Change of paradigms in nephrology--a view back and a look forward.

Nephrol Dial Transplant (1998) 13: 556–563
Personal Opinion
Nephrology
Dialysis
Transplantation
Change of paradigms in nephrology—a view back and a look forward
Adalbert Bohle
Department of Pathology, Tübingen, Germany
‘Verlernen ist oft schwerer als lernen.’ (Arthur normal science is based upon a set of concepts or
Koestler)
paradigms. Normal science has cumulative character,
(To forget is often more difficult than to learn.) i.e. involves collecting facts and unravelling problems
which confirm the consistency and correctness of
Scientific revolutions by change of paradigms
existing paradigms. By necessity, groundbreaking
innovations will shake the foundations of such tenets
and are usually suppressed by normal science.
In his book The structure of scientific revolutions, T.S. Nevertheless, according to Kuhn, on certain occasions
Kuhn [1] distinguished between great and small revolualous
normal science yields results which appear to be anomtions
in science and subjected such revolutions to more
and cannot be reconciled with the predictions
detailed analysis. He came to the conclusion that based on existing paradigms. Such anomalies can then
Fig. 1 a–d. Glomerular changes in decompensated benign nephrosclerosis. (a) Segmental sclerosis and hyalinosis; (b) advanced sclerosis
and pseudotubules; (c) almost complete obliteration of all the glomerular capillaries, pseudotubules, and fat droplets in the hyalinized
capillaries and endothelial cells; (d) complete destruction of the glomerulus, fat droplets still being identifiable in the areas of subendothelial
hyalinosis. Periodic acid–Schiff reaction.
Correspondence and offprint requests to: Professor Adalbert Bohle,
Professor Emeritus Pathology, Im Kleeacker 10, D-72072 Tübingen,
Germany.
© 1998 European Renal Association–European Dialysis and Transplant Association

Change of paradigms in nephrology—a view back and a look forward 557
Fig. 3. Interlobular artery from the case described in Table 1. There
is stenosing intimal oedema and early subendothelial accumulation
of smooth-muscle cells and fibrocytes. Goldner’s trichrome stain.
competing paradigms. Thus, an anomaly in the results
does not lead to a crisis, but to the birth of a new
paradigm.
My scientific career was dedicated to renal pathomorphology.
More than 45 000 renal biopsies were
interpreted in the context of clinical data. Based on
this experience I wish to propose some thoughts that
illustrate how in-depth clinicomorphological analysis
forced changes in some paradigms concerning the
pathogenesis of renoparenchymal diseases.
Fig. 2. Decompensated benign nephrosclerosis, uraemic stage (24-
year-old male, blood pressure 195/120 mmHg). There is destruction
of juxtamedullary glomeruli and marked interstitial fibrosis and
tubular atrophy in the middle layers of the cortex.
no longer be disregarded and suppressed. According
to Kuhn, this is the point of departure of extraordinary
science, which puts science on a new basis. Such
episodes which involve a change of concepts he called
‘scientific revolutions’. They destroy traditions, in contrast
to normal science which is based on maintenance
of traditions. Because of such scientific revolutions
scientists are forced to refute time-honoured scientific
theories in favour of new paradigms which apparently
are more successful in answering problems than the
The change in concepts concerning vascular lesions
of the kidney
Franz Volhard [2] and Theodor Fahr [3] analysed
vascular diseases of the kidney and distinguished
between benign nephrosclerosis and malignant
nephrosclerosis. Later, Theodor Fahr distinguished
between a compensated and a decompensated form of
benign nephrosclerosis. By decompensation he implied
that renal function was altered [4]. As a case in point,
he described a 55-year-old hypertensive patient who
presented with proteinuria and erythrocyturia and died
in renal failure. Apart from benign nephrosclerosis
Table 1. Clinical data of a case of primary malignant nephrosclerosis. Patient, a 39-year-old male, suffered 10 days of cough, nosebleeds,
and diarrhoea
Dialysis Histology I Histology II
Haemoglobin (g %) 4.5
Reticulocytes (%) 35
LDH 1380
Platelets 108 000
Bilirubin (mg %) 1 3
Creatinine (mg %) 28
BUN (mg %) 187
Blood pressure 140/80 140/80 165/85 170/120 230/140
Urine output (ml ) 100 Anuria Onset of 1710
polyuria
Follow-up (days) 1 28 40 56 61 +

558
A. Bohle
(a)
(b)
(c)
(d)
Fig. 4 a–d. Interlobular arteries in secondary malignant nephrosclerosis. (a) Early changes, stenosing intimal oedema; (b) stenosing intimal
oedema and subendothelial necrosis; (c) late changes, intimal fibrosis and fresh haemorrhages; (d) Late changes, obliterating endarteritis.
All the blood vessels are from the same kidney. Goldner’s trichrome stain.
(a)
(b)
(c)
(e)
(d)
(f)
Fig. 5 a–f. Examples of severe isolated glomerulopathy with
normal serum creatinine concentration. (a) Control kidney; serum
creatinine concentration 1.2 mg%, periodic acid–Schiff reaction.
(b) Acute endocapillary glomerulonephritis with mild acute renal
failure; serum creatinine concentration 1.29 mg%, urinary output
2450 ml/24 h. (c) Membranoproliferative glomerulonephritis
type I; serum creatinine concentration 1.2 mg%, creatinine
clearance 90 ml/min, periodic acid–Schiff reaction. (d)
Membranoproliferative glomerulonephritis type I plus chronic
membranous glomerulonephritis; serum creatinine concentration
1.06 mg%, creatinine clearance 172 ml/min, periodic acid–Schiff
reaction. (e) Severe nodular diabetic glomerulosclerosis; serum
creatinine concentration 1.2 mg%, periodic acid–Schiff reaction.
(f ) Perireticular renal amyloidosis; serum creatinine concentration
0.8 mg%, periodic acid–Schiff reaction.

Change of paradigms in nephrology—a view back and a look forward 559
attention to this entity [5], which is not infrequent.
We have observed this constellation in more than 250
cases [6]. Figures 1 and 2 show the characteristic
lesions. The glomerular changes (Figure 1) first appear
in the juxtamedullary glomeruli, which are not subject
to autoregulation (Figure 2) and spread out into the
cortex. Interstitital fibrosis and tubular atrophy are
also seen. With the aging of the population and longer
survival of patients with hypertension, this lesion will
be recognized more frequently, particularly since end-
Fahr found glomerular lesions with focal thickening
of capillaries without cellular proliferation and circumscribed
adhesions between glomerular capillaries and
Bowman’s capsule. Glomerular obsolescence was strikingly
frequent and the renal interstitium was fibrosed,
particularly in the deep cortex.
Despite such precise description by this pioneer of
renal pathology, decompensated benign nephrosclerosis
had been completely forgotten as a cause of
renal dysfunction. We had great difficulty in drawing
Fig. 6. Diabetic glomerulosclerosis. Correlations between relative area of postglomerular capillaries and serum creatinine concentration at
time of biopsy.
Fig. 7. Renal amyloidosis. Correlations between relative area of postglomerular capillaries and serum creatinine concentration at time
of biopsy.

560
A. Bohle
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
Fig. 8. (a) Control kidney stained with anti-pancytokeratin. The
parietal epithelial cells are non-reactive. (b) Immunoreactivity of the
podocytes with an antibody against vimentin is demonstrated.
(c) Extracapillary glomerulonephritis. The crescents are reactive
with anti-vimentin. (d) Extracapillary glomerulonephritis stained
with anti-pancytokeratin. The crescents are non-reactive. (e)
Extracapillary glomerulonephritis. The cells of the crescents are
reactive with anti-pancytokeratin. (f ) Extracapillary glomerulonephritis.
The basement membrane of the glomerular capillaries is
reactive for laminin. (g) Extracapillary glomerulonephritis with
crescent-shaped collections of monocytes/macrophages that stain
with the monocyte/macrophage marker PGM 1. (h) Extracapillary
glomerulonephritis. The crescents are non-reactive with the proliferation-associated
antibody MiB1 (Ki67).
had the characteristic histological lesions of malignant
nephrosclerosis (Table 1 and Figure 3). Currently we
have collected more than 200 cases of malignant
nephrosclerosis who had documented normotension at
the start of their disease. We designated this entity
‘primary malignant nephrosclerosis’ to distinguish it
from ‘secondary malignant nephrosclerosis’ as a con-
sequence of malignant hypertension [9–11]. We feel
that primary and secondary malignant nephrosclerosis
can be distinguished in that in primary malignant
nephrosclerosis all preglomerular vascular lesions are
of similar age, whereas in secondary malignant
nephrosclerosis the lesions are of different ages (Figure
4). It has meanwhile become apparent that primary
malignant nephrosclerosis is the morphological counterpart
of the haemolytic–uraemic syndrome. It is more
frequent in females (particularly after hormonal con-
traception) in contrast to secondary malignant hyper-
tension, which is more frequent in males [12].
stage renal failure is by no means infrequent in the
elderly hypertensive patient [7].
Another example for revision of past pathogenetic
concepts is malignant nephrosclerosis. F. Volhard had
assumed that the characteristic obliterative preglomer-
ular vascular lesions are exclusively the result of malig-
nant hypertension. Based on one single case, i.e. a
young normotensive patient with normal cardiac size
who exhibited malignant nephrosclerosis at autopsy,
Fahr [8] proposed that the vascular lesions were
primary and that hypertension was a secondary
consequence. Volhard’s opinion was based on the
observation of numerous patients with malignant
hypertension and appeared perfectly convincing. We
were satisfied with Volhard’s concept until we examined
the renal biopsy of a 39-year-old patient who had been
normotensive as documented by daily blood-pressure
measurements over a 4-week period. Nevertheless he
Fig. 9. Corrosion specimen of a human kidney after methacrylate
injection under normal pressure. The glomerular capillaries and the
dense network of postglomerular capillaries are filled with methacrylate.
Scanning electron-microscopy.
What have we learned from the above?
Clinicopathological analysis of the vascular lesions has
forced us to change the paradigms to explain the

Change of paradigms in nephrology—a view back and a look forward 561
aetiology and the pathogenesis of benign nephrosclerosis.
We have also come to learn that in malignant
nephrosclerosis the same morphological lesion may be
caused by two entirely different aetiologies.
The role of glomerular lesions in progression
Fig. 10. Decompensated benign nephrosclerosis with fatty degeneration
of endothelial cells. Between the podocyte and the basement
membrane from which it has become detached there are basement
membrane lamellae (arrowhead) that are not in contact with the
blood-stream but are in contact with Bowman’s space. ×5000.
Fig. 11. Proximal tubule with intraepithelial lymphocytes, mainly
T-helper cells (arrowhead).
Classical opinion held that in glomerular diseases renal
failure develops when the primary glomerular lesions
have substantially reduced the filtration surface and
thus led to a reduction of glomerular filtration rate.
As a result, histological studies of the diseased kidney
focused almost exclusively on the alterations of the
glomerulus [13–18]. Morphometric analysis of the
kidneys in diabetic nephropathy by Mauer et al.[19]
and Osterby et al. [20,21] reported a tight correlation
between the filtering surface on the one hand and GFR
on the other.
Very early on [22–24] we were struck by the paradoxical
observation that even severe glomerular lesions in
glomerular diseases such as endocapillary glomerulonephritis,
membranoproliferative glomerulonephritis,
diabetic glomerulosclerosis, and perireticular renal
amyloidosis, were not consistently accompanied by an
elevation of serum creatinine if the postglomerular
capillary bed was not compromised [23] (Figure 5).
Such glomerulopathies were associated with progressive
renal failure only when the postglomerular capil-
laries were obliterated by interstitial fibrosis, thus
presumably interfering with postglomerular blood flow
[23,25,26]. A highly significant inverse correlation was
noted between the relative capillary surface of the renal
cortex on the one hand and serum creatinine concentration
on the other. We wish to illustrate this diabetic
glomerulosclerosis (Figure 6) and perireticular renal
amyloidosis (Figure 7).
Do these observations of ours [22,23,25,26] and of
others authors [27,28] force us to change past paradigms
to explain chronic progressive loss of renal
function in glomerular disease? I think yes. In contrast
to many statements in literature I do not feel that
glomerular lesions by themselves are sufficient to
explain impaired renal function in the majority of
cases. To take an extreme case: rapidly progressive
extracapillary glomerulonephritis with renal failure. I
do not feel that renal failure is explained by obliteration
of Bowman’s space by crescents [22]. Crescents consist
not of visceral epithelial cells [29,30] (which are postmitotic
and cannot proliferate) [31–33], but of pluripotent
blood-derived mononuclear cells (Figure 8) [34]
which are trapped and accumulate in Bowman’s space.
Even in its initial stage extracapillary glomerulonephritis
may be accompanied by acute renal failure
[22]. Our belief is strengthened by the observation that
glomerular reserve capacity is so enormous that even
advanced reduction of the filtration surface, as documented
by light and electron microscopy, does not
necessarily lead to an increase in serum creatinine [24].
The Achilles’ heel for renal function is not the glomerulus,
but the postglomerular capillary. If this tenet is

562
A. Bohle
correct, it is the lesion of the postglomerular vessel Anatomie und Histologie, Vol. 6, 2, 1934, Springer, Berlin;
that leads to renal failure (Figure 9).
900–934
5. Bohle A, Ratschek M. The compensated and decompensated
This concept raises the issue of which process dam- form of benign nephrosclerosis. Pathol Res Pract 1982; 174,
ages the postglomerular capillaries. Obviously an inter- 357–367
play between altered tubular epithelial cells and 6. Bohle A, Christensen J, Grossmann T, Inniger R, Cavalcanti de
interstitial cells leads to interstitial fibrosis [35,36 ]. Oliveira V, Wehrmann M. Über benigne und maligne
Several hypotheses on the genesis of interstitial fibrosis Nephrosclerosen. Pathologe 1988; 10: 263–271
7. Klag MJ, Whelton PK, Randall BL et al. Blood pressure and
have been proposed, e.g. exposure of the luminal end-stage renal disease in men. N Engl J Med 1996; 334: 13–18
surface of tubular epithelial cells to serum proteins as 8. Fahr Th. Über atypische Befunde aus dem Kapitel des Morbus
a result of glomerular leakage [37,38]. Several factors Brightii nebst anhangsweisen Bemerkungen zur Hypertonie-
may be involved and we proposed one further hypo- frage. Virchows Arch Path Anat u Physiol 1924; 248: 101–102
thesis. According to Hawkins and Cochran [39] and 9. Bohle A, Helmchen U, Meyer D et al. Über die primäre und
sekundäre maligne Nephrosklerose. Klin Wochenschr 1973; 51:
to McCluskey [40] lesioned glomeruli release an 841–857
autoantigen into Bowman’s space. We presume that 10. Bohle A, Grund KE, Helmchen U, Meyer D. Primary malignant
the site of origin of this autoantigen is the glomerular nephrosclerosis. Clin Sci 1976; 51: 23–25
epithelial cell [41], which desquamates and releases 11. Bohle A, Helmchen U, Grund KE et al. Malignant nephroscl-
GBM material from the basal membrane (Figure 10). erosis in patients with hemolytic uremic syndrome (Primary
malignant nephrosclerosis). Curr Top Pathol 1977; 65: 81–113
Such material, which is normally sequestered and 12. Bohle A. Paradigmawechsel in der Nephrologie. Nieren
cannot be recognized by the immune system, may Hochdruckkrankh 1994; 23: 45–57
become accessible to the immune system after pro- 13. Kincaid-Smith P, Dowling JP, Mathews DC. Atlas of Glomerular
cessing by tubular epithelial cells and by lymphocytes Disease. Morphological and Clinical Correlations. Aids Health
in intertubular location (Figure 11). This hypothesis Science Press, Sydney 1985
14. Meadows K. Renal Histopathology. A Light Microscopic Study
requires confirmation; if it is correct, it would imply of Renal Disease. Oxford University Press, New York, 1978
that interstitial fibrosis is the result of an autoaggres- 15. Spargo BH, Seymour E, Ordonez NE. Renal Biopsy Pathology
sion process. with Diagnostic and Therapeutic Implications. Wiley, New
York, 1980
16. Rotter,W. Farbatlas der Nierenbiopsie. Pathologie glomerulärer
Conclusions
Erkrankungen. Schattauer, Stuttgart, 1983
17. Lawler,W. Glomerular Pathology. Churchill Livingstone,
What have we learned from a life-time of work on
London 1991
18. Steffes MW, Osterby R, Shavers B, Mauer SM. Mesangial
glomerular pathomorphology?
expansion as a central mechanism for loss of kidney function in
It is obvious that intellectual economy requires diabetic patients. Diabetes 1989; 38: 1077–1081
19. Mauer SM, Steffes MW, Ellis EN, Sutherland ER, Brown GM,
working hypotheses to handle the data provided by
Goetz FC. Structural–functional relationship in diabetic nephroobservational
or interventional studies. The above pathy. J Clin Invest 1984; 74: 1149–1185
examples illustrate, however, that we have to keep an 20. Osterby R, Gunderson HJG. Glomerula size and structure in
open mind and have to be prepared for the possibility diabetes mellitus: early abnormalities. Diabetes 1975; 25: 872–875
that our concepts and paradigms may be wrong and 21. Hrose K, Tsudica H, Osterby,R, Gunderson HJG. A strong
correlation between glomerular filtration rate and filtration
may have to be modified in the light of new data (or
surface in diabetic kidney hyperfunction. Lab Invest 1980; 43:
new interpretations). I feel that the example of 434–437
nephrosclerosis and primary glomerulopathies is illus- 22. Bohle A, Gärtner HV, Laberke G, Krück, F. Die Niere. Struktur
trative and serves to make investigators humble. The und Funktion. Schattauer Verlag, Stuttgart, 1984; 11
change of paradigms, according to Kuhn, is a painful 23. Bohle A, Wehrmann M, Mackensen-Haen S. Pathogenesis of
chronic renal failure in primary glomerulopathies. Nephrol Dial
process and may explain why such new concepts often
1994; 3 [Suppl ]: 4–12
cause controversy and are met by fierce opposition. 24. Bohle A, Aeikens B, Eenboom A et al. The structure of the
human renal glomerulus with particular reference to the length
Acknowledgements. This article is dedicated to Professor Ernst and surface area of the glomerular capillaries and the filtration
Jünger, a great poet, philosopher, and morphologist on the occasion area under physiological conditions and severe isolated glomerulof
his 103rd birthday. The text is the written version of the Farewell opathies. Am J Nephrol (in press)
Lecture at the Eberhard Karl University, Tübingen on 16 February 25. Bohle A, Gise H v. Mackensen-Haen S, Stark-Jakob B. The
1990. obliteration of the postglomerular capillaries and its influence
upon the function of both glomeruli and tubuli. Functional
interpretations of morphologic findings. Klin Wochenschr 1981;
References
59: 1043–1051
26. Bohle A, Mackensen-Haen S, Wehrmann M. Significance of
1. Kuhn TS. Die Struktur wissenschaftlicher Revolutionen. postglomerular capillaries in the pathogenesis of chronic renal
Suhrkamp Tagebuch Wissenschaft, 1969
failure. Kidney Blood Press Res 1996; 19: 191–195
2. Volhard F. Die doppelseitigen hämatogenen Nierenkrankheiten. 27. Risdon FC, Sloper SA, de Wardener HE. Relationship between
Springer, Berlin, 1918
renal function and histological changes found in renal biopsy
3. Fahr Th. Pathologische Anatomie des Morbus Brightii. In: specimens from patients with persistent glomerular nephritis.
Henke F, Lubarsch O. Handbuch der speziellen pathologischen Lancet 1968; 2: 363–366
Anatomie und Histologie, Vol. 6, 1, 1925, Springer, Berlin; 28. Schainuck LI, Striker GE, Luther RE, Benditt EP.
177–178
Structural–functional correlations in renal diseases. The correla-
4. Fahr Th. Pathologische Anatomie des Morbus Brightii. In: tions. Hum Pathol 1970; 1: 631–641
Henke F, Lubarsch O. Handbuch der speziellen pathologischen 29. Gabbert H, Thoenes,W. Formation of basement membrane in

Change of paradigms in nephrology—a view back and a look forward 563
extracapillary proliferates in rapidly progressive glomerulo- GA. Expression of HLA-DQ-DR, and -DP antigens in normal
nephritis. Virchchows Arch [B] 1977; 25: 265–269
kidney and glomerulonephritis. Kidney Int 1989; 35: 116–124
30. Adler S. Glomerular epithelial cells. In: Neilson EG, Couser 36. Müller GA. Mechanismen der chronischen Niereninsuffizienz
WG, ed. Immunologic Renal Diseases. Lippincott-Raven, und Möglichkeiten zu deren Prävention. Nieren- und
Philadelphia, 1997; 635
Hochdruckkrankh 1977; 26 (6) S231–S239
31. Nagata M, Schärer, K, Kriz W. Glomerular damage after 37. Müller GA, Markovic–Lipkovski J. Muller CA. Intercellular
uninephrectomy in young rats. I. Hypertrophy and distortion of adhesion molecule-l expression in human kidneys with glom-
capillary architecture. Kidney Int 1992; 42: 130–147
erulonephritis. Clin Nephrol 1991; 36: 203–208
38. Remuzzi G, Ruggenenti P, Benigni A. Understanding the nature
32. Kriz W, Hähnel B, Rösener S. Elger M. Long term treatment
of renal disease progression. Kidney Int 1997; 51: 2–15
of rats with FGF-2 results in focal segmental glomerulosclerosis.
39. Hawkins D, Cochran CG. Glomerular basement membrane
Kidney Int 1995; 48: 1435–1440
damage in immunological glomerulonephritis. Immunology 1968;
33. Kriz W. et al. Frequent pathway to glomerulosclerosis. 14: 665–681
Determination of tuft architecture. Podocyte damage and seg- 40. McCluskey RT. In: Heptinstall RH, ed. Pathology of the Kidney.
mental sclerosis. Kidney Blood Press Res 1996; 19: 245–253 Little Brown, Boston, 1992; 169–260
34. Heptinstall RH, ed. Pathology of the Kidney. 4th edn.Vol. 1. 41. Bohle A, Haussmann P, Vogt W. Über Beziehungen zwischen
Little Brown, Boston, 1992; 647
Lymphocyten und Harnkanälchenepithelien in der Niere des
35. Müller CA, Markovic-Lipkovski J, Risler T, Bohle A, Müller Menschen. Klin Wochenschr 1970; 48: 1323–1326