histopathological changes
were present in heart, skeletal muscle and liver.
Histopathological changes in myocardial spongy
layer were characterized by widespread vacuolation,
degeneration and subsequent cavitation of cardiac
myocytes, with loss of striation, increased eosinophilia,
karyorhexis of myocyte nuclei, and infiltration by
a limited number of neutrophils and macrophages,
some of the latter being ceroid-laden. In conjunction
with these changes there was karyomegaly of myocyte
nuclei, and in two fish ÔnestsÕ of smaller nuclei were
present within the substance of the myocyte (Fig. 2).
In most cases these nests of nuclei were loosely
aggregated, although in a few instances they were
arranged in a row within the centre of the cell. Both
endocardium and sub-endocardial cells were intact
and prominent but not greatly hypertrophic,
although there was some evidence of mobilization
of sub-endocardial cells into the middle of myocytes.
Changes in the atrium were similar, but in most cases
were less severe. Nuclear nests in the atrium were
correspondingly lower in number.
Changes in the compact layer were not so severe,
but there was still degeneration and necrosis,
characterized by hypereosinophilia and loss of
striation, and limited hypercellularity. Karyomegaly
was also limited and no nuclear nests were seen. Of
interest, however, was the presence in a few fish of
elongated Anitschkow-like nuclei with a prominent
wavy chromatin pattern (Fig. 3). These were most
obvious in those fish in which nuclear nests in the
spongy layer were present. Many fish had mild to
Figure 1 Atlantic salmon showing
proteinaceous blood-tinged ascitic fluid
within abdominal cavity. The heart in this
fish is also mis-shapen.

Ó 2005
Blackwell Publishing Ltd
Journal of Fish Diseases 2005, 28, 119–123 H W Ferguson et al. Heart disease in Atlantic salmon
Figure 2 Micrograph of ventricle from
salmon showing multinuclear nests (arrows)
within myocardium of spongy layer
(H&E, ·45).
moderate accumulation of mostly mononuclear
cells on the epicardial surface. One fish had a few
small aggregates of microsporidia within compact
myocardium, but there was no associated host
response and they were regarded as incidental.
Skeletal muscle lesions were present in both red
and white fibres. In some fish they were more
pronounced in red than in white, or vice versa, while
in other fish they were equally severe in both
(Fig. 4). Acute degeneration and necrosis with
myophagia were seen, as was regeneration, all
present within one field of view. Lesions in other
tissues included severe heart-failure-type periacinar
zonal hepatocellular necrosis and limited myodegeneration
and necrosis in oesophageal muscle. By
contrast, there were no significant lesions in
pancreas, nor in any other tissue.
Bacteriology revealed no significant growth from
most of the moribund and dead fish sampled. Fish
Figure 3 Micrograph of myocardial
compact layer from salmon showing
presence of Antischkow-like nucleus (arrow)
(H&E, ·450).
121
with more obvious signs of external abrasion tended
to have variable mixed growth of Vibrio-like
bacteria on culture; these were regarded as incidental
to the outbreak in general. Neutralizing antibodies
to SPDV were found in all sera at a titre of
‡1/40. None of the sera was viraemic.
There is little doubt that fish with lesions as
severe as these would be reluctant to move, and
would be suffering clinically from a failing cardiovascular
system. This contention is supported by
the gross observations of ascites and the presence of
zonal liver lesions that are most probably the result
of reduced cardiac output and hence anoxia. The
significance of the titres to SPDV is unknown, but
at the very least they indicate that the fish had prior
exposure to SPDV at some stage of the production
cycle. While there is considerable variation in the
extent of the heart and skeletal muscle lesions
described in field cases of SPD (McLoughlin,

Ó 2005
Blackwell Publishing Ltd
Journal of Fish Diseases 2005, 28, 119–123 H W Ferguson et al. Heart disease in Atlantic salmon
Nelson, McCormick, Rowley & Bryson 2002) the
rhabdomyopathy lesions in the present case were
much more severe than would normally be
expected, especially those seen in the heart. The
hepatic lesions alone bore testimony to the
functional severity of these lesions; such liver lesions
are not normally seen in SPD, although they are
seen in CMS. There were also no pancreatic lesions
in any fish, as might be expected if SPD were
involved in the disease outbreak, although this
could indicate that fish were in the chronic stages of
SPD, when cardiac and skeletal muscle lesions
predominate and pancreatic lesions are repaired
(Murphy, Rodger, Drinan, Gannon, Kruse &
Korting 1992; McLoughlin et al. 2002).
The changes in these fish were very similar to
those described recently from Norway for HSMI
(Kongtorp, Taksdal & Lyngøy 2004). They differ
from classical SPD in their severity and in having
no significant involvement of pancreas. They also
differ from CMS in having marked skeletal muscle
lesions. The involvement of compact myocardium
was not as pronounced as described from Norway,
but in many cases it was still more obvious than
would be normal for CMS, thereby representing yet
another differentiating feature.
The pathology of the spongy layer was very
similar, however, to that described for CMS. As
well as the formation of hollow tubes delimited
by endocardium, basement membrane and subendocardial
cells, greatly enlarged cardiac myocyte
nuclei were also present. Presumably this represents
an attempt at regeneration or compensatory
hypertrophy in a failing heart, one in which a large
number of myocytes are non-functional due to loss
122
Figure 4 Low-power micrograph of axial
muscle of salmon showing limited degeneration
and necrosis of white fibres (arrow) but
pronounced degeneration and necrosis with
ongoing regeneration in the red fibres
(H&E, ·45).
of contractile elements. The one interesting
difference from CMS was the presence in some
fish of multinucleate ÔnestsÕ within cardiac myocytes.
These have not been described before and
they may represent abortive cell division responding
to the influence of some toxin or other
pathogen. Viruses like aquareovirus can result in
multinucleated syncytial cell formation (Cusack,
Groman, MacKinnon, Kibenge, Wadowska &
Brown 2001; Ferguson, Millar & Kibenge
2003). It seems also possible, however, that they
represent normal attempts at regeneration by
myocytes. It is well recognized that cardiac
myocytes of lower vertebrates, including fish, are
capable of regeneration, even in relatively mature
animals (Rumyantsev 1979; Ferguson et al. 1990;
Poss, Wilson & Keating 2002), but the mechanism
is not described. There is no mention of such
nuclear aggregates in diseases such as CMS and
SPD where there is substantial damage to
myocytes of spongy myocardium. This may
suggest a different pathogenesis or merely fortuitous
timing of infection and/or sampling, as
they were seen in only a few fish (two of 10) even
in the present case. The origin of the nuclei is
unknown, but the apparent mobilization of subendocardial
cells would seem to be one possible
explanation, as it has always been assumed that
these are reserve ‘satellite’ cells.
Similarly, in the compact layer, the presence of
Anitschkow-like nuclei suggests myocardial repair
or regeneration, but utilizing a mechanism(s)
different from that seen in spongy myocardium.
Considering the very different architectures, phagocytic
capability, metabolic profiles and detoxifica-

Ó 2005
Blackwell Publishing Ltd
Journal of Fish Diseases 2005, 28, 119–123 H W Ferguson et al. Heart disease in Atlantic salmon
tion capabilities of these two types of heart tissue in
teleosts in general (Ferguson 1975; Farrell 1990;
Stegeman, Miller & Hinton 1990), it should be no
real surprise that they respond differently to insult
or in a situation where regeneration is underway. In
mammals, Anitschkow cells are associated with
immune-mediated myocardial disease such as rheumatic
fever. The current consensus of opinion is
that the cells are derived from histiocytes, although
Anitschkow himself believed that they represented
modified cardiac muscle cells (Satoh & Tsusumi
1999). If we can trust in gleaning pathophysiological
clues from phylogeny, the present evidence
suggests that Anitschkow may have been correct in
his original interpretation.
If the multinucleate nests do indeed represent
regeneration, recovery of the heart might have been
anticipated in affected fish, and with little of the
scarring and reduced efficiency that would be
expected in a mammal with similar lesions. On the
contrary, the ongoing degeneration and necrosis of
myocardium in the face of possible regeneration
suggests continuing insult and a guarded prognosis.
The latter possibility was obviously the one that
weighed the heavier with the farmer who decided to
slaughter out the affected site. Attention to biosecurity
in this case was also important to help limit
spread to adjacent sites, as HSMI has been spreading
throughout the Norwegian salmon farming industry.
At a time when humans are being encouraged to
eat fish to help combat a range of conditions
including coronary disease, it seems somewhat
ironic that heart disease seems to be such a problem
in the fish themselves. Given the potential economic
impact of the outbreak described here, there is a
clear need for further work to determine the cause of
HSMI, and the role (if any) of SPDV.
Acknowledgements
We would like to thank H. Jewhurst for the
serological analysis and D. Faichney for the histotechnology.
References
Cusack R.R., Groman D.B., MacKinnon A.-M., Kibenge F.S.B.,
Wadowska D. & Brown, N. (2001) Pathology associated with
an aquareovirus in captive juvenile Atlantic halibut Hippoglossus
hippoglossus and an experimental treatment strategy for a
concurrent bacterial infection. Diseases of Aquatic Organisms
44, 7–16.
123
Farrell A.P. (1990) From hagfish to tuna: a perspective on cardiac
function in fish. Physiological Zoology 64, 1137–1164.
Ferguson H.W. (1975) Phagocytosis by the endocardial lining
cells of the atrium of plaice (Pleuronectes platessa). Journal of
Comparative Pathology 85, 561–569.
Ferguson H.W., Roberts R.J., Richards R.H., Collins R.O. &
Rice D.A. (1986) Severe degenerative cardiomyopathy associated
with pancreas disease in Atlantic salmon, Salmo salar L.
Journal of Fish Diseases 9, 95–98.
Ferguson H.W., Poppe T.T. & Speare D.J. (1990) Cardiomyopathy
syndrome in farmed Norwegian salmon Salmo
salar. Diseases of Aquatic Organisms 8, 225–231.
Ferguson H.W., Millar S.D & Kibenge F.S.B. (2003) Reoviruslike
hepatitis in farmed halibut (Hippoglossus hippoglossus).
Veterinary Record 153, 86–87.
Graham D.A., Jewhurst V.A., Rowley H.M., McLoughlin M.F.
& Todd D. (2003) A rapid immunoperoxidase-based virus
neutralization assay for salmonid alphavirus used for a
serological survey in Northern Ireland. Journal of Fish
Diseases 26, 407–413.
Jewhurst V.A., Todd D., Rowley H.M., Walker I.W., Weston
J.H., McLoughlin M.F. & Graham D.A. (2004) Detection
and antigenic characterization of salmonid alphavirus isolates
from sera obtained from farmed Atlantic salmon, Salmo salar
L., and farmed rainbow trout, Oncorhynchus mykiss
(Walbaum). Journal of Fish Diseases 27, 143–149.
Kongtorp R.T., Kjerstad A., Taksdal T., Guttvik A. & Falk K.
(2004) Heart and skeletal muscle inflammation in Atlantic
salmon, Salmo salar L.: a new infectious disease. Journal of Fish
Diseases 27, 351–358.
Kongtorp R.T., Taksdal T. & Lyngøy A. (2004) Pathology of heart
and skeletal muscle inflammation (HSMI) in farmed Atlantic
salmon Salmo salar. Diseases of Aquatic Organisms 59, 217–224.
McLoughlin M.F., Nelson R.N., McCormick J.I., Rowley H.M.
& Bryson D. (2002) Clinical and histopathological features of
naturally occurring pancreas disease in farmed Atlantic
salmon, Salmo salar L. Journal of Fish Diseases 25, 33–43.
Murphy T.M., Rodger H.D., Drinan E.M., Gannon F., Kruse P.
& Korting W. (1992) The sequential pathology of pancreas
disease in Atlantic salmon farms in Ireland. Journal of Fish
Diseases 15, 401–408.
Poss K., Wilson L. & Keating M. (2002) Heart regeneration in
zebra fish. Science 298, 2188.
Rumyantsev P.P. (1979) Some comparative aspects of myocardial
regeneration. In: Muscle Regeneration (ed. by A. Mauro,
B.M. Carlson & B. Lipton). Raven Press, New York, USA.
Satoh F. & Tsusumi Y. (1999) Anitschkow cells in the human
heart. Pathology International 49, 85–87.
Stegeman J.J., Miller M.R & Hinton D.E. (1990) Cytochrome
p450IA1 induction and localization in endothelium of vertebrate
(teleost) heart. Molecular Pharmacology 36, 723–729.
Received: 21 September 2004
Revision received: 8 November 2004
Accepted: 24 November 2004