Annals of Diagnostic Paediatric Pathology

Annals of Diagnostic Paediatric Pathology 

Official Journal of the Polish Paediatric Pathology Society 

and Section of Oncological Surgery of Polish Association of Paediatric Surgeons 

Editor-in-Chief 

Co-Editors 

Associate Editors 

ProductIon Editor 

B. M. WoŸniewicz, Warsaw 

B. M. Cukrowska, Warsaw 

A. I. Prokurat, Bydgoszcz 

J. Cielecka-Kuszyk, Warsaw 

E. Czarnowska, Warsaw 

W. T. Dura, Warsaw 

W. Grajkowska, Warsaw 

M. Liebhardt, Warsaw 

A. Wasiutyñski, Warsaw 

G. Napiórkowski, Warsaw 

A. Bysiek, Cracow 

P. Czauderna, Gdansk 

J. Godziñski, Wroc³aw 

J. Niedzielski, Lodz 

W. WoŸniak, Warsaw 

M. Wysoki, Bydgoszcz 

Editorial Office 


Department of Pathology 

The Children´s Memorial Health Institute 

Aleja Dzieci Polskich 20 

04 736 Warszawa, Poland 

Tel.: +48-22-815-19-72 

Fax: +48-22-815-19-73 

E-mail: ptpd@czd.waw.pl 

Editorial Board 

J. P. Barbet, Paris 

L. A. Boccon-Gibod, Paris 

P. E. Campbell, Melbourne 

A. Chilarski, Lodz 

J. Czernik, Wroclaw 

E. Gilbert-Baarness, Tampa 

A. A. Greco, New York 

M. D. Haust, Londyn 

A. Hinek, Toronto 

J. Huber, Utrecht 

C. G. Gopalakrishnam, Trivandrum 

S. Gogus, Ankara 

A. Jankowski, Poznan 

P. Januszewicz, Warsaw 

B. Jarz¹b, Gliwice 

B. A. Kakulas, Perth 

R. O. C. Kaschula, Rondebosch 

W. Kawalec, Warsaw 

J. Kobos, Lodz 

G. Karpati, Montreal 

J. Las Heras, Santiago de Chile 

K. Madaliñski, Warsaw 

D. M. F. Menezes, Rio de Janeiro 

W. A. Newton, Jr., Johnstown 

B. Otte, Brussels 

S. A. Pileri, Bologna 

J. Plaschkes, Berne 

F. Raafat, Birmingham 

S. W. Sadowinski, Mexico City 

K. Sawicz-Birkowska, Wroclaw 

J. Stejskal, Prague 

Cz. Stoba, Gdansk 

G. Thiene, Padova 

S. Variend, Shieffield 

T. H. Wyszyñska, Warsaw 

A. Zimmermann, Berne 

The journal is supported by the State Committee for Research. 

Aims and scope 

The Annals of Diagnostic Paediatric Pathology is an international peer-reviewed journal. The focus of the journal is current 

progress in clinical paediatric pathology and surgical oncology in both basic and clinical applications. Experimental studies and 

clinical trials are accepted for publication, as are case reports supported by literature review. The main policy of the Annals is 

to publish papers that present practical knowledge that can be applied by clinicians. 

© Copyright by Polish Paediatric Pathology Society, 2001 

ISSN 1427-4426

Review 

Papers 


Official Journal of the Polish Paediatric Pathology Society 

and Section of Oncological Surgery of Polish Association of Paediatric Surgeons 

Volume 8 Number 3-4 Winter 2004 

CONTENTS 

Genetics and biology of neuroblastoma – an overview ....................................................................................69 

Peter F. Ambros, Ruth Ladenstein, Helmut Gadner, Ingeborg M. Ambros 

Neuroblastoma: histoprognostic classification and handling of material for biology 

– a challenge for the pathologist .........................................................................................................................73 

Klaus Beiske 

Original 

Papers 

Intracranial ependymomas in children: correlation of selected factors and event-free 

survival in the group of 142 consecutive patients treated at one institution. .................................................83 

S³awomir Barszcz, Wies³awa Grajkowska, Danuta Perek, Marcin Roszkowski, Monika Drogosiewicz, 

Marta Perek-Polnik, El¿bieta Jurkiewicz, Pawe³ Daszkiewicz 

Pathology of fatty liver in children with LCHAD deficiency...........................................................................89 

Katarzyna Iwanicka, Janusz Ksi¹¿yk, Monika Pohorecka, Maciej Pronicki 

Isolation of urothelial cells for tissue engineered bladder augmentation ....................................................... 95 

Tomasz Drewa, Przemys³aw Galazka, Zbigniew Wolski, Andrzej I. Prokurat, Jan Sir 

Status of oral hygiene and dentition in children exposed to anticancer treatment .......................................99 

Dorota Olczak-Kowalczyk, Marta Daszkiewicz, Barbara Adamowicz-Klepalska, El¿bieta Czarnowska, Agnieszka Sowiñska, 

Bo¿ena Dembowska-Bagiñska, Danuta Perek, Pawe³ Daszkiewicz 

Preliminary review of the treatment results of hepatoblastoma and hepatocarcinoma 

in children in Poland in the period 1998-2002 ................................................................................................107 

Czes³aw Stoba, Katarzyna Nierzwicka, Piotr Czauderna, Barbara Skoczylas-Stoba , Stefan Popadiuk, Jolanta Bonar, 

Krystyna Sawicz-Birkowska, Hanna Kaczmarek-Kanold, Przemys³aw Mañkowski, Krzysztof K¹tski, Wojciech Madziara, 

Magdalena Rych³owska, Anna Sitkiewicz, Anna Szo³kiewicz, Sabina Szymik-Kantorowicz, Violetta Œwi¹tkiewicz 

Liver transplantation for primary malignant liver tumors in children – single center experience ...........111 

Piotr Kaliciñski, Hor Ismail, Andrzej Kamiñski, Danuta Perek, Joanna Paw³owska, Przemys³aw Kluge, Joanna Teisserye, 

Marek Szymczak, Tomasz Drewniak, Pawe³ Nachulewicz, Bo¿ena Dembowska-Bagiñska, Marek Stefanowicz, 

Andrzej Byszewski, Ma³gorzata Manowska 

Hepatocyte ultrastructure in non-hemolytic hyperbilirubinemias ...............................................................115 

El¿bieta Czarnowska, Bogdan WoŸniewicz 

Congenital Heart Diseases Database: analysis of occurrence, 

diagnostics and coexistence with anomalies of other systems. Single institution experience .....................119 

Anna Rybak, Aneta Wójcik, Bogdan WoŸniewicz 

Case 

Reports 

Invited 

Lectures 

Solid pseudopapillary tumor of pancreas. Case report and the review of literature ..................................121 

Andrzej Chilarski, Anna Piaseczna-Piotrowska, Jan Strzelczyk, Stanis³aw £ukaszek, Andrzej Kulig 

Liver tumors of childhood - pathology. Report of the Kiel Pediatric Tumor Registry................................125 

Dieter Harms

Annals of Diagnostic Paediatric Pathology 2004, 8(3-4): 

© Copyright by Polish Paediatric Pathology Society 

Genetics and biology of neuroblastoma – an overview 

Peter F. Ambros 1 , Ruth Ladenstein 2 , Helmut Gadner 1, 2 , Ingeborg M. Ambros 1 

Annals of 

Diagnostic 

Paediatric 

Pathology 

1 

CCRI, Children’s Cancer Research Institute, 

St. Anna Children’s Hospital, 

A-1090 Vienna, Austria 

2 

St. Anna Children’s Hospital, 


Abstract 

Neuroblastoma, the most common extra cranial solid tumor in early childhood, is a heterogeneous disease 

with different clinical and biological behavior. Unique features are the spontaneous regression and maturation 

which both require similar genetic prerequisites and, in the case of maturation, also the presence of the 

non-neoplastic Schwann cells. However, over the half of the tumors display an aggressive clinical course. 

Besides the stage and age at diagnosis, the MYCN oncogene (amplified versus non-amplified) has gained 

fundamental importance for therapy stratification, particularly in patients with localized disease and in infants. 

Further prognostic markers associated with an unfavorable outcome are DNA di-tetraploidy, deletion 

at chromosomal regions 1p36, 3p and 11q and gain of 17q, besides others. Most of these genetic markers 

have been shown to predict outcome in patients with localized disease while their impact in stage 4 patients 

is discussed controversially. Recent data indicate that not even MYCN amplification can separate prognostic 

subgroups in the latter patient group. However, a rapid BM-clearing seems to be associated with a far more 

favorable outcome even in the group of patients with unfavorable genetic features, thus expressing the response 

to chemotherapy and possibly providing an excellent prognostic marker. Furthermore, recent gene 

expression studies highlighting the whole transcriptome may provide also valuable information on the dignity 

of neuroblastic tumors. We can conclude that MYCN amplification still represents an excellent way to 

discriminate aggressive from non- or less aggressive tumors especially in patients with localized disease or 

infants. However, new biological and genetic information is necessary to be able to reach our final goal 

which is patient tailored therapy based on the genetic and biological profile of the tumor. 

Key words: genetics, MYCN, oncogene 

Genetic hallmarks in neuroblastic tumors 

Over the last few years the amplification of the MYCN 

oncogene [11, 32] has taken shape as the most important 

marker for the assessment of the dignity of neuroblastic tumors. 

Amplifications of this oncogene can be observed in approximately 

20% of tumors and are often associated with a 

deletion at the short arm of chromosome 1, del1p36 [10]. Both 

aberrations can be found in both genetic subgroups (Fig. 1) 

which are based on the different DNA-content (di-/tetraploid 

versus near-triploid or penta-/hexaploid). What up to now is 

entirely unclear is why MYCN amplifications and/or aberrations 

of the chromosomal region 1p36.3 occur in only about 

15% of the near-triploid neuroblastic tumors while the same 

aberrations can be found in approx. 56% of the diploid group 

(unpublished data). The genetic aberration which is most frequently 

detected in neuroblastic tumors is gain of the long 

arm of chromosome 17 (in 66-69% of all cases) [9, 20, 39]. 

The prognostic value of this aberration has however not yet 

been completely clarified and was already opposed [36]. Also 

other chromosome aberrations like del11q23 [13, 15, 22, 35, 

40] and del3p26 [13, 40] have been described frequently. 

Whereby, loss of sequences at 11q may indicate a rather promising 

genetic marker especially in MYCN non amplified tumors. 

However, one has to keep in mind that genetic/biological 

studies will only give clear results when the material to be 

tested is of good quality and is present in sufficient amount. 

Address for correspondence 

Assoc. Prof. Dr. Peter F. Ambros 

CCRI, Children's Cancer Research Institute, St. Anna Kinderspital 

Kinderspitalgasse 6 


Phone: +43 1 40470 411 

Fax: +431 408 7230 

E-mail: ambros@ccri.univie.ac.at

70 

Therefore, the sampling and storage of tumor, blood and bone 

marrow should be done in the appropriate way to guarantee 

the highest possible quality of the results [5]. Furthermore, 

quality assessment studies and a uniform nomenclature to 

describe the results of the genetic/biological studies provide 

additional security on the high quality of the results [1]. 

Subgroups of neuroblastic tumors based on different 

DNA-content 

Based on a difference in DNA-content, neuroblastic tumors 

can be divided into two major groups (Fig. 1). Slightly more 

than half of the tumors (55%) have a near-triploid DNA-content 

(DNA-index between 1.26 and 1.74) while in approx. 45% 

of neuroblastic tumors a diploid (or tetraploid or even combined 

diploid/tetraploid) DNA content can be observed. Spontaneous 

regression or maturation, not triggered by therapy, 

and which can often be found in neuroblastic tumors, is very 

likely limited to the near-triploid subgroup [2, 7]. In approx. 

15% of the cases, however, aggressive tumor behavior can 

also be observed in this subgroup (Ambros, unpublished). In 

these cases structural chromosome aberrations (MYCN amplification, 

1p-deletion) may be found. Most of the diploid 

tumors, on the other hand, are aggressive even without MYCN 

amplification or 1p36.3 aberration, irrespective of the patient’s 

age at diagnosis or the stage of the tumor [19, 21]. Alterations 

of the genome can be observed in almost all diploid tumors. 

Genesis of neuroblastic tumors and genetic intratumoral 

heterogeneity (focal MYCN amplification) 

When focusing on the two different ploidy groups, diploid 

versus near-triploid neuroblastic tumors, we can identify different 

frequency of structural chromosomal aberrations (i.e. 

deletions, gains or even amplifications). The majority of neartriploid 

tumors exclusively have only numerical chromosome 

aberrations – which seems to be sufficient for (often self-limited) 

tumor growth. Succeeding secondary changes, as for 

example MYCN-amplification or 1p36.3 aberration, lead, on 

the other hand, to malignant transformations (Fig. 1). As regards 

the diploid (2n) tumor group (without gain of additional 

chromosomes), the initiating event has not been identified so 

far, even though other chromosome aberrations like del11q23, 

del3p26 and 17q-gain have been described frequently [10]. 

What all these diploid tumors have in common, however, is 

their aggressive cell growth (Fig.1). 

3n 

trkA 

expression 

3n 

no NGF 

Schwann cell 

regression 

maturation 

(GNB, GN) 

neuroblasts 

del1p, MNA 

progression 

2n 

LOH 3p, 11q, etc. 

progression 

Fig. 1 Spontaneous regression and spontaneous maturation are restricted to those tumors with a near triploid (+ 3n) or near pentaploid/hexaploid (+5n/6n) 

DNA content. In tumors of patients under 1 year of age, lack of nerve growth factor (NGF) is a common finding [18, 27] and is probably involved in the 

phenomenon of spontaneous regression. Tumors undergoing spontaneous maturation processes, on the other hand, occur in patients over 1 year of age and are 

characterized by the presence of a reactive, i.e. non neoplastic cell population, namely the Schwann cells, leading to a mixed population of near triploid 

(penta-, hexaploid) neuroblastic/ganglionic cells and diploid Schwann cells [2]. The tumor cells in both groups are usually characterized by whole chromosome 

gains, structural aberrations are found only occasionally. However, aggressive neuroblastomas frequently display diploid tumor cells, which can either 

show amplification of the MYCN oncogene and/or deletions at the short arm of chromosome 1, or display losses at the chromosomal regions 3p, 11q, 14q or 

other aberrations and high telomerase activity. 17q gain was found in both groups. Interestingly, also triploid tumors can acquire MYCN amplifications or 1p 

deletions, or other aberrations thus transforming this benign tumor into a highly aggressive one (dotted line). (MNA = MYCN amplification, LOH = loss of 

heterozygosity, NGF = nerve growth factor). 

Figure modified from the article: 'Neuroblastom' by Peter F. Ambros and Inge M. Ambros in Medizinische Genetik 2:119-124 (2002)

71 

Until recently, it has been assumed that the individual neuroblastic 

tumors are genetically homogenous and that for this 

reason the investigation of a very small area of the tumor (e.g. 

biopsy) is genetically representative of the entire tumor. However, 

various facts led to quite different results. New and more 

sensitive techniques, i.e. serial genetic investigations of individual 

tumors used according to therapy-protocols and neuroblastoma 

screening, discovered also genetically in-homogenous 

tumors. Thus it could be shown that in up to 15% 

of all MYCN-amplified tumors the MYCN-amplification is 

focal, i.e. not ubiquitously detectable in the tumor tissue [4]. 

The discovery of the focal MYCN-amplification implies that 

this genetic change occurs only in the course of tumor progression 

– at least with some tumors – and causes, at least 

temporarily ,heterogeneous’ cell populations within a tumor. 

It may furthermore be assumed that the amplified cells “overgrow” 

the non-amplified cells because of their advanced proliferation 

rate. These observations regarding heterogeneous 

tumors may be unexpected for neuroblastic tumors, but are 

not at all unusual in other tumor types in which amplified and 

non-amplified areas can be found side by side. 

Since the first time that the familial neuroblastomas 

have been described in the middle of the last century, only a 

very few cases have been investigated and the involvement of 

the chromosomal region 16p seems to be important [24]. 

Spontaneous tumor regression 

Complete spontaneous regression of tumors (even of “metastasized” 

tumors) without cytotoxic treatment is a phenomenon 

which scientists have not yet been able to explain. And even 

though spontaneous regression most often occurs in 4s stage 

tumors [12, 14], it is by no means limited to this stage. It can 

be observed in localised tumors [41] in 6 month old patients 

who have been diagnosed through the urinary mass screening 

and even in “classical” stage 4 tumors in the first year of life. 

In tumors which have the ability to regress spontaneously, 

near-triploidy (or polyploidy) was found but no MYCN-amplification 

or 1p-deletion [7] (Fig.1) was detected. The latter 

two genetic aberrations represent markers of aggressive tumor 

behavior. Moreover, the telomerase activity in aggressive 

neuroblastomas is increased as is the case in many other human 

neoplasias [16, 29]. 4s tumors and spontaneously regressing 

tumors however had low or absent telomerase activity. 

The only exceptions are patients with a stage 4s tumor and 

lethal outcome who did actually show an increase in telomerase 

activity [16, 29]. 

Spontaneous tumor maturation 

Spontaneous tumor maturation (which is also not induced by 

cytotoxic treatment) does not occur in patients younger than 

12 months but is observed in patients over one year of age. 

Fully matured tumors, i.e. ganglioneuromas, are usually only 

diagnosed in patients over 3 or 4 years of age. The genetic 

change consists of triploidization of the genome (or penta- or 

hexaploidization respectively) and a lack of chromosome 1p36 

and MYCN-amplification [2]. In all ganglioneuroblastomas 

and ganglioneuromas a diploid cell population can be found 

besides a triploid cell population. Via in situ-hybridization it 

could then be shown that this diploid cell population is made 

up of Schwann cells which usually occur in mature tumors 

[2]. Until then, these Schwann cells, the amount of which increases 

during maturation, were considered to be tumor cells 

(like the ganglionic cells in the tumor) which led to the opinion 

that neuroblastic tumors evolve from “pluripotent” neural 

crest cell [37]. It is however, quite unlikely that a triploid neuroblast 

differentiates into a triploid ganglionic cell and a diploid 

Schwann cell. According to the current model [3], it is 

the neuroblastoma cell which, in the course of their cellular 

differentiation process, start to express chemotactic, mitogenic 

and also differentiation-inducing factors and thus “attract” 

Schwann cells, prompting them to proliferate and differentiate 

[23, 25, 26]. In addition, Schwann cells have an 

antiproliferative effect on NBs while at the same time stimulating 

differentiation [3, 4, 30, 34]. These findings confirm 

the central role of Schwann cells in the maturation process. 

Prognostic impact of bone marrow clearing in stage 4 

patients 

Different approaches were used so far to investigate the prognostic 

value of tumor cell clearing in neuroblastoma patients. 

The techniques mainly used are based on immunological detection 

of GD2 stained cells using different detection systems. 

They range from conventional immunohistochemical detection 

assays [33], fluorescence microscopy [31] and a combined 

approach using fluorescence detection of the immunological 

target and subsequent FISH of unclear results [1, 

Modritz et al. in preparation]. Furthermore, RT-PCR was used 

to detect TH mRNA [17]. The different reports imply a similarity 

to the data in leukemia with their correlation between 

rapid bone marrow clearing and an increased probability of 

survival [28, 38]. However, we have to keep in mind that the 

ideal time point to obtain insights into the dynamics of the 

disease has to be determined in large multi-centre studies. 

References 

1. Ambros IM, Benard J, Boavida M, et al 

(2003) Quality assessment of genetic 

markers used for therapy stratification. 

J Clin Oncol 21: 2077-2084 

2. Ambros IM, Zellner A, Roald B, et al 

(1996) Role of ploidy, chromosome 1p, 

and Schwann cells in the maturation of 

neuroblastoma [see comments]. N Engl 

J Med 334:1505-1511 

3. Ambros IM, Ambros PF (2000) The role 

of Schwann cells in neuroblastoma. In: 

Brodeur GM, Sawada T, Tsuchida Y, 

Voute PA (eds) Neuroblastoma, Elsevier, 

Amsterdam, The Netherlands, pp 229- 

239

4. Ambros IM, Attarbaschi A, Rumpler, S, 

Luegmayr A, Turkof E, Gadner H, Ambros 

PF (2001): Neuroblastoma cells provoke 

Schwann cell proliferation in vitro. Med 

Pediatr Oncol 36:163-168 

5. Ambros PF, Ambros IM (2001) Pathology 

and biology guidelines for resectable and 

unresectable neuroblastic tumors and bone 

marrow examination guidelines. Med 

Pediatr Oncol 37:492-504 

6. Ambros PF, Ambros IM, Kerbl R, et al 

(2001): Intratumoural heterogeneity of 1p 

deletions and MYCN amplification in neuroblastomas. 

Med Pediatr Oncol 36:1-4 

7. Ambros PF, Ambros IM, Strehl S, et al 

(1995) Regression and progression in 

neuroblastoma. Does genetics predict 

tumour behaviour Eur J Cancer 

31A:510-515 

8. Ambros P F, Mehes G, Ambros I M, 

Ladenstein R (2003) Disseminated tumor 

cells in the bone marrow - chances and 

consequences of microscopical detection 

methods. Cancer Lett 197:29-34 

9. Bown, N, Cotterill S, Lastowska M, et 

al (1999) Gain of chromosome arm 17q 

and adverse outcome in patients with 

neuroblastoma [see comments]. N Engl 

J Med 340:1954-1961 

10. Brodeur G M, Ambros PF (2000) Genetic 

and biological markers of prognosis in 

neuroblastoma. In: Brodeur GM, Sawada 

T, Tsuchida Y, Voute PA (eds) Neuroblastoma, 

Elsevier, Amsterdam, The Netherlands, 

pp 355-365 

11. Brodeur GM, Seeger RC, Schwab M, 

Varmus HE, Bishop JM (1984) Amplification 

of N-myc in untreated human neuroblastomas 

correlates with advanced disease 

stage. Science 224:1121-1124 

12. D’Angio GJ, Evans AE, Koop CE 

(1971) Special pattern of widespread 

neuroblastoma with a favourable prognosis. 

Lancet 1:1046-1049 

13. Ejeskar K, Aburatani H, Abrahamsson 

J, Kogner P, Martinsson T (1998) Loss 

of heterozygosity of 3p markers in neuroblastoma 

tumours implicate a tumoursuppressor 

locus distal to the FHIT gene. 

Br J Cancer 77:1787-1791 

14. Evans AE, Chatten J, D’Angio GJ, 

Gerson JM, Robinson J, Schnaufer L 

(1980) A review of 17 IV-S neuroblastoma 

patients at the Children’s Hospital 

of Philadelphia. Cancer 45: 833-839 

15. Guo C, White PS, Weiss MJ, Hogarty 

MD, et al (1999) Allelic deletion at 

11q23 is common in MYCN single copy 

neuroblastomas. Oncogene 18:4948- 

4957 

16. Hiyama E, Hiyama K, Ohtsu K, 

Yamaoka H, Ichikawa T, Shay JW, 

Yokoyama T (1997) Telomerase activity 

in neuroblastoma: is it a prognostic 

indicator of clinical behaviour Eur J 

Cancer 33:1932-1936 

17. Horibe K, Fukuda M, Miyajima Y, 

Matsumoto K, Kondo M, Inaba J, 

Miyashita Y (2001) Outcome prediction 

by molecular detection of minimal residual 

disease in bone marrow for advanced 

neuroblastoma. Med Pediatr 

Oncol 36:203-204 

18. Kogner P, et al (1994) Trk mRNA and 

low affinity nerve growth factor receptor 

mRNA expression and triploid DNA content 

in favorable neuroblastoma tumors. 

Prog Clin Biol Res 385:137-145 

19. Ladenstein R, Ambros IM, Potschger U, 

et al (2001) Prognostic significance of 

DNA di-tetraploidy in neuroblastoma. 

Med Pediatr Oncol 36:83-92 

20. Lastowska M, Cotterill S, Bown N, et 

al (2002) Breakpoint position on 17q 

identifies the most aggressive neuroblastoma 

tumors. Genes Chromosomes Cancer 

34:428-36 

21. Look AT, Hayes FA, Shuster JJ, et al. 

(1991) Clinical relevance of tumor cell 

ploidy and N-myc gene amplification in 

childhood neuroblastoma: a Pediatric 

Oncology Group study. J Clin Oncol 

9:581-591 

22. Luttikhuis ME, Powell JE, Rees SA, et 

al (2001) Neuroblastomas with chromosome 

11q loss and single copy MYCN 

comprise a biologically distinct group 

of tumours with adverse prognosis. Br J 

Cancer 85:531-537 

23. Marchionni MA, Goodearl AD, Chen 

MS, et al (1993) Glial growth factors 

are alternatively spliced erbB2 ligands 

expressed in the nervous system [see 

comments]. Nature 362:312-318 

24. Maris JM, Weiss MJ, Mosse Y, et al 

(2002) Evidence for a hereditary neuroblastoma 

predisposition locus at chromosome 

16p12-13. Cancer Res 

62:6651-6658 

25. Minghetti L, Goodearl AD, Mistry K, 

Stroobant P (1996) Glial growth factors 

I-III are specific mitogens for glial cells. 

J Neurosci Res 43:684-693 

26. Morrissey TK, Levi AD, Nuijens A, 

Sliwkowski, MX, Bunge RP (1995) 

Axon-induced mitogenesis of human 

Schwann cells involves heregulin and 

p185erbB2. Proc Natl Acad Sci USA 

92:1431-1435 

27. Nakagawara A, et al (1993) Association 

between high levels of expression of the 

TRK gene and favorable outcome in 

human neuroblastoma. N Engl J Med 

328:847-854 

28. Panzer-Grumayer ER, Schneider M, 

Panzer S, Fasching K, Gadner H (2000) 

Rapid molecular response during early 

induction chemotherapy predicts a good 

outcome in childhood acute lymphoblastic 

leukemia. Blood 95:790-794 

29. Poremba C, Willenbring H, Hero B, et 

al (1999) Telomerase activity distinguishes 

between neuroblastomas with 

good and poor prognosis [In Process Citation]. 

Ann Oncol 10:715-721 

30. Reynolds ML, Woolf CJ (1993) Reciprocal 

Schwann cell-axon interactions. 

Curr Opin Neurobiol 3:683-693 

31. Saarinen UM, Wikstrom S, Makipernaa 

A, et al (1996 In vivo purging of bone 

marrow in children with poor-risk neuroblastoma 

for marrow collection and 

autologous bone marrow transplantation. 

J Clin Oncol 14:2791-2802 

32. Seeger RC, Brodeur GM, Sather H, 

Dalton A, Siegel SE, Wong KY, 

Hammond D (1985) Association of 

multiple copies of the N-myc oncogene 

with rapid progression of neuroblastomas. 

N Engl J Med 313:1111-1116 

33. Seeger RC, Reynolds CP, Gallego R, 

Stram DO, Gerbing RB, Matthay KK 

(2000) Quantitative tumor cell content 

of bone marrow and blood as a predictor 

of outcome in stage IV neuroblastoma: 

a Children’s Cancer Group Study. 

J Clin Oncol 18:4067-4076 

34. Snider WD (1994) Functions of the 

neurotrophins during nervous system 

development: what the knockouts are 

teaching us. Cell 77:627-638 

35. Spitz R, Hero B, Ernestus K, Berthold 

F (2003) Deletions in chromosome arms 

3p and 11q are new prognostic markers 

in localized and 4s neuroblastoma. Clin 

Cancer Res 9:52-58 

36. Spitz R, Hero B, Ernestus K, Berthold 

F (2003) Gain of distal chromosome arm 

17q is not associated with poor prognosis 

in neuroblastoma. Clin Cancer Res 

9:4835-4840 

37. Tsokos M, Scarpa S, Ross RA, Triche TJ 

(1987) Differentiation of human neuroblastoma 

recapitulates neural crest development. 

Study of morphology, neurotransmitter 

enzymes, and extracellular matrix proteins. 

Am J Pathol 128:484-96 

38. van Dongen JJ, Seriu T, Panzer- 

Grumayer ER, et al (1998) Prognostic 

value of minimal residual disease in 

acute lymphoblastic leukaemia in childhood. 

Lancet 352:1731-1738 

39. van Roy N, Laureys G, Van, GM, et al 

(1997) Analysis of 1;17 translocation 

breakpoints in neuroblastoma: implications 

for mapping of neuroblastoma 

genes. Eur J Cancer 33:1974-1978 

40. Vandesompele J, Van RN, Van GM, et 

al (1998) Genetic heterogeneity of neuroblastoma 

studied by comparative genomic 

hybridization. Genes Chromosomes 

Cancer 23:141-152 

41. Yamamoto K, Hanada R, Kikuchi A, et al 

(1998) Spontaneous regression of localized 

neuroblastoma detected by mass 

screening. J Clin Oncol 16:1265-1269



Neuroblastoma: histoprognostic classification and handling of 

material for biology – a challenge for the pathologist 

Klaus Beiske 





Department of Pathology, 

Rikshospitalet, 

Oslo, Norway 

Abstract 

Peripheral neuroblastic tumors originate from immature elements of the sympathetic nerve system and are 

characterized by a striking morphological, biological and clinical heterogeneity. For about 100 years, pathologists 

could not reach an agreement on a comprehensible and prognostic significant classification system. 

This paper summarizes the contributions made by Hiroyuki Shimada, who proposed to link morphological 

differentiation and cellular turn-over (measured as the number of mitotic and karyorrhectic nuclei) to 

the patients age, and describes the further development of his system into the International Neuroblastoma 

Pathology Classification (INPC) of 1999 with a recent revision of 2003. Pathologists play now a key role in 

the handling of neuroblastic tumors based on macroscopic inspection. They are responsible for an adequate 

sampling and saving of material for morphological, biological and genetic investigations, while the results of 

their microscopic assessment create the basis for a histoprognostic categorization beyond the subsets which 

presently are defined by biological/genetical markers. 

Key words: classification, genetical markers, neuroblastoma, pathology 

Introduction 

Neuroblastomas belong, together with ganglioneuro-blastomas 

and ganglioneuromas, to the group of peripheral neuroblastic 

tumors (pNTs) derived from sympa-thicoblasts (sympathetic 

neuroblasts), i.e. immature precursors of the sympathetic 

nervous system. 

They are considered as embryonic tumors originating 

either before or after birth from immature neural crest cells which 

are “left over” after normal organogenesis. Thus, the primary 

tumors are found in the same anatomical localization as their 

normal counterparts, i.e. in the adrenal medulla, and in sympathetic 

ganglia and paraganglia. 

PNTs are the most common extracranial solid malign 

tumors of childhood (1/10.000 births) and represent 15% 

of all malignant neoplasms occurring during the first four 

years of life [22]. 90% are diagnosed under the age of five 

years [18]. 

The challenge of heterogeneity 

Since the sympathetic neuroblasts of embryonic origin display 

a restricted maturation capacity in individual tumors, 

pNTs are typified by an amazing heterogeneity, both with regard 

to the specter of histological phenotypes, biological properties 

and clinical behavior. 

Already the very first reports which were published 

during the three last decades of the 19 th and the beginning of 

the 20 th century, outlined these neoplasms to comprise features 

of ganglioneuroma [16], neurocytoma/neuroblastoma 

[32] and ganglioneuroblastoma [21]. In 1934, Scherer reported 

on fully differentiated ganglioneuromas which contained nodules 

of either undifferentiated neuroblasts or differentiated 

ganglion cells, which he called “development centers” [23]. 

Thirteen years later, Stout confirmed these observations and 

suggested the designation “composite” ganglioneuromas/neuroblastomas 

for this variant of pNT which today would be 

categorized as nodular ganglioneuroblastoma [28]. 


Klaus Beiske, MD, PhD 


Rikshospitalet 

Sognsvannsveien 20 

N-0027 Oslo, Norway 

Phone: +47 23 07 40 76 

Fax: +47 23 07 14 10 

E-mail: klaus.beiske@labmed.uio.no

74 

Biologically, subsets of pNTs, like neuroblastomas of stage 

4S, can show complete spontaneous regression by massive 

apoptosis [10], while others, e.g. those with MYCN amplification, 

display aggressive proliferation and are frequently diagnosed 

at advanced stages of disease [4]. A third variant could 

be defined as those which grow slowly achieving more or less 

complete maturation. As early as in 1926, Cushing and 

Wolbach proposed a sympathetic neuroblastoma to be able to 

transform into a ganglioneuroma after comparison of tumor 

material removed from the same patient with a time interval 

of 10 years [5]. This idea was later supported by a number of 

reports [6, 7, 12, 30]. Several authors pointed out that more 

mature lesions were associated with higher age of the patient 

[9, 31]. It is not difficult to understand that tumors of such 

great morphological and biological diversity will present with 

various clinical pictures. 

Histoprognostic classification of neuroblastic tumors 

Histopathological classification systems are designed to define 

categories and subtypes of diseases on the basis of commonly 

accepted and recognizable morphological features. To 

make them applicable to surgical pathology, these categories/ 

subtypes should ideally be equivalent to clinicopathological 

entities implying essential clinical information about prognosis 

and appropriate treatment. Considering the above outlined 

heterogeneity of pNTs, it is obvious that a comprehensible 

histoprognostical classification system, based on reproducible 

morphological criteria and conveying valuable clinical 

information, is not created in a trice for these tumors. 

The dilemma of ganglioneuroblastomas. Previous attempts 

to delimit prognostic categories of neuroblastomas and 

ganglioneuroblastomas, focused on the differentiation of 

neuroblasts and proposed to list parameters as nuclear and 

nucleolar size, relative amount of eosinophilic cytoplasm, and 

formation of clumps, rosettes and/or cytoplasmic (nerve) processes 

in an either quantitative [2] or qualitative manner [17]. 

Although ganglioneuromas were widely accepted as the most 

mature form within the specter of neuroblastic tumors, the 

relative amount of Schwannian stroma was not appreciated as 

a parameter of differentiation for less mature tumors like 

ganglioneuroblastomas before the early eighties of the 20th 

century. Nodules of undifferentiated neuroblasts or differentiated 

ganglion cells had indeed been recognized in 

ganglioneuroblastomas (see above), but rather vague designations 

as “development centers” or “composite” tumors indicated 

that the maturational relationship of these nodules to 

the surrounding ganglioneuromatous tissue was poorly understood. 

Some were even interpreted as a sign of focal partial 

or focal complete differentiation [2]. The search for a basic 

agreement to the entity of ganglioneuroblastoma was further 

hampered by a lack of commonly accepted histological 

criteria for this tumor: Some included neuroblastomas with a 

high degree of gangliocytic differentiation, while others required 

a substantial amount of ganglioneuromatous tissue 

between foci of differentiating neuroblasts [12]. Thus, a 

histoprognostic classification of ganglioneuroblastomas fell 

between two stools, i.e. immature neuroblastomas on one side 

and mature ganglioneuromas on the other. 

The Shimada Classification 

The credit for the first histoprognostical classification system 

which includes all peripheral neuroblastic tumors, goes to 

Hiroyuki Shimada [26]. 

Histomorphological differentiation and MKI 

Shimada recognized the amount of Schwannian stroma in neuroblastic 

tumors as a major criterium to distinguish between 

immature (i.e. Schwannian stroma-poor) and mature (i.e. 

Schwannian stroma-rich) lesions. 

The stroma-poor group was further subdivided into undifferentiated 

and differentiating tumors depending on whether 

the tumor contains less or more than 5% gangliocytic differentiated 

neuroblasts. This distinction represented a modification 

of the prognostic system originally proposed by Beckwith 

and Martin for neuroblastic tumors. As a measure of aggressive 

behavior of tumor cells in the stroma-poor group, Shimada 

proposed to analyse the mitosis-karyorrhexis index (MKI), 

i.e. the percentage of mitotic and karyorrhectic nuclei per 5000 

tumor cells. He found that an MKI of 4% as high. Mitotic and 

karyorrhectic cells were summed up in one index because it 

was not always possible to distinguish between both phenomena 

on light microscopic grounds (Fig. 1). 

The stroma-rich group contained three subgroups: (a) 

stroma-rich well differentiated tumors consisting mainly of 

ganglioneuromatous tissue with few immature neuroblastic 

cells, (b) stroma-rich intermixed tumors with nests of variably 

differentiated neuroblastic cells, and (c) stroma-rich nodular 

tumors with often grossly visible, hemorrhagic nodules of 

stroma-poor neuroblastoma displaying microscopically pushing 

borders towards the surrounding stroma-rich tissue. 

Shimada pointed out that his stroma-poor group overlaps with 

tumors called by others “classical neuroblastoma” and “diffuse 

ganglioneuroblastoma”, while the stroma-rich group corresponds 

to tumors previously called “ganglioneuroblastoma” and 

“immature ganglioneuroma”. 

Fig. 1 Undifferentiated neuroblastoma with area of high MKI. Karyorrhectic 

cells often show condensed eosinophilic cytoplasm and hyperchromatic, 

fragmented nuclei. Round nuclei with a regular outer border might be lymphocytes 

and are not accepted as karyorrhectic.

75 

Age 

In his original paper of 1984, Shimada applied these morphological 

criteria on histological sections from 295 neuroblastomas 

and ganglioneuroblastomas recruited from two American, 

one Japanese and one British center. When looking at the 

distribution of the stroma-poor and stroma-rich groups and 

subgroups according to the patients age at diagnosis, he found 

that stroma-rich tumors had their debut at a relatively higher 

age, i.e. the stroma-rich intermixed and nodular ones mainly 

between two and five years, and stroma-rich well differentiated 

ones at the age of five years and older. These findings 

confirmed previous observations (see above) and supported 

the idea that the immature elements in the non-aggressive variants 

of these embryonic tumors have the capability to mature 

into benign ganglioneuromas over a certain number of years. 

On the other hand, a disturbance (e.g. delay) of this 

“normal” sequence of maturation could probably forebode a 

tumor which because of its restricted maturation capacity will 

persist at an immature, biologically more aggressive stage. 

Shimada found strong support also for this hypothesis when 

he compared the age distribution of patients with stroma-poor 

tumors to MKI and clinical outcome. The majority of undifferentiated 

and differentiated stroma-poor tumors with low 

MKI diagnosed under the age of 1,5 year had a favorable outcome, 

while almost all undifferentiated tumors with low MKI 

diagnosed at >1,5 years killed the patients. Stroma-poor tumors 

with high MKI had a bad outcome, and the majority 

occurred in children older than 1,5 year. All children with 

stroma-poor tumors diagnosed at >5 years died. 

The stroma-rich intermixed and well differentiated subgroups 

occurred in children older than 2 and 5 years, respectively, 

and almost all showed a favorable outcome. In contrast, 

most of the stroma-rich nodular tumors had an unfavorable 

prognosis, irrespective of their occurrence at the “correct” 

age of 2-5 years. It seemed probable that the stromapoor 

nodules could represent a malignant outgrowth of biologically 

more aggressive clones within an otherwise maturing 

stroma-rich tumor. 

Taken together, Shimada was the first to prove the usefulness 

of a classification based on a combination of 

histomorphology (i.e. the relative amount of Schwannian 

stroma and gangliocytic differentiation), MKI and age at diagnosis 

and concluded that 

“1) the prognosis of the patients under 1,5 years old is directly 

influenced by the MKI regardless of the degree of maturation, 

2) the prognosis of the patients between 1,5 and 5 years old is 

influenced both by the degree of maturation and the MKI, 

3) the prognosis of the patients over 5 years old is poor regardless 

of the degree of maturation and/or the MKI.” 

Modifications of the Shimada classification 

In 1992, Joshi [12] searched to combine the conventional (pre- 

Shimada) terminology of neuroblastic tumors with the prognostic 

categories of the Shimada classification and suggested 

the following changes/additions: 

(1) The stroma-poor tumors of Shimada should be called 

neuroblastomas and defined as tumors with a neuroblastic 

component of >50% of the total tumor area. 

(2) Shimadas stroma-poor differentiated tumors should be 

splitted up into poorly differentiated neuroblastomas 

(containing neuropil and 5% differentiated 

neuroblasts). 

(3) Shimadas stroma-rich tumors should be called 

ganglioneuroblastoma, nodular, intermixed and borderline 

and should be defined as tumors containing >50% 

ganglioneuromatous tissue. The borderline type was 

meant to replace Shimadas stroma-rich well differentiated 

type. 

(4) Tumors with equal amounts of neuroblastic and 

ganglioneuromatous components should be called “transitional 

neuroblastic tumors”. The term “unclassifiable 

neuroblastic tumor” should be reserved for tumors where 

the histological assessment was hampered by necrosis, 

hemorrhage, calcification, crush artifacts, cystic changes, 

and poor fixation/processing. 

Interestingly, Joshi also performed a histoprognostic categorization 

of the nodules in nine nodular ganglioneuroblastomas 

(see below), without being able to show any significant impact 

on the survival of these few patients. 

In other publications, he proposed a histological grading 

system based on the presence/absence of calcification and 

level of mitotic rate (MR) [13] which he later modified by 

replacing MR with MKI [15]. By combining these (modified) 

histological grades with the patients age, he defined (modified) 

low and high risk groups independent of the morphological 

differentiation. 

The International Neuroblastoma Pathology 

Classification (INPC) 

In 1994, the International Neuroblastoma Pathology Committee 

was established with the goal to achieve a “standardization 

of terminology and morphologic criteria of neuroblastic 

tumors and establishment of a morphologic classification that 

is prognostically significant, biologically relevant, and reproducible” 

[24]. Both Shimada and Joshi were members of this 

committee, and the proposed classification represents by and 

large the original Shimada system of 1984 with a few modifications, 

some proposed previously by Joshi (see above), and 

some performed by the committee. 

Morphological categorization 

The following categories and subtypes of peripheral neuroblastic 

tumors were listed: 

1. Neuroblastoma (Schwannian Stroma-Poor) 

Definition: PNT with 50% or less Schwannian stroma 

a. undifferentiated subtype 

consists of small-to-medium sized neuroblasts without 

neuritic processes (neuropil) or gangliocytic 

differentiation (Fig. 2). Scattered or foci of large or 

pleomorphic-anaplastic cells occur infrequently.

76 

The diagnosis of this subtype of pNT may pend on ancillary 

techniques as e.g. immunohistology (NSE, CD56, tyrosine 

hydroxylase, synaptophysin, chromogranin A and markers to 

exclude myogenic tumors, primitive neuroectodermal tumors/ 

Ewing sarcoma and lymphomas), cytogenetics and elevated 

urinary levels of metabolites of catecholamine synthesis. 

b. poorly differentiated subtype 

consists of neuroblasts and neuropil. Less than 5% of 

neuroblasts show gangliocytic differentiation (Fig. 3). 

Scattered or foci of large or pleomorphic-anaplastic 

cells occur infrequently. 

c. differentiating subtype 

consists of neuroblasts and neuropil. 5% or more of 

neuroblasts show synchronous gangliocytic 

differentiation (Fig. 4), i.e. enlarged, eccentric nucleus, 

vesicular chromatin pattern, prominent nucleolus and 

increased amount of cytoplasm (total cell diameter 

larger than twice the nuclear diameter). Areas of 

Schwannian stroma can be seen at the periphery of tumor 

(transitional zone), but do not exceed 50% of the 

tumor area. 

2. Ganglioneuroblastoma, intermixed (Schwannian 

Stroma-Rich) 

Definition: pNT with >50% ganglioneuromatous tissue which 

includes sharply defined nests of neuroblastic tissue (Fig. 5) 

containing neuropil and a mixture of differentiating neuroblasts 

and and maturing ganglion cells (Fig. 6). 

3. Ganglioneuroma (Schwannian Stroma-Dominant) 

a. Ganglioneuroma, maturing subtype 

consists of >50% ganglioneuromatous tissue with scattered 

differentiating neuroblasts, maturing and mature 

ganglion cells which never lie in sharply demarcated 

nests (Fig. 7). This subtype corresponds to Shimadas 

stroma-rich, well differentiated [26] and Joshis 

ganglioneuroblastoma, borderline. 

b. Ganglioneuroma, mature subtype 

consists of mature Schwannian stroma with fascicular 

neuritic processes and mature ganglion cells 

(Fig. 8). Neuroblasts are no longer present. 

Fig. 2 The undifferentiated subtype of neuroblastoma consists of small-tomedium 

sized cells without neuritic processes or gangliocytic differentiation. 

Fig. 3 In the poorly differentiated subtype, the neuroblasts are surrounded by 

a fibrillary meshwork of neuritic processes (neuropil). A few cells (50% Schwannian 

stroma which includes sharply defined nests of neuroblastic tissue.

77 

4. Ganglioneuroblastoma, Nodular (Composite Schwannian 

Stroma-Rich/Stroma-Dominant and Stroma- 

Poor) 

Definition: pNT with a stroma-rich (like in ganglioneuro-blastoma, 

intermixed) or strom-dominant (like in ganglioneuroma, 

maturing) component which surrounds macroscopically visible, 

often hemorrhagic nodules of a neuroblastic, stroma-poor 

component (Fig. 9 a and b), regarded as a more aggressive 

clone. The nodule(s) show either pushing borders (Fig. 10) or 

a more infiltrative pattern towards the stroma-rich component. 

It is important to be aware of the possibility that a surgical 

biopsy may only contain a small rim of stroma-rich tissue 

around the nodule, i.e. the stroma-rich part will not necessarily 

make up >50% of the visible tumor area. 

5. Terminology for not completely classifiable pNTs 

a. Neuroblastoma (Schwannian Stroma-Poor), NOS 

is reserverd for tumors which clearly belong to the category 

of Schwannian stroma-poor pNT, but which can 

not be allocated to any subtype because of poor quality 

of the sections, bleeding, cystic degeneration, necrosis, 

crush artifact or diffuse calcification. 

Fig. 6 The neuroblastomatous nests of ganglioneuroblastoma, intermixed, 

contain differentiating neuroblasts, maturing ganglion cells and neuropil. 

b. Ganglioneuroblastoma, NOS 

is applied to a stroma-rich tumor which extensive calcification 

which may obscure a stroma-poor nodule. 

c. Neuroblastic tumor, unclassifiable 

is recommended for a pNT which cannot be allocated 

to any of the categories listed above. This may e.g apply 

to very small biopsies. 

Prognostic categorization 

In order to (1) test the consensus between the committee members 

to recognize the above listed morphological categories 

and (2) identify the most powerful system for prognostic 

categorization, the committee reviewed 227 cases of pNTs 

[25] and compared the original Shimada system (based on 

morphology, MKI and age) with Joshi’s histological grades 

(based on MR and calcification), risk groups (based on histological 

grade and age) and their modifications replacing MR 

with MKI, which can only be used for prognostic analysis of 

stroma-poor tumors. The combination of morphology, MKI 

and age, originally proposed by Shimada, had the strongest 

impact on prognosis, measured as event free survival (EFS). 

The statistical analysis also confirmed the age of 1,5 years as 

the borderline of greatest prognostic significance. Beyond the 

age of 5 years, all stroma-poor tumors had a bad outcome. 

Consequently, the International Neuroblastoma Pathology 

Committee included age and MKI, but not MR and calcification 

as relevant prognostic data for stroma-poor tumors 

into the new classification. 

The MKI can be determined applying a method previously 

proposed by Joshi [14] It is based on the initial estimation 

of cell density in different areas of a given tumor (Low: 

100-300, Moderate: 300-600, High: 600-900 cells per high 

power field (HPF) of 400x magnification) and the subsequent 

definition of the number HPFs which are required to evaluate 

5000 tumor cells. The MKI is reported as one of three classes 

(low, intermediate, high) according to the definitions originally 

given by Shimada [26]. 

Fig. 7 Ganglioneuroma, maturing subtype, contains >50% Schwannian stroma 

and single or small aggregates of differentiating neuroblasts, and maturing/mature 

ganglion cells, but no neuropil. 

Fig. 8 Ganglioneuroma, mature subtype, consists of abundant Schwannian 

stroma and scattered mature ganglion cells.

78 

Fig. 9 In ganglioneuroblastoma, 

nodular, a ganglioneuromatous 

component surrounds nodule(s) of 

neuroblastic stroma-poor tumor 

tissue (a). Note lack of protein S- 

100-positive Schwannian stroma 

innside the nodule (b). 

Fig. 10 The neuroblastic nodule (left) shows pushing borders by compressing 

the surrounding ganglioneuromatous tissue (right). 

Table 1 shows the prognostic allocation of the morphological 

categories and subtypes of pNTs as recommended in the INPC. 

The process of reaching a final histoprognostic conclusion 

according to differentiation, MKI and age may appear confusing 

to the inexperienced. For newcomers, it can be recommended 

to learn the favorable categories by heart because their 

number is quite low: poorly differentiated and differentiating 

tumors with low and intermediate MKI below 1,5 years, and 

differentiating neuroblastoma with low MKI between 1,5 and 

5 years. All other neuroblastomas are actually unfavorable. 

Similar to the Shimada classification of 1984, Schwannian 

stroma-rich and dominant tumors have a favorable prognosis 

unrelated to the patient’s age, except for nodular 

ganglioneuroblastomas which all are ascribed an unfavorable 

outcome. 

Revision of the International Neuroblastoma Pathology 

Classification (INPC) 

Both Joshi [12] and the International Neuroblastoma Pathology 

Committee [24, 25] mentioned the possibility of performing 

a prognostic grading of nodular ganglioneuroblastoma 

(GNBn) on the basis of the stroma-poor nodule(s) since this 

component of the tumor had been regarded as the malignant 

one already many years ago [28]. However, the number of 

GNBns in their reviews was too small. 

Umehara et al. [29] were the first to review 70 GNBns 

and to perform a prognostic categorization by analyzing the 

nodule(s) according to the above detailed criteria of the INPC 

for neuroblastomatous tumors (morphology and MKI) and 

relating their data to the patients’ age. These authors also described 

GNBns with single or multiple nodules and variant 

forms which (1) consisted mainly of the stroma-poor component 

surrounded only by a thin, microscopically visible rim 

of stroma-rich/-dominant tissue, and (2) only showed 

ganglioneuromatous tissue in the primary tumor (probably due 

to poor sampling), but a neuroblastomatous component at the 

metastatic site. Depending on the presence of nodule(s) of the 

favorable or unfavorable category, favorable (32,4%) and unfavorable 

(67,7%) subsets of GNBn were established which 

showed statistically significant differences in event free survival 

and survival, quite similar to neuroblastic tumors. 

In 2003, The International Neuroblastoma Pathology Committee 

reviewed the same cohort of tumors and confirmed these 

results [20]. A revision of the INPC was proposed including 

the following terms: GNBn, classic, containing one nodule; 

and variant forms containing either multiple nodules, large 

nodules or no nodule, but a metastatic stroma-poor tumor. 

Unfavorable prognosis was defined by the demonstration of 

at least one nodule displaying features of the unfavorable INPC 

categories of neuroblastomatous tumors. 

The acceptance of a large nodule as a variant of GNBn 

touches on one of the principles of the former INPC, which 

claimed that a tumor belonging to the stroma-rich or -dominant 

category has to be made up by >50% Schwannian stroma 

(see above). It is important for pathologists to be aware of all 

variants of GNBn because they were two times more frequent 

than the classic form among the cases reviewed by Umehara 

and Peuchmaur. 

The greatest advantage of this INPC revision, however, 

is brought by the recognition of a subset (about 30%) 

of nodular ganglioneuroblastomas which, in contrast to previous 

thinking, belong to a prognostic favorable category 

which might benefit from a milder therapeutic regime.

79 

Table 1 

International Neuroblastoma Pathology Classification (INPC). Assignment of morphological categories/subtypes to prognostic 

categories according to MKI and age (modified from Shimada et al. [25]) 

Neuroblastoma 

Schwannian stroma-poor 

favorable 

5 yrs all tumors 

Ganglioneuroblastoma, intermixed 

favorable independent from age 

Ganglioneuroma 

maturing 

mature 

favorable independent from age 

Ganglioneuroblastoma, nodular 

Figure 11 shows the morphologic and prognostic categories 

of the revised INPC. 

Handling of tumor material for biological investigations 

Schwannian stroma-rich 

Schwannian stroma-dominant 

composite Schwannian stroma-rich/ 

stroma-dominant and stroma-poor 

Independent from the above outlined endeavor of pathologists 

towards a clinically informative histoprognostic classification 

system for pNTs, a growing number of biological and genetic 

alterations has been discovered in these tumors during the past 

two decades which today allow an assignment to low-, intermediate- 

and high-risk groups [3]. The prognostic most important 

ones are still MYCN/1p status and ploidy. It is not 

possible to enter a child with a neuroblastic tumor into any of 

the current European clinical protocols without at least knowledge 

about MYCN, analyzed by Southern blot and/or fluorescence 

in situ hybridization (FISH). 

Since many of these investigations either require or 

are most easily performed on unfixed tumor tissue, additional 

tasks have to be assumed by pathologists who usually are the 

first in the hospital to receive fresh tumor tissue from the operating 

theatre. 

The various procedures for handling samples of neuroblastic 

tumors were in detail described by Ambros PF and 

Ambros IM [1], and will not be repeated here. The most central 

functions of the pathologist can be summed up in the following 

three key words: sampling, saving and microscopical 

evaluation. 

Sampling. Taking into account the great morphological 

heterogeneity of these tumors, it is obvious that the pathologist, 

prior to any fixation, should cut the tumor into slices 

searching for e.g. nodules or hemorrhagic areas. It is important 

to take samples from every macroscopic variant, and this 

should proceed under sterile conditions if some of the material 

is planned to be used for culturing. 

The further sub-sampling of any macroscopic area, 

being of interest for special investigations, follows a back-toback 

principle, i.e. a representative piece designated for paraffin 

embedding should be used to produce at least 10 imprints 

from a fresh cut surface (for e.g. FISH or static ploidy 

analysis) before fixation in formalin. The two most closely 

neighbored (“opposite”) tissue slices should be snap frozen 

(for e.g. extraction of DNA, RNA and proteins, immunohistology 

requiring unfixed tissue and interphase cytogenetics if 

imprints were not freshly taken) and put into a sterile culture 

medium (for e.g flow cytometry, conventional cytogenetics 

and culturing), respectively. 

Saving. The adequate saving of the samples depends 

on the individual pathologists knowledge about the techniques 

which are required for the most essential biological investigations. 

Thus, imprints should be air dried over night and then 

kept frozen in air-tight containers at -80º. Snap frozen tissue 

should also be stored at -80º, or in liquid nitrogen if preservation 

of RNA is intended. Frozen tissue cannot be used for 

conventional cytogenetics. Tissue planned to be analyzed by 

cytogenetics and flow cytometry, should be brought as soon 

as possible in phosphate buffered saline or culture medium to 

analysis, although it is possible to recover nuclei for FISH 

and flow cytometry from paraffin embedded tissue if fresh 

material is not available.

80 

Fig. 11 International Neuroblastoma Pathology Classification (INPC) (revised form). Morphological and prognostic categories. FH: favorable histology; UH: 

unfavorable histology; GNBn: ganglioneuroblastoma, nodular; MKI: mitosis-karyorrhexis index. The prognostic categorization of GNBn is performed by 

evaluating the neuroblastomatous component in the nodule(s) according to the criteria for neuroblastomas (modified from Peuchmaur et al., 2003). 

Microscopical evaluation. The pathologist should not only 

analyze morphology and MKI for the histoprognostic classification, 

but also use the microscope for estimation of the tumor 

cell content in the material designed for biological investigations. 

Thus, microscoping a paraffin section from the tissue 

sample which was used for imprinting prior to fixation, 

will provide important information about the possible percentage 

of tumor cells, non-tumor cells (e.g. lymphocytes or 

stroma) and necrosis which might influence the representativity 

of the imprint. A microscopic analysis of the tumor cell content 

performed on frozen sections from the material which is 

intended to be used e.g. for PCR of the 1p36.3 region, is even 

more demanded because the correct interpretation of the PCR 

result requires a tumor cell content of at least 70%. 

It is obvious that the pathologist who is in charge of 

handling samples from neuroblastic tumors, plays a central 

role in providing adequate material for biological analyses. 

Prognostic categorization by INPC and biological 

markers – competing or complementary systems 

It has been known for many years that MYCN amplification 

in neuroblastomas is a strong indicator of aggressive tumor 

behavior [4], and often associated with undifferentiated morphology 

and increased MKI [27]. One might therefore be 

tempted to ask whether the histoprognostic categorization provided 

by the INPC could be replaced by the analysis of MYCN. 

The correlation between INPC categorization and 

MYCN status was investigated by Goto et al. [8] in a large 

cohort of 628 patients with neuroblastic tumors. As expected, 

97,7% of tumors with favorable histology (FH) were MYCN 

non-amplified, while 93,2% of MYCN amplified tumors 

showed an unfavorable histology (UH). However, the comparison 

of MYCN status and INPC categories revealed a third 

and surprisingly large group of tumors with UH but without 

MYCN amplification. As shown in Fig. 12, the prognosis of 

this group was intermediate to FH/non-amplified and UH/ 

amplified tumors, and the difference in EFS between all three 

groups was statistically significant. 

The majority of UH, non-amplified tumors were observed 

at stage 3 and 4 in children older than one year. The 

results of this investigation demonstrate that the 

histoprognostic INPC categorization is able to identify a substantial 

number of patients who, in spite of normal MYCN 

status, belong to a group of inferior prognosis, and it appears 

relevant to ask whether this category of patients should be 

offered a modified form of therapy. 

Further support to this notion comes from the recently 

completed histopathological review of 124 stage 2A and 2B 

MYCN non-amplified trial patients who were treated with 

surgery alone according to the LNESG (Localised Neuroblastoma 

European Study Group) protocol 94.01 (Navarro et al., 

in preparation). The review was performed according to the 

INPC by the seven members of the SIOP European Neuroblastoma 

(SIOPEN) Pathology Reference Panel (Gabriele 

Amann, Klaus Beiske, Catherine Cullinane, Emanuele 

D’Amore, Claudio Gambini, Sam Navarro, Michel 

Peuchmaur), and assigned 20% of these MYCN-negative tumors 

to the unfavorable category. The difference in event free 

and overall survival between the favorable and unfavorable 

subset was statistically significant and underlines the importance 

of the INPC as a valuable tool for prognostic categorization 

of localized, MYCN non-amplified tumors. 

In the forthcoming LNESG2 study, these results have 

already influenced the therapeutic design in case of relapse of 

the localized tumor. It has been decided that the relapse will

81 

receive chemotherapy if the histoprognostic categorization of 

the primary tumor was unfavorable, while the relapse of a 

histologically favorable tumor will be treated with surgery 

alone. However, according to the new protocol, the primary 

tumor’s histoprognostic category will only be implied into 

therapeutic decisions for a relapse, if the histological slides of 

the primary have been reviewed by one of the members of the 

SIOPEN Pathology Reference Panel within six weeks after 

the local pathologist finished his/her report. It is mandatory 

that the decision about favorable or unfavorable histoprognosis 

of the localized tumor is based on at least two pathologists 

from different institutions. If the local pathologist can not agree 

to the conclusion made by the first reviewer, another member 

of the SIOPEN Pathology Reference Panel will be contacted 

for a third opinion. 

Summary 

Pathologists are today ascribed a key role in the diagnostic 

work up of pNTs: 

1. After macroscopic inspection, pathologists are responsible 

for an adequate sampling, saving and quality control of tumor 

material for morphological, biological and genetic investigations. 

2. The result of the pathologist’s microscopic analysis assigns 

a tumor to one of five morphological categories which in case 

of stroma-rich/-dominant tumors (except for nodular 

ganglioneuroblastomas) render the analysis of biological markers 

irrelevant. 

The determination of morphological differentiation and 

MKI enables the pathologist to categorize the 

neuroblastomatous tumors and nodular ganglioneuroblastoma 

into favorable or unfavorable subsets which will influence the 

choice of therapy of a possible relapse in those patients who 

are enrolled into the forthcoming European protocol for localized, 

non- MYCN amplified peripheral neuroblastic tumors. 

Finally, the great importance of obtaining biopsy material, 

which is adequate for histoprognostic assessment, cannot 

be stressed enough. Cytological specimens of tumor cells, 

produced from fine needle aspirates, can be used for biological 

analyses - provided that a pathologist has controlled their 

tumor cell content -, but such samples preclude any 

histoprognostic evaluation. The biological information of a 

MYCN non-amplified status, obtained from cytological 

samples, should not make the clinician live in the belief that 

this is a tumor of good prognosis. Data by Goto et al [8] and 

Navarro et al [19], summarized in this paper, provide clear 

evidence that MYCN non-amplified tumors may “hide” a 

considerable proportion of cases with unfavorable outcome. 

Fig. 12 Event free survival for patients grouped according to FH (favorable 

histology), UH (unfavorable histology) and MYCN status (modified from 

Goto et al., 2001). 

References 

1. Ambros PF, Ambros IM (2001) Pathology 

and biology guidelines for respectable 

and unresectable neuroblastic 

tumors and bone marrow examination 

guidelines. Med Pediatr Oncol 

37:492-504 

2. Beckwith JB, Martin RF(1968) Observations 

on the histopathology of neuroblastoma. 

J Pediatr Surg 3:106-110 

3. Brodeur GM (2003) Neuroblastoma: 

biological insights into a clinical 

enigma. Nat Rev Cancer 3(3):203-216 

4. Brodeur GM, Seeger RC, Schwab M, 

Varmus HE, Bishop JM (1984) Amplification 

of N-myc in untreated human 

neuroblastomas correlates with advanced 

stage of disease. Science 

224:1121-1124 

5. Cushing H, Wolbach SB (1927) Transformation 

of malignant paravertebral 

sympathicoblastoma into benign ganglioneuroma. 

Am J Pathol 3:203-216 

6. Everson TC (1964) Spontaneous regression 

of cancer. Ann N Y Acad Sci 

114:767-770 

7. Fox F, Davidson J, Thomas LB (1959) 

Maturation of sympathicoblastoma into 

ganglioneuroma. Report of two patients 

with 20- and 46-year survicals, respectively. 

Cancer 12:108-116 

8. Goto S, Umehara S, Gerbing RB, et al 

(2001) Histopathology (International 

Neuroblastoma Pathology classification) 

and MYCN status in patients with 

peripheral neuroblastic tumors. Cancer 

92:2699-2708 

9. Grosfeld JL, Skinner MA, Rescorla FJ, 

West KW, Scherer LR (1994) Mediastinal 

tumors in children: experience with 

196 cases. Ann Surg Oncol 1:121-127 

10. Haas D, Albin AR, Miller C, Zoger S, 

Matthay KK (1988) Complete pathologic 

maturation and regression of stage 

IV-S neuroblastoma without treatment. 

Cancer 62:818-825 

11. Horn RC, Koop CE, Kiesenwetter WB 

(1956) Neuroblastoma in childhood. 

Lab Invest 5:106-119 

12. Joshi VV, Cantor AB, Altshuler G, Larkin 

EW, Neill JSA, Shuster JJ, Holbrook CT, 

Hayes FA, Castleberry RP (1992) Recommendations 

for modification of terminology 

of neuroblastic tumors and prognostic 

significance of Shimada classification: 

a clinicopathologic study of 213 

cases from the Pediatric Oncology 

Group. Cancer 69:2183-2196

13. Joshi VV, Cantor AB, Altshuler G, 

Karkin EW, Neill JSA, Shuster JJ, 

Holbrook CT, Hayes FA, Nitschke R, 

Duncan MH, Shochat SJ, Talbert J, 

Smith EI, Castleberry RP (1992) Agelinked 

prognostic categorization based 

on a new histologic grading system of 

neuroblastomas. Cancer 69:2197-221 

14. Joshi VV, Chatten J, Sather HN, 

Shimada H (1991) Evaluation of 

Shimada classification in advanced neuroblastoma 

with special reference to the 

mitosis karyorrhexis index: a report 

from the Children’s Cancer Study 

Group. Mod Pathol 4:139-147 

15. Joshi VV, Rao PV, Cantor AB, Altshuler 

G, Shuster JJ, Castleberry RP (1996) 

Modified histologic grading of neuroblastoma 

by replacement of mitotic rate 

with mitosis karyorrhexis index: a clinicopathologic 

study of 223 cases from 

the Pediatric Oncology Group. Cancer 

77:1582-1588 

16. Loretz W. Ein Fall von gangliosem 

Neurom (Gangliom) (1870) Arch Pathol 

Anat Physiol Klin Med 49:435-437 

17. Mäkinen J (1972) Microscopic patterns 

as a guide to prognosis of neuroblastomas 

in childhood. Cancer 29:1637-1646 

18. Monforte-Munoz H, Kawakami T, 

Stram D, Gerbing R, Matthay KK, 

Lukens J et al (1997) Neuroblastoma in 

children over 5 years of age: recognition 

of a rare and aggressive subset with 

“unconventional” morphology. Mod 

Pathol 10:4P 

19. Navarro S, Amann G, Beiske K, 

Cullinane C, D’Amore E, Gambini C, 

Mosseri V, De Bernardi B, Michon J, 

Peuchmaur M (2004) Prognostic value 

of International Neuroblastoma Pathology 

Committee Classification in 

localised resectable peripheral neuroblastic 

tumors. In preparation 

20. Peuchmaur M, d’Amore ESG, Joshi VV, 

et al (2003) Revisión of the International 

Neuroblastoma Pathology Classification. 

Confirmation of favorable and unfavorable 

prognostic subsets in 

Ganglioneuroblastoma nodular. Cancer 

298:2274-2281 

21. Robertson HE (1915) Das Ganglioneuroblastom, 

ein besonderer Typus im 

System der Neurone. Virchows Arch 

[Pathol Anat] 63:147-168 

22. Ross JA, Severson RK, Pollock BH, 

Robison LL (1996) Childhood cancer in 

the United States: a geographic analysis 

of cases from the Pediatric Cooperative 

Clinical Trials Group. Cancer 

77:201-207 

23. Scherer HJ (1934) Bemerkungen zur 

Empfindlichkeit der Ganglienzelle. 

Virchows Arch Path Anat 293:429-437 

24. Shimada H, Ambros IM, Dehner LP, 

Hata J, Joshi VV, Roald B (1999) Terminology 

and morphologic criteria of 

neuroblastic tumors: recommendations 

by the International Neuroblastoma Pathology 

Committee. Cancer 86:349-363 

25. Shimada H, Ambros IM, Dehner LP, et 

al (1999) The International Neuroblastoma 

Pathology Classification (the 

Shimada system). Cancer 86:364-372 

26. Shimada H, Chatten J, Newton WA Jr, 

Sachs N, Hamoudi AB, Chiba T, 

Marsden HB, Misugi K (1984) Histopathologic 

prognostic factors in neuroblastic 

tumors: definition of subtypes of 

ganglioneuroblastoma and an age-linked 

classification of neuroblastomas. J Natl 

Cancer Inst 73:405-416 

27. Shimada H, Stram DO, Chatten J, et al 

(1995) Identification of subsets of neuroblastomas 

by combined histopathologic 

and N-myc analysis. J Natl Cancer 

Inst 187:1470-1476 

28. Stout AP (1947) Ganglioneuroma of the 

sympathetic nervous system. Surg 

Gynecol Obstet 84:101-110 

29. Umehara S, Nakagava A, Matthay KK, 

Lukens JN, Seeger RC, Stram DO, 

Gerbing RB, Shimada H (2000) Histopathology 

defines prognostic subsets of 

ganglioneuroblastoma, nodular. Cancer 

89:1150-1161 

30. Visfeldt J (1963) Transformation of 

sympathicoblastoma into ganglioneuroma. 

Acta Path Microbiol Scand 

158:414-428 

31. Wahl HR (1914) Neuroblastoma: with 

study of a case illustrating the three types 

that arise from the sympathetic nervous 

system. J Med Res 30:205-260 

32. Wright JH (1910) Neurocytoma or neuroblastoma, 

a kind of tumor not generally 

recognized. J Exper Med xii:556



Intracranial ependymomas in children: correlation of selected 

factors and event-free survival in the group of 142 consecutive 

patients treated at one institution. 





S³awomir Barszcz 1 , Wies³awa Grajkowska 2 , Danuta Perek 3 , Marcin Roszkowski 1 , Monika 

Drogosiewicz 3 , Marta Perek-Polnik 3 , El¿bieta Jurkiewicz 4 , Pawe³ Daszkiewicz 1 

1 

Department of Neurosurgery, 

2 


3 

Department of Oncology, 

4 

Department of Radiology, 

The Children’s Memorial Health Institute, 

Warsaw, Poland 

Abstract 

Ependymomas are among the most frequent intracranial tumors in children. Nevertheless, the identification 

of risk factors and the role of various treatment modalities are still far from standarization. The aim of this 

study was to evaluate the impact of patients’ age, tumor location, extent of surgical resection, degree of 

histological malignancy, type of radiotherapy (RT), and chemotherapy (CHT) on event free survival in the 

group of 142 consecutive patients treated at one institution. Retrospective analysis and comparisons of eventfree 

survival (EFS) probabilities were done in subgroups of patients selected on the basis of various treatemt 

protocols. For the entire group, age of patients, extent of surgical resection, and histological malignancy 

significantly affected EFS. Extent of surgical resection significantly influenced EFS in benign tumors and 

did not in anaplastic ependymomas. Neuroaxis RT did not improve EFS for anaplastic lesions. Treatment 

results were better after 1997 in anaplastic ependymomas (potential role of CHT) and finally CHT significantly 

improved control of small tumor residues 

Key words: chemotherapy, ependymoma, prognosis, radiotherapy, surgery 

Introduction 

Ependymomas are among the most frequent intracranial tumors 

in children, accounting for approximately 10% of primary 

central nervous system neoplasms [35]. Treatment of 

patients with ependymomas includes neurosurgical resection, 

radiotherapy (RT), chemotherapy (CHT), and radiosurgery 

[16, 33]. Despite introduction of various treatment regimens, 

treatment outcomes are still unsatisfactory [27, 31, 36]. 

The aim of the study was to analyse the influence of 

selected factors on event-free survival (EFS) in the group of 

intracranial ependymomas treated at our institution. 

Material and methods 

In the years since 1981 thru 2003, a total number of 154 children 

suffering from intracranial ependymomas were treated 

at the Children’s Memorial Health Institute in Warsaw. Twelve 

patients were lost to follow-up, and the remaining group of142 

children (74 boys and 68 girls) were used for further analysis. 

Retrospective analysis of the data was performed to 

evaluate the influence of patients’age, tumor location, extent 

of surgery, tumor malignancy, type of radiotherapy (RT) and 

use of chemotherapy (CHT) on event-free survival (EFS), 

according to various treatmet protocols. The group of patients 

was heterogenous (various treatment protocols, mainly before 

and after 1997). Age of patients, extent of resection, and tumor 

malignancy were the main factors influencing selection 


S³awomir Barszcz, MD, PhD 

Department of Neurosurgery 

Children's Memorial Health Institute 

Al. Dzieci Polskich 20 

04-853 Warsaw, Poland 

Phone: +48 22 8157560 

E-mail: slawomirbarszcz@wp.pl

84 

of treatment options. In such circumstances the mentioned 

factors with potential influence on survival were not independent 

verieties, and multivariate analysis could not be performed. 

We made a 2-step statistical analysis. First, comparability 

of the subgroups of benign and anaplastic ependymomas 

was ascertained, according to age, extent of surgery, type 

of RT, and use of CHT. Then, we made a few separate comparisons 

for comparable subgroups of patients in order to 

analyse the potential influence of risk factors on actuarial EFS. 

EFS probability curves were drawn, and their stratification 

was studied using the log-rank test. 

We did not study overall survival (OS), because recurrences 

of tumors as well as their treatments were not analysed. 

Results 

Age of the patients ranged from 5 months to 18 years and 6 

months (median 5 years 0 months). In 63 patients ependymomas 

were located in the infratentorial space, and in 59 children 

supratentorial tumors were found. In 61 patients the tumors 

were totally, in 38 subtotally, and in 38 children partially 

resected. In the remaining 5 cases only biopsies were 

performed. In the years since 1981 thru 1999, the extent of 

surgical tumor removal was evaluated on the basis of surgical 

reports and postoperative control CTs. Since 1999, the intraoperative 

estimation as well as postoperative MRI studies have 

been used to determine the extent of surgical resection. All 

pathological specimens were reviewed and devided into two 

subgroups of 89 benign and 53 malignant (anaplastic) ependymomas. 

Anaplastic tumors matched the following critera: 

marked nuclear pleomorphism, high mitotic activity, necrosis, 

vascular proliferation. 118 patients were treated with radiotherapy. 

Radiotherapy significanty changed over the described 

period of time, due to the developement of cobalttherapy 

and linac technology as well as introduction of new 

treatment protocols. To simplify the variety of problems concerning 

radiotherapy, for further analysis we defined only two 

main subgroups: 1. “local” radiotherapy (29 patients) and 2. 

neuroaxis radiotherapy (89 patients). Untill 1996, irrespective 

of the extent of surgical resection, all patients underwent 

whole neuroaxis irradiation. Since1996 thru 2000, patients 

after total removal of benign ependymomas were treated with 

“local” radiotherapy. In 2000 the “wait and see” approach was 

introduced after total removal of benign tumors. Partial resections 

of benign ependymomas were followed by chemotherapy 

and “local” RT. Anaplastic ependymomas after surgery were 

treated with adjuvant chemotherapy followed by tumor bed 

and ventricles irradiation, and maintenance chemotherapy. Two 

different treatment approaches were tailored for children under 

3 years of age. Total resection of benign ependymoma 

was the only treatment modality. The rest of patients below 3 

years of age (partial resection of benign tumor and all patients 

with anaplastic ependymomas) were treated with 18-months’ 

chemotherapy without RT. The use of the whole neuroaxis 

RT was reserved only for disseminated tumors. Chemotherapy 

was used for the treatment of 58 patients. The median followup 

period was 3 years and 2 months (ranging from 2 months 

to 18 years and 6 months). 

In the study group there were no significant differences 

between patients with benign and anaplastic tumors according 

to age and extent of resection. The use of neuroaxis vs. local RT 

were close to the limit of signifficance (Table 1). The entire 

study group was devided into 3 age-dependent subgroups. There 

were 41 children below 3 years of age. In 75 patients, age ranged 

from 3 to 10 years, and 26 patients were older than 10 years. 

The list of subsequent comparisons of potential risk factors, 

outcomes and statistical analysis are shown in Table 2. 

The study revealed that age of patients significantly influenced 

EFS.The lower the age, the worse the prognosis (Fig. 

1). No differences were found concernig event-free survival 

probabillities for the group of 83 patients with infratentorial 

tumors vs. 59 patients with supratentorial ependymomas. Primary 

tumor location did not influence EFS. Diagnosis of malignant 

(anaplastic) ependymoma according to the mentioned 

morphologic criteria was a marker of poor prognosis in the study 

group (Fig. 2). Event-free survival probability for the group of 

99/142 patients after total and subtotal tumor resection was better 

than for the group of 43/142 patients after partial resection and 

biopsy. Extent of surgical resection significantly influenced EFS 

for the entire study group. Extent of surgical removal significantly 

influenced EFS for benign tumors. The prognosis was 

better for patients after total and subtotal resection of benign 

ependymomas (Fig. 3). Event-free survival probability after total 

and subtotal resection of grade III ependymomas was slightly 

better than after less radical resection, but stratification of survival 

curves was not significant. The comparison of EFS probabilities 

for benign ependymomas treated with neuroaxis and 

“local” RT revealed that EFS was better for the subgroup after 

whole neuroaxis irradiation. It is rather confusing, taking into 

account everyday clinical practice. Before 1996, all the cases 

of intracranial ependymomas irrespective histological malignancy 

and extent of surgical resection received whole neuroaxis 

RT, whereas after 1996 neuroaxis irradiation was reserved for 

disseminated lesions only. There was no statistically significant 

difference between EFS probabilities for neuroaxis vs. “local” 

(tumor bed or tumor bed + ventricles) RT for the patients 

with malignant ependymomas. It was shown that introduction 

(1997) of the standardized treatment protocols including: surgery, 

adjuvant chemotherapy, tumor bed or tumor bed + ventricles 

radiotherapy, and finally maintenance chemotherapy significantly 

improved EFS probability for malignant ependymomas 

(Fig. 4). Statistically significant stratification of EFS curves 

for the patients treated with chemotherapy vs. patients treated 

without chemotherapy after total or subtotal tumor removals 

indicates that chemotherapy improved control of small tumor 

residues (Fig. 5). 

Discussion 

Identification of clinical and pathological factors influencing 

survival has a key role in improvement of treatment outcomes 

in children suffering from intracranial ependymomas. Various 

risk factors were proposed, and because of relatively small 

and heterogenous groups of patients and different treatemt 

protocols, the results of numerous studies differ considerably

85 

Table 1 

Comparison between ependymoma grade I - II and ependymoma grade III patients according to: age, extent of surgery, type of 

RT, and use of CHT (NS=not significant) 

Table 2 

Feature Benign vs. anaplastic (ch 2 ) Significance 

Age P > 0.05 NS 

Surgery P> 0.05 NS 

RT (CNS vs. local) P = 0.06 +/- 

CHT P < 0.0005 ∗∗∗ 

List of comparisons of the studied factors, and their significance (NS=not significance, * , ** , *** = degrees of statistical 

significance) 

Comparisons of EFS probabilities for the studied parameters Log-rank Significance 

Three age groups p= 0.0039 ∗∗ 

Infratentorial vs. supratentorial ependymomas p=0.37 NS 

Benign vs. anaplastic ependymomas. p=0.0037 ∗∗ 

Total and subtotal resections vs. partial resections and biopsies 

p=0.0001 

∗∗∗ 

(entire study group). 


p=0.00005 

∗∗∗ 

(benign ependymomas). 


p=0.15 

NS 

(anaplastic ependymomas). 

Neuroaxis RT vs. “local” RT 

(benign ependymomas) 

p=0.007 

∗∗ 

Neuroaxis RT vs. “local” RT 

(anaplastic ependymomas) 

p=0.6326 

Anaplastic ependymomas treated before 1997 vs. after 1997. p=0.043 ∗ 

Totally and subtotally removed tumors CHT (+) vs. CHT (-) p=0.047 ∗ 

NS 

[2, 6, 9, 19, 25]. Lower age of patients was found to be a poor 

prognostic factor in most series [3]. It was not clear wheather 

poorer prognosis in very young children was due to different 

tumor biology, different treatement protocols, higher morbidity 

and poorer response to various treatment modalities especially 

radiotherapy [34]. Our results confirmed that observation. 

EFS probabilities were lower for the subgroups of 

younger children. The influence of primary tumor location on 

survival was difficult to evaluate reliably. Predominance of 

infratentorial ependymomas in the first decade of life and 

higher frequency of supratentorial tumors in adolescents and 

adults as well as precise tumor location have led to different 

treatment outcomes in particular subgroups of patients. No 

differences were found in the study group concerning two main 

localization groups of supra- and infratentorial ependymomas. 

However, it should be remembered that surgical mortality and 

morbidity as well as extent of surgical resection are more favorable 

in supratentorial extraaxial and midline posterior fossa 

ependymomas as compared to midline supratentorial and posterior 

fossa paramedian (cerebellopontine angle) tumors [4, 

20]. Extent of surgical resection was the most consistent and 

strongest prognostic factor in almost all of series as well as in 

our group of patients [24, 35]. The main problem in tailoring 

treatment and predicting prognosis is to define reliable features 

in the diagnosis of malignant (anaplastic) ependymomas. 

Uncritical use of the morfologic criteria of histologic 

malignancy accepted for astrocytomas is doubtful. In our group 

of patients, actuarial event-free survival was significantly better 

for the so called benign ependymomas, but a lot of authors 

indicate that there is no strict correlation between tumor morphology 

and prognosis in ependymomas [11, 29, 30]. Much 

time and effort were spent to study various immunohistochemical 

and genetic features as well as their correlations with tumor 

biology, but there are still no reliable markers of malig-

86 

Fig. 1 Statistically significant stratification of event-free survival probability 

curves for various age groups of patients (log-rank: p= 0.0039) 

Fig. 2 Significant stratification of EFS probabnility curves for benign (B - 89 

patients) and anaplastic (A- 53 patients) tumors (log-rank: p=0.0037) 

Fig. 3 Significantly higher EFS after total and subtotal resections (T+ST) vs. 

partial resections and biopsies (P+B) of benign ependymomas (log-rank: 

p=0.00005) 

Fig. 4 Significant improvement of treatment outcome of anaplastic ependymomas 

after 1997 (introduction of complex treatment protocols: surgery + 

adjuvant CHT + RT + maintenance CHT; log-rank: p=0.043) 

Fig. 5 Better EFS after chemotherapy (CH +) in patients in whom total or 

subtotalal tumor resections were performed (log-rank: p=0.047)

87 

nancy in ependymomas [7, 21]. Opinions on adjuvant therapy 

differ from series to series. It is generally accepted that radiotherapy 

and chemotherapy can delay tumor recurrence but their 

influence on overall survival is still unclear [1, 5, 23, 28]. The 

rationale for agressive removal of ependymomas in certain 

locations is debatable. Devastating late complications of 

neuroaxis radiotherapy, especially in the youngest age group 

are well known [17, 18, 32]. That is why there is a trend to 

eliminate radiotherapy in children under 3 years of age and 

after total removal of histologically benign tumors. The fields 

of irradiation were restricted to tumor bed or tumor bed and 

ventricles for partially resected benign and anaplastic lesions 

respectively. The neuroaxis irradiation is usually reserved for 

disseminated ependymomas [12, 26]. This tendency is supported 

by the observations that reccurences are limited to primary 

tumor location in almost all cases and CSF dissemination 

is relatively rare [13]. Patterns of failure indicate, that 

final treatment outcome depends mostly on the local control 

of disease. These statements were the reason for the concept 

of “second-look” surgery, whenever reoperation was possible 

in cases of residual disease [8]. Chemotherapy was introduced 

in the treatment of children under 3 years to delay or 

eliminate the need for radiotherapy as well as in the treatment 

of malignant lesions and tumor recurrences and disseminations 

and finally in the treatment of partially resected benign 

lesions [15, 37]. Variable response rates to the same treatment 

protocols created the need for potential identification of subpopulations 

sensitive or resistent to chemotherapy, 

developement of novel drugs and intensification of treatment 

with subsequent bone marrow reconstruction [10, 14, 22]. As 

mentioned above, in our series neuroaxis irradiation correleated 

with better actuarial EFS for benign ependymomas. Significant 

change of treatment protocols over the years resulted in 

fairly confusing data. Before 1996, all cases were treated with 

neuroaxis irradiation; after 1996, RT was reserved only for 

disseminated ependymomas. It was clearly shown that 

neuroaxis RT did not improve EFS in anaplastic ependymomas 

in our series. An introduction of standarized treatment 

protocols (including chemotherapy) in 1997 improved treatment 

outcomes in malignant tumors. Chemotherapy also improved 

control of small tumor residues after surgery. Treatment 

response of tiny remnants seems to be an interesting 

clinical model for evaluation of CHT efficacy in the treatment 

of ependymomas. 

Conclusions 

1. For the entire group of patients, age, extent of surgical 

resection, and histological malignancy, significantly influenced 

EFS; 

2. Extent of surgical resection significantly influenced EFS in 

benign tumors and did not influence EFS in anaplastic 

ependymomas; 

3. Neuroaxis RT did not influence EFS in patients with anaplastic 

ependymomas; 

4. Treatment results in anaplastic ependymomas improved after 

1997 (a potential role of CHT); 

5. CHT significantly improved control of small tumor residues. 

References 

1. Bouffet E, Foreman N (1999) Chemotherapy 

for intracranial ependymomas 

(review). Child’s Nerv Syst 15:563-323 

2. Bouffet E, Perilongo G, Canete A, et al 

(1998) Intracranial ependymomas in 

children: a critical review of prognostic 

factors and a plea for cooperation. Medical 

and Pediatric Oncology 30:319-331 

3. Duffner PK, Krischer JP, Sanford RA, 

et al (1998) Prognostic factors in infants 

and very young children with intracranial 

ependymomas. Pediatr Neurosurg 

28:215-222 

4. Ernestus R, Schroder R, Stutzer H, et al 

(1996) Prognostic relevance of localization 

and grading in intracranial ependymomas 

of childhood. Child’s Nerv Syst 

12:522-526 

5. Evans AE, Anderson JR, Lefkowitz- 

Boudreaux IB, et al (1996) Adjuvant 

chemotherapy of childhood posterior 

fossa ependymoma: caraniospinal irradiation 

with or without adjuvant CCNU, 

vincristine, and prednisone: a Childrens 

Cancer Group study. Med Pediatr Oncol 

27:8-14 

6. Figarella-Branger D, Civatte M, Bouvier-Labit 

C, et al (2000) Prognostic factors 

in intracranial ependymomas in 

children. J Neurosurg 93:605-613 

7. Figarella-Branger D, Gambarelli D, 

Dollo C, et al (1991) Infratentorial 

ependymomas of childhood. Correlation 

between histological features, immunohistological 

fenotype, silver nucleolar 

staining values and post-operative survival 

in 16 cases; Acta Neuropathol 

82:208-216 

8. Foreman NK, Love S, Gill SS, Coakham 

HB (1997) Second-look surgery for incompletely 

resected fourth ventricle 

ependymomas: technical case report. 

Neurosurgery 40:856-860 

9. Foreman NK, Love S, Thorne R (1996) 

Intracranial ependymomas: analysis of 

prognostic factors in a population-based 

series. Pediatr Neurosurg 24:119-125 

10. Geddes JF, Vowles GH, Ashmore SM 

(1994) Detection of multidrug resistance 

gene product (P-glycoprotein) expression 

in ependymomas. Neuropathol 

Appl Neurobiol 20(2):118-121 

11. Gerszten PC, Pollack IF, Martinez AJ 

(1996) Intracranial ependymomas of 

childhood. Lack of correlation of histopathology 

and clinical outcome. Path 

Res Pract 192:515-522 

12. Goldwein JW, Corn BW, Finlay JL, et 

al (1991) Is craniospinal irradiation required 

to cure children with malignant 

(anaplastic) intractranial ependymomas 

Cancer 67:2766-2771 

13. Goldwein JW, Glause TA, Packer RJ, 

et al (1990) Recurrent intracranial 

ependymomas in children. Survival, 

patterns of feilure, and prognostic factors. 

Cancer 66: 557-563 

14. Grill J, Kalifa C, Doz F, et al (1996) A 

high-dose busulfan thiotepa combination 

followed by autologous bone marrow 

transplantation in childhood recurrent 

ependymoma. A phase II study. 

Pediatr Neurosurg 25:7-12 

15. Grill J, LeDeley MC, Gambarelli D, et 

al (2001) Postoperative chemotherapy 

without irradiation for ependymoma in 

children under 5 years of age: a 

multicenter trial of the French Society 

of Pediatric Oncology. J Clin Oncol 

19:1288-1296

16. Grill J, Pascal C, Chantal K (2003) 

Childhood ependymoma a systemic review 

of treatment options and strategies. 

Paediatr Drugs 5(8):533-543 

17. Grill J, Renaux VK, Bulteau C, et al 

(1999) Long-term intellectual outcome 

in children with posterior fossa tumors 

according to radiation dose and volumes. 

Int J Radiation Oncology Biol 

Phys 45:137-145 

18. Hoppe-Hirsch E, Brunet L, Laroussinie 

F, et al (1995) Intellectual outcome in 

children with malignant tumors of posterior 

fossa: influence of the field of irradiation 

and quality of surgery. Child’s 

Nerv Syst 11(6):340-345 

19. Horn B, Heideman R, Geyer R, et al 

(1999) A multiple-institutional retrospective 

study of intracranial ependymomas 

in children: identification of risk 

factors. J Pediatr Hematol Oncol 

21:203-211 

20. Ikezaki K, Matsushima T, Inoue T, et al 

(1993) Correlation of microanatomical 

localization with postoperative survival 

in posterior fossa ependymomas. Neurosurgery 

32(1):38-44 

21. Kordek R, Liberski PP (2001) 

Nowotwory gleju wyœció³kowego. Pol 

J Pathol 53, 4 (suppl):39-53 (in Polish) 

22. Mason WP, Goldman S, Yates AJ, et al 

(1998) Survival following intensive chemotherapy 

with bone marrow reconstruction 

for children with recurrent intracranial 

ependymoma – a report of the 

Children’s Cancer Group. J Neurooncol 

37:135-143 

23. McLaughlin MP, Marcus RB, Buatti 

JM, et al (1998) Ependymoma: results 

prognostic factors and treatment recommendations. 

Int J Radiation Oncology 

Biol Phys 40:845-850 

24. Palma L, Celli P, Mariottini A (2000) 

The importance of surgery in supratentorial 

ependymomas. Child’s Nerv Syst 

16:170-175 

25. Paulino AC, Bouffet E, Perilongo G, et 

al (1998) Intracranial ependymomas in 

children: a critical review of prognostic 

factors and a plea for cooperation. Medical 

and Pediatric Oncology 30:319-331 

26. Paulino AC (2001) The local field in 

infratentorial ependymoma: does the 

entire posterior fossa need to be treated 

Int J Radiation Oncology Biol Phys 

49:757-761 

27. Pollack IF, Gerszten PC, Martinez AJ, 

et al (1995) Intracranial ependymomas 

of childhood: long-term outcome and 

prognostic factors. Neurosurgery 

37(4):655-666 

28. Robertson PL, Zeltzer PM, Boyett JM, 

et al (1998) Survival and prognostic factors 

following radiation therapy and chemotherapy 

for ependymomas in children: 

a report of the Children’s Cancer Group. 

J Neurosurg. 88:695-703 

29. Rorke LB (1987) Relationship of morphology 

of ependymomas in children to 

prognosis. Prog Exp Tumor Res 30:170- 

174 

30. Ross GW, Rubinstein LJ (1989) Lack 

of histopathological correlation of malignant 

ependymomas with postoperative 

survival. J Neurosurg 70:31-36 

31. Schiffer D, Giordana MT (1998) Prognosis 

of ependymoma. Child’s Nerv 

Syst 14:357-361 

32. Seaver E, Geyer R, Sulzbacher S, et al 

(1994) Psychosocial adjustment in longterm 

survivors of childhood 

meduloblastoma and ependymoma 

treated with craniospinal irradiation. 

Pediatr Neurosurg 20(4):248-253 

33. Smyth MD, Horn BN, Russo C, et al 

(2000) Intracranial ependymomas of 

childhood: current management strategies. 

Pediatr Neurosurg 33:138-150 

34. Suc E, Kalifa C, Brauner R, et al (1990) 

Brain tumors under the age of tree. The 

price of survival. Acta Neurochir 

106:93-98 

35. Sutton LN, Goldwein JW, Schwartz D 

(1999) Ependymoma. In: Albright AL, 

Pollack IF, Adelson PD (eds) Principles 

and practice of pediatric neurosurgery. 

Thieme, New York, Stuttgart, pp 609- 

628 

36. Teo Ch, Nakaji P, Symons P, et al (2003) 

Ependymoma. Child’s Nerv Syst 

19:270-285 

37. White L, Johnston H, Jones R, et al 

(1993) Postoperative chemotherapy 

without radiation in young children with 

malignant non-astrocytic brain tumors. 

A report from the Australia and New 

Zeland Childhood Cancer Study Group 

(ANZCCSG). Cancer Chemoter 

Pharmacol 32(5):403-406



Pathology of fatty liver in children with LCHAD deficiency 

Katarzyna Iwanicka 1 , Janusz Ksi¹¿yk 2 , Monika Pohorecka 3 , Maciej Pronicki 1 





1 


2 

The Clinic of Pediatrics, 

3 

Department of Metabolic Diseases, 



Abstract 

Deficiency of long chain 3-hydroksyacyl-CoA dehydrogenase (LCHADD) impairs the process of mitochondrial 

beta-oxidation of fatty acids. LCHADD may be clinically benign up to the moment of first symptoms, 

i.e. fasting inducing the first acute life-threatening episode (ALTE). Patients present vomiting, lethargy, convulsions, 

arrhytmia and hypoketotic hypoglycemia, which is essential for established a diagnosis. Thus, after 

fasting, healthy looking child may die with serious symptoms of acute illness. The aim of this study was the 

histopathological assessment of liver in our own LCHADD patients with particular attention to characteristic 

changes including intensity, type and localization of liver steatosis. Seven liver samples were assessed: 3 

biopsies and 4 autopsy specimens obtained from 6 patients, aged from 3 months to one year. In 3 autopsies 

we found severe macrovacuolar fatty liver, in one biopsy moderate mixed steatosis, and in remaining 3 cases 

mild fatty liver of macrovacuolar and mixed character. All samples but one displayed diffuse distribution of 

steatotic hepatocytes. In 4 cases steatosis was accompanied by fibrosis and cirrhosis, in 3 by inflammatory 

infiltrates. Intensity of steatosis was related to the severity of clinical symptoms. Fibrosis was another morphological 

feature characteristic for LCHAD deficiency. The results showed that liver steatosis found during 

autopsy of patients who died of unexplained reason should attract attention to LCHADD as a possible cause 

of death. 

Key words: LCHAD deficiency, liver steatosis 

Introduction 

Clinical symptoms of long chain 3-hydroksyacyl-CoA dehydrogenase 

(LCHAD) deficiency usually appear after prolonged 

fasting or could be evoked by upper respiratory tract infections, 

vaccination, physical exercises and other stress. Clinically 

healthy child presents hypoketotic hypoglycemia, which 

may be accompanied by vomiting, abdominal pain, convulsion, 

muscle weakness and cardiomegaly. In the course of the 

disease sensorimotor neuropathy and pigmentary retinopathy 

develop [5, 9, 11]. The age of onset ranges from one day to 40 

months. Treatment involves the diet with long chain fatty acids 

restriction and increased supply of carbohydrates. 

Fatty acids are the most important source of energy 

during fasting [5, 11]. They are transported into the inner 

membrane of mitochondria to be used in beta-oxidation process 

[5, 9, 11]. This process has especially important value 

for muscle, heart and liver cells as their own source of energy. 

Beta-oxidation process has two steps: transport and following 

oxidation. Three enzymes take parts in transport of fatty 

acids into the mitochondria, creating the pathway called carnitine 

transporting system [2, 3, 5, 12]. Proper beta-oxidation 

pathway depends on activity of several enzymes (four for long, 

four of medium and four for short chain fatty acids). The names 

of appropriate enzymes differ only by the name of length of 

the chain. The basis of beta-oxidation process is shortening of 

fatty acid chain by cutting acetyl-CoA to introduce it to tricarboxylic 

acids cycle. Deficiency of 3-hydroksyacyl-CoA dehydrogenase 

(LCHAD) impairs the process and leads to toxic 

accumulation of fatty acids and their acylcarnitines derivatives 

[2, 5, 7, 9, 11]. Accumulation of lipids in the cells may 

be a site effect of this metabolic decompensation. 


Iwanicka Katarzyna 


The Children's Memorial Health Institute 

Aleja Dzieci Polskich 20, 


Phone: +48 22 815 16 14 

E-mail: kasiai@yahoo.com

90 

Since 1994, twenty two unrelated children with LCHAD deficiency 

have been identified in Poland. Diagnosis was done 

on the analysis of GC-MS organic acid profile in urine and 

than confirmed by genetic investigation. Common protocol 

of treatment and monitoring (CPTM) has been introduced in 

2000. There were 8 LCHAD-deficient children identified before 

CPTM. Diagnosis was markedly delayed in all cases, 

rarely established at first acute life-threatening episode 

(ALTE), frequently post mortem. Five of them died, only three 

patients were in good condition. In second period (since 2000) 

14 new LCHAD deficient cases were identified. The diagnosis 

was established usually at first ALTE, only in 4 patients 

too late (post mortem or post-episodic neurological sequels). 

As in the literature there are only few reports concerning 

histological pattern of the liver from patients with LCHAD 

deficiency [5, 7, 9, 11], the aim of our study was to analyse 

morphological findings in the liver of 6 patients with LCHAD 

deficiency to establish the spectrum of morphological changes 

and compare our results with the described reports. 

Results 

Case No 1 

The liver specimen has been obtained on autopsy. It revealed 

the diffused localization of mixed pattern of steatosis concerning 

of 100 % of hepatocytes. Apart of congection any 

other pathological change was not detected in the liver. Steatosis 

of cardiomyocytes and lipid accumulation in renal proximal 

tubules was observed as well (Fig. 1). 


Material consists of seven liver samples obtained from six 

LCHAD deficient patients (4 boys and 2 girls). One of the 

patients underwent liver biopsy at the age of 3 months and 

post mortem examination at 6 months. There were three biopsy 

samples obtained from the LCHAD-deficient boys and 

four autopsy specimens derived from affected patients (two 

boys and two girls). Patients were aged from 3 to 12 months. 

Clinical diagnosis of LCHAD deficiency was based on urine 

organic acid profile analysis using GC-MS method (Dept of 

Laboratory Diagnostics, CMHI), and confirmed by molecular 

study. 

The liver samples obtained during biopsies were prepared 

for light and electron microscopy. Slides were stained 

with hematoxylin and eosin, PAS, diastase digested PAS, for 

collagen and reticulin fibres – AZAN and Gomori stains, respectively. 

Semi-thin slides of plastic embedded (epon) liver 

samples post fixed in osmium tetroxide and stained with toluidine 

blue were used for steatosis detection. Liver tissue samples 

obtained during autopsies were processed routinely for histology 

and stained with hematoxylin and eosin. 

During microscopical assessment we assumed three 

grades of steatosis intensity: mild, moderate and severe, when 

lipid droplets comprised up to 30%, 30-70%, and over 70% 

of liver parenchyma, respectively. Two types of distribution 

of steatotic hepatocytes were recorded: diffuse or focal. Type 

of steatosis included three categories: macrovacuolar, 

microvacuolar, and mixed. Semi-thin slides are very helpful 

in correct lipid detection, especially of microvacuolar type. 

Accompanying inflammatory infiltrates and fibrosis were assessed 

according to routine grading and staging system. 

Fig. 1 Liver biopsy of patient no 1 reveals severe mixed steatosis. Presence 

of lipid accumulation in renal proximal tubules (of patient no 1). 

Case No 2 

The liver specimen has been obtained on autopsy. It revealed 

the typical pattern of fatty liver. Macrovesicular lipid droplets 

are present in almost all hepatocytes. Mild fibrosis was present 

only in periportal areas. Inflammatory infiltrates and 

cholestasis were not detected. There was found vacuolisation 

corresponding probably to lipid accumulation in proximal renal 

tubules (Fig. 2).

91 

Fig. 2 Liver biopsy of patient no 2 shows severe diffuse macrovesicular steatosis. 

Liver biopsy of patient no 3 shows diffuse, moderate, mixed steatosis. 

Mild intralobular inflamatory infiltrates of lymphoid cells and single necrotic 

hepatocytes are observed. 

Case No 3 

The liver tissue obtained during biopsy showed fatty changes 

in approximately 60 % of hepatocytes. They revealed diffuse 

distribution. Lipid droplets in hepatocytes were of mixed pattern: 

macrovacuoles and microvacuoles were present. The mild 

intralobular and portal inflamatory infiltrates of lymphoid cells 

were found. Single necrotic hepatocytes were observed. Fibrosis 

and cholestasis were not detected (Fig. 2). 

Case No 4 

The liver biopsy specimen obtained prior to autopsy revealed 

precirrhotic pattern of liver injury. Macrovesicular steatotis were 

seen in 30 % of hepatocytes. Their localization was rather focal, 

within some of the nodules. Severe portal inflammatory 

infiltrates were present. Cholestasis was not found (Fig. 3). 

Case No 5 (the same patient as above) 

Subsequent liver specimen obtained during autopsy revealed 

cirrhosis. Severe macrovesicular fatty changes concerned all 

hepatocytes. Mild portal inflammatory infiltrates remained 

similar to those of the biopsy specimen. No cholestasis was 

present (Fig. 3). 

Fig. 3 Precihrrotic pattern of moderate, focal, macrovesicular steatosis of 

patient 4. Subsequent, post mortem examination of the same patient reveals 

severe, macrovesicular fatty changes. 

Case No 6 

The liver biopsy differed substantially from other patients. 

Diffuse, mixed steatotic changes were seen only in about 5- 

10% of hepatocytes. No inflammatory infiltrates, fibrosis or 

cholestasis were detected in the liver. 

Case No 7 

The girl was referred to autopsy examination with diagnosis 

of cardiac insufficiency due to hypertrophied cardiomyopathy. 

Appropriate diagnosis was established retrospectively, 

some years after death on the basis on genetic family study. It 

was performed after her younger brother was diagnosed to 

have LCHAD deficiency. On autopsy examination features of 

excentric hypertrophy of left cardiac ventricle with accompanying 

signs of mitral insufficiency and endocardial thickening 

was observed. Steatosis of the liver was also reported in 

autopsy protocol, but no morphological characteristic was 

made. Microscopically, macrovesicular steatosis of diffuse localization 

consisted of 10 % of hepatocytes accompanied by 

mild fibrosis was found. No inflammation and cholestasis was 

observed.

92 

Summary 

Liver steatosis of severe degree involving 100% of liver parenchyma 

was found in three patients (No 1, 2, and 5) who 

died during ALTE. Moderate steatosis involving approximately 

60% of liver was present in one biopsy ( No 3). Mild steatosis 

affecting 30% and 15% of liver parenchyma was found in two 

biopsies coming from cases 4 and 6, respectively. In one autopsy 

sample liver steatosis involved 10% of parenchyma (No 

7). Steatosis was macrovacuolar or mixed. The first type was 

found in two sequential examinations of the same patient (biopsy 

– No 4 and autopsy – No 5) and in two autopsies (samples 

No 2 and 7). Macrovacuolar steatosis was severe in two cases 

(No 2 and 5) and mild in two others (No 4 and 7). Mixed 

steatosis was seen in two biopsies (No 3 and 6) and one autopsy 

(No 1). Its degeree was mild (sample No 6), moderate 

(sample No 3), and severe (sample No 1). None of the patients 

had pure microvacuolar fatty liver. In six cases distribution 

of steatotic hepatocytes was diffuse, in one focal (sample 

No 4). Inflammation was present in three cases. It was mild 

(cases No 3 and 5) or severe (case No 4). Fibrosis was disclosed 

in four examined samples derived from three patients: 

in two cases its intensity was mild (No 2 and 7) in two severe 

(samples No 4 and 5). Liver biopsy samples available for assessment 

from the same patient disclosed progression from 

precirrhosis to cirrhosis. Data concerning patients and their 

liver morphology are presenting in Table 1. 

Discussion 

It is known that the process of mitochondrial fatty acid oxidation 

can be divided in two stages. First involves transport 

through mitochondrial membrane, the second consists of cutting 

off two-carbon moieties from fatty acid chain by several 

enzymes for long, medium, and short chain fatty acids, respectively 

[5, 9, 11]. Fatty liver may develop in both patients 

with defects of the first stage involving carnitine dependent 

transport [2, 3, 12], and with defects of the proper beta-oxidation 

[2, 4, 6-8, 10, 11]. Baldridge in his study of 82 patients 

with liver steatosis reported only one child in whom the 

LCHAD deficiency was established [1]. Our group comprises 

six patients with this diagnosis. Roe and Coates provided general 

description of liver morphology in LCHAD deficiency 

[9]. They pointed lipid accumulation and sometimes accompanying 

fibrosis as characteristic features of this disorder. Our 

findings are consistent with this observation. However, they 

have not provided particulars concerning intensity, type, and 

distribution of steatosis. Gilbert-Barness and Barness have 

presented more detailed description of liver microscopical 

pathology in LCHAD deficient patients [5]. They considered 

diffuse or zonal macrovacuolar severe fatty liver to be characteristic 

of the majority of LCHAD-deficient patients. In our 

group this type of morphological pattern was found in three 

patients always when the liver biopsy was performed at the 

acute clinical presentation. 

In a number of described patients with LCHAD deficiency 

the liver discloses only focal steatosis of mild intensity 

[5]. We observed focal fatty changes only in one case out of 

seven (sample No 4). In remaining patients the degree of steatosis 

was mild to moderate and was almost invariably diffusely 

distributed. 

Moore described morphology of the liver in six months 

old girl with LCHAD deficiency [7]. There were mixed steatosis 

accompanied by mild cholestasis. Author has not provided 

data concerning intensity and localization of fatty 

changes. In our series of seven cases we did not find any case 

with the features of cholestasis. 

Tyni described liver morphology including microscopical 

findings of 16 patients coming from 12 families, who died 

with diagnosis of LCHAD deficiency [11]. All patients in her 

study had features of macrovacuolar fatty liver. In our patients 

steatosis was not only macrovacuolar, but also mixed. Tyni 

concluded that steatosis was typical for fatty acid beta oxidation 

defects, and emphasized that accompanying fibrosis was 

Table 1 

Details of morphological changes of the liver in patients with LCHADD 

Type of 

examination 

Sample 

No 

Age 

(mo) 

Intensity of 

steatosis (%) 

Type of 

steatosis 

Localization 

of steatosis 

Inflammation 

Fibrosis Cholesta 

A 3299 4 100 mixed diffuse no no no 

A 3289 3 100 macro diffuse no mild no 

B 30573 12 60 mixed diffuse mild no no 

B 30760 3 30 macro focal severe precirrhosis no 

A 2701 6 100 macro diffuse mild cirrhosis no 

B 51605 7 15 mixed diffuse no no no 

A 3070 6 10 macro diffuse no mild no

93 

especially typical for LCHAD deficiency [11]. In our series 

the fibrosis was found in four out of seven liver samples (obtained 

from three children). Tyni suggested presence of relationship 

between the severity of clinical course and the intensity 

of morphological changes in the liver [11]. Our observations 

are in agreement with this statement. 

Our observations strongly underline that liver steatosis 

of any degree detected at autopsy of a child died without 

clearly established diagnosis obliges to include fatty acids 

oxidation disorders into the differential diagnosis [2]. One of 

our patient, died many years ago with the diagnosis of hypertrophied 

cardiomyopathy, was detected just recently as 

LCHAD-deficient during genetic family study, after establishing 

LCHAD deficiency in her younger brother. Lack of acute 

life threatening episode in her medical history and only mild 

degree of liver steatosis in her post mortem examination, preliminarily 

did not paid our attention and missed from correct 

diagnosis. 

In general our results are consistent with observations of other 

authors and warrant drawing the following conclusions: 

1. Fatty liver is characteristic of LCHAD deficiency. 

2. Intensity of steatosis is related to severity of clinical symptoms 

3. In patients with mitochondrial beta-oxidation defects liver 

steatosis may be accompanied by fibrosis which is suggestive 

of LCHAD deficiency. 

4. Fatty liver found in a patient who died of undiagnosed disease 

should suggest the possibility of LCHAD deficiency 

as a cause of death. 

5. Semi-thin plastic embedded tissue section postfixed in osmium 

tetroxide and stained with toluidine blue is very helpful 

in reliable assessment of liver steatosis 

Acknowledgments 

The study was supported by a statutory grant No S 86 of The 

Children’s Memorial Health Institute 

References 

1. Baldridge AD, Perez-Atayde AR 

Graeme-Cook F, Higgins L, Lavine JE 

(1995) Idiopathic steatohepatitis in 

childchood A multicenter study. J 

Pediatr 127:700-704 

2. Chace DH, DiPerna JC, Mitchell BL, 

Sgroi B, Hofman LF, Naylor EW (2001) 

Electrospray Tandem Mass Spectrometry 

or Analysis of Acylcarnitines in Dried 

Postmortem Blood Spcimens Collcted at 

Autopsy from Infants with Unexplained 

Cause of Death. Clin Chemist 

47(7):1166-1182 

3. Chalmers RA, Stanley CA, English N, 

Wigglesworth JS (1997) Mitochondrial 

carnitine-acylcarnitine translocase deficiency 

presenting as sudden neonatal 

death. J Pediatr 131:220-225 

4. Dawson DB, Waber L, Hale DE, 

Bennett MJ (1995) Transient organic 

aciduria and persistent lacticacidemia in 

a patient with short-chain acyl-coenzyme 

A dehydrogenase deficiency. J 

Pediatr 126:69-71 

5. Gilbert-Barness and Barness (2000) 

Fatty acids beta-oxidation defects. In: 

Gilbert-Barness and Barness (eds) 

Metabolic Diseases Eaton Publishing, 

Natick, pp113-136 

6. Iafolla AK, Thompson RJ, Roe CR 

(1994) Medium-chain acyl-coenzyme A 

dehydrogenase deficiency: Clinical 

course in 120 affected children. J Pediatr 

124:409-415 

7. Moore R, Glasgow JFT, Bingham MA, 

Dode JA, Pollitt RJ,Olpin SE, Middleton 

B, Carpentier K (1993) Long chain 3- 

hydroxyacyl-coenzyme A dehydrogenase 

deficiency – diagnosis, 

plasmacarnitine fractions and management 

in a further patient. Eur J Pediatr 

152:433-6 

8. Roe CR Millington DS, Maltby DA 

Kinnebrew P (1986) Recognition of medium-chain 

acyl-CoA dehydrogenase 

deficiency in asymptomatic siblings of 

children dying of sudden infant death 

or Reye-like syndromes. J Pediatr 

108:13-18 

9. Roe CR, Coates PM (1995) Mitochondrial 

fatty acids oxidation disorders. In: 

Scriver CR, Beaudet AL, Sly WS, Valle 

D (eds) The metabolic and molecular 

basis of inherited disease. McGraw-Hill, 

New York, pp 1501-1535 

10. Santer R, Schmidt-Sommerfeld E, 

Leung YK, Fischer JE, Lebental E 

(1990) Medium-chain acyl-CoA dehydrogenase 

deficiency: electron microscopic 

differentiation from Reye syndrome. 

Eur J Pediatr 150:111-114 

11. Tyni T, Rapola J, Paetau A, Palotie A, 

Pihko H (1997) Patology of long chain 

3-hydroxyacyl-CoA dehydrogenase deficiency 

causes by the G1528C mutation. 

Ped Path Lab Med 17:427-447 

12. Winter SC, Szabo-Aczel S, Curry CJR, 

Hutchinson HT, Hogue R, Shug A 

(1987) Plasma carnitine deficiency 

Clinical Observations in 51 Pediatric 

Patients. AJDC 141:660-665



Isolation of urothelial cells for tissue engineered 

bladder augmentation 

Tomasz Drewa 1, 2 , Przemys³aw Galazka 1, 3 , Zbigniew Wolski 2 , Andrzej I. Prokurat 3 , Jan Sir 4 





1 

Department of Biology, 

2 

Department of Urology, 

3 

Department of Pediatric Surgery, 

The L. Rydygier Medical University in Bydgoszcz, 

Bydgoszcz, Poland 

4 


Oncological Center in Bydgoszcz, 

Bydgoszcz, Poland 

Abstract 

Objectives. In case of certain human congenital anomalies such as bladder extrophies, there may not be 

enough residual bladder for closure. Tissue engineering is a promising method for bladder augmentation. 

The aim of our study was to compare isolation efficiency of various digestive mixtures and establish isolation 

and cell culture methodology of urothelium. Different variants of growth media and their influence on proliferative 

capacity was assessed. Methods. Primary Rabbit Urothelial Cells (PRUC) culture was established 

from 8 months old male rabbit urinary bladder wall. 5 different digesting mixtures for isolation of urothelial 

cells were tested. Cells obtained from 0.1% collagenase I digestion after 1 and 4 hours were checked for their 

growth potential in 3 different media. Results. The best digestion solution for urothelial cells isolation was 

0.1% collagenase I. It can be noticed that the cell number of all secondary cultures was higher after 4 h 

digestion than after 1 h. The medium substrates also influenced total cell number in urothelial subcultures. 

The best foetal bovine serum (FBS) supplementation was 5%, when the medium was additionally enriched 

with micro-supplements. Conclusions. Despite of enzymes used for isolation, the digestion time and medium 

enriched with properly amount of serum both deeply influence on the outcome of cell therapy. Autologous 

cells could be used for urinary tract reconstruction using tissue engineering methods. 

Key words: tissue engineering, urinary tract reconstruction, urothelial cells 

Introduction 

In case of certain human congenital anomalies such as bladder 

extrophies, there may not be enough residual bladder for closure 

[7]. The autologous materials and techniques have been 

proposed for bladder augmentation with small and large bowel, 

stomach [1, 13]. Unfortunately none of used materials has been 

enough satisfactory [9, 17]. A number of biomaterials have been 

used in experimental models including acellular tissue matrix 

grafts, synthetic biodegradable or non-biodegradable polymers 

[2]. Many reports revealed failure of these attempts because of 

recurrent urinary tract infections, marked contracture of the graft 

or graft rejection and stones formation [5, 12, 15, 17]. 

Tissue engineering is a promising method for bladder augmentation 

[7]. The use of an autologous cultured urothelium for 

bladder reconstruction could prevent many of the problems associated 

with applied materials [4]. Tissue engineered bladder 

wall consisting of a adequate shaped biodegradable polymer 

used as a scaffold, can deliver autologous urothelial cells [11]. 

This device may restore bladder capacity more effective than 

biomaterials alone. Atala et al demonstrated that urothelial cells 

seeding on the bioabsorbable scaffold can allow creation of new 

functional urinary bladder wall [3]. 


Tomasz Drewa, MD, PhD 

Dpt. of Medical Biology, 

the L. Rydygier Medical University in Bydgoszcz 

Ul. Karlowicza 24 

85-094 Bydgoszcz, Poland 

E-mail: tomaszdrewa@wp.pl

96 

The aim of our study was to compare isolation efficiency of 

various digestive mixtures and establish proper isolation methodology 

of urothelial cells. Also different variants of growth 

medium and its influence on proliferative capacity of growing 

cells was assessed. In the future these cells can be used for urinary 

tract reconstruction using tissue engineering methods. 

Methods 

Primary Rabbit Urothelial Cells (PRUC) culture was established 

from 8 months old male rabbit urinary bladder wall. 

After ketamine (25 mg/kg m.c. i.m.) and scoline (50 mg/kg) 

overdosage bladder wall sample (5 x 5 cm) was obtained. Fragment 

was immersed in sterile Phosphate Buffered Saline (PBS, 

Biomed). In the next step urinary mucosa was mechanically 

separated from muscle layer. At this stage samples were cut 

into 12 equal sized parts (1 cm 2 each). Mucosa samples were 

cut into 2 mm 3 pieces. 

First of all 5 different digesting mixtures for isolation 

of urothelial cells were tested. Incubation time was 4 hours. 

Digesting solutions were as follow: 

1. 0.125% trypsin and 0.01% EDTA 

2. 0.08 % trypsin and 0.006% EDTA 

3. 0.06 % trypsin and 0.005% EDTA 

4. 0.05% trypsin and 0.02% EDTA 

5. 0.1% collagenase-I 

After incubation time the cell pellet was centrifuged 

(800 g/5 min), resuspended in each of the tested culture media 

and seeded on 25 cm 2 culture dishes (Greiner). Cells were 

identified as epithelial cells by morphological criteria as well 

as presence of cytokeratins (Anti-Cytokeratin, Clone: MNF 

116, DAKO), after three passages. 

In the next step 2 nd PRUC was established using the 

most efficient digestive solution: 0.1% collagenase I. Digestion 

efficiency was assessed using two incubation times (1 

and 4 hours). Cells obtained from 0.1% collagenase I digestion 

after 1 and 4 hours were checked for their growth potential 

in 3 different media: 

I- DMEM supplemented with 10% FBS and antibiotics: penicillin 

(100 IU/ml) (Polfa) and streptomycin (100 ug/ml) (Polfa) 

II- 3:1 DMEM and F-12 mixture supplemented with 5% FBS, 

epidermal growth factor (EGF, Sigma), bovine pituary extract 

(Sigma), cholera toxin (30 ng/ml) (Sigma), penicillin (100 

IU/ml) (Polfa) and streptomycin (100 ug/ml) (Polfa) 

III- 1:1 DMEM and F-12 mixture supplemented with 2.5% 

FBS, epidermal growth factor (EGF, Sigma), bovine pituary 

extract (Sigma), cholera toxin (30 ng/ml) (Sigma), penicillin 

(100 IU/ml) (Polfa) and streptomycin (100 ug/ml) (Polfa). 

Cells were seeded on 25 cm 2 culture dishes (Greiner) 

and grown in 37 o C in a humidified atmosphere with 5% CO 2 

. 

Cells were counted using trypan blue exclusion test after four 

days of cultivation. Neubauer«s cytologic chamber was used 

for counting. Morphology was examined under inverted microscope. 

Photographic documentation was done. 

Mean values (+/-SD) from 10 counts for each culture were 

compared to each other. The Student test was used for statistical 

analysis. A p value

97 

3 

2 

1 

0 

x10 6 cells 

1 hour 4 hours 

Fig. 2 Digestion efficiency using 0.1% collagenase I solution was assessed 

using two incubation times (1 and 4 hours) 

4 

3 

2 

1 

0 

x10 6 cells 

A B C D E F 

Fig. 3 Cells obtained from 0.1% collagenase I digestion after 1 hour (a,b,c) 

and 4 hours (d,e,f) were checked for their growth potential in 3 different 

media as: - DMEM with 10% FBS (a,d); - 3:1 DMEM and F-12 mixture 

enriched with microsupplements and 5% FBS (b,e); - 1:1 DMEM and F-12 

mixture enriched with microsupplements and 2.5% FBS (c,f). Cells were 

counted after 4 days of culturing (A vs D, p

References 

1. Adams MC, Mitchell ME, Rink RC 

(1882) Gastrocystoplasty: an alternative 

solution to the problem of urological reconstruction 

in the severely compromised 

patient. J Urol 140:1152-1156 

2. Atala A (1997) Tissue engineering in 

the genitourinary system. In: Tissue engineering, 

Atala A, Mooney D (eds) 

Birhauser Press, Boston, MA, pp. 149- 

164 

3. Atala A (2002) Future trends in bladder 

reconstructive surgery. Semin Pediatr 

Surg 11:134-142 

4. Atala A, Vacanti JP, Peters CA, Mandell 

J, Retik AB, Freeman MR (1992) Formation 

of urothelial structures in vivo 

from dissociated cells attached to biodegradable 

polymer scaffolds in vitro.J 

Urol 148:658-662 

5. Cain MP, Casale AJ, Kaefer M, Yerkes 

E, Rink RC (2002) Percutaneous 

cystolithotomy in the pediatric augmented 

bladder. J Urol 168(4 Pt 

2):1881-1882 

6. Drewa T, Wolski Z, Galazka P, Sir J, 

WoŸniak A, Czaplewski A (2004) Preliminary 

study on alginate scaffold application 

for chondrocytes transplantation 

into the rat bladder wall. 

Biotechnologia 3(66):213-224 

7. Fauza DO, Fishmann SJ, Mehegan K, 

Atala A (1998) Videoscopically assisted 

fetal tissue engineering: bladder augmentation. 

J Ped Surg 33(1):7-12 

8. Follmann W, Guhe C, Weber S, Birkner 

S, Mahler S (2000) Cultured porcine urinary 

bladder epithelial cells as a screening 

model for genotoxic effects of aromatic 

amines: characterisation and application 

of the cell culture model. Altern 

Lab Anim 28(6):833-854 

9. Gough DC (2001) Enterocystoplasty. 

BJU 88(7):739-743 

10. Hobbiesiefken J, Ehlers ME, Behrens 

P, Wunsch L (2003) Urothelial mesh- a 

new technique of cell culture on 

biomaterials. Eur J Pediatr Surg 

13(6):361-366 

11. Kim BS, Baez CE, Atala A (2000) 

Biomaterials for tissue engineering. 

World J Urol 18(1):2-9 

12. Mingin GC, Stock JA, Hanna MK 

(1999) Gastrocystoplasty: long-term 

complications in 22 patients. J Urol 

162(3 Pt 2):1122-1125 

13. Motley RC, Montgomery BT, Zollman 

PE, Holley KE, Kramer SA (1990) Augmentation 

cystoplasty utilizing de-epithelialized 

sigmoid colon: a preliminary 

study. J Urol 143:1257-1260 

14. Rebel JM, De Boer WI, Thijssen CD, 

Vermey M, Zwarthoff EC, Van der 

Kwast TH (1994) An in vitro model of 

urothelial regeneration: effects of 

growth factors and extracellular matrix 

proteins. J Pathol. 173(3):283-291 

15. Sugasi S, Lesbros Y, Bisson I, Zhang YY, 

Kucera P, Frey P (2000) In vitro engineering 

of human stratified urothelium: analysis 

of its morphology and function. J Urol 

164(3 Pt 2):951-957 

16. Wechselberger G, Schoeller T, Stenzl A, 

Ninkovic M, Lille S, Russell RC (1998) 

Fibrin glue as a delivery vehicle for autologous 

urothelial cell transplantation 

onto a prefabricated pouch. J Urol 

160(2):583-586 

17. Woodhouse CR, Redgrave NG (1996) 

Late failure of the reconstructed exstrophy 

bladder. Br J Urol 77(4):590-592 

18. Zhang YY, Ludwikowski B, Hurst R, 

Frey P (2001) Expansion and long-term 

culture of differentiated normal rat 

urothelial cells in vitro. In Vitro Cell Dev 

Biol Anim 37(7):419-429



Status of oral hygiene and dentition in children 

exposed to anticancer treatment 





Dorota Olczak-Kowalczyk 1 , Marta Daszkiewicz 1 , Barbara Adamowicz-Klepalska 2 , El¿bieta Czarnowska 3 , 

Agnieszka Sowiñska 3 , Bo¿ena Dembowska-Bagiñska 4 , Danuta Perek 4 , Pawe³ Daszkiewicz 5 

1 

Department of Oral Pathology, 

3 


4 

Department. of Oncology, 

5 

Department of Neurosurgery, 



2 

Department of Developmental Age Stomatology, 

Medical Academy, 

Gdañsk, Poland 

Abstract 

Background: Deleterious effects of radiation and chemotherapy on oral cavity tissues are well known, but 

little data exist on the effect of combined anticancer treatment on developing teeth. Additionally, there are no 

established standards for dental care in children undergoing complex anticancer treatment. The aim of study: 

Evaluation of oral hygiene and dentition status in children subjected to combined anticancer treatment. Material 

and method: 88 children (age range 3-18 years), who underwent combined chemo-radiotherapy or 

chemotherapy alone due to a malignant neoplasm, were included to the study. Follow-up time was from 6 

months to 5 years. Patients were assessed in the dental out-patients’ clinic and their medical documentation 

reviewed. Results and conclusions: An inadequate oral hygiene was found in all patients as well as a severe 

caries and an increased incidence of non-caries-mediated enamel defects. A deleterious effect of combined 

chemo-radiotherapy on mineralized tooth tissues was found, both in primary and in permanent teeth. Ionizing 

radiation had a stronger negative impact than chemotherapy alone. In all children analyzed, there was a 

discrepancy between actual requirements for dental care and currently performed dental treatment. 

Key words: children, chemotherapy, enamel structure, oral hygiene, permanent dentition, primary dentition, 

radiotherapy 

Introduction 

Anticancer treatment, including chemotherapy and radiotherapy 

of the facial region, may enhance both the development 

of dental caries and also of non-caries dependant lesions 

on smooth surfaces of teeth [4-6, 10, 11]. The etiology of these 

phenomena is multi-factorial and still poorly understood. Both 

chemo- and radiotherapy increase the susceptibility of teeth 

to noxious factors and affect adversely the ecosystem of oral 

cavity [9, 10]. In patients undergoing combined anticancer 

treatment, these adverse effects may enhance each other [7]. 

In general opinion, the direct effect of ionizing radiation 

on teeth leads to denaturation of protein matrix of the dentin 

and to damage of its structure due to degeneration and dysfunction 

of dental pulp. On the other hand, there is much less 

information concerning the pathogenesis of enamel lesions in 


Dorota Olczak-Kowalczyk 

Department of Oral Pathology 




Phone : +4822 815 13 15; +4822 815 16 31 

E-mail : dok.@bobas.czd.pl

100 

available literature. Enamel is described as poorly mineralized, 

less hard and more brittle, or as more susceptible to damage by 

acids while preserving normal hardness. Independent on the 

mechanism of radiation-induced lesions in mineralized tissues 

of teeth, they become more susceptible to harmful factors [5]. 

There is little data concerning the effects of chemotherapy not 

combined with radiotherapy on teeth. Nevertheless, they show 

its deleterious effect on the function of dental pulp, thereby 

increasing susceptibility of teeth to various pathologic processes 

[10]. The effect of particular cytostatics on tooth pulp is difficult 

to ascertain. This is due to general use of multi-drug chemotherapy 

regimens. Toxic effects of vincristin are known the 

best. Histopathologic examination of teeth extracted from patients 

undergoing multi-drug chemotherapy revealed linear lesions, 

most probably resulting from disturbed production of 

collagen matrix by odontoblasts. An important point is finding 

of correlation between the number and distribution of these lines 

and chemotherapy cycles, when one of administered drugs was 

vincristin [6]. It is also well known that colchicin and vinblastin 

block dentin formation in incisors in rats, while trethanomelamin 

and cyclophosphamide delay cutting of molars in these animals 

[6, 7]. 

Adverse change of oral eco-system in patients undergoing 

anticancer treatment is caused mainly by qualitative and 

quantitative disturbances of salivary secretion [9]. Facial irradiation 

often leads to dysfunction of serous cells of parotid 

gland. Depending on total absorbed dose of radiation (over 

60 Gy), this may lead to total (irreversible) or partial (reversible) 

xerostomia and to qualitative change of saliva, such as 

increased viscosity, decreased buffering properties and decreased 

IgA level. A decreased volume of secreted saliva may 

accompany administration of many cytostatic drugs [2, 4, 9]. 

Increased acidity of oral cavity may be further enhanced by 

vomiting, which often accompanies chemotherapy. An extremely 

unfavorable aspect of chemotherapy is its prolonged 

administration, ranging from 6 months to 3 years, depending 

on the kind of neoplasm. 

Unfavorable alteration of oral eco-system in patients undergoing 

oncologic treatment may enhance demineralization 

of teeth, thus increasing the susceptibility of teeth to cariespromoting 

bacteria, leading to the development of caries and 

erosion of enamel. Particular form of pathologic process (caries 

and/or erosion) depends mainly on acid pH and duration of 

its contact with mineralized tissues of teeth [2]. Increased susceptibility 

of teeth to the influence of deleterious factors (chemical, 

bacterial and mechanical) may be an important factor particularly 

in children and adolescents, where it is caused not only 

by the neoplasm itself and anti-cancer therapy, but also by morphologic 

and functional immaturity of teeth [15, 16]. 

When discussing the etiology of damage to mineralized 

tissues of teeth observed in patients undergoing anti-cancer 

treatment, we must also take into account the effect of 

neurologic deficits, e.g. weakness of chewing muscles leading 

to “lazy” chewing and compromising self-purification of 

oral cavity. Other important factors are: co-existing stomatitis, 

disturbed taste, lack of appetite and general malaise, leading 

to hygienic neglect and dietary mistakes. 

There is no published data concerning the profile of structural 

changes in mineralized tissues of teeth, developing in children 

undergoing chemotherapy combined with facial irradiation. 

Aim of the study 

The aim of this study was clinical assessment of lesions in 

mineralized tissues of teeth in children subjected to chemotherapy 

combined with radiotherapy to the facial region and 

an electron-microscopic analysis of structure of atypical defects 

present on smooth surface of primary teeth. 

Material and method 

Patients. The study included 88 children (57 boys and 31 girls) 

aged from 3 to 18 years, who underwent combined anti-cancer 

treatment. Time since completion of oncologic treatment 

ranged from 6 months to 5 years. Patients’ characteristics concerning 

basic diagnosis, total dose of radiation and location 

of irradiated area (face or neurocranium), are presented in Table 

1. Children included in the study have been subdivided into 3 

subgroups: those with primary dentition, those with mixed 

dentition and those with permanent dentition (Table 2). 

Stomatologic assessment of dentition. Dental examinations 

were performed in the setting of a dental office, using 

a shadowless lamp, dental probe and dental mirror. Primary 

assessed parameters included: oral hygiene and dental status, 

presence of caries, fillings, non-caries-mediated enamel defects 

and missing teeth. At the same time, prophylactic and 

therapeutic requirements were defined and secondary endpoints 

were assessed, such as frequency and severity of caries, 

treatment index and incidence of atypical non-caries-mediated 

defects. 

Oral hygiene. Oral hygiene was assessed using the 

“Oral Hygiene Index” according to Green and Vermillion 

(OHI-S). Dental plaque stained with eosin was evaluated on 

buccal/labial and palatal/lingual surfaces of 6 representative 

teeth: 55 (15), 53 (13), 51 (11), 75 (36), 73 (13), 71 (31). If 

the desired tooth was missing, the adjacent tooth was examined. 

On the basis of mean values of the OHI-S index, oral 

hygiene was considered good if the score was 0-1,0; adequate 

if the score was >1,0-2,0 and poor if the score was >2,0-3,0. 

Severity of caries. Evaluation was performed according 

to the DMF index for permanent teeth and the dmf index for 

primary teeth. The following lesions of mineralized tooth tissues 

were considered as atypical: white smooth stains, white or 

rusty rough stains, enamel defect within a white or a rusty stain. 

Structure of defect surface. Teeth were cut along the longitudinal 

axis and transversely at the defect level. They were 

fixed in 4 % glutaraldehyde solution (Serva) in 0.1 M cacodyle 

buffer with pH value of 7.3 (BDH Chemicals Ltd, Poole, England) 

and next they were dehydrated using increasing concentrations 

of ethyl alcohol and acetone. Tooth fragments were attached 

to brass rods with silver paste (AGAR, England) and 

vacuum-coated with charcoal and gold powder and finally analyzed 

using a scanning electron microscope (JSM 35C, Jeol).

101 

Table 1 

Diagnosis, mean total dose and field of radiation in patients subjected to combined chemo-radiotherapy. Radiation fields 

included teeth and teeth buds 

Diagnosis 

Number of 

children (N) 

Mean dose of 

radiation (cGy) 

Irradiated field 

Medulloblastoma 18 3500 

4500 

Neural axis (brain and spinal cord) 

Posterior fossa 

Brainstem tumors 7 5400 Tumor site 

Ependymoma 5 5400 Tumor site 

Astrocytoma III 5 5400 Tumor site 

Glioblastoma multiforme 2 5400 Tumor site 

Carcinoma plexus chorioidei 3 5400 Tumor site 

Nasopharyngeal 

9 5400 Tumor site 

rhabdomyosarcoma 

Non-Hodgkin lymphoma 1 2400 Neural axis 

Neuroblastoma stage IV 1 4500 Brain metastases 

Retinoblastoma 1 4500 Lymph node metastases in head and 

Table 2 

Distribution of dentition types in the analyzed population of 

children subjected to anti-cancer treatment 

Dentition type 

Primary dentition 10 

Mixed dentition 30 

Permanent dentition 25 

Total 65 

Results 

Number of children (N) 

Clinical evaluation 

The incidence of caries in the studied population of children 

after combined chemo-radiotherapy was high and approached 

100 %. Oral hygiene in the whole group studied was considered 

adequate (mean OHI-S score = 1,88), while in the subgroup 

of children with primary teeth oral hygiene was poor 

(mean OHI-S score = 2,25) (Table 3). 

An analysis of severity of caries is presented in Table 

4. The mean dmf score was 14,6; 7,06 and 6,0 in subgroups of 

children with primary, mixed and permanent teeth respectively, 

indicating extremely severe caries both in primary and in permanent 

dentition. In all subgroups, the proportion of primary 

and permanent teeth with active caries were a significant component 

of the mean score of dmf and DMF indices. 

An analysis of treatment indices, both for primary and 

for permanent teeth, showed significant neglect concerning 

both treatment and prevention (Table 5). Obtained values were 

very small, ranging from 0,02 to 0,43, depending on sex and 

type of dentition. 

In the area of dental therapeutic requirements (Table 6), the 

need for conservative treatment significantly outnumbered all 

other parameters. On the basis of proportion of teeth with active 

caries, need for conservative treatment was observed in 

71,42 % of teeth in the subgroup of children with primary dentition, 

in 48,82 % of primary teeth and 89.,10 % of permanent 

teeth in the subgroup with mixed dentition and in 95,22 % of 

teeth in the subgroup of children with permanent dentition. 

In the subgroup of children with primary dentition, the 

mean number of teeth with active caries was 13,3 out of 20,0 

teeth examined, where the mean number of teeth qualifying 

for extraction was 3,8 (Table 7). In children with permanent 

dentition, the mean number of teeth with caries was 10,88 out 

of 26,88 teeth examined and the mean number of teeth qualified 

for extraction was 0,52. In the subgroup of children with 

mixed dentition, mean numbers of teeth with active caries and 

teeth qualified for extraction were elevated too. Interpretation 

of results obtained in this subgroup is difficult due to on-going 

process of exchange of dentition. 

Non-caries-mediated enamel defects were noted in 49 

children (75,38 % of the whole group) (Table 8). The proportion 

of children with non-caries-mediated enamel defects were 

70,0 %, 76,6 % and 76,0 % in the subgroups of children with 

primary, mixed and permanent dentition, respectively. Noncaries-mediated 

enamel defects were located on buccal or labial 

surfaces, in pericervical area or encompassed more than 

one tooth surface. However, most non-caries-dependant defects 

were localized on labial or buccal surfaces, taking the 

form of white-smooth (173 out of 229 lesions seen in 39 subjects) 

or rusty-rough patches (156 out of 254 lesions seen in 

33 subjects). Enamel defects were noticed in 22 patients. Labial 

or buccal location of lesions was seen in 104 out of 182 

lesions (Table 8).

102 

Table 3 

Correlation of oral hygiene with age and sex in children subjected to chemio- and radiotherapy (OHI-S mean value) 

Dentition type Sex N (children) OHI-S (mean value) 

Primary dentition Boys 7 2,51 

Girls 3 1,63 

Total 10 2,25 

Mixed dentition Boys 19 1,90 

Girls 11 2,01 

Total 30 1,94 

Permanent dentition Boys 17 1,70 

Girls 8 1,55 

Total 25 1,65 

Total Boys 43 1,92 

Table 4 

Girls 22 1,79 

Total 65 1,88 

Correlation of Severity of caries, sex and type of dentition in children subjected to chemotherapy combined with radiotherapy 

(mean scores of dmf and DMF indices) 

N (teeth) 

Indem sc 

Dentition 

type 

Sex 

with caries removed filled dmf D 

N 

(children) d D m M f F 

primary boys 7 105 - 0 - 2 - 15,20 - 

girls 3 28 - 0 - 11 - 13.0 - 

total 10 133 - 0 - 13 - 14.6 - 

mixed boys 19 104 102 13 0 20 15 7,21 6. 

girls 11 66 54 2 0 7 9 6,81 5, 

total 30 170 156 15 0 27 24 7,06 6, 

permanent boys 17 - 191 - 1 - 106 - 17 

girls 8 - 81 - 3 - 62 - 20 

total 25 - 272 - 4 - 168 - 17 

d/D - caries in primary /permanent dentition 

m/M - removed primary teeth/permanent 

f/F - filled primary teeth/permanent 

duf/DUF - dental caries severity index in primary/permanent dentition

103 

Table 5 

Correlation of teeth-treatment index, dentition type and sex in children subjected to combined chemo-radiotherapy 

Teeth-treatment index 

Dentition type Sex N (children) Primary Permanent 

Primary boys 7 0,02 - 

girls 3 0,28 - 

total 10 0,09 - 

Mixed boys 19 0,16 0,12 

girls 11 0,09 0,14 

total 30 0,14 0,13 

permanent boys 17 - 0,36 

girls 8 - 0,43 

total 25 - 0,38 

Table 6 

Correlation of requirement for conservative and surgical treatment with sex and dentition type in children subjected to chemotherapy 

combined with radiotherapy 

Dentition 

type 

Sex 

N 

(children) 

examined 

N (teeth) 

with caries 

n=100% For conservative treatment For extraction 

M S M S M S M S 

n % n % n % n % 

Primary boys 7 140 - 105 - 71 67,6 - - 34 32,4 - - 

girls 3 60 - 28 - 24 85,7 - - 4 14,28 - - 

total 10 200 - 133 - 95 71,42 - - 38 28,57 - - 

Mixed boys 19 163 271 104 102 53 50,96 94 92,15 51 49,03 8 7,84 

girls 11 102 155 66 54 30 45,45 45 83,33 36 54,54 9 16,66 

total 30 265 426 170 156 83 48,82 139 89,10 87 51,17 17 10,89 

Perma-nent boys 17 - 451 - 191 - - 180 94,24 - - 11 5,75 

girls 8 - 221 - 81 - - 79 97,53 - - 2 2,46 

total 25 - 672 - 272 - - 259 95,22 - - 13 4,77 

M- primary teeth 

S- permanent teeth

104 

Table 7 

Total mean number of teeth with caries and total mean number of teeth requiring conservative and surgical treatment depending 

on dentition type 

Dentition 

type N Examined with caries 

N (teeth) (mean value) 

requiring 

conservative 

treatment 

requiring 

extraction 

M S M S M S M S 

Primary 10 20 - 13,30 - 9,50 - 3,80 - 

Mixed 30 8,83 14,20 5,66 5,20 2,76 4,63 2,90 0,56 

Permanent 25 - 26,88 - 10,88 - 10,36 - 0,52 

Table 8 

Non-caries-dependent enamel defects in children subjected to chemo- and radiotherapy dependent on dentition type (w-labial/ 

buccal surface, sz-cervical area) 

White smooth stain Rough stain Enamel defects within s 

Dentition 

Nt Nt Nt 

type Nc Nc w sz >1 total Nc w sz >1 total Nc w sz >1 

Primary 7 2 2 4 - 6 6 35 8 - 43 5 49 - 15 

Mixed 23 21 106 1 27 134 16 77 3 35 115 6 19 - 6 

Perma-nent 19 16 65 1 24 89 13 40 4 58 96 11 36 36 21 

Total 49 39 173 6 51 229 33 156 15 93 254 22 104 36 42 

Nc = number of children 

Nt = number of teeth 

Fig. 1 Scanning electron microscopic view of tooth surface with cracked 

enamel and altered enamel prisms within an area of enamel defect. Microscopical 

magnification x 1100 

Fig. 2 Scanning electron microscopic view of the tooth dentin canals. Microscopical 

magnification x 1500

105 

Electron-microscopic examination 

Scanning electron microscopical analysis of the enamel of teeth 

showed its lack or crack in number of areas. Derangement of 

enamel prisms and their clump was observed (Fig.1). No alterations 

of the pattern of dentin canals were noticed (Fig.2). 

Discussion 

Intensification of chemotherapy and it’s combination with 

surgery and/or radiotherapy have made possible an increasingly 

effective treatment of head and neck malignancies [1, 

2]. However, such an aggressive treatment is associated with 

an increased risk of many adverse side-effects, concerning 

many different components of the chewing apparatus. Particularly 

severe complications may accompany anticancer 

treatment of children and adolescents. Anticancer treatment 

proves effective in a growing number of patients, resulting in 

longer survival time. This in turn contributes to an increased 

incidence of late adverse effects of therapy, e.g. destruction of 

mineralized tissues of teeth and various developmental abnormalities 

of chewing apparatus [3, 4, 5]. Therefore, oral care 

professionals will be increasingly often confronted by such 

problems. Unfortunately, available literature provides little data 

concerning dental problems in patients undergoing anticancer 

treatment in childhood or guidelines for their effective 

prophylactic, therapeutic and rehabilitative management. 

Results of clinical trials indicate significant hygienic 

neglect and poor dentition status of children subjected to chemotherapy 

combined with radiotherapy. We noticed alarmingly 

high incidence of caries and it’s extreme severity. In our 

study, the mean dmf score was 14,6, largely exceeding values 

observed in children in the risk group for infectious endocarditis, 

where dmf score was 9,25 and DMF score was 10,42, as 

well as in children undergoing chemotherapy alone, where it 

equaled to 2,5 and 14,1 respectively [13]. We have noticed a 

large disproportion between requirements for dental treatment 

and dental treatment actually carried out. Values of treatment 

indices in children after anticancer treatment and those in the 

risk group for infectious endocarditis were similar, in the range 

of 0-0,34 and 0,03-0,34 respectively. 

An analysis of the scope of required dental treatment 

in children after chemotherapy combined with radiotherapy 

revealed significant needs in the area of conservative treatment 

and relatively smaller needs in the field of surgical treatment, 

possibly indicating high speed of development of new 

caries-dependent defects. 

In 69 % of children analyzed we observed atypical noncaries 

dependent enamel lesions in the form of patches and 

enamel defects. They occurred twice as often than in children 

undergoing chemotherapy alone. Children after chemotherapy 

combined with radiotherapy presented mostly white, rusty or 

rough patches, while children after chemotherapy alone had 

usually white lesions. 

Deleterious effect of radiotherapy on dentition is well known 

since a long time. The so-called “post-radiation caries” in patients 

undergoing radiotherapy in the area of head and neck, 

is a phenomenon well established in the literature [2, 3, 5, 8, 

11, 12, 14]. Observed thereby enamel defects are at the beginning 

probably non-bacterial in nature and are caused by chemical 

and mechanical factors acting upon a primarily more susceptible 

enamel. It would seem, that subsequent development 

of caries in these patients is secondary to the above described 

phenomena. 

Conclusions 

1. High incidence and severity of caries and atypical non-caries 

mediated enamel defects observed in patients subjected to 

chemotherapy combined with radiotherapy in the facial region, 

support the thesis about deleterious effect of anti-cancer 

on dentition. 

2. Significant disproportion between requirements for dental 

treatment and actually performed conservative and surgical 

treatment indicate, that dental prophylactic and therapeutic 

management should be instituted as early as possible in these 

patients. 

References 

1. Cioch M (2001) Uszkodzenie bariery 

sluzówkowej (Mucosal barier injury- 

MBI) w nastêpstwie intensywnego 

leczenia cytostatycznego. Onkol Pol 4,2: 

85-89 (in Polish) 

2. Darczuk D (1999) Zmiany w jamie 

ustnej wywo³ane napromienianiem 

nowotworów g³owy i szyi. Stomat 

Wspó³czesna 6,1:23-25 (in Polish) 

3. Groetz KA, Wagner W, Duschner H 

(2001) Chronische Strahlenfolgen an 

den 

Zahnhartgeweben 

(“Strahlenkaries”). Klassifikation und 

Behandlungsansaetze. Strahlentherapie 

und Onkologie-Abstract 177,2 :96-10 

4. Jaffe N, Toth BB, Hoar RE, Ried HL, 

Sullivan MP, McNeese M (1984) Dental 

and maxillofacial abnormalites in 

long-term survivors of childhood cancer: 

effects of treatment with chemotherapy 

and radiation to the head and 

neck. Pediatrics 73,6:816-823 

5. Jankowska K (1992) Próchnica 

popromienna (opis przypadku). 

Wroc³awska Stomatologia 235-238 (in 

Polish) 

6. Macleod RI, Welbury RR, Soames J 

(1987) Effects of cytotoxic chemotherapy 

on dental development. J Royal 

Soc Med 80:207-209 

7. Maguire A, Welbury R (1996) Long-term 

Effects of antineoplastic chemotherapy 

and radiotherapy on dental development. 

Dental Update 23 (5):188-194 

8. Manguire A, Craft AW, Evans RGB, 

Amineddine H, Kehrnahan J, Macleod 

RI (1987) Long-term effects of treatment 

on the dental condition of children 

surviving malignant disease. Cancer 

60:2570-2575 

9. Markitziu A, Zafiropolos G, Tsalikis L, 

Cohen L (1992) Gingival health and salivary 

function in head- and neck– 

irradited patients. Oral Surg Oral Med 

Oral Pathol 73:427-433

10. Nunn JH, Welbury RR, Gordon PH, 

Kernahan J, Craft AW (1991) Dental 

caries and dental anomalies in children 

treated by chemotherapy for malignant 

disease: a study in the north of England. 

Int J Paediatr Dent 1:131-135 

11. Olczak-Kowalczyk D, Dobies K, 

Daszkiewicz M (2002) Zastosowanie 

lakieru Seal & Protect w profilaktyce 

próchnicy popromiennej u dzieci po 

przebytej terapii przeciwnowotworowej. 

Magazyn Stomatologiczny 

11,133,12:10-13 (in Polish) 

12. Olczak-Kowalczyk D, Daszkiewicz M 

(2002) Wspó³czesne metody 

profilaktyki próchnicy w grupie 

wysokiego ryzyka jej wystêpowania. 

Standardy Medyczne 4,6 (32):402-411 

(in Polish) 

13. Olczak-Kowalczyk D, Bedra B (2003) 

Stan zdrowia jamy ustnej i 

stomatologiczne potrzeby lecznicze w 

badanej populacji dzieci z ryzykiem 

infekcyjnego zapalenia wsierdzia. Nowa 

Stomat 3,25:111-119 (in Polish) 

14. Olczak-Kowalczyk D, Perek D, 

Daszkiewicz M, Adamowicz-Klepalska 

B, Dembowska-Baginska B, 

Daszkiewicz P (2003) Problemy 

stomatologiczne u dzieci z chorobami 

nowotworowymi. Doœwiadczenia 

wlasne. Nowa Stomat 4;26:175-179 (in 

Polish) 

15. Rosenberg SW, Kolodney H, Wong 

GY, Murphy ML (1987) Altered dental 

root development in long-term survivors 

of pediatric acute lymphoblastic 

leukemia. A review of 17 cases. 

Cancer 59:1640-1648 

16. Welbury RR, Craft AW, Murray JJ, 

Kehrnahan J (1984) Dental health of survivors 

of malignant disease. Archiv Dis 

Child 59:1186-1188

Annals of Diagnostic Paediatric Pathology 2004, 8(3-4): 106-110 


Preliminary review of the treatment results of hepatoblastoma and 

hepatocarcinoma in children in Poland in the period 1998-2002 

Czes³aw Stoba 1 , Katarzyna Nierzwicka 1 , Piotr Czauderna 1 , Barbara Skoczylas-Stoba 2 , 

Stefan Popadiuk 3 , Jolanta Bonar 4 , Krystyna Sawicz-Birkowska 5 , Hanna Kaczmarek-Kanold 6 , 

Przemys³aw Mañkowski 7 , Krzysztof K¹tski 8 , Wojciech Madziara 9 , Magdalena Rych³owska 10 , 

Anna Sitkiewicz 11 , Anna Szo³kiewicz 12 , Sabina Szymik-Kantorowicz 13 , Violetta Œwi¹tkiewicz 14 





1 


3 

Department of Pediatrics, Gastroenterology and Oncology 

12 

Department of Pediatric Oncology 

Medical University of Gdansk, Gdansk,Poland 

2 

Department of Anesthesiology and Intensive Therapy, 

Regional Copernicus Hospital, Gdansk, Poland 

4 

Department of Pediatrics, Hematology, Oncology 

and Endocrinology, 

5 


Medical University of Wroclaw, Wroclaw, Poland 

6 

Department of Pediatric Hematology and Oncology, 

7 


Medical University of Poznan, Poznan, Poland 

8 

Department of Pediatric Hematology and Oncology, 

Medical University of Lublin, Lublin, Poland 

9 


Medical University of Katowice, Katowice, Poland 

10 

Department of Pediatric Surgery and Oncology, 

Institute of Mother and Child, Warszawa, Poland 

11 

Department of Pediatric Surgery and Oncology, 

Medical University of Lodz, Lodz, Poland 

13 


Medical University of Krakow, Krakow, Poland 

14 

Department of Hematology and Oncology, 

Medical University of Bydgoszcz, Bydgoszcz, Poland 

Abstract 


Objectives: Hepatoblastoma (HB) and hepatocellular carcinomas (HCC) are most common of all primary 

malignant liver tumors in children. This series of HB and HCC represents a preliminary report of treatment in 

11 Polish pediatric oncology centers. Treatment protocols were based on two consecutive SIOPEL group 

studies: 2 and 3. Material and methods: From 1998 to 2002 there were 34 children with HB and HCC 

diagnosed (28 were HB and 6 were HCC). All HB and HCC patients received preoperative chemotherapy 

based on SIOPEL 3 protocols (CDDP alone, PLADO or SuperPLADO). Among HB 23 tumors were unifocal 

and 5 multifocal. Eight HB cases were qualified to the high risk group according to the SIOPEL protocol 

definition. Alphafetoprotein (AFP) was elevated in all cases. Lung metastases were present in one HB and in 

three HCC cases. In 5 cases diagnosis was made on the clinical ground, in most of the patients initial biopsy 

was performed. Standard risk tumor were treated either Cisplatin monotherapy or PLADO regimen. High 

risk group was treated with SuperPLADO regimen. Twenty-one HB cases were operated in a delayed setting. 

Five tumor resections were incomplete: 3 macroscopically and 2 microscopically. In 4 cases information on 

surgery is missing. Results: Among 28 HB patient 19 are alive with no evidence of disease, 2 are alive with 

disease, 6 children died and one is lost to follow-up. Thus 3-year overall survival of HB patient was 75%, 

while 3-year event-free-survival was 64%. Median follow-up time is 40 mouths. Overall survival among 

standard risk group was 90% in comparison with 66% in the high risk group. All children with macroscopic 

residuum post resection died. Overall survival in HCC cases at 3 years is 33%, but of the two alive patients 

one currently experienced recurrence. Conclusions: Reported treatment for hepatoblastoma are similar to 

those achieved by other study groups and constitute an improvement in comparison with our retrospective 

series (OS=75% vs. 44%). High risk HB patients require modification of the current treatment strategy due to 

inferior survival than in standard risk HB. Treatment results of HCC (OS=33%) were clearly inferior to HB 

being in line with experience of other international study groups. Importance of the complete tumor resection, 

both in HB and HCC was confirmed. 

Keywords: children, hepatoblastoma, hepatocellular carcinoma, surgery, treatment 

Piotr Czauderna, PhD 

Dept. of Pediatric Surgery, 

Medical University of Gdansk, 

Ul. Nowe Ogrody 1-6, 

80-803 Gdansk, Poland. 

Phone/fax: +48 58 302 64 27 

E-mail: pedsurg@amg.gda.pl

108 

Introduction 

In the last 15 years a steady increase in survival in pediatric 

liver tumors has been noted. First observations on improvement 

in survival with the use of adjuvant chemotherapy came 

from Children’s Cancer Study Group (CCSG) and Pediatric 

Oncology Group (POG) in the late 70’s. Later in unresectable 

cases neoadjuvant chemotherapy was applied [3, 9]. In 1986 

Belgian and Canadian centers introduced in hepatoblastoma 

standard pre- and postoperative chemotherapy consisting of 

cisplatin and adriamycin. This chemotherapy protocol, called 

later PLADO, has become a golden standard of 

hepatoblastoma (HB) treatment in many centers, including 

Mother and Child Memorial Hospital in Warsaw. Efficacy of 

further modified protocols was tested by the multinational 

SIOPEL group in three subsequent trials: SIOPEL 1-3 [9-11]. 

It was shown that cisplatin alone is as effective as PLADO in 

standard risk HB [9, 10]. 

Despite chemotherapy progress complete tumor resection 

remains a goal of treatment and prerequisite for cure. If it 

cannot be achieved it is better to abandon any surgical attempt 

and resort to another treatment modality and/or consult the 

case with referral center. Our remarks on principles of the 

surgical treatment of liver tumors were presented during Surgical 

Workshop in Szklarska Porêba, Poland in 2002 [13]. It 

was reported that progress in surgical management of liver 

tumors had been associated with better knowledge of its 

anatomy, improved surgical technique and equipment (ultrasonic 

and water-jet dissectors, argon beam coagulation, 

thrombostatic materials), as well as new imaging techniques, 

new method of anesthesia and improved perioperative care. 

Presented results emerged from multicenter cooperation 

of 13 Polish institutions within the study devoted to surgical 

treatment of malignant epithelial tumors of childhood 

including patients’ stratification according to risk groups in 

cooperation with the International Childhood Liver Tumors 

Strategy Group (SIOPEL). This study is coordinated in Poland 

by the Department of Pediatric Surgery of the Medical 

University of Gdansk. 


In the period 1998-2002 all together 34 cases of primary epithelial 

liver tumors were treated in 11 Polish paediatric oncology 

centers. Twenty-eight were HB and 6 were HCC including 

one transitional (HB/HCC) liver tumor. Patients’ characteristics 

and stage are shown in Tab.1 and 2. Maximal tumor 

diameter ranged from 5 to 18 cm. Among HB 23 tumors (82%) 

were unifocal and 6 – multifocal. In HCC multifocal tumors 

prevailed: 4 out of 5. Eight HB cases (29%) were qualified to 

the high risk group according to the SIOPEL protocol definition. 

Alphafetoprotein (AFP) was elevated in all but one HB 

case (from 150 ng/ml to 2.480.000 ng/ml), in which AFP level 

was < 100 ng/ml. In all HCC cases AFP was elevated (from 

150 ng/ml to 2.500.000 ng/ml). Lung metastases were present 

in one HB case (3,6%) and in 3 HCC cases (50%). In one 

HCC case metastases involved also mediastinal and retroperitoneal 

lymph nodes. 

Treatment protocols were based on consecutive SIOPEL studies: 

SIOPEL 2 and 3. Standard imaging methods included: 

abdominal ultrasonography (US), chest X-ray (AP and lateral), 

computed tomography (CT) of the abdomen (with and 

without i.v. contrast) and chest (to detect eventual pulmonary 

metastases) and/or magnetic resonance imaging (MRI). Also 

complete blood count (including platelet level) and serum AFP 

were obtained at diagnosis. AFP, if elevated, was used to monitor 

response to treatment. The protocol required preoperative 

assessment of tumor extent according to PRETEXT classification, 

which is described in details elsewhere [12]. Closed 

needle biopsy was performed in 2 cases, open biopsies were 

done in 21 cases (18 – wedge and 3 – open needle biopsies). 

In one case liver biopsy was done laparoscopically. In 5 cases 

diagnosis was made on the clinical ground. In 4 patients data 

on biopsy are missing. Patients with biopsy proven HB and 

HCC were qualified to two risk groups on the basis of PRE- 

TEXT grouping. High risk tumors were those PRETEXT 4 

(involving the whole liver) or belonging to any PRETEXT 

category with the following features: significant intravascular 

involvement (V or P), extrahepatic extension and/or distant 

metastases (M). Hence high risk tumors were primarily 

unresectable. All others formed standard risk group. In tumors 

occurring between 6 months and 3 years of age with 

unequivocal imaging and increased AFP level diagnosis of 

hepatoblastoma on clinical ground was allowed. Before definite 

surgery Doppler US and helical CT were required. All 

patients with hepatoblastoma received preoperative chemotherapy 

which was dependent on the risk group assignment. 

Standard risk tumors were treated either cisplatin monotherapy 

or PLADO regimen (those registered in the SIOPEL 3 trial 

were randomized), however few patients were not randomized. 

High risk patients were treated with SuperPLADO regimen, 

which included 2-weekly cisplatin alternating with 

doxorubicin administered together with carboplatin. Four triweekly 

chemotherapy courses were given. Postoperatively two 

more courses of the same chemotherapy were administered. 

For details of the treatment protocol check elsewhere [9, 10, 

11]. Cisplatin alone was used in 10 children, PLADO in 7 

cases and SuperPLADO in 9 cases. In 2 cases details on preoperative 

chemotherapy are missing. Twenty-one HB cases 

were operated in a delayed setting. Type and completeness of 

performed surgery is shown in Table 3. Treatment of operable 

hepatocellular carcinoma (HCC) patients was started with tumor 

resection followed by 6 postoperative courses of 

SuperPLADO regimen. Two HCC cases were primarily operated 

including one orthotopic liver transplantation Treatment 

of unresectable and/or metastatic HCC was identical with high 

risk HB. One child with HCC was operated in a delayed setting 

after partial response to preoperative chemotherapy. In 

non-responders (2 cases) chemoembolization was used.

109 

Results 

Among 28 HB patients 19 are alive with no evidence of disease, 

2 are alive with disease, 6 children died and one is lost to 

follow-up. Thus 3-year overall survival of hepatoblastoma 

patients was 75%, while 3-year event-free-survival was 64%. 

Three deaths resulted from disease progression and 3 from 

treatment complications: one was surgery-related (intestinal 

necrosis), one resulted from chemotherapy complications (severe 

neutropenia and typhlitis) and one was unexplained (possible 

anthracycline-related cardiomyopathy). Median followup 

time is 40 months (range from 24 to 72 months). Overall 

survival among standard risk patients was 90% (17 out of 19) 

in comparison with 66% (6 out of 9) in the high risk group. 

Preliminary analysis showed that 75% of standard risk 

patients (12 out of 16; in 3 cases data are not known) responded 

to preoperative chemotherapy (CDDP or PLADO), while from 

the high risk group 55% (5 out of 9) responded to 

SuperPLADO protocol. 

All together in HB cases 5 incomplete tumor resections 

were done: in 3 macroscopic residuum resulted and in 

two microscopic residuum was left. All patients with macroscopic 

residuum died, while one patients after microscopically 

incomplete resection is alive with no evidence of disease 

while another died due to surgical complications. In 4 

cases information on surgery is missing. 

Overall survival in HCC cases at 3 years is 33% but 

one of the two alive patients currently experienced recurrence. 

Among 6 patients with HCC in 2 cases primary tumor resection 

was done (including one orthotopic liver transplantation) 

was performed and 3 patients were treated with induction chemotherapy. 

In 2 non-responders chemombolization was used. 

Two patients were never made operable and in one data on 

surgery are missing. 

Table 1 

Localisation of liver tumors 

Localisation 

Hepatoblastoma 

Both lobem 12 4 

Right lobe 11 2 

Left lobe 5 - 

Table 2 

Liver tumors' stages 

PRETEXT 


I 1 - 

II 14 1 

III 7 2 

IV 6 3 

Hepatocellular 

carcinoma 

Hepatocellular 

carcinoma 

Discussion 

Reported treatment results for HB are similar to those achieved 

by other study groups (OS=75%). Introduction of efficient 

chemotherapy has contributed to a significant progress in the 

treatment of hepatoblastoma as nowadays overall survival 

reaches 70%. Before this era only 25% of HB patients survived 

and tumor resection was possible in not more than 50% 

of patients, while half of them died later due to metastases 

[3]. Similarly retrospective analysis of the Polish treatment 

results for the period 1985-1995 showed inferior outcome with 

only 44% long term survival in hepatoblastoma due to the 

scarcely used preoperative chemotherapy and few extended 

liver resections done [1, 6]. In that period as many as 30% of 

analyzed tumors were never made operable. Thus, currently 

achieved treatment results of hepatoblastoma in Poland constitute 

a significant improvement in comparison with our past 

series [1]. However despite overall improvement prognosis of 

the high risk patients remains uncertain – with 66% overall 

survival vs. 90% for the standard risk group. Similar results 

were achieved in the SIOPEL 2 study, in which 3-year eventfree-survival 

for high risk patients was 47% vs. 89% for standard 

risk [9, 10]. Most failures resulted from tumor’s 

unresectability and insufficient or complete lack of response 

to preoperative chemotherapy. This confirms that high risk 

patients require some modifications in treatment strategy, 

which should be aimed at increase of resection rate. This includes 

more efficient induction chemotherapy, as well as more 

frequent use of liver transplantations in hepatoblastoma cases 

unresectable by conventional means [7, 8]. 

Treatment results of hepatocellular carcinoma (33% 

overall survival) were clearly inferior to hepatoblastoma is 

confirmed by most international study groups, in which survival 

remained in the range of 20-30% [2, 5]. Complete tumor 

resection should be an essential treatment step, both in 

HB and HCC. In the reported series all patients after macroscopically 

incomplete tumor resections died, while all HB 

cases, who underwent complete resection, had been alive and 

well. Due to the relatively short follow-up of some patients, 

incomplete data and limited number of cases presented analysis 

has a preliminary character only. 

Conclusions 

1. Overall survival achieved in hepatoblastoma – 75%, as well 

as EFS = 64% are similar to other international reports and 

constitute a significant improvement in comparison with 

the Polish past series (1985-1995): OS = 44%. 

2. Complete tumor resection remains an essential step to 

achieve cure in primary epithelial liver tumors (HB and 

HCC). It is however difficult in high risk HB, as well as in 

HCC patients. 

3. High risk HB patients require modification of the current 

treatment strategy due to inferior survival than in standard 

risk HB. 

4. HCC are associated with poor prognosis and require completely 

different therapeutic strategy than HB.

Table 3 

Characteristics of surgical treatment 

Type of surgery Hepatoblastoma Hepatocarcinoma Radicality 

RT Hemihepatectomy 9 1 7rad. / 2 non-rad. 

LT Hemihepatectomy 3 - 3 rad. 

Extended Hemihepatectomy 7 - 5 rad. / 2 non-rad. 

Atypical 3 - 2 rad. /1 non-rad. 

Central resection 1 - 1 rad. 

Liver transplantation - 1 1 rad. 

Never operable 1 3 - 

Missing data 4 1 - 

References 

1. Czauderna P, Popadiuk S, Korzon M, 

et al (2001) Multicenter retrospective 

analysis of various primary pediatric 

malignant hepatic tumours management 

in the series of 47 Polish patients (1985- 

1995). Eur J Pediatr Surg 11:82-85 

2. Czauderna P, Mackinlay G, Perilongo 

G, et al (2002) Hepatocellular carcinoma 

in children – Results of the first prospective 

study of the International Society 

of Paediatric Oncology. J Clin Oncol 

20:2798-2804 

3. Ehrlich PF, Grereenberg ML, Filler 

RM (1997) Improved long-term survival 

with preoperative chemotherapy 

for hepatoblastoma. J Pediatr Surg 

32:999-1002 

4. Katzenstein HM, London WB, 

Douglass EC, Reynolds M, Plaschkes 

J, Finegold MJ, Bowman LC (2002) 

Treatment of unresectable and metastatic 

hepatoblastoma: a pediatric oncology 

group phase II study J Clin Oncol 

20:3438-3444 

5. Katzenstein HM, Krailo MD, 

Malogolowkin MH, et al (2002) Hepatocellular 

carcinoma in children and 

adolescents: results from the Pediatric 

Oncology Group and the Children’s 

Cancer Group Intergroup Study. J Clin 

Oncol 20: 2789-2797 

6. Korzon M, Popadiuk S, Stoba C, et al 

(1997) Analiza retrospektywna 

zlosliwych guzów watroby u dzieci w 

Polsce w latach 1985-1995. Pediatr Pol 

11 (Suppl):37-42 (in Polish) 

7. Ortega JA, Douglass EC, Feusner JH, 

et al (2000) Randomized comparison of 

Cisplatin/Vincristine/5-Fluorouracil and 

Cisplatin/continuous infusion Doxorubicin 

for the treatment of paediatric 

hepatoblastoma (HB): a report from the 

Children’s Cancer Group and the Paediatric 

Oncology Group. J Clin Oncol 

18:2665-2675 

8. Otte JB, Aronson DC, Brown J, 

Czauderna P, et al (2004) Liver transplantation 

for hepatoblastoma. results 

from the International Society of Pediatric 

Oncology (SIOP) study siopel-1 

and review of the world experience. 

Pediatr Blood Cancer 42:74-83 

9. Perilongo G, Shafford E, Plasches J 

(2000) SIOPEL trials using preoperative 

chemotherapy in hepatoblastoma. The 

Lancet Oncology 1:94-100 

10. Perilongo G, Shafford E, Maibach R 

(2004) Risk-adapted treatment for childhood 

hepatoblastoma. Final report of the 

second study of the International Society 

of Pediatric Oncology - SIOPEL 2. 

Eur J Cancer 40:411-421 

11. Pritchard J, Brown J, Shafford E (2000) 

Cisplatin, doxorubicin and delayed surgery 

for childhood hepato-blastoma: a 

succesful approach - Results of the first 

prospective study of the International 

Society of Paediatric Oncology J Clin 

Oncol 18:3819-3828 

12. Schnater JM, Aronson DC, Plaschkes 

J (2002) Surgical view of the treatment 

of patients with hepatoblastoma: results 

from the first prospective trial of 

the International Society of Pediatric 

Oncology Liver Tumor Study Group 

(SIOPEL-1). Cancer 94:1111-1120 

13. Stoba C, Czauderna P, Narozanski 

(2001) Pierwotne nowotwory watroby 

u dzieci – aspekty chirurgiczne. In: 

Wybrane nowotwory jamy brzusznej 

u dzieci. A. Jankowski, P. Mankowski 

(Eds) Stowarzyszenie Pomocy 

Dzieciom Wymagajacym Leczenia 

Chirurgicznego, Poznan, pp. 12-30 (in 

Polish)



Liver transplantation for primary malignant liver tumors 

in children – single center experience 

Piotr Kaliciñski 1 , Hor Ismail 1 , Andrzej Kamiñski 1 , Danuta Perek 2 , Joanna Paw³owska 3 , 

Przemys³aw Kluge 4 , Joanna Teisserye 1 , Marek Szymczak 1 , Tomasz Drewniak 1 , 

Pawe³ Nachulewicz 1 , Bo¿ena Dembowska-Bagiñska 2 , Marek Stefanowicz 1 , 

Andrzej Byszewski 5 , Ma³gorzata Manowska 5 





1 

Department of Pediatric Surgery and Organ Transplantation 

2 

Department of Oncology 

3 

Department of Gastrology, Hepatology and Immunology 

4 

Department of Patomorphology 

5 

Department of Anesthesiologu and Intensive Care 



Abstract 

The aim of the study was to analyse results of treatment of primary liver tumors in children in a center 

performing the liver transplant program. Clinical material consisted of 42 children treated for hepatoblastoma 

or hepatocarcinoma since 1990, since liver transplantant program was established. Retrospective comparison 

of results was performed according to type of the tumor and different treatments protocols. Survival of 

patients with hepatoblastoma is similar after primary conventional treatment (chemotherapy and resection) 

and primary transplantation after chemotherapy, however number of transplanted patients is to low to draw 

any conclusions. Patients with tumor recurrence after resection probably should not be qualified for transplantation 

as their prognosis is very poor. In hepatocarcinoma group overall results after chemotherapy and 

primary transplantation seem to be much better than after conventional therapy. This observation should be 

confirmed however with longer follow-up of these patients. In conclusion liver transplantation should be 

considered as an option early in the process of treatment of primary malignant tumors in children particularly 

with hepatocarcinoma, and transplantation as a primary surgical treatment in all patients with unresectable or 

marginally resectable tumors and without extrahepatic extension. 

Key words: children, hepatoblastoma, hepatocellular carcinoma, liver transplantation, primary liver tumors 

Introduction 

Primary hepatic malignancies are main indication for liver 

transplantation in 2-4% pediatric recipients. This is well less 

than in adult population, where patients with hepatocellular 

carcinoma account for 11-12% of liver transplantations [3-5]. 

In contrast to adult group, there are no established strict criteria 

for qualification for transplantation in children with primary 

liver tumors, particularly that various types of hepatic 

tumors may develop in childhood (hepatoblastoma, 

hepatocarcinoma, undifferentiated sarcoma, angiosarcoma 

etc.) [1, 7, 12]. The aim of this study is to summarize our 

experience with this particular group of liver transplant recipients. 


Liver transplant program was initiated in our institution in 

march 1990. Since then 222 liver transplants were performed 

till July 2004. We have analysed retrospectively data of 42 

children treated for hepatoblastoma (HBL) or hepatocarcinoma 

(HCC) within the same period in our institution. 

Patients with HBL 

Between 1990 and 2004, 23 patients aged 5 months to 16 

years, most with advanced tumors, were treated for HBL. 

Twenty children with HBL were treated initially with chemotherapy 

and then by conventional surgery – partial liver resection. 

In two of them recurrence was suspected despite of mul- 


Piotr Kalicinski MD, PhD 



Phone: +48-22-8151360 

Fax: +48-22-8151450 

E-mail: piokal@czd.waw.pl

0000000000000000000 

0000000000000000000 

0000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

000000000000000000 

000000000000000000 

000000000000000000 

000000000000000000 

e 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000000 

0000000000000000000 

0000000000000000000 

0000000000000000000 

ea 

00000000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000 

0000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000 

0000000000000 

00000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000 

00000000000000000 

00000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

0000000000000000 

000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 

0000000000000000 

0000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000 

00000000000000000 

00000000000000000 

00000000000000000 

00000000000000000 

e 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

0000000000000000000000000000000000000000000000000000 

00000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

00000000000000000 

00000000000000000 

00000000000000000 

ea 

000000000000000000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

00000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

00000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

00000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

00000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

00000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

00000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

00000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000000000000000000 

00000000000000000 

00000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000000000000000000 

000000000000000000000000000000000000000000000000000000000000000000 

112 

tiple postoperative chemotherapy courses, and were qualified 

for total hepatectomy of the remaining liver and orthotopic 

liver transplantation. Another 2 children were transplanted 

immediately after initial chemotherapy. In 1 patient only palliative 

treatment (chemotherapy) was administered. 

Patients with HCC 

There were 19 patients aged 3 to 18 years treated for HCC. In 

9 children initial chemotherapy was administered, followed 

by partial liver resection, in 2 of them despite multiple resections 

tumor recurred and salvage transplantation was done in 

both of them. In 5 children liver transplantation was performed 

as primary surgical treatment. All of them exceeded Milano 

criteria for transplantation in patients with HCC. Initial chemotherapy 

was given in 4 of them. In 4 pts liver transplantation 

was performed due to suspected tumor development in 

cirrhotic liver (on the basis of CT or ultrasound findings or 

growing serum alphafetoprotein concentration). No 

pretransplant chemotherapy was administered to these patients. 

In 1 child only palliative chemotherapy was given. 

reccurence. One patient is alive after extensive resection, however 

with poor prognosis due to extrahepatic invasion of the 

tumor at the time of surgery. 

Among patients transplanted, 9/11 (81,8%) are alive, 

while 2 deaths were not related to tumor reccurence (early 

graft failure in 1 and chronic rejection in 1 (Fig. 2). 

In all 4 patients with cirrhosis and suspected tumor 

small HCC was found in explanted liver. One of these patients 

died of graft primary non-function. One of two patients 

who underwent multiple resections before transplantation presented 

pulmonary metastases almost 2 years after transplantation 

and extensive metastasectomy was performed, however 

prognosis for this patient is rather poor. No recurrence or 

metastatic dissemination was noted until now among 5 patients 

who underwent primary liver transplantation after total 

hepatectomy. 

Results 


Twelve out of 23 (52,2%) children with HBL are alive and 

without active disease, while 11 patients died (47,8%). Best 

results were obtained with standard therapy consisting of chemotherapy, 

tumor resection and postoperative chemotherapy 

with long term survival of 11/20 pts (55,0%), including two 

patients who died after transplantation performed due to tumor 

recurrence. Among 2 patients who were transplanted primarily 

- 1 died 6 months after transplantation due to 

posttransplant lymphoproliferative disease. Another one is 

alive and well 25 months after Tx.Two additional children 

transplanted for tumor recurrence after previous resections died 

of metastatic disease despite of posttransplant chemotherapy. 

No graft involvement was observed in these patients (Fig. 1). 

Hepatocellular carcinoma 

Overall survival among 19 children with HCC is 12/19 

(63,2%), 42,9% (3/7) children are alive and free of tumor 

after CHT and resection, while all other died of tumor 

3 

3 

00000000000000000 

alive&well 

e e ea 

e a e alive wi 

Fig. 2 Hepatocarcinoma - actual results according to type of treatment (1990- 

2004) 

Posttransplant observation is still relatively short in survivors 

(range: 5 - 40 months). AFP concentration remains within 

normal limits in all, however it is necessary to say that it was 

not elevated in 2 of 11 before transplantation (Fig. 3). Chemotherapy 

(1-3 courses) was given after transplantation to 5 

patiens. It had to be stopped in most of them after 1-2 courses 

due to complications. Liver graft function is good in all but 

one patient who developed severe cholangitis after chemotherapy, 

further complicated by intrahepatic bile duct strictures. 

1 

1 

11 

0000 

000 

AFP (ng/ml) 

0000 

000 

0000 

1 1 

1 

000 

0 

before treatment before a t 

0000000000000000000 

alive&well 

e e ea 

e a e 

Fig. 1 Hepatoblastoma - results according to type of treatment (1990-2004) 

Fig. 3 Alphafetoprotein (AFP) in 9 patients with HCC after liver transplantation 

(follow up 5-40 months)

113 

Discussion 

Total hepatectomy with liver transplantation became an important 

therapeutic option for children with malignant primary 

hepatic tumors [1, 5, 9, 12]. Overall long term survival in pediatric 

liver transplant recipients is now reaching in leading 

centers 90-95% independently on indication for transplantation, 

while survival of children with hepatoblastoma and 

hepatocarcinoma after conventional treatment (chemotherapy 

+ resection) is much lower, 60-80% and 30-40% respectively 

[2, 4, 5, 11, 12]. Various protocols and trials have been tried 

to improve results of conventional treatment, and much success 

was achieved independently by SIOPEL group, German 

group or Japanese group, mostly in HBL, with less optimistic 

results in patients with HCC [9, 10]. The place of total hepatectomy 

and liver transplantation have been not clearly defined 

in pediatric population, although many patients benefited 

from this type of treatment on the individual qualification basis 

[1, 9, 12]. Standard qualification criteria for transplantation 

in patients with primary hepatic tumors included in most centers: 

unresectable tumor, no extrahepatic tumor extension, no 

distant metatases (or in case of hepatoblastoma pulmonary 

metastases which cleared with chemotherapy), [5, 9]. It is 

obvious that conventional complex treatment is best for most 

patients with hepatoblastoma, however in our opinion transplantation 

should be considered as first surgery in these children 

in which tumor is unresectable, but also in cases of successful 

response of large, unresectable tumors to chemotherapy 

which allows downstaging of the tumor and complete extensive 

but marginal resection. Although not all agree that narrow 

margin free of HBL in resected liver specimen is important 

for long term survival, but one should keep in mind that 

recurrence of HBL in these patients indicates almost uniformly 

poor prognosis also after transplantation [2, 5, 9, 12]. 

Conventional treatment can not offer such good results in children 

with hepatocarcinoma despite intensive chemotherapy 

and improvements in resection techniques. In adult population 

large groups of patients with HCC were qualified to liver 

transplantation since 25 years, however no criteria for transplantation 

in these patients existed until study done by 

Mazzaferro et all in 1996 [7, 8]. They defined so called Milano 

criteria which have been accepted by most adult transplant 

centers. According to these criteria liver transplantation should 

be offered to patients with: one single tumor with diameter 

less than 5 cm, maximum number of 3 tumors with diameters 

less than 3 cm each no vascular invasion and no extrahepatic 

metastases. With these criteria 4 years survival after liver transplantation 

in adults achieve 75%, which is impossible to 

achieve by conventional treatment. One can not directly transfer 

Milano criteria to pediatric population, as tumor behaviour 

and response to chemotherapy in children are different in children 

and adults. There are no large enough groups to perform 

such analyses in pediatric patients however. Since all our patients 

with “non-accidentaloma” did not fit into Milano criteria 

one should expect very early tumor recurrence and dissemination 

in patients on immunosuppression. We did not 

observe such events in any patient till now with medium length 

follow-up. Multicenter analysis should be performed probably 

to try to define pediatric criteria for liver transplantation 

in patients with HCC, as single center can not collect number 

of patients necessary for such analysis. 

In conclusion, liver transplantation should be considered 

as an option early in the process of treatment of primary 

malignant tumors in children, particularly with 

hepatocarcinoma, and transplantation should be considered 

as a primary surgical treatment in all patients with unresectable 

or marginally resectable tumors and without extrahepatic extension. 

Living related donor transplantation is one of important 

solution in best timing of transplantation within treatment 

protocol [6]. 

References 

1. Cillo U, Ciarleglio FA, Bassanello M, 

et al (2003) Liver transplantation for the 

management of hepatoblastoma. Transplant 

Proceed 35:2983-2985 

2. Dicken BJ, Bigam DL, Lees GM 

(2004) Association between surgical 

margins and long-term outcome in advanced 

hepatoblastoma. J Pediatr Surg 

39:721-725 

3. European Registry of Liver Transplantation 

Report (ELTR), 2003. 

4. Kaliciñski P (1997) Przeszczep w¹troby. 

Wybrane zagadnienia. Fundacja Polski 

Przegl¹d Chirurgiczny (Ed), pp 1-138 

5. Kamiñski A, Kaliciñski , Ismail H, et 

al (2003) Liver transplantation in children 

with hepatic tumors. Pediatr Transplant 

7(suppl):124 

6. Lo CM, Fan ST, Liu CL, Chan SC, 

Wong J (2004) The role and limitation 

of living donor liver transplantation for 

hepatocellular carcinoma. Liver Transpl 

10:440-447 

7. Mazzaferro V, Regalia E, Doci R, et al 

(1996) Liver transplantation for the 

treatment of small hepatocellular carcinomas 

in patients with cirrhosis. N Engl 

J Med 334:693-699 

8. Nart D, Arikan C, Akyildiz M, et al 

(2003) Hepatocellular carcinoma in liver 

transplant era: a clinicopathologic analysis. 

Transplant Proceed 35:2986-2990 

9. Otte JB, Pritchard J, Aronson DC, et al 

(2004) Liver transplantation for 

hepatoblastoma: results from the International 

Society of Pediatric Oncology 

(SIOP) study SIOPEL-1 and review of 

the world experience. Pediatr Blood 

Cancer 42:74-83 

10. Perilongo G, Shafford E, Maibach R, et 

al (2004) Risk-adapted treatment for 

childhood hepatoblastoma. Final report 

of the second study of the International 

Society of Paediatric Oncology – 

SIOPEL 2. Eur J Cancer 40:411-421 

11. Prokurat A, Kluge P, Chrupek M, 

Koœciesza A (2002) Possibilities and 

limitations in the treatment of hepatocellular 

carcinoma in children. Ann 

Diagn Paediatr Pathol 6:51-57 

12. Srinivasan P, McCall J, Pritchard J, et al 

(2002) Ortotopic liver transplantation 

for unresectable hepatoblastoma. Transplantation 

74:652-655



Hepatocyte ultrastructure in non-hemolytic hyperbilirubinemias 

El¿bieta Czarnowska, Bogdan WoŸniewicz 





Departament of Pathology, 

The Children’s Memory Health Institute, 


Abstract 

The non-hemolytic hyperbilirubienias comprise a group of syndromes that are clinically and some in cases 

biochemically and genetically well characterized, and ultrastructural investigations of biopsy samples have 

still diagnostic applications. In this study utrastructural analysis of biopsy tissues in order to identify pattern 

of features characteristic for different non-hemolytic hyperbirubinemias among group of patients with clinical 

presentation of hyperbilirubinemia and normal histopathology were done. A group of 62 patients was 

selected on the basis of mild unconjugated hyperbilirubinemia and, in some cases, on the basis of a mild 

unconjugated or conjugated hyperbilirubinemias in the presence of repeatedly normal liver function tests and 

absence of hyperhemolysis. Electron microscopic investigation distinguished between patients in three groups 

of primary hyperbilirubiemia: Gilbert syndrome (21 pts), Dubin-Johnson syndrome (5 pts), Rotor syndrome 

(3 pts) and a group with non specific changes (33 pts). The ultrastructural features of hepatocytes that univocally 

distinguish primary hyperbilirubinemias from secondary diseases were presented. Electron microscopic 

methods permit on recognition of non-specific changes in the ultrastructure of liver tissue among samples 

obtained from patients with non-hemolytic hyperbilirubinemias and can be still usefull in diagnosis of 

jaundince. 

Key words: congenital hyperbilirubinemia, Dubin-Johnson disease, Gilbert disease, Rotor disease, ultrastructure 

Introduction 

Idiopathic hyperbilirubinemias may encompass several primary 

and secondary diseases of bilirubin metabolism [2]. Bilirubin, 

an oxidative product of the hem group of hemoglobin, 

myoglobin and cytochrome P-450, in the liver is conjugated 

with glucuronic acid and than excreted in bile via ATPdependent 

transporter [5]. Currently, injuries of molecular 

bases of this metabolic pathway are known [5]. Although, this 

knowledge allow for precision diagnosis of non-hemolytic 

hyperbilirubinemias in most cases, but still ultrastructural investigation 

of biopsy samples have diagnostic application. 

Clinically, conjugated and unconjugated hyperbilirubinemia 

can be distinguished on the basis of blood bilirubin level, normal 

liver tests and exclusion of hemolytic disease. In conjugated 

hyperbilirubinemias i.e Gilbert, Crigler-Najjar syndrome, 

transport of bilirubin from plasma into liver cells is 

impaired, while in conjugated hyperbilirubinemias, i.e. Dubin- 

Johnson, Rotor syndrome, microsomal enzymes transferase, 

bilirubin transport from hepatic cells into bile caniculi system 

are disturbed [2, 5]. In non-hemolytic hyperbilirubinemias, 

liver histopathology may present normal feature or cholestasis. 

These investigations were aimed to identify ultrastructural 

pattern of characteristic features for different nonhemolytic 

hyperbirubinemias among group of patients with 

clinical presentation of hyperbilirubinemia and normal histopathology. 


Material 

62 biopsy samples from patients (age 6-17 years, mean 10,2 

± 4,6) with marked liver disfunction and clinical symptoms 

of non-hemolytic hyperbilirubinemia were selected for electron 

microscopic studies among samples collected in the registry 

of Pathology Department of CMHI. Tissue samples were 

routinely preserved for histological, histochemical and ultrastructural 

examination. All of the examinated patients had total 

bilirubin in the range of 1,3-5,1 mg% (Table 1). 


Elzbieta Czarnowska, PhD 


The Children's Memory Health Institute 

Al. Dzieci Polskich 20, 04-730 Warsaw, Poland 

Phone: +48 22 815 48 15 

E-mail: czar@czd.waw.pl

116 

Table 1 

Total and direct bilirubin level in serum in distinguished groups of primary hyperbilirubinemia and with non-specific changes 

Disease 

Gilbert 

Compl. 

Incompl. 

Number of patients Age (years) 

10 

11 

10-17 

6-7 

Bilirubin (mg%) 

Total Average Direct Ave 

1,5-4,6 2,56 0,5-0,9 0,62 

1,2-5,1 2,48 0,3-1,1 0,95 

Dubin-Johnson 5 9-17 1,3-4,7 2,54 0,6-2,0 1,05 

Rotor 3 16-17 1,6-4,5 2,10 0,4-0,8 0,55 

Non specific 33 7-17 1,2-5,7 2,62 0,2-2,8 1,45 

Histology 

Five mm thick paraffin sections were stained in a routine manner 

with hematoxylin-eosin, azan, silver, PAS, PAS with 

diastaze digestion and than reviewed to exclude other disease 

than hyperbilirubinemia. 

Electron microscopy 

Samples were fixed in 2.5% glutaraldehyde in cacodylate 

buffer at pH 7.3, postfixed in 2% OsO 4, 

dehydrated in alcohols 

and embedded in Epon 812. Thick sections stained with toluidine 

blue were analysed under light microscope. Ultrathin 

sections stained with uranyl acetate and lead citrate were examined 

by electron microscopy. 

A 

B 

Results 

Examinations of paraffin and semi-thin sections showed in all 

cases an apparently normal liver, except for Dubin-Johnson 

that shown the existence of dark pigment predominantly in 

centrolobular areas. 

On the basis of EM analysis morphological equivalents 

of primary congenital hyperbilirubinemias were confirmed 

in 29 cases and secondary reactive symptomatic 

hyperbilirubinemias were recognized in 33 cases. 

Gilbert syndrome 

Gilbert syndrome was diagnosed in 21 cases. There were observed 

shortening and mal-developed vascular pole microvilli 

or loss of microvilli, and increase of collagen in the Disse 

space in various intensity. In all patients an increase of the 

endoplasmic reticulum profiles with coexisting dilatation of 

rough endoplasmic reticulum with degranulation and pigment 

granules variable in size containing both electron dense material 

which alternate with less dense material and lipid droplets 

was observed (Fig. 1A). In 10 biopsy samples enlarged or 

giant, oval or polymorphic mitochondria with paracrystalline 

inclusions (Fig. 1B) were present. In all cases the biliary pole 

of hepatocytes shown normal caniculi and microvilli, but with 

numerous pigment granules around. Lipofuscin granules were 

Fig. 1 Gilbert syndrome: (A) vascular pole of hepatocyte with single microvilli. 

Mag. x 40 000, (B) mitochondrium containing paracrystallines, mag. x 

22 000 

common finding. The remaining group of 11 patients showed 

uncompleted ultrastructural features of Gilbert disease, i.e. lack 

of giant mitochondria and no paracristalline in mitochondrial 

matrix. 

Dubin-Johnson syndrome 

Dubin-Johnson syndrome was defined in 5 cases. There 

were not seen alterations in the vascular pole of hepatocytes, 

which exhibit normal microvilli. Slightly increased number 

of smooth endoplasmic reticulum profiles that formed small 

vesicles, normal mitochondria with some variation in size, 

well developed Golgi profiles located around bile canicules 

were common features. Dilation of bile caniculi with loss of 

microvilli and characteristic pigment granules around were 

seen (Fig. 2). Pigment granules were composed mainly of 

granular material, however vesicle-like or lipid components 

were also seen. They were polygonal in shape and delineated 

by a single membrane. The lack of lipofuscin in hepatocytes 

was common feature.

117 

Fig. 2 Dubin-Johnson syndrome: a bile pole of hepatocytes with pigment 

granules around the bile canicule. Mag.x 30 000 

Rotor Syndrome 

Rotor syndrome was recognised in 3 cases. Only slight changes 

in hepatocytes were observed: some of the bile canicules were 

dilated and exhibited reduced number of microvilli, the 

interhepatocytic junctional interspaces were dilatated and cell 

membranes along the lateral cell surface formed microvilli 

(Fig. 3a), cell membranes facing the Disse space showed reduced 

number of microvilli. In side of hepatocytes a few pleomorphic 

mitochondria or megamitochondria with 

paracristalline were seen. The most characteristic feature was 

lack of lipid component in pigment granule that composed of 

electron dense and less dense granular material. They were 

distributed mainly around bile canicules (Fig. 3b). 

Non characteristic lesions 

Non characteristic ultrastructural fine lesions accompanying 

secondary hyperbilirubinemia were found in 33 cases. Reactive 

changes were recognized in 9 cases. There were increased 

number and elongation of microvilli on the vascular pool (Fig. 

4), polymorphic mitochondria (sometimes with paracrystalline 

inclusions), increased number of smooth endoplasmic reticulum 

profiles, the presence of single lipid droplets and pigment 

or lipofuscin granules in cytoplasm, and collagen fibrils 

in Disse spaces. 

Fig. 4 Non specific changes. Vascular pole of hepatocyte with numerous, 

elongated microvilli. Mag.x 40 000 

Discussion 

In about 50% of the patients with clinical symptoms of hyperbilirubinemia 

electron microscopy confirmed primary congenital 

hyperbilirubinemia. Among of them, complete ultrastructural 

features of Gilbert syndrome were presented in 10 cases, 

uncomplete Gilbert in 11 cases, Dubin-Johnson in 5 cases 

and Rotor in 3 cases. 

In seventies the ultrastructural observations were helpful 

to establish the morphological equivalent of primary hyperbilirubinemia 

and distinguish combination of defects impairing 

primary hyperbilirubinemias, particularly in Gilbert 

syndrome when complete and incomplete features coexisted 

[1, 4, 6]. Our patients with Gilbert disease clinically were characterized 

by relatively low-grade direct bilirubin and they 

reached a higher proportion of unconjugated bilirubin. These 

abnormalities were correlated with decrease or the lack of microvilli 

at sinusoidal pole. 

It is known that the liver in Dubin-Johnson syndrome 

might be characterized by slight and inconsistant changes. In 

our patients, the canicular pole exhibited normal microvilli or 

disappearing microvilli coexisting with a slightly dilated lumen. 

These results suggest that disturbances of bile flow were 

caused by different disfunctions of the canicular area. These 

are in agreement of with distinct mechanism of bilirubin glucuronide 

transport involving ATP-dependent and ATP-independent 

system that have been proposed [7]. 

Fig. 3 Rotor syndrome: (A) lateral hepatocytes 

surface with number of microvilli. Mag. x 

50 000; (B) pigment granules composed of 

dense granular and lipid material, mag.x20 000 

A 

B

Patients with Rotor syndrome ultrastructurally presented fairly 

similar features to Gilbert and to Dubin-Johnson simultaneously: 

some injury of vascular and canicular pole concomitant 

with an increase of smooth endoplasmic reticulum and 

pleomorphic mitochondria in hepatocytes. These similarities 

between some ultrastructural features of Gilbert and Dubin- 

Johnson syndrome were signaled in literature before [8]. The 

presence of microvilli along the lateral surface of hepatocytes 

was a particular feature, which distinguished patients with 

Rotor syndrome from other patients. Patients with Rotor syndrome 

exhibited lower level of direct bilirubin than 

unconjugated, and this observation does not agree with main 

idea of liver defect [5]. However, similar data on a direct bilirubin 

in serum have been mentioned in literature [3]. 

Electron microscopic methods permit on recognition 

of non-specific changes in the ultrastructure of the liver tissue 

among samples obtained from patients with non-hemolytic 

hyperbilirubinemias and can be still usefull in diagnosis of 

jaundince. 

References 

1. Bethelot P, Dhumeaux D (1978) New 

signts in the classification and mechanism 

of hereditiary, chronic, nonhaemolytic 

hyperbilirubinemias. Gut 

19:474-480 

2. Chowdhury JR, Wolkoff AW, Wolkoff 

N, Chodhury R, Arias JM (1995) 

Hereditiary jaundice and disorders of 

bilirubin metabolism. In: Scriver ChR, 

Beaudet AL, Sly WS, Walle D (eds) The 

metabolic basis of inherited diseases, 

New York, Vol 1, pp 2161-2208 

3. Czarnecki J, Cichosz H, Slomke M, 

Siezieniewska G, Czechowska G (1988) 

Clinical, biochemical and morphologic 

analysis of congenital non-hemolytic 

hyperbilirubinemia. Wiad Lek 

11(17):1151-1156 (in Polish) 

4. Dawson J, Carr-Loske DL, Talbot JC, 

Rossenthal FD (1978) Hepatic ultrastructure 

in Gilbert syndrome: evidence 

two populations. Gut 19A:995-999 

5. Iyanagi T, Emi Y, Ikushiro S (1998) Biochemical 

and molecular aspects of genetic 

disorders of bilirubin metabolism. 

Biochem Biophys 1407:173-184 

6. McGee JOD, Allan JG, Russel RJ, 

Patrick RS (19750 Liver ultrastructure 

in Gilbert syndrom. Gut 16: 220-224 

7. Nishida T, Gatmaitan Z, Roy Chodhry 

R, Arias IM (1992) Two distinct mechanisms 

for bilirubin glucuronide transport 

by rat ble membrane vesicles. Demonstration 

of defective ATP-dependent 

transport in rats (TR-) with inherited 

conjugated hyperbilirubinemia. J Clin 

Invest 90:21-2135 

8. Okolicsanyi L, Torrado A, Nussle D, 

Gardid D, Gautier A (1969) Etude 

clinique et ultrastructurale d’un cas de 

syndrome de Rotor. Revue Int Hepat 

19:393-398



Congenital Heart Diseases Database: analysis of occurrence, 

diagnostics and coexistence with anomalies of other systems. 

Single institution experience 





Anna Rybak, Aneta Wójcik, Bogdan WoŸniewicz 


The Children’s Memorial Health Institute 


Abstract 

The Congenital Heart Diseases Database was created in the Children’s Memorial Health Institute during 

the period of 2001 – 2003. It contains over 1000 preparations of congenital heart defects of children aged 

from 0 till 14 years. According to our data, the most common defect was ventricular septal defect. The 

compatibility of the clinic and pathological diagnoses was high (about 80% of the diagnoses were confirmed), 

however significantly higher after 1990. Although in our Database plenty of coexisting anomalies 

and diseases of the other systems or organs occurred, we found no significant relations between them and 

congenital heart defects. 

Key words: congenital heart disease, database, heart defects 

Introduction 

Since 1980 until now, the Children’s Memorial Health Institute 

of Warsaw has assembled over 1200 congenital heart defects 

preparations from patients aged 0-14 years. Between 

September 2001 and March 2003 we have made the repeated 

sections of those preparations and created “The Congenital 

Heart Diseases Database” (CHDDb) that includes 1037 records 

with full descriptions of the hearts’ and defects’ anatomy, clinical 

and sectional diagnoses, anomalies of other systems and 

performed operations. This is to present the direct conclusions 

that have arisen from this Database, including the frequency 

of occurrence of the particular heart defects, compatibility of 

the clinical and autopsy diagnoses and finally, regularity in 

coexistence of the congenital heart defects with the anomalies 

of the other systems. 


The Congenital Heart Diseases Database is composed of 1037 

preparations of the hearts with the congenital defects. They 

are fully described in four files which contain personal patients’ 

data, clinical diagnoses with the descriptions of defects 


Bogdan Wozniewicz 



Aleja Dzieci Polskich 20, 04-736 Warsaw, Poland 

of the other systems and types of performed operations, pathologic 

diagnoses of the heart defect and coexisting diseases. 

The last file shows complete measurement of all the hearts 

including external dimensions as well as internal measurements 

of the valves, arteries, veins or myocardium. 

The patients’ age was between 1 day and 14 years. 

There were 437 female and 600 male patients described and 

they were all admitted to Children’s Memorial Health Institute 

during the period 1980 – 2003 and almost half of them 

were operated because of the heart defect. 

A nomenclature for the congenital heart disease was 

based on the International Nomenclature for Congenital Heart 

Surgery that was officially adopted at the Annual Meeting of 

the EACTS in Glasgow on September 6, 1999 [3]. 

Results and discussion 

Occurrence of the congenital heart defects 

The most common anomaly that appeared in our material was 

ventricular septal defect (VSD) with the frequency just under 

30% (Fig.1). In details, there were 31 cases of VSD in muscular 

part of the septum (VSD single–29; VSD multiple-2 cases) 

and 259 cases of this defect in other parts of the interventricular 

septum (VSD; NOS). 

Phone: +4822-815-1960 

E-mail: b.wozniewicz@czd.waw.pl

% 

30 

25 

20 

15 

10 

5 

0 

VSD PDA TGA ASD TOF AVC CoA HLHS PA HRV PS DORVTAPVC 

Fig. 1 Occurrence of the congenital heart diseases 

The frequency of patent arterial duct (PDA) was about 19%. 

Similarly, other congenital heart defects like transposition of 

the great arteries (TGA) and atrial septal defect (ASD) are 

about 18%. Although there were over 160 ASD diagnosed, 

the largest group was due to persisted foramen ovale (PFO ) 

type (50%). ASD secundum type appeared in 69 cases that 

was just over 40% of all atrial defects. Other atrial defects, 

like ASD type sinus venosus, coronary sinus type or single 

atrium, appeared rarely (14 cases). 

Tetralogy of Fallot (TOF) was diagnosed in about 12% 

of all cases. Atriovertricular canal (AV canal) as well as coarctation 

of aorta (CoA) concerned 9% of all heart defects. Hypoplastic 

left heart syndrome (HLHS) likewise pulmonary atresia 

(PA) were visible in 7% of the examined preparations. Similarly 

hypoplastic right ventricle (HRV) or pulmonary stenosis 

(6,5%). Double outlet from the right ventricle (DORV) and total 

anomalous pulmonary venous connection (TAPVC) appeared 

in less than 5% of all cases. The rarest were aortic atresia (3 

patients), aortic valve hypoplasia (1 patient), pseudotruncus (3 

patients) and inversion of the ventricles (2 patients). 

There is also some regularity with reference to the gender 

of the patients. The difference is most significant in the 

group of diseases like partial anomalous pulmonary venous 

connection or right ventricle outlet obstruction, which occurred 

two to five times more often in the female group than in the 

male group. However aortic atresia, Bland-White-Garland syndrome 

and cor triatriatum occurred in our material only in 

reference to the male group. Such defects like truncus arteriosus, 

aortic stenosis, HLHS, TGA or mitral atresia occurred 

over two times more often in the group of boys than in the 

group of girls. 

Clinical and autopsy diagnosis 

There are 1037 congenital heart diseases in the Database, however 

clinical diagnostics and diagnoses were established due 

to 867 patients. In this group, autopsy diagnoses differed from 

clinical diagnoses in about 19% (166 cases). 

Clinical diagnostics was based on physical examination, ECG, 

blood pressure measurement, echocardiography and invasive 

cardiology, which was used from the beginning 1990. Because 

of that fact, we decided to divide our data into two groups: in 

the first group there were patients that had been admitted to 

the Children’s Memorial Health Institute till the end of 1989 

(533 patients); the second group included rest of the patients 

(334 patients). The results of a comparison between clinical 

and pathologic diagnoses were as follows: 166 children (22%) 

from the first group were misdiagnosed during their hospitalization. 

In the second group, where the patients were additionally 

diagnosed by means of the invasive cardiology, the 

number of misdiagnoses has fallen to 15% (Table 1). Among 

166 incorrectly diagnosed congenital heart defects, the most 

often was HLHS (about 13%), than TGA and AV canal (just 

under 9%). In contrast, the greatest conformity between clinic 

and autopsy was due to interrupted aortic arch (over 90%) 

and TAPVC or PDA (about 80%). 

Table 1 

Compatibility between clinical and autopsy diagnoses 

Period 

1980-1989 78% 

1990-2003 85% 

Compatibility 

Coexistence of the congenital heart defects with the 

anomalies of other systems 

In reference to the CHDDb there were no significant relations 

between heart defects and other anomalies. The most common 

were malformations of the alimentary tract, polysplenia, 

asplenia, urinary tract defects and additional superior caval 

vein which occurred mainly in reference to heart defects like 

AV canal, coarctation of aorta, TAPVC and TOF. There were 

eight patients with situs inversus, although with different heart 

defects. 

In the group of patients with chromosome aberrations 

there were 24 patients with Down syndrome. Over 62% of 

these patients suffered from atrioventricular canal and about 

34% from VSD. There was one patient with Patau syndrome 

that had coarctation of aorta and one patient with Edward’s 

syndrome who suffered from VSD. 

In our Database occurred six patients with developmental 

multiply malformations. Three patients with Pierre- 

Robin sequence suffered from ASD, VSD and TOF. Two children 

with DiGeorge syndrome had TOF and tricuspid valve 

atresia, while in literature, the most common are the anomalies 

of the aortic arch [1, 2]. The congenital heart disease of 

one patient with Noonan syndrome was AV canal (in literature, 

over 50% patients with Noonan syndrome suffer from 

pulmonary stenosis) [2]. 

References 

1. Becker AE, Anderson RH (1985) Pathologie des Herzens. Georg 

Thieme Verlag 

2. Connor M, Ferguson-Smith M (1998) Podstawy genetyki 

medycznej, PZWL (in Polish) 

3. Lacour-Gayet F, Maruszewski B, Mavroudis C, Jacobs JP, Elliott 

MJ (2000) Presentation of the International Nomenclature for 

Congenital Heart Surgery. Eur Journ Card-Thor Surg 18:128- 

135



Solid pseudopapillary tumor of pancreas. 

Case report and the review of literature 





Andrzej Chilarski 1 , Anna Piaseczna-Piotrowska 1 , Jan Strzelczyk 2 , Stanis³aw £ukaszek 3 , Andrzej Kulig 3 

1 

Depart. of Pediatric Surgery 

3 

Depart. of Pathology, 

Polish Mother’s Health Institute, 

Lodz, Poland 

2 

Depart. of General Surgery and Transplantology, 

Medical University, 

Lodz, Poland 

Abstract 

Authors describe the case of 14 year-old girl operated on because of the neoplasm of tail of pancreas proved 

to be solid pseudopapillary pancreatic tumor. Radical resection was curative. After surgery she made uneventful 

recovery being symptoms’ free for 8 months. The review of literature on this rare, relatively low, 

malignant tumor is presented. 

Key words: pancreas, pseudopapillary tumor, solid tumors 

Introduction 

Solid pseudopapillary tumors of pancreas (SPTP) first described 

by Frantz in 1959 are rare findings among other pancreatic 

tumors [5, 6, 9, 10, 11, 13]. To date a bit more then 500 

cases have been described in the literature [16]. SPTP occurs 

predominantly in teenage girls and young women although 

all age groups may be affected and also males are not disease 

– free [12, 15, 17, 18]. 

The tumors are reckoned as a low-grade malignant papillary 

– cystic neoplasms of the pancreas of unclear histogenetic 

origin, but with highly characteristic, distinct histology and 

usually quite benign biologic behaviour [3, 9, 10, 13, 15, 18]. 

Their clinical presentation is vague and non characteristic. The 

most common symptoms are abdominal pains, nausea, emesis, 

sometimes palpable abdominal mass and jaundice – if the 

head of pancreas is involved [1]. Asymptomatic course is not 

uncommon and diagnosis can be established incidentally following 

blunt abdominal trauma [3, 9, 19]. Complete resection 

is usually curative and so the radical surgery is the treatment 

of choice. No adjuvant therapy is recommended. 

Case report 

Patient 

Girl 14-year-old girl (J. P.; files 054774B) was transferred to 

Department of Pediatric Surgery, Polish Mother’s Health Institute 

from local hospital because of suspected pancreatic tumor. 

She gave the history of year-long, moderate, occasional 

abdominal pain and nausea and single incident of syncope 

just before admission to the hospital. Her general condition 

was quite good. Physical examination revealed palpable mass 

in upper abdomen on left side. No other abnormalities have 

been detected. Laboratory investigations were all within normal 

range. In abdominal ultrasonography a solid tumor in left 

upper quadrant, originating probably from pancreating tail and 

reaching spleen, was found. CT scan confirmed this finding 

(Fig. 1) showing the tumor of pancreas tail about 6 cm in 

diameter in the vicinity of spleen and its vessels. The indications 

for surgical treatment were established. Exposure of the 

pancreas revealed the presence of the well vascularized mass 

arising from the tail of pancreas and reaching to the hilum of 

the spleen. Main splenic vessels were surrounded by the tumor 

and could not be liberated. The resection of the tumor 


Andrzej Chilarski, 

Depart. of Pediatric Surgery, 

Polish Mother's Health Institute, Lodz 

Rzegowska St. 281/289 

Lodz, Poland 

Phone: + 48 42 2712136 

E-mail: klinikachirdziec@poczta.onet.pl

122 

In the cystic part of the tumor there were foci of degeneration 

and haemorrhage resulting in creation of pseudocystic and 

pseudopapillary pattern, particularly around small vessels (Fig. 

3). No vascular invasion was found. On the borderline of the 

tumor and pancreatic parenchyma a few otherwise normal 

glands were found with the lining partly consisting of tumor 

cells (Fig. 4). The tumor cells showed: positive immunostaining 

for á 1 

antitrypsin (Fig. 5), chromogranin, vimentin, neuron – 

specific enolase (NSE) and synaptophysin. The final histopathologic 

diagnosis was - solid pseudopapillary tumor of pancreas 

(SPTP). 

Fig. 1 CT scan: tumor of the tail of pancreas 

was performed “en bloc” with the spleen and distal part of the 

pancreas 1 cm from the borderline – within healthy tissue. 

The hemostasis was achieved and the line of resection of pancreatic 

tissue protected with continuous 4.0 nonabsorbable 

suture. No other abnormalities were detected in abdominal 

cavity and parietes were closed in layers. 

Because of splenectomy that has been performed together 

with the tumor’s resection - vaccinations against Streptococcus 

pneumoniae, Haemophilus influenzae and Neisseria 

meningitidis were carried out. Now, eight months postoperatively 

girl is well, free of symptoms. The patient has made an 

uneventful recovery and was discharged home on 12 th postoperative 

day. 

Histopathology 

The specimen consisted of tumor – 6 cm in diameter with margin 

of normal pancreatic tissue attached to the hilum of the spleen, 

without its infiltration, encompassing splenic vessels. The tumor 

was well demarcated, its architecture was partly solid and 

partly cystic. The solid part was built of unifrom, medium-size 

cells with faintly eosinophilic cytoplasm and round nuclei. Mitoses 

were not numerous, but foci of increased mitotic activity 

were present as well. The tumor’s cells formed solid structures, 

cords and trabeculae (Fig. 2). 

Fig. 3 Pseudopapillary pattern, particularly around small vessels. (HE 100x) 

Fig. 4 Tumor's cells within the lining of some glands. (HE 200x) 

Fig. 2 General architecture of the tumor - the solid part. Two mitoses are 

present. (HE 100x) 

Fig. 5 Positive immunostaining for alfa1 antitrypsin. (HE 100x)

123 

Discussion 

Preoperative diagnosis of pancreatic tumor, putting aside uncharacteristic 

clinical symptoms as far as its size, localization, 

detailed anatomy and topography is concerned, is based 

upon ultrasonography, computed tomoghraphy, magnetic resonance 

image [4, 7, 15, 17]. Still diagnostic failures, such as 

pseudocyst or ruptured hepatic mass are described in literature 

[17, 19]. Some authors advocate preoperative fine needle 

biopsy to establish exact histopatology of neoplasm [3, 17]. 

However such a policy is controversial as needle biopsy may 

not allow an undisputed diagnosis and the indications for operation 

exist, this way or another. This was a case in our patient, 

too. Aggressive surgery and complete resection produces 

the specimen to be meticulously examined [1, 5, 13, 15, 16]. 

In 1996 solid pseudopapillary tumor was introduced in 

WHO classification of tumors of the exocrine pancreas [8, 11]. 

It is usually included into cystic group of exocrine neoplasm of 

pancreas that encompasses [8, 20]: 

- serous cystis neoplasms, 

- mucinous cystic neoplasms, 

- solid papillary neoplasms, 

- intraductal pancreatic mucinous neoplasm. 

Prognosis for the patients with these tumors is much 

better than that of patients with adenocarcinoma, although in 

about 5% of patients with SPTP malignant course with metastases 

and peritoneal spread can develop [2, 6]. It is estimated 

that after surgical resection the outcome is excellent 

resulting in long – term survival in about 90% of patients [5, 

9, 15, 16, 18]. Complete radical resection may need extension 

to neighbouring organs e. g. transverse colon or spleen – as 

was a case in our patient. Recurrence has been described in 

approximately 10% of cases [2, 5, 10, 15]. Solid 

pseudopapillary tumor of pancreas should be always taken 

into consideration in the differential diagnosis of pancreatic 

neoplasm. 

Acknowledgements 

Authors greatly appreciate the assistance of Tomasz Krawczyk, 

MD in preparing microscopic pictures. 

References 

1. Akiyama H, Ono K, Takano M, Sumida 

K, Ikuta K, Miyamoto O (2002) Solidpsudopapillary 

tumor of the pancreatic 

head causing marked distal atrophy: a 

tumor originated posterior to the main 

pancreatic duct. Int Gastrointest Cancer 

32(1):29-31 

2. Andronikou S, Moon A, Usser R (2003) 

Peritoneal metastatic disease in a child 

after excision of a solid pseudopapillary 

tumor of the pancreas: a unique case. 

Pediatr Radiol 33(4):269-271 

3. Asthon J, Sutherland F, Nixon J, Nayak 

V (2003) A case of solid-pseudopapillary 

tumor of the pancreas: preoperative cyst 

fluid analysis and treatment by 

enueclation. Hepatogastroenterology 

50(54):2239-2241 

4. Cantisani V, Mortele KJ, Levy A, 

Glickman JN, Ricci P, Passariello R, Ros 

PR, Silverman SG (2003) MR imaging 

features of solid pseudopapillary tumor 

of the pancreas in adult and pediatric 

patients. AJR Am J Roentgenol 

181(2):395-401 

5. Cervantes-Montiel F, Florez-Zorrilla C, 

Alvarez-Martinez I (2002) Solid-cystic 

pseudopapillary tumor of the pancreas: 

acute post-traumatic presentation. Case 

report review of the literature. Rev 

Gastroenterol Mex 67(2):93-96 

6. Illert B, Luhrs H, Timmermann W, 

Muller JG (2003) Solid-pseudopapillary 

tumor of the pancreas as differential diagnosis 

in pancreatic tumors of the 

young. Zentralbl Chir 128(5):438-442 

7. Kato T, Egawa N, Kamisawa T, Tu Y, 

Sanaka M, Sakaki N, Okamoto A, Bando 

N, Funata N, Isoyama T (2002) A case 

of solid psudopapillary neoplasm of the 

pancreas and tumor doubling time. 

Pancreatology 2(5):495-498 

8. Kloppel G, Bogomoletz WV (2003) 

Diagnostic histopathology of tumors. 

Ch. Fletcher (Ed), Churchill, 

Livingstone, p. 470 

9. Mancini GJ, Dudrick PS, Grindstaff 

AD, Bell JL (2004) Solidpseudopapillary 

tumor of the pancreas: 

two cases in male patients. Am Surg 

70(1):29-31 

10. Martin RC, Klimstra DS., Brennan MF, 

Conlon KC (2002) Solid 

pseudopapillary tumor of the pancreas: 

a surgical enigma Ann Surg Oncol 

9(1):35-40 

11. Meyer S, Aliani S, Graf N, Bonkhoff 

H, Schneideer G, Reinhard H (2002) 

Solid-pseudopapillary tumor of the pancreas 

in a 9-year-old girl. Klin Pediatr 

214(5):316-318 

12. Molino D, Perriotti P, Napoli V, 

Antropoli C, Gargano E, Antropoli M, 

D’Antonio D, Vicenzo L, Boscaino A, 

D’Anna L, Guidi G (2003) Solid 

pseudopapillary tumour of the pancreas. 

Report of a case. Minerva Chir 

58(6):815-821 

13. Nachulewicz P, Kalicinski P, Polnik D, 

Chy¿yñska A, Broniszczak, D (2002) 

Solid and papillary epithalial tumor of 

the pancreas. A case report of surgically 

curable tumor. Surg Childh Intern 3:148- 

150 

14. Nadler EP, Novikov A, Landzberg BR, 

Pochapin MB, Centeno B, Fahey TJ, 

Spigland N (2002) The use of endoscopic 

ultrasund in the diagnosis of solid 

pseudopapillary tumors of the pancreas 

children. J Pediatr Surg 37(9):1370-1373 

15. Patel VG, Fortson JK, Weaver WL, 

Hammami A (2002) Solidpseudopapillary 

tumor of the pancreas 

masquerading as a pancreatic 

pseudocyst. Am Surg 68(7):631-632 

16. Pettinato G, Di Vizi D, Manivel JC, 

Pambuccian SE, Somma P, Insabato L 

(2002) Solid-pseudopapillary tumor of 

the pancreas: a neoplasm with distinct 

and highly characteristic cytological 

features. Diagn Cytopathol 27(6):325- 

334 

17. Pezzolla F, Lorusso D, Caruso ML, 

Demma I (2002) Solid pseudopapillary 

tumor of the pancreas. Consideration of 

two cases. Anticancer Res 22(3):1807- 

1812 

18. Pedevin J, Triau S, Mirallie E, Le 

Borgne J (2003) Solid-pseudopapillary 

tumor of the pancreas: a study of five 

cases, and review of the literature. Ann 

Chir 128(8):543-548 

19. Potrc S, Kavalar R, Horvat M, Gadzijev 

EM (2003) Urgent Whipple resection for 

solid pseudopapillary tumor of the pancreas. 

J Hepatobiliary Pancreat Surg 

10(5):386-389 

20. Sheehan MK, Beck K, Pickleman J, 

Aranha GV (2003) Spectrum of cystic 

neoplasms of the pancreas and their surgical 

management. Arch Surg 

138(6):657-660



The invited lecture 

Liver tumors of childhood - pathology. Report of the Kiel Pediatric 

Tumor Registry* 

Dieter Harms 





Department of Pediatric Pathology, 

University of Kiel, 

Kiel, Germany 

Dear Colleagues, 

Over the years I have collected a series of 522 primary liver 

tumors. Hepatoblastomas, accounting for 51.3% of the cases, 

are by far the most frequent liver tumors in the files of the 

Registry, followed by infantile hemangioendothelioma and 

hepatocellular carcinoma (15.5%, each). All other liver tumors, 

including mesenchymal hamartoma, focal nodular hyperplasia, 

embryonal rhabdomyosarcoma and undifferentiated 

sarcoma, are rare. 

Hepatoblastoma has the peak incidence in the first two years 

of life, and up to 90% of the cases occur before the age of five 

years. As it is well known, the wide majority of 

hepatoblastomas present as a single mass involving in decreasing 

frequency the right lobe, both lobes and the left lobe. Approximately 

20% of the cases are multifocal. Microscopically, 

the more frequent epithelial hepatoblastomas (56%) and the 

less frequent (44%) mixed epithelial and mesenchymal 

hepatoblastomas have to be distinguished. With regard to the 

subtypes mixed epithelial/mesenchymal tumors without teratoid 

features are most frequent (34%), followed by pure fetal 

(31%), fetal/embryonal (19%) and mixed teratoid 

hepatoblastomas (10%). The macrotrabecular and the small 

cell/anaplastic hepatoblastoma variants are very rare (3%, 

each). Notably, a pure embryonal variant of epithelial 

hepatoblastoma does virtually not exist. 

Nevertheless, data from the Liver Tumor Study HB 94 

show that the histologic subtype does not influence the prognosis 

significantly. By contrast, significant prognostic factors 

are the tumor growth pattern in the liver, vascular invasion, 

distant metastases, surgical radicality, initial levels of AFP, 

and the response to chemotherapy. The overall survival rate 

of patients with hepatoblastoma in different multicenter studies 

is approximately 75%.The prognostically much more unfavorable 

hepatocellular carcinoma (HCC) is significantly less 

frequent than hepatoblastoma except in countries with high 

hepatitis B virus infection rates. In the files of the Kiel Pediatric 

Tumor Registry 81 cases of HCC including 14 cases of the 

fibrolamellar variant are collected. The proportion of HCC to 

hepatoblastoma is 0,3:1. Most cases of HCC did occur clini- 

cally in the second decade of life, this in contrast to the wide 

majority of hepatoblastomas. Microscopically, pediatric HCC 

are similar to HCC of the adults. Usually the tumor cells are 

larger than the surrounding normal hepatocytes. The tumor 

cells can be arranged in a pseudoglandular, trabecular, or 

pseudoacinar pattern. In most cases the cytoplasm shows an 

intensive eosinophilia. Rarely HCC displays tumor cells with 

clear cytoplasm similar to clear-cell renal carcinoma and to 

some fetal-epithelial hepatoblastomas. Mitotic activity and 

nuclear atypia are variable from case to case and may be variable 

even in the same tumor. In this case at least moderate 

atypia and some mitotic figures can be seen. 

Immunohistochemically, tumor cells express AFP, and vessel 

invasions are very common. Since many HCC are hepatitis B 

virus (HBV)-associated, it is important to look for features of 

preceding HBV infections.The fibrolamellar variant of HCC 

is not associated with virus infections and, moreover, is not 

associated with other conditions like metabolic diseases or 

biliary atresia, which are well-known risk factors for the development 

of classic liver cell carcinoma. In our files 17% of 

the liver cell carcinomas belong to the fibrolamellar variant. 

Fibrolamellar carcinoma occurs predominantly in older children, 

adolescents and young adults and usually presents as a 

solitary, well-circumscribed tumor mass with a predilection 

to the left lobe. Microscopically, cords of relatively large tumor 

cells are separated by paucicellular collagen bands. The 

cytoplasm is intensively eosinophilic, and frequently contains 

lumina. The proliferation activity is comparably low indicating 

slow tumor growth. Late tumor relapses can occur, and in 

so far the final prognosis of fibrolamellar carcinoma is not so 

favorable as was supposed initially. 

The most important mesenchymal tumors of the liver 

are infantile hemangioendothelioma, mesenchymal hamartoma, 

embryonal rhabdomyosarcoma, and undifferentiated 

(embryonal) sarcoma with 15.5%, 4.8%, 5.6%, and 2.9%, respectively, 

of the cases collected in our Registry. With regard 

to the age at clinical manifestation, significant differences 

between these tumors become overt. Thus, most cases of infantile 

hemangioendothelioma are diagnosed in very young 

children. Referring to our publication on tumors of the newborn 

and the very young infant [3], 86% of the 

* The lecture was presented during the Meeting of the Polish Pediatric Group for the Solid Tumors’ Treatment (Liver Tumors), 

22-24.04.2004, Gdansk, Poland

126 

hemangioendotheliomas occured in children below the age of 

6 months. For comparison, 9 hepatoblastomas have manifested 

in this age class, and the proportion of hemangioendothelioma 

to hepatoblastoma is 2.1 to 1 in early infancy. Mesenchymal 

hamartoma can be present at birth, like hemangioendothelioma. 

The wide majority of mesenchymal hamartomas occurs 

in patients under the age of 2 years. Embryonal rhabdomyosarcoma 

of the liver has a broad age distribution, but 

the majority of rhabdomyosarcomas develop within the first 

5 years of life, this by contrast to the probably rhabdomyosarcoma-related 

undifferentiated (embryonal) sarcoma (malignant 

mesenchymoma of the liver in the old nomenclature), 

for which the peak incidence is between five and ten years of 

age. Infantile hemangioendothelioma occurs as a solitary mass 

or consists of multiple tumor nodes involving large parts of 

the liver. Under the microscope infantile hemangioendotheliomas 

are composed of multiple, sometimes dilated vascular 

channels lined by plump endothelial cells. Large or multicentric 

lesions can be complicated clinically by consumption 

coagulopathy or disseminated intravascular coagulation 

(Kasabach-Merritt-syndrome) or, due to intralesional vascular 

anastomoses, by severe congestive heart failure. Most infantile 

hemangioendotheliomas have to be classified according 

to DEHNER’s proposal [2] as type I-hemangioendotheliomas.The 

rare type II-hemangioendothelioma shows a more 

distinct papillary pattern and endothelial cells with more atypical 

nuclei than in type I -tumors. Type II-infantile 

hemangioendotheliomas are now considered to be pediatric 

angiosarcomas, but malignant behavior of infantile 

hemangioendotheliomas is generally rare. The majority of the 

tumors are principally benign lesions. The prognosis depends 

on size and distribution of the tumor throughout the liver, on 

the severity of complications, and the resectability, which may 

be impossible in multifocal disease. Mesenchymal hamartomas 

are generally solitary, no encapsulated and can be very 

large. Up to 30% of the tumors are pedunculated.Typically, 

mesenchymal hamartomas show cysts containing a clear fluid. 

Microscopically, in typical cases a so-called ductal-plate pattern 

with tortous bile ducts, surrounded by loose connective 

tissue, and at the periphery by a more dense fibrous tissue, 

can be seen, so that ductal-plates are reminiscent on fibrocystic 

disease of the female breast. Some bile ducts may show slight 

nuclear atypia. Fluid accumulations in the tumor tissue my 

cause lymphangioma-like spaces and cysts, which are lined 

by flattened non-endothelial, fibroblastic cells. Taking together 

these macroscopic and microscopic findings and with regard 

to the very young age of the patients the diagnosis of mesenchymal 

hamartoma is easy, except in small biopsies, in which 

it may be impossible to exclude undifferentiated sarcoma of 

the liver. The prognosis of mesenchymal hamartomas is very 

favorable. Tumor relapses are rare. 

Embryonal rhabdomyosarcoma of the liver developes 

either in the extrahepatic bile ducts or intrahepatic. Occlusion 

of larger bile ducts induces obstruction jaundice. Microscopically, 

rhabdomyosarcomas of the liver are in most cases embryonal 

rhabdomyosarcomas. Particularly, they present as the 

botryoid variant of embryonal RMS exhibiting the characteristic 

subepithelial cambium layer. This layer consists of parallel 

to the surface arranged, at least moderately differentiated, 

intensivly desmin-positive rhabdomyoblasts. This diagnostic 

cambium layer the botryoid RMS usually shows small 

stellate-like cells arranged in a loose pattern. By contrast to 

botryoid rhabdomyosarcomas of other locations, the surface 

of the tumor proliferations can be covered by cylindrical or 

cuboidal epithelial cells and not by metaplastic squamous 

epithelial cells. Treatment should be done according to the 

well-established soft-tissue sarcoma therapy protocols. The 

same is true for undifferentiated sarcoma of the liver. 

Undifferentiated (embryonal) sarcoma of the liver usually 

develops in children between five to ten years of life. The 

median age of 17 patients with undifferentiated sarcoma treated 

in a cooperative Italian/German study [1] was seven years. 

Macroscopically, this undifferentiated sarcoma (this specimen 

stems exceptionally from an adult patient) shows a grayishwhite 

to red and brown cut surface with intensive necrosis. 

Under the microscope the tumor consists of spindle and seellate 

cells arranged within a myxoid stroma. In addition to these 

comparably small cells some larger and pleomorphic tumor 

cells can be seen. Typically, the periphery of the tumor shows 

entrapped bile ducts with slight nuclear atypia. These bile ducts 

can appear with moderately dilated lumina. 

Immunohistochemically, the tumor cells express vimentin intermediate 

filaments. In addition, some tumor cells are positive 

for desmin too, but true rhabdomyoblasts, particularly 

more differentiated rhabdomyoblasts with cross striations do 

not occur, this in contrast to embryonal rhabdomyosarcoma. 

Nevertheless, the histologic features of undifferentiated sarcoma 

and embryonal rhabdomyosarcoma may overlap. However, 

basically, the undifferentiated sarcoma is a more primitive 

sarcoma than embryonal rhabdomyosarcoma. The prognosis 

of undifferentiated sarcoma of the liver was very poor in 

the past (mortality about 80 % - STOCKER and ISHAK 1978). 

Introduction of multimodal soft-tissue sarcoma therapy strategies 

has improved the prognosis dramatically. According to 

data of the forementioned Italin/German study 12 from 17 

patients (approximately 70 % of the cases) have survived with 

follow-up ranging from 2,4 to 20 years [1]. 

Liver tumor pathology covers a wide variety of malignant 

and benign, epithelial and mesenchymal tumor entities. 

The wide majority of principally epithelial tumors 

(hepatoblastomas and liver cell carcinomas) are malignant, 

whereas the majority of mesenchymal tumors (except undifferentiated 

sarcoma and embryonal rhabdomyosarcoma) are 

morphologically benign. The prognosis of malignant liver tumors 

has improved dramatically during the recent few decades. 

Improvement of prognosis is due to interdisciplinary cooperation 

in multimodal therapy studies. Another milestone was 

the development of new pediatric surgery strategies and techniques, 

which have contributed significantly to the prognostical 

success in malignant and in surgically problematic benign tumors 

of the liver too.

1. Bisogno G, Pilz T, Perilongo G (2002) Undifferentiated 

sarcoma of the liver in childhood: a curable disease. Cancer 

94(1):252-257 

2. Dehner LP (1978) Hepatic tumors in the pediatric age group: 

a distinctive clinicopathologic spectrum. Perspect Pediatr 

Pathol 4:217-268 

3. Harms D, Schmidt D, Leuschner I (1989) Abdominal, retroperitoneal 

and sacrococcygeal tumours of the newborn and 

the very young infant. Report from the Kiel Paediatric Tumour 

Registry. Eur J Pediatr 148(8):720-728 

127

Annals of Diagnostic Paediatric Pathology

Create successful ePaper yourself

Delete template?

Save as template?