Morphological and Cytogenetic Aspects of Renal-Cell Carcinoma1
H.R. Burger Institute of Pathology, University of Zurich, Switzerland
Key Words. Renal tumor . Renal-cell carcinoma . Hyponephroma . Cytogenetics of renal tumors
Abstract. The medical history of ‘hypernephroma’ is characterized by errors which arose from the notion of ‘disseminated adrenocortical foci’. Its origin from tubule epithelia of the kidney is on the other hand to be substan- tiated both in morphological and in immunological terms. Many attempts to arrive at a clinically serviceable estab- lishment of the prognosis on the basis of the morphology have proved to be inadequate for the large group of the classical forms. Only the sarcomatous tumors and the oncocytes could be classified prognostically with sufficient certainty. An attempt is made in the present study to propagate the idea that besides the histogenetic, morphologi- cally more or less tangible characteristics, molecular biological features are also important for the biological behavior of a tumor and should be used to appraise the prognosis.
Introduction
The medical history of ‘hypernephroma’ is character- ized by a series of errors in the exploration of this diversely manifested neoplasia of the kidney tissue. When in 1883 Grawitz [1] described the tumors, which had been regarded as lipomas up to that time, he substi- tuted the misunderstanding entailed in the lipoma the- ory with the misunderstanding of the ‘disseminated adrenocortical remnants’ with consideration of Virchow ideas. The name ‘hypernephroma’ was derived from this interpretation. It was subsequently used as a synomyn for the designation ‘Grawitz tumor’. Endocrine adreno- cortical symptoms are associated with a ‘hypernephro- ma’, at least in a few case descriptions [2, 3]. According- ly, it cannot be ruled out that cells with the endocrine
properties of the adrenal cortex can indeed give rise to a tumor within the kidneys in rare cases. The origin of renal-cell carcinomas was soon looked for in the tubule epithelia, and the hypothesis of the `disseminated adre- nocortical foci’ was abandoned [4]. The term was ex- tended and reference was made to ‘hypernephroid renal cell carcinoma’ [5]. In the classical theory of germinal layers [6], the term ‘carcinoma’ is applied to the epithe- lia, i.e. the ectoderm and the entoderm. However, the kidney tissue is at least partly of mesodermal origin, and tumors deriving from this part of the kidney tissue would have to be classified as mesodermal neoplasia. The spe- cial position of the adrenocortical tumors can be ex- plained exclusively on this basis. They indeed may share characteristics (e.g. the biphasic structure) with typical mesodermal tumors (c.g. mesothelioma). However, this interpretation of the ontogenesis has been rendered rela- tive by molecular biological discoveries of recent years, and terms such as ‘histogenesis’ have declined somewhat in importance.
Paper held on the occasion of the Annual Meeting of the Swiss Urological Society in 1989.
In view of this development, there has been no dearth of attempts to establish a classification of kidney tumors [7-16]. Such a classification should not only take into consideration the biology of the tumors, but above all also the practical requirements, in particular establish- ment of the prognosis, so that it provides a guide to ther- apeutic action. However, any classification of biological phenomena is a simplification, and does not do justice to the nature of such biological phenomena. Despite this restriction, new knowledge in molecular biology pro- vides insights into tumorigenesis permitting a new bio- logical interpretation and a classification in terms of new criteria. Before examining these criteria, we shall recall the morphological characteristics of classical kidney-cell cancer and collate these with biological behavior.
Morphology and Attempts at Morphological Classification
In the classical case, the tumors under discussion have a nodular macroscopic structure with a whitish to yellowish cut surface and more or less sharply delineated boundaries. Hemorrhages frequently occur which give the tumor a variegated external appearance on the cut surface, together with the often myxoid organization of the necrotic areas.
Histologically, the tumor cell clusters are often very near to vessels. They are often only separated from the bloodstream by a thin basement membrane and the endothelium of the capillary wall, so that they retain the microanatomical constellation of the renal tubule. This closeness to the vessels explains the pronounced ten- dency to bleeding and the tendency to hemorrhagic necroses.
On the one hand, the relatedness to the tubule epithe- lial cells and on the other hand the mesenchymal charac- ter of the tissue are also manifested in the histological growth form: besides organoid growth forms, a pro- nounced sarcomatous pattern [18] in which the classifi- cation of the tumor as kidney-cell cancer can only be inferred on the basis of transitions to typical structures is to be found in about 2-15% of cases [16, 17]. This posi- tion between the epithelium and mesenchyma is also shown in the expression of cytoskeletal markers. It could be shown that both normal renal tissue and renal-cell cancers can express both epithelial and mesenchymal markers [19-21].
The tumor cells of classical hypernephroma also show their relatedness to the tubule epithelia of the kidneys
both biochemically with the storage of fats and glycogen and ultrastructurally with the formation of microvilli and often abundant mitochondria [22, 23]. They may also express renal tubular antigens [24].
In accordance with the normal renal tissue, tumors may also have endocrine effects. Erythropoietin-produc- ing tumors [25-29], tumors with a hormone production resembling that of the parathyroids [30, 31] and with further hormone-producing properties [32-34] have been described.
Besides many other researchers, Thoenes et al. [16] have recently suggested a classification based on cytolog- ical, ultrastructural and histological findings with refer- ence to more than 600 kidney tumors. Their classifica- tion goes beyond the rather perfunctory, relatively undif- ferentiated classification of the WHO [15]. They distin- guish various cytological and histological characteristics which can be combined with each other as well as with cytological criteria of malignancy. Cytologically, clear- cell, chromophobe, chromophilic, fusiform and onco- cytic forms are differentiated. Histologically, renal-cell carcinomas are subdivided in three groups of growth forms, namely compact (solid, acinar, trabecular), tubu- lopapillary and cystic. Three percent of the cases in this study show a mixed cell type with different growth forms and cannot be unequivocally classified. A further 4% are not classified precisely, being accommodated in the cat- egory ‘other forms’.
Reis and Faria [17] contributed an additional inter- esting idea to the discussion. They introduced a category
‘stratified pattern’ with a squamous epithelial pattern in his material. This category shows a similar poor progno- sis to that of the sarcomatous forms. Here, the occasion- ally difficult differentiation of kidney carcinomas from transitional epithelial carcinomas plays a role: various authors regard the latter as duct of Bellini carcinomas or as carcinomas of the collecting tubules [35-37].
If the attempts to draw prognostic conclusions from the morphology of kidney-cell cancers from this and other classifications [17, 21, 38-41] are compared, it can be observed that only limited prognoses are possible. A reliable prognosis cannot be made especially for the main group of tumors, the classical ‘hypernephromas’, which account for a total of about 40-50%. In addition, many contradictions are to be found. For example, chro- mophobic carcinoma evidently has a better prognosis [21] compared to the clear-cell cancer, but this does not apply to the pure, well-differentiated forms of the solid clear-cell type, which may also have a good prognosis [17]. There is agreement with regard to the fusiform, sar- comatous forms which have an unfavorable course [18].
The oncocytomas constitute a homogenous group. These are tumors which exclusively comprise character- istic oncocytes. They have a more favorable prognosis [14. 42-44]. They are neoplasias, the uniform cell ele- ments of which are congested with substantially enlarged mitochondria [16]. Although metastatic oncocytomas have also been reported [45-49], these are frequently designated as ‘benign’.
Besides the histological characteristics, the size of the tumor is also of prognostic significance [50]. In particu- lar, statements on metastatic spreading are noteworthy. Even neoplasms of more than 10 cm diameter are to be found without metastases in more than 20% [51]. On the other hand, metastatic spreading may occur even in small tumors of less than 2 cm diameter. This size- dependence is thus a statistical and not an absolute cor- relation. The probability of metastatic spreading in- creased with the size of the primary tumor, but is never 100% nor on the other hand quite 0%. For reasons of practicability, it has therefore been established that tu- mors in excess of a certain size (more than 2 cm in our hospital) are to be designated as carcinomas, whereas tumors with a diameter of less than 2 cm are to be regarded as adenomas. The biology of the kidney-cell tumor does not accord with this ruling. On the contrary, the special mesenchymal character of the tumor is once more reflected in these statistical interrelationships. In contrast to kidney-cell cancer, a classical carcinoma is
defined by its tendency to infiltration and its cytological malignancy, irrespective of whether a microscopic tumor is involved or not. Both criteria are very much less defin- itive in kidney tumors, and they often do not permit a diagnosis of malignancy, for example in a small histolog- ical section.
The size of the tumor on discovery is in turn corre- lated with the histological type [17]. On discovery, tu- mors with a poor prognosis such (as sarcomatous forms) are frequently already larger than clear-cell carcinomas. On the other hand, oncocytomas are more than 10 cm size in the majority of cases. If carcinomas over 10 cm in diameter are compared, it is striking that two prognostic groups can be distinguished: oncocytomas, papillary car- cinomas and well-differentiated clear-cell carcinomas as compared to sarcomatous and undifferentiated variants [14].
Finally, it is to be observed that a great histological variability occurs in a considerable number of tumors. This makes it difficult to assign them unequivocally to a specific histological classification form.
To summarize, it must be observed that a purely mor- phological appraisal of a concrete neoplasia with consid- eration of the size, histology and cytology only permits a reliable appraisal of its biological behavior with limita- tions. The numerous further studies not quoted here make no difference to this. It is admissible to undertake statistical appraisals as a pathologist working in the field of diagnosis, but a reliable prognosis cannot be made for individual cases, in particular on the basis of a tiny biopsy in the majority of cases. An oncocytoma can be considered to be a largely benign tumor. At all events, a sarcomatous form is associated with a poor prognosis. However, the prognosis is relatively imprecise for the remainder, a group comprising more than 80%. This is unsatisfactory in particular in terms of a conservative surgical procedure. However, additional criteria are therefore called for.
Cytogenetic and Molecular Biological Aspects: Oncogenes and Chromosome Analyses
Molecular biology has contributed ideas on oncogen- esis [52-54] and has introduced the concept of onco- genes. The central notion in this concept is that there are genes which may cause a cell to become neoplastic. In narrow terms, these oncogenes may induce neoplasms directly.
The second group of ‘tumor suppressor’ genes or antioncogenes constitute a kind of counterweight. These are factors which are recessive and only have an onco- genic action when all alleles are inactivated. They pro- vide an especially good illustration for the concept of carcinogenesis in several steps [55-57]. Finally, a third group of genes has a modifying action on the tumor-host relationship, and may thus permit tumor proliferation. For example, the reduced expression of genes of the main histocompatibility complex plays a role in tumor development and progression [58]. With modern tech- niques, these oncogenes can also be used as markers for particular tumors. In addition, they enable a distinction to be made between various tumor types and etiologies [59].
With regard to kidney cell tumors, in this context it is interesting that hereditary forms have already been known for a long time [60]. The most familiar manifes- tation is the increased occurrence of renal-cell carcino- mas in Hippel-Lindau disease, in which cytogenetic changes in the form of characteristic chromosomal aber- rations could also be demonstrated [61, 62].
If molecular biological, cytogenetic or chromosome- analytical investigations can be correlated with specific histological patterns, a new approach is made possible for a broader-based consideration of various human tumors. This is precisely what Kovacs et al. [63-67] tried to do for kidney neoplasms in that they carried out sys- tematic chromosomal analyses and thus arrived at a new classification of these tumors.
Cytogenetic and Molecular Biological Aspects: An Approach to a New Classification
We have seen that the prognosis both of the rare onco- cytomas and of the sarcomatous forms of renal-cell car- cinoma can already be adequately defined in morpholog- ical terms. This is not substantially affected by their size and is an adequate basis for therapeutic action. The former category (the oncocytomas) constitutes a very uniform and discrete group, even in morphological terms.
Apart from these two groups, the group of papillary neoplasms (comprising about 10%) strikes the clinical pathologist’s eye. They have by no means been recog- nized everywhere as independent tumors, but have some uniform typical morphological features of their own. The relatively hetergeneous tumors of the classical nonpapil- lary ‘hypernephroma’ type contrast with this group.
The hereditary forms of nonpapillary renal-cell carci- noma often show a multiple and bilateral occurrence in patients aged 40-50 years. Chromosomal analysis in families with an increased occurrence of renal-cell carci- nomas of the nonpapillary type have shown that lesions occur here in the region of the short arm of chromo- some 3. These cannot only be demonstrated in tumor cells, but also in germinal cells. These alterations, which are to be localized in the region 3p13-14.2 [60, 67]. might lead to disturbance of a suspected suppressor gene. the one allele of which is thus inactivated. If this is trans- mitted genetically (as in the case of hereditary predispo- sition), only a further inactivation of the remaining sup- pressor gene is necessary in the carrier in order to con- vert a tubule cell into a neoplastic cell. Similar changes can also be suspected in hereditary forms in which no chromosomal aberrations can be seen in the germinal cells [68]. The gene responsible for the Hippel-Lindau disease could also be located on the short arm of chro- mosome 3 [69]. Loss of the 3p segment could also be demonstrated in renal-cell carcinomas of patients with Hippel-Lindau disease [61, 62]. It is not clear whether the same gene is involved in other familial kidney can- cers as in Hippel-Lindau disease. However, it is probable that a gene or a gene cluster which is located here is involved in all hereditary forms of kidney cancer.
In contrast to this, sporadic nonpapillary kidney cell cancers occur later, at an age of 60-70 years on average. In more than 90% of the cases of sporadic, nonpapillary renal-cell carcinomas investigated, Kovacs and Frisch [65] have also found a chromosomal aberration on chro- mosome 3p, usually with a break point in the region 3p14, and have thus confirmed the findings of other authors [70-72]. In addition, a rearrangement of chro- mosome 3p in 81 out of 85 nonpapillary tumors investi- gated could be demonstrated with DNA analyses [73]. Furthermore, it could be shown by means of DNA anal- yses that in renal-carcinoma cells alleles are absent in the segment of the particular short arm of chromosome 3 under consideration [64, 74, 75]. Among these, the ab- sence of an allele of the c-raf and c-erbAbeta oncogenes could be demonstrated [71]. These characteristics can be used as diagnostic markers for nonpapillary renal cell tumors: when cytogenetic or DNA analyses of kidney tumors show a rearrangement or loss of heterozygosis in the region of the short arm of chromosome 3, the tumor can be designated as nonpapillary kidney cancer.
In addition, Kovacs et al. [63] reported numerical and structural changes on chromosome 5 in nonpapillary cancers. A trisomy status of region 5q22-qter is demon-
*
| Morphology | Characteristic | Prognosis | |
|---|---|---|---|
| Oncozytoma | mitochondrial DNA changed | good | |
| Papillary | trisomy 7 100% | good? | |
| adenoma | trisomy 17 100% | ||
| Papillary | trisomy 7 | 20% | moderate (?) |
| carcinoma | trisomy 17 80% | ||
| Nonpapillary | deletion 3P >90% | variable | |
| carcinoma | +5Q ~ 50% | depending on histology and grading | |
| Carcino- sarcoma | mesenchymal pattern | poor | |
| Carcinoma of collecting duct | morphologically defined | poor | |
strated in more than 50% [65]. A loss of chromosome 14 in about 50% can also be found [65]. The c-fos protoon- cogene, the product of which play a role in proliferation, is located on the long arm of this chromosome [76]. In consequence of this, it appears to be probable that tumo- rigenesis in nonpapillary renal-cell carcinoma is in- fluenced by several chromosomal regions. The cascade theory of carcinogenesis supports this assumption [57].
The concept of carcinogenesis in nonpapillary kid- ney-cell cancer thus postulates a suppressor gene on chromosome 3 both in the hereditary and in the nonher- editary case. If the alteration of an allele of this gene is already present in the germ plasm, a further step (hit) is necessary in order to destroy the healthy allele as well and to set off oncogenesis. For this reason, there is an earlier and multiple occurrence of the carcinoma in the hereditary lines. In the sporadic case, two mutations (each somatic) are necessary in the coding genes: both alleles must be altered before tumorigenesis of a nonpa- pillary kidney-cell cancer can commence.
The carcinogenesis of papillary renal tumors is re- lated to the organogenesis of the kidney itself. The tumor originates in nondifferentiated tubule epithelia, whereas the nonpapillary kidney-cell cancer derives from mature tubule cells. Persistent tissues of the nephrogenic blas- tema are known as nodular blastema [40] or as meta- nephrogenic hamartoma. They may become manifest as nephroblastomas or as other hamartomatous tissue for- mations. Transitions between nephroblastomas and pa-
pillary carcinomas are known [77]. A coincidence be- tween various of these dysontogenetic tumors has been described [78].
Papillary kidney-cell tumors display a very uniform pattern of papillae which is often interspersed by macro- phages. A tendency to necrosis has been described in the literature [79, 80], but in our experience this is not more pronounced than in nonpapillary renal-cell carcinoma. About 70% of papillary kidney tumors display trisomy of chromosome 17, which is not found in nonpapillary tumors. A trisomy 7 is found more frequently than in nonpapillary tumor. In contrast to nonpapillary carcino- ma, there are no changes on the short arm of chromo- some 3 and on chromosome 5.
Chromosome analysis of oncocytomas of the kidneys does not show any regular alterations of the caryotype [48, 81, 82 as opposed to 83]. However, only relatively small numbers of tumors have been investigated. On the other hand, a mosaic with clonal or nonclonal chromo- somal alterations is often found, which is not manifested in other kidney tumors. Ultrastructurally, oncocytomas are characterized by enlarged, densely packed mitochon- dria. Restriction analyses of mitochondrial DNA re- vealed a typically altered pattern [73] compared to papil- lary and nonpapillary tumors. It is possible that this altered DNA is responsible for the characteristic, atypi- cal morphology of mitochondria. It is also not to be ruled out that it is connected with the pathological neoplastic proliferation.
Discussion
If an attempt is made to summarize what has been said up to now, it must be observed that very many ques- tions are still open in kidney tumors in both practical and theoretical terms. The incorporation of molecular biological and cytogenetic methods in their exploration has opened up new perspectives which are not only man- ifested in the suggestion of new classifications, but also contribute to a better understanding of the biology of these tumors and will have therapeutic consequences.
In particular the question of multicentricity calls pos- sibly for the modification of the strategy of surgery. Do the synchronous kidney cancers which are manifested contralaterally in about 3% (or more) [84] of patients, but which occur more frequently in the ipsilateral kid- ney, mostly belong to the papillary forms?
A classification which considers the data and factors mentioned does not only have to incorporate morpho- logical criteria, but also other (e.g. molecular biological) parameters. According to the present results, a practi- cally applicable classification might be suggested in ac- cordance with table 1.
The papillary tumors, which are subdivided into ade- nomas and carcinomas in table I, do not only have a uniform morphology in accordance with this suggestion, but are also clearly defined in molecular biological terms. However, triploidy of the chromosome 7 or chro- mosome 17 is only maintained in adenomas, and the question arises as to how adenomas can be identified in the investigations concerned. Such a distinction is also unsuitable for practical purposes, since it entails more time and effort than that required for routine diagnos- tics. It remains to be investigated to what extent there is a relationship between tumor size and malignancy in this homogeneous group.
It is apparent that the majority of malignant tumors is still subject to a relatively undefined, variable prognosis. It cannot be decided in individual cases how the tumor will respond. Histological information must correspond- ingly be applied here. A further subclassification of these tumors may be possible with molecular biological meth- ods.
The predominant mesenchymal sarcomatous tumors might be regarded as an extreme variant. This is indi- cated by the observation that sarcomatous regions may be present in nonpapillary morphologically multiform tumors and that morphological transitions are to be observed from the one to the other form. In morpholog- ical terms, a separation is only possible on an arbitrary
basis. However, an additional or indeed independent molecular biological lesion cannot be ruled out in these sarcomatous variants.
A tumor which often gives rise to difficulties in differ- ential diagnosis from urothelial carcinoma is the rare carcinoma of the collecting tubules. These are reports [35-37] alleging that such tumors can be exactly classi- fied on the basis of immunohistochemical characteris- tics. On the other hand. it is asserted that the oncocy- toma derives from the collecting tubules [85].
This illustrates that a detailed histogenesis of the tumors as the sole criterion for classification will have to be abandoned in the future, not only in kidney carcino- mas. Within limits, the tissue of origin of a tumor con- stitutes an important basis for its biological behavior. so that to a certain extent a kidney tumor will imitate a tubule cell of the kidney. On the other hand, it is to be assumed that these properties of the tumors are superim- posed by pathogenetic processes on the molecular bio- logical level, so that quite different properties of the tumor tissue may be manifested. An exact inference (for example from immunohistologically demonstrable gene products) to the underlying tissue is thus not possible in principle. This does not imply that this may not be help- ful in a first approximation.
It remains to be hoped that the biology of kidney tumors can be better understood on the basis of a multi- tiered approach, and thus contribute to a more differen- tiated therapy.
Acknowledgment
We thank Prof. G. Kovacs for his excellent support. above all for giving us his original diapositives as a loan.
References
I Grawitz P: Die sogenannten Lipome der Niere. Virchows Arch Pathol Anat 1883;93:39-62.
2 Klimpel, K: Multiple endokrine Erscheinungen bei einem re- nalen Hypernephrom, die nach Nephrektomie schwanden. J Urol 1954;47:618-634.
3 Hannah J, Lippe B, Lai Goldman M. Bhuta S: Oncocytic carci- noid of the kidney associated with periodic Cushing’s syndrome. Cancer 1988:61:2136-2140.
4 Stoerk O: Zur Histogenese der Grawitz’schen Nierenge- schwülste. Beitr Pathol Anta 1908:43:393-437.
5 Rosenfeld E: Die histogenetische Ableitung der Grawitz’schen Nicrengeschwülste. Frankf Z Pathol 1913:14:151-184.
6 Ortmann R: Grundriss der Entwicklungsgeschichte des Men- schen. Berlin, Springer, 1966.
7 Apitz K: Die Geschwülste und Gewebsmissbildungen der Niere.
III. Die Adenome. Virchows Arch Pathol Anat 1943:311:328- 359.
8 Apitz K: Die Geschwülste und Gewebsmissbildungen der Niere. IV. Strumen, Krebse und Carcinosarkome. Virchows Arch Pa- thol Anat 1943;311:360-431.
9 Largiadèr F: Morphologie, Histogenese und Klassifikation der Nierentumoren. Urol Int 1958;6:273-367.
10 Zollinger HU: Niere und ableitende Harnwege; in Doerr W. Uehlinger E (eds): Spezielle pathologische Anatomie, Bd 3. Ber- lin, Springer, 1966.
11 Warter AL: Recent progress in the pathological anatomy of can- cers in the kidney. Proc Surg 1983;17:32-57.
12 Leder LD, Richter JH, Stambolis CHR: Pathology of renal and adrenal tumors; in Löhr E (ed.): Berlin, Springer, 1979.
13 Mostofi FK: Pathology and spread of renal cell carcinoma; in King JS (ed): Renal Neoplasia. Boston, Little, Brown. 1967. pp 41-85.
14 Mostofi FK: Tumors of the renal parenchyma; in Churg J, Spargo BH. Mostofi FK, Abell MR (eds): Kidney Disease: Present Status, pp 356-412.
15 Mostofi FK (ed): Histological Typing of Kidney Tumors. Inter- national Histological Classification of Tumors. Geneva, WHO. No 25, 1979.
16 Thoenes W, Störkel S. Rumpelt HJ: Histopathology and classifi- cation of renal cell tumors (adenomas, oncocytomas and carci- nomas). Path Res Pract 181: 125-143 (1986).
17 Reis M, Faria V: Renal carcinoma, Reevaluation of prognostic factors. Cancer 1988;61:1192-1199.
18 Macke RA. Hussain MB. Imray TJ, Wilson RB. Cohen SM: Osteogenic and sarcomatoid differentiation of a renal cell carci- noma. Cancer 1985:56:2452-2457.
19 Herman CJ, Moesker O, Kant A, Huysmans A, Voojs GP. Ramaekers FCS: Is renal cell (Grawitz) tumor a carcinosarcoma? Virchows Arch [B] 1983:44:73-83.
20 Waldherr R. Schwechheimer K: Co-expression of cytokeratin and vimentin intermediate-sized filaments in renal cell carcino- mas. Comparative study of the intermediate-sized filament dis- tribution in renal cell carcinoma and normal human kidney. Vir- chows Arch [A] 1985;408:15-17.
21 Thoenes W. Storkel ST. Rumpelt HJ. Moll R. Baum HP, Werner S: Chromophobe cell renal carcinoma and its variants - A report on 32 cases. J Pathol 1988:155:277-287.
22 Oberling CH, Ribière M. Haguenau FR: Ultrastructure des épi- théliomas à cellules claires du rein (hypernéphromes ou tumeurs de Grawitz) et son implication pour l’histogénèse de ces tu- meurs. Bull Assoc Fr Cancer 1959;46:356-381.
23 Seljelid R, Ericsson JL: Electronmicroscopic observations on specializations of the cell surface in renal clear cell carcinoma. Lab Invest 1967;14:435-447.
24 Wallace AC, Nairn RC: Renal tubular antigens in kidney tu- mors. Cancer 1972;29:977-981.
25 Saito T. Saito K, Trent DJ, Draganac PS, Andrews RB. Farkas WR, Dunn CD, Etkin LD, Lange RD: Translation of messenger RNA from a renal tumor into a product with the biological prop- erties of erythropoietin. Exp Hematol 1985;13:23-28.
26 Nielsen OJ, Jepersen FF. Hilden M: Erythropoietin-induced polycythemia in a patient with a renal cell carcinoma. APMIS 1988;96:688-694.
27 Shouval D, Anton M, Galun E, Sherwood JB: Erythropoietin- induced polycytemia in athymic mice following transplantation
of human renal carcinoma cell line. Cancer Res 1988:48:3430- 3434.
28 Ueno M. Brookins J, Beckman BS, Fisher JW: Effects of reactive oxygen metabolites on erythropoietin production in renal carci- noma cells. Biochem Biophys Res Commun 1988:154:773- 780.
29 Tanaka S. Yasumoto R. Yuki K, Iseki T. Nakatani T. Maekawa M: A case of renal tumor with erythrocytosis. Acta Urol Jpn 1989;35:469-474.
30 Strewler GJ. Stern PH. Jacobs JW. Eveloff J. Klein RF, Leung SC, Rosenblatt M. Nissenson RA: Parathyroid hormone-like protein from human renal carcinoma cells. J Clin Invest 1987; 80:1803-1807.
31 Thiede MA, Strewler GJ, Nissenson RA. Rosenblatt M. Rodan GA: Human renal carcinoma expresses two messages encoding a parathyroid hormone-like peptide. Proc Natl Acad Sci USA 1988:85:4605-4609.
32 Watanobe H. Yoshioka M, Takebe K: Ectopic ACTH syndrome due to Grawitz tumor. Horm Metab Res 1988;20:453-456.
33 Cook HT, Taylor GM. Malone P. Risdon RA: Renin in meso- blastic nephroma: An immunohistochemical study. Hum Pathol 1988;19:1347-1351.
34 Gray HW. Smith ML. Chung FF. Yogorajah S. Beastall GH: Hyperinsulinism associated with primary renal cell carcinoma. Urology 1989:33:238-240.
35 Becht E, Muller SC, Storkel S, Alken P: Distal nephron carcino- ma: A rare kidney tumor. Eur Urol 1988;14:253-254.
36 Storkel S. Steart PV, Drenckhahn D. Thoenes W: The human chromophobe cell renal carcinoma: Its probable relation to inter- calated cells of the collecting duct. Virchows Arch [B] 1989:56: 237-245.
37 Kobari T. Machida T. Ohishi A. Kotera S. Onishi T. Aizawa S. Kikuchi Y: A case of renal carcinoma possibly originating from Bellinis duct. Jap J Urol 1988:79:1108-1113.
38 Claes G: Concerning the relationship between the morphology and the symptomatology of hypernephroma. Urol Int 1963;15: 266-279.
39 Arner O, Blanck C, Von Schreeb T: Renal adenocarcinoma. Acta Chir Scand 1965:346(suppl): 1-15.
40 Bottiger LE: Prognosis in renal carcinoma. Cancer 1970;26:780- 787.
41 Lieber MM, Tomera FM, Taylor WF. Farrow GM: Renal adc- nocarcinoma in young adults: Survival and variables affecting prognosis. J Urol 1981;125:164-168.
42 Störkel S, Thoenes W. Rumpelt HJ. Becht E. Düber C: Zur Mor- phologie und Klinik des Onkozytoms der Niere. Verh Dtsch Ges Pathol 1985;69:615.
43 Walt van der JD, Reid HAS, Riedon RA. Shaw JHF: Renal oncocytoma. A review of the literature and report of an unusual multicentric case. Virchows Arch [A] 1983:398:291-304.
44 Wenzel M, Alles JU, Fischer HP. Kracht J, Pascu F. Stambolis C: Onkozytom der Niere. Fallberichte und Literaturübersicht. Pathologe 1986;7:101-109.
45 Fairchild TN, Dail DH, Brannen GE: Renal oncocytoma - Bilat- cral, multifocal. Urology 1983;22:355-359.
46 Jockle GA, Toker C, Shamsuddin AM: Metastatic renal onco- cytic neoplasm with benign histologic appearance. Urology 1987:30:79-81.
47 Young RH, Dunn J. Dickersin GR: An unusual oncocytic renal
tumor with sarcomatoid foci and osteogenic differentiation. Arch Pathol Lab Med 1988;112:937-940.
48 Psihramis KE. Dal Cin P. Dretler SP. Prout GR Jr. Sandberg AA: Further evidence that renal oncocytoma has malignant potential. J Urol 1988:139:585-587.
49 Bisceglia M. Tardio B. Bosman C: Renal oncocytoma: A clinico- pathologic study of a case with review of the literature. Appl Pathol 1989:7:61-69.
50 Bennington JL, Beckwith JB: Tumors of the kidney, renal pelvis and ureter; in Firminger HI (cd): Atlas of Tumor Pathology. Armed Forces Institute of Pathology, 1975.
51 Bell ET: Renal Diseases, ed 2. Philadelphia, Lea & Febiger, 1950, p 435.
52 Knudson AG: Hereditary cancer, oncogenes and anti-oncogenes. Cancer Res 1985;45:1437-1443.
53 Bishop JM: The molecular genetics of cancer. Science 1987;235: 305-311.
54 Klein G. Klein E: Evolution of tumours and the impact of molecular oncology. Nature 1985;315:190-195.
55 Knudson AG: Mutation and cancer: Statistical study of retino- blastoma. Proc Natl Acad Sci USA 1971:60:820.
56 Knudson AG, Strong LC: Mutation and cancer: A model for Wilms’ tumor of the kidney. J Natl Cancer Inst 1972; 38:313.
57 Knudson AG: A two mutation model for human cancer. Adv Viral Oncol 1987;7:1-17.
58 Festenstein H, Garrido F: MHC antigens and malignancy. Na- ture 1986;322:502-506.
59 Fujita J. Kraus MH, Onue H, Srivastava SK, Ebi Y. Kitamura Y. Rhim JS: Activated H-ras oncogenes in human kidney tumors. Cancer Res Arch 1988;48:5251-5255.
60 Cohen AJ, Li FP, Berg S, Marchetto DJ, Tsai J. Jacobs SC, Brown RS: Hereditary renal-cell carcinoma associated with chromosomal translocation. N Engl J Med 1979:301:592-595.
61 King CR, Schimke RN, Arthur T, Davoren B, Collins D: Proxi- mal 3p deletion in renal cell carcinoma cells from a patient with von Hippel-Lindau disease. Cancer Genet Cytogenet 1987;27: 345-348.
62 Decker HJ. Neuman HPH, Walter TA. Sandberg AA: 3p in- volvement in a renal cell carcinoma in von Hippel-Lindau syn- drome. Cancer Genet Cytogenet 1988;33:59-65.
63 Kovacs G, Scucs S, DeRiese W. Baumgartel H: Specific chromo- some aberration in human renal carcinoma. Int J Cancer 1987; 40:171-178.
64 Kovacs G, Erlandsson R, Boldog F, Invarrsson S, Muller-Brech- lin R. Klkein G, Sumegi J: Consistent chromosome 3p deletion and loss of heterozygosity in renal cell carcinoma. Proc Natl Acad Sci USA 1987;85:1571-1575.
65 Kovacs G, Frisch S: Clonal chromosome abnormalities in tumor cells from patients with sporadic renal cell carcinomas. Cancer Res 1989;49:651-659.
66 Kovacs G: Papillary renal cell carcinoma. A morphologic and cytogenetic study of 11 cases. Am J Pathol 1989;134:27-34.
67 Kovacs G, Brusa P. DeRiese W: Tissue-specific expression of a constitutional 3:6 translocation: Development of multiple bilat- eral renal cell carcinomas. Int J Cancer 1989:43:422-427.
68 Li FP, Marchetto D, Brown RS: Familial renal carcinoma. Can- cer Genet Cytogenet 1982;7:271-275.
69 Selzinger BR. Rouleau GA, Ozelius LJ, et al: Von Hippel-Lindau disease maps to the region 3 associated with renal cell carcino- ma. Nature 1988; 332:268-269.
70 Tajara EH, Berger CS, Hecht BK, Gemmill RM. Sandberg AA. Hecht F: Loss of common 3p14 fragile site expression in renal cell carcinoma with deletion breakpoint at 3p14. Cancer Genet Cytogenet 1988:31:75-82.
71 Kristoffersson U, Jin YS, Lundgren R: Chromosome analysis of a newly established renal carcinoma cell line. Cancer Genet Cytogenet 1988;33:133-138.
72 Nordenson I, Ljunaberg B. Roos G: Chromosomes in renal car- cinoma with reference to intratumor hetereogeneity. Cancer Genet Cytogenet 1988:32:35-41.
73 Kovacs G, Wilkens L. Papp T, DeRiese W: Differentiation between papillary and nonpapillary renal cell carcinomas by DNA analysis. J Natl Cancer Inst 1989:81:527-530.
74 Zbar B, Brauch H, Talmadge C, Linehan M: Loss of alleles of loci on the short arm of chromosome 3 in renal carcinoma. Nature 1987;327:721-724.
75 Bergerheim U, Nordenskjold M. Collins VP: Deletion mapping in human renal cell carcinoma. Cancer Res 1989:49:1390- 1396.
76 Muller R: Cellular and viral fos genes: Structure. regulation of expression and biological properties of their encoded products. Biochim Biophys Acta 1986:823:207-225.
77 Kodet R. Marsden HB: Papillary Wilms’ tumour with carcino- ma-like foci and renal cell carcinoma in childhood. Histopathol- ogy 1985;9:1091-1102.
78 Huang JK, Ho DM, Wang JH, Chou YH, Chen MT. Chang SS: Coincidental angiomyolipoma and renal cell carcinoma - Re- port of 1 case and review of literature. J Urol 1988;140:1516- 1518.
79 Mancilla-Jimenez R. Stanley RS: Papillary renal cell carcinoma: A clinical, radiologic and pathologic study of 34 cases. Cancer 1986;38:2469-2480.
80 Orain I, Buzelin F. Ferry N: Tubulopapillary tumours of the kidney. Apropos of 20 new cases and a review of the literature. J Urol 1987;139:585-587.
81 Kovacs G. Scucs S, Eicher W. Maschek HJ, Wahnschaffe U, DeRiese W: Renal oncocytoma. A cytogenetic and morphologic study. Cancer 1987:59:2071-2077.
82 Kovacs G, Welter L, Blin N. DeRiese W: Renal Oncocytoma. A phenotypic and genotypic entity of renal parenchymal tumors. Am J Pathol 1989:134:967-971.
83 Walter TA. Pennington RD. Decker HJH, Sandberg AA: Trans- location t(9; 11) (p23; q12): A primary chromosomal change in renal oncocytoma. J Urol 1989; 142:117-119.
84 Colt H, Wholey MH: Bilateral renal carcinoma. Am Fam Physi- cian 1988:37:121-125.
85 Zerban H, Noggueria E, Riedasch G. Bannasch P: Renal onco- cytoma: Origin from the collecting duct. Virchows Arch [B] 1987;52:375-388.
Dr. H.R. Burger Institute of Pathology University of Zurich CH-8091 Zürich (Switzerland)