INITIATING GENETIC EVENTS IN SMALL RENAL NEOPLASMS DETECTED BY COMPARATIVE GENOMIC HYBRIDIZATION

0022-5347/98/1604- 1557$03.00/0 THE JOURNAL OF UROLOGY Copyright 0 1998 by AMERICAN UROLOGICAL ASSOCIATION, INC.

Vol. 160, 1557-1561, October 1998 Printed in U S A .

INITIATING GENETIC EVENTS IN SMALL RENAL NEOPLASMS DETECTED BY COMPARATIVE GENOMIC HYBRIDIZATION

JOSEPH C. PRESTI, JR.,* HOLGER MOCH, ARNOLD B. GELB, DANH HUYNH AND

FREDERIC M. WALDMAN From the Department of Urology, Division of Molecular Cytometry, Department of Laboratory Medicine and Department of Pathology,

University of California, San Francisco, California, and the Institute of Pathology, University of Basel, Basel, Switzerland

ABSTRACT

Purpose: To identify the genetic alterations associated with renal adenomas. Materials a n d Methods: We analyzed 37 renal adenomas obtained at autopsy (23 papillary and

14 non-papillary) by comparative genomic hybridization. Results: In papillary tumors, the median number of gains and losses of genetic material per

tumor was 2.0 and 1.0, respectively. Papillary tumors were characterized predominantly by gains of genetic material on chromosomes 7 (57%), 17 (35%), 16 (26%), 12 (26%), 3 (22%), 20 (22%) and loss of a sex chromosome (83%). In 6 papillary tumors less than or equal to 5 mm. in diameter, gain of chromosome 7 occurred in 4 specimens. Initiating events for papillary renal adenomas include gain of chromosome 7 a n d loss of a sex chromosome. In non-papillary tumors, the median number of gains and losses of genetic material per tumor was 1.0 and 1.0, respectively. Non- papillary tumors were characterized by loss of genetic material on chromosome 3p (50%), loss of a sex chromosome (36%) and a gain of chromosome 5 (43%). The initiating event for non-papillary renal adenomas is the loss of chromosome 3p.

Conclusions: Renal adenomas demonstrate similar genetic alterations as clinically detected renal cell carcinomas. Their clinically indolent course may, in part, be a result of the lower number of genetic alterations per tumor than their clinically detected counterparts. Renal adenomas are thus small carcinomas which have not yet acquired the necessary genetic alterations leading to tumor progression.

KEY WORDS: renal cell carcinoma, renal adenoma

The term renal adenoma is sometimes used to describe small tumors of the kidney which are detected as incidental findings at autopsy. The biology of these tumors is not well defined, yet they are usually considered indolent as they are typically of low grade and stage when detected. Much con- troversy exists over whether a benign epithelial tumor of the kidney (adenoma) can be readily characterized and distin- guished from renal cell carcinoma (RCC). Adenomas are typically less than 2.5 cm. in size, lack nuclear anaplasia and are usually found incidentally at autopsy. It must be recognized that all tumors, both benign and malignant, are small at some time during their development. Thus whether these tumors are truly benign adenomas or small RCC remains controversial. The frequency of small renal cortical neoplasms in autopsy series has been as high as 22% and current pathologic classification systems include renal cortical adenomas as distinct entities.'

Unlike clinically detected RCC, the genetic alterations associated with renal adenomas have not been previously described. The genetic alterations of the two main histopatho- logic subgroups of RCC are well characterized. The hallmark alteration of non-papillary RCC is a loss of chromosome 3p. Other genetic alterations of non-papillary RCC involve chromosomes 5 ,6 ,9 , 10, 11,13, 17,18, and 19.'-13 Papillary RCC are genetically characterized by a normal chromosome 3p and frequently have additional copies of chromosomes 7, 12, and 17.9,'4, l5 The objective of the present study was to characterize the genetic alterations of incidentally detected renal tumors and correlate these alterations with histopathology.

Accepted for publication May 1, 1998. * Requests for reprints: Department of Urology, University of

Supported in part by a VA Merit Review Grant. California, 533 Parnassus, U-575, San Francisco, CA 94143-0738.

Our hypothesis was that these tumors would demonstrate similar genetic alterations to clinically detected RCC and might provide information about the initiating genetic events in the pathogenesis of RCC. This was accomplished with comparative genomic hybridization (CGH).

MATERIALS AND METHODS

Tissue and histopathological diagnosis. Tumor specimens were obtained from autopsy specimens from 37 patients. In 15 cases, paraffin-embedded tissue from normal kidney was also obtained. Diagnostic specimens were formalin-fixed and paraffin embedded. Five-micron thick sections were stained with hematoxylin and eosin for review prior to sectioning for DNA extraction. Each tumor was characterized by a single pathologist (AG) for cell type, pathological grade using the Fuhrman System and pathological stage according to the tumor-node-metastasis classification system.l6. l7

DNA isolation and CGH analysis. DNA was isolated from twenty-five 10 micron-thick paraffin sections. Sections were deparaffinized and suspended in DNA extraction buffer containing 0.5 mg./ml. proteinase K. Additional proteinase K was added at 24 and 48 hours later, for a total incubation time of 72 hours." CGH was done as described before.lg Briefly, DNA from tumor and a sex matched normal control were directly labeled by nick translation (NT) with fluorescein -14-dUTP (tumor) or Texas Red-11-dUTP (normal). They were mixed with unlabeled Cot-1 DNA to block binding to repetitive sequences, denatured, and hybridized to a normal lymphocyte metaphase spread for 2 to 3 days at 37C. Slides were washed to remove unbound probe and DNA was counterstained with 4, 5-diamino-2-phenyl-indole (DAF'I, blue) producing a banding pattern used for chromosome identifi-

1557

1558 INITIATING GENETIC EVENTS IN SMALL RENAL NEOPLASMS

cation. The relative binding of tumor (green) and normal (red) DNA’s along each metaphase chromosome reflected the relative abundance of DNA sequences in the tumor. Thus, DNA sequences that were over-represented in the tumor showed relatively increased green fluorescence, and those regions that were under-represented in the tumor appeared with relatively decreased green fluorescence. The labeled normal DNA served as a control for regional variations in the ability to hybridize to the target chromosomes. This DNA does not have to come from the same person as the tumor DNA. Inverse labeling CGH was performed on all samples (tumor DNA nick translated with Texas Red dUTP and normal DNA with fluorescein dUTP) for confirmation of all alterations.

In 15 specimens (#1 to 4, 7, 20 to 22, 25 to 311, CGH was also performed on DNA extracted from the normal kidney tissue from the same patient (normal-normal hybridization).

Digital image analysis. A quantitative image processing system was used for the acquisition and processing of se- quential three-color images using filters for green (fluorescein), red (Texas Red), and blue (DAPI) fluorescence as previously described. Software programs based on the SCIL- image software package were used to display and overlay multicolor images in pseudocolors and to contrast-stretch the images to enhance green and red color differences and the blue DAPI banding.” Green to red ratios of each chromosome were then plotted as a function of distance from the p terminus to q terminus (left to right). Generally, four ratio profiles were averaged for each chromosome (two separate metaphases) to reduce noise. Green-to-red ratios greater than 1.20 were considered as gains of genetic material while ratios less than 0.8 were considered as losses of genetic material. Imbalances were scored by CGH only if they were present in both hybridizations (CGH and inverse CGH).

Degenerate oligonucleotide primer (DOP) -PCR CGH. In 6 specimens less than or equal to 5 mm. in diameter (“microadenomas”, #18 to 231, DOP-PCR CGH analysis was used. For each tumor, two adjacent 5 micron sections were cut from the paraffin block. One section was stained with hematoxylin and eosin and the second section was stained with methyl green (0.1%). All sections were photographed. Using the adjacent hematoxylin and eosin section for orientation, the pathologist identified on the photographs of the methyl green stained sections the tumor area for CGH analysis. The se- lected areas were recovered by carefully scraping away sur- rounding tissue using a disposable #11 scalpel blade. A new #15 blade with a droplet of PCR DNA extraction buffer [ lo mM Tris-HC1 (pH 8.0); 1.5 mM MgC1,; 50 mM KC1; 0.5% Tween-20; 0.4 mg./ml. proteinase K (Sigma)] was used t o transfer desired regions into sterile 0.5 ml. tubes containing 15 p L (more for larger areas) of PCR DNA extraction buffer. Mineral oil (20 pL) was placed over the sample and incubated at 55C overnight in a shaking water bath. Fresh proteinase K (0.3 pL of 20 mg./ml. stock) was added daily for two more days. Proteinase K was inactivated at 95C for 15 minutes. DNA was separated from the oil and transferred into new tubes. All steps were handled carefully to avoid cross over contamination for subsequent PCR reactions.

The amplification of the microdissected DNA was done based on the DOP-PCR protocol of Guan et aLZ1 Samples were amplified in duplicate, in separate PCR reactions. Briefly, a 1 to 2 pL aliquot of microdissected DNA was added to 5 pL of l x PCR buffer (10 mM Tris-HC1 (pH 8.0); 1.5 mM MgCl,; 50 mM KC1; 0.2 mM dNTPs; 2 pM DOP primer) containing 0.2 pL 50 mM MgCl, and 0.1 pL TOPOisomerase I (Promega), covered with oil. The samples were incubated at 37C for 30 minutes, then at 96C for 10 minutes. The TOP0 pretreatment was followed by 5 cycles of sequenase treat- ment (1 minute denaturing at 94C; 2 minutes annealing at 30C, 3 minutes extension at 37C). The sequenase (USB) (0.3 pL of a 1:8 dilution) was added through the oil during the

annealing step for each cycle. All samples had at least 2 minutes of annealing, but no more than 5 minutes. Pream- plification was followed by 1 cycle at 95C for 10 minutes, and 45 pL of l x PCR buffer with 2 Units of Taq DNA Polymerase (Boehringer) was added during the last 5 minutes through the oil, followed by additional oil overlay. This was followed by 35 cycles at 94C for 1 minute, 56C for 1 minute, 72C for 3 minutes, with a final extension at 72C for 5 minutes.

PCR amplified DNA (3 pL) was run on 1% agarose ethidium bromide stained gel to determine the resulting PCR product size. Each PCR experiment included samples of normal female genomic DNA (isolated from peripheral blood), MPE6OO (breast cancer cell line with known CGH aberrations), and a blank to check for PCR contamination.

Fifty ng. of reference or MPEGOO cell line DNA resulted in approximately 2 to 3 pg. of amplified DNA, ranging in size from 200 bp-6 kB. Microdissected DNA yielded up to 1 pg. of PCR product, with sizes averaging around 600 bp in size (ranging from 200 bp-2 kB).

PCR amplified DNA was labeled in duplicate by NT. Nor- mal reference DNA was labeled with Texas Red-5-dUTP or with fluorescein-12-dUTP (Dupont). MPE6OO cell line was labeled with fluorescein or with digoxigenin-11-dUTP (Boehringer), and amplified microdissected DNA was labeled with fluorescein or digoxigenin. The NT reaction was for 60 minutes at 15C, and stopped by heat inactivation at 75C for 15 minutes. The probe was stored at -2OC. The size of the product was tested by electrophoresis through 1% agarose. The usual size for CGH is 500 bp-2000 bp. Nick translated PCR products were close to the original PCR product size.

Hybridizations were done as previously described. Each sample was hybridized in duplicate with different fluoro- chromes. Digoxigenin NT PCR amplified samples were stained with anti-digoxigenin rhodamine (Boehringer). All of the NT microdissected PCR products were hybridized with 20 pg. human Cot-1 DNA (Life Technologies) and either 15 pL (-300 ng.) of the Texas Red NT or 20 pL of the fluorescein isothiocyanate NT female reference PCR probe. For each set of samples a NT PCR normal control and MPE6OO were hybridized against the same reference PCR probe as was used against the test samples. This was done for each PCR reaction set. The samples were hybridized for 2 to 3 days at 37C, washed as previously described and counterstained with 0.1 pg./ml. DAPI in antifade. The red digoxigenin labeled microdissected samples were generally brighter and smoother than the fluorescein samples. Chromosomal alterations were interpreted as described above.

Statistical analysis. Correlations between the number of genetic alterations (chromosomal arm gains, losses, and total number of alterations) and tumor grade, size and patient age were assessed with linear regression or Spearman Rank Or- der coefficients for the papillary tumors. The small number of non-papillary tumors precluded this analysis.

RESULTS

Papillary tumors. Histopathological and CGH data are pre- sented in the table. All tumors were stage pT1NO. In 23 papillary renal tumors the median patient age was 75.0 years (range 55 to 93) and median tumor size was 1.2 cm. The mean and median number of gains per tumor was 2.0 and 2.0, (range 0 to 6) and the mean and median number of losses was 1.0 and 1.0 (range 0 to 31, respectively. Gains involving chromosomes 7 (n = 13) and 17 (n = 8) were the most common followed by gains of chromosome 16 (n = 61, chromosome 12 (n = 61, chromosome 3 (n = 51, and chromosome 20 (n = 5). Losses were commonly seen on the sex chromosomes (n = 19), but were otherwise rare events for papillary RCC. No areas of high level amplification were seen in any papillary tumor.

- I A subset of the papillary tumors, less than or equal to 5

INITIATING GENETIC EVENTS IN SMALL RENAL NEOPLASMS

Histopathological and CGH data of 37 renal adenomas Papillary Renal Adenomas

1559

CGH Losses Age Gradeb Size' CGH Gains Specimen # Sex"

1 f 61 2 1.8 X 2 m 83 2 1.2 7, 12, 13q, 17, 20 Y 3 m 59 2 1.2 7 Y 4 m 79 2 1.2 Y 5 m 86 2 2.3 13q 16p, 20% Y 6 m 71 3 1.5 7, 16, 17, 20 Y 7 m 82 2 1.5 7, 16, 17 Y 8 f 71 1 1.5 X 9 m 75 2 2.5 12,20 Y

11 m 63 2 1.5 7, 12, 17 Y 12 f 77 2 1.0 13q, 16, 17, 20 X 13 m 93 2 1.0 Y 14 m 75 1 1.0 7 Y 15 f 75 1 1.3 3, 7, 16, 17 21 16 m 78 2 2 3, 7, 12, 16, 17, 20 Y 17 m 86 2 1.5 3, 7, 16, 17 Y 18 f 67 1 0.4 19 m 58 1 0.5 Y 20 m 69 1 0.5 7, 12 Y 21 f 55 2 0.3 3q23qter, 7 22 m 64 2 0.4 3, 7, 12 Y 23 f 71 1 0.3 7

m 81 1 0.6 X 19q, y 10

24 25 26 27 28 29 30 31 32 33 34 35 36 37

f 97 f 56 m 74 m 63 f 71 f 82 f 83 f 71 f 94 rn 93 m 76 m 69 f 82 m 65

m, male; ffemale. Fuhrman grade. Diameter in cm.

Non-papillary Renal Adenomas 2 2.0 3 2.4 1 2.4

21 9

1 1.5 5, 7 3P

1 2.0 3q 3P, 8P. x 2 1.5 5 3P 1 1.5 3p, 6q22-qter 1 1.2 5q, 7, 12 1 1 2 2.5 5q, 12q22-pter 3p14.3-pter, Y 1 2 5, 7 2 2 5 3P, x 2 0.6 7 Y

1 2.4 13q21.3-qter, 18 lp31-pter, 3p, 13centq14.1, X

mm. in diameter were analyzed by DOP-PCR CGH. Four of 6 of these microadenomas demonstrated alterations of the au- tosomes, all included gains of chromosome 7.

For the papillary tumors, the total number of genetic alterations (gains and losses) significantly correlated with tumor grade (r = 0.44, p = 0.04) and size (r = 0.41, p = 0.05) and approached significance with patient age (r = 0.39, p = 0.06). If only gains of genetic material were considered, a significant correlation was only seen with tumor grade (r = 0.42, p = 0.05). If only loss of genetic material was considered, significant correlations were observed with tumor size (r = 0.61, p = 0.002) and patient age (r = 0.58, p = 0.004).

Nonpapillary tumors. In 14 non-papillary tumors the median patient age was 75.0 years (range 56 to 97) and median tumor size was 2.0 cm. The mean and median number of gains per tumor was 1.1 and 1.0 (range 0 to 3) and the mean and median number of losses was 1.3 and 1.0 (range 0 to 4), respectively. Seven tumors demonstrated loss of material from chromosome 3p and 5 demonstrated loss of material from the sex chromosomes. Gains were most commonly seen on chromosome 5 (n = 6) and chromosome 7 (n = 4). NO areas of high level amplification were seen in any non- papillary tumor.

In the 15 normal-normal hybridizations (8 obtained from kidneys with papillary tumors including 3 microadenomas, and 7 obtained from kidneys with non-papillary tumors), no alterations were detected by CGH.

DISCUSSION

The term renal "adenoma" may be a misnomer as these tumors, although detected as incidental findings at autopsy,

probably represent small carcinomas which may develop late in life in an older host with a limited life span and thus are clinically silent. To test this hypothesis, we studied 37 s m d (5 2.5 cm.) renal tumors incidentally detected at autopsy and characterized their genetic alterations by CGH.

In 23 papillary tumors, the most common finding was loss of a sex chromosome in 19 specimens (83%). The significance of this loss is unknown and is seen in many malignancies, as well as in the bone marrow cells of healthy elderly male^.^'.'^ It may thus be a result of the aging process rather than being of biological significance in tumor pathogenesis. However, one cannot ignore that loss of a sex chromosome in non- papillary tumors was not as common (36%). We did observe gains involving chromosomes 7 (57%), 17 (35%), 12 (26%), and 16 (26%). These small tumors thus demonstrate similar findings to those seen in clinically detected papillary RCC. As gain of chromosome 7 was the most common gain in these tumors and the only gain noted in 3 tumors, it is possible that this genetic event is the earliest detectable event in the development of papillary RCC. Additional support for the hypothesis that gain of chromosome 7 may be the earliest detectable event in the development of papillary RCC comes from the analyses of the microadenomas. Four of 6 of these tumors demonstrated gain of chromosome 7. As the gene for epidermal growth factor receptor resides on chromosome 7, it is possible that over-expression of this receptor may be im- portant in the early development of papillruy RCC.

One must be cautious in interpreting the correlations which were observed between genetic alterations and clini- copathological parameters in papillary tumors. While statis-

1560

tically significant correlations were observed, these relation- ships are weak, as manifested by the magnitude of the correlation coefficients. While a simificant correlation with

INITIATING GENETIC E W N T S IN SMALL RENAL NEOPLASMS

dolent course may, in part, be a result of the lower number of genetic alterations per tumor compared to their clinically detected counterparts. Renal adenomas are thus small car-

tumor grade was observed for botg total number of genetic alterations and for gains alone, the former probably holds true because almost all tumors demonstrated a relatively constant number of losses (one, usually a sex chromosome). We hypothesize that it is the gain of genetic material which drives the progression to a higher grade in papillary tumors. The association between loss of genetic material (predominantly a sex chromosome) and patient age has been previously described.

In 14 non-papillary tumors, loss of chromosome 3p was seen in 7 specimens. This alteration is considered the hallmark of clinically detected non-papillary RCC. The lack of finding a loss of chromosome 3p in the remaining 7 tumors might be explained by mechanisms relating to sensitivity (the loss of chromosome 3p does exist but is too small to be detected by CGH), or perhaps, a yet unidentified genetic alteration is responsible for the initiation of tumorigenesis in these small lesions. The frequency of loss of chromosome 3p in the non-papillary tumors in this series is comparable to what we observed previously in a series of locally advanced RCC analyzed by CGH.24 Regarding sensitivity, the alteration might be below the resolution of CGH which can identify large alterations in the genome usually on the order of megabase-pairs in size. CGH does not detect reciprocal trans- locations or other balanced rearrangements resulting in no net change in genetic material. In addition, contamination of tumor DNA with normal DNA is a well-recognized problem when analyzing primary tumor tissue and will decrease the sensitivity of detection of genetic abnormalities by any mo- dality. The DNA of lymphocytes found in primary tumors results in the "dilution" of tumor DNA. Partial allele losses by restriction fragment length polymorphism analysis in RCC resulting from lymphocytes has previously been dem- ~ n s t r a t e d . ~ ~ An alternative, but less likely hypothesis regarding lack of chromosome 3p abnormalities, is that although the loss of 3p occurs early, it may not be the initiating event. However following loss of chromosome 3p, genomic instability occurs and numerous genetic alterations are observed. I t is of interest that in the 7 specimens which did not demonstrate a loss on chromosome 3p (#24,25,26,32,33,35, 37), alterations were rare (median number of alterations per tumor was 1.0) in comparison to the specimens demonstrating 3p deletions (median number of alterations per tumor was 3.0). This observation supports both of our hypotheses.

In a recent report of locally advanced non-papillary RCC (pTSNO), we demonstrated that clinical outcome was related to the number of DNA sequence losses or loss of chromosome 9p observed in the primary tumor.26 Tumors demonstrating fewer than 3 losses had a 5-year recurrence-free survival of approximately 90% while survival in patients with 3 or more DNA sequence losses was less than 50%. In the current series of 14 clinically silent, non-papillary RCC, the median number of losses was only 1.0 per tumor. Thus, these clinically indolent tumors demonstrate few DNA sequence losses by CGH which may contribute to their low malignant potential. Only one tumor in the present study demonstrated a loss of chromosome 9. No information is currently available regarding genetic alterations and clinical outcome with papillary RCC. However, it is interesting to note that gains, rather than losses of DNA sequences, seem to characterize papillary tumors and that the losses, when they do occur, typically involve the sex chromosomes. The biologic significance, if any, of the latter is unclear.

CONCLUSIONS

Renal adenomas demonstrate similar genetic alterations to clinically detected renal cell carcinomas. Their clinically in-

cinomas which have not yet acquired the necessary genetic alterations leading to tumor progression.

REFERENCES

1. Murphy, W. M., Beckwith, J. B. and Farrow, G. M.: Tumors of the kidney, bladder and related urinary structures. In: Atlas of Tumor Pathology. 3rd Series, Fascicle 11. Washington D. C.: Armed Forces Institute of Pathology, 1994.

2. Kovacs, G., Szucs, S., DeRiese, W. and Baumgartel, H.: Specific chromosome aberrations in human renal cell carcinoma. Int. J. Cancer, 4 0 171, 1987.

3. Pathak, S. and Dhaliwal, M. K.: Cytogenetic lesions in renal cell carcinoma. Cancer Bulletin, 41: 324, 1989.

4. Kovacs, G. and Frisch, S.: Clonal chromosome abnormalities in tumor cells from patients with sporadic renal cell carcinomas. Cancer Res., 4 9 651, 1989.

5. m a r , B., Brauch, H., Talmadge, C. and Linehan, M.: Loss of alleles of loci on the short arm of chromosome 3 in renal cell carcinoma. Nature, 327: 721, 1987.

6. Kovacs, G., Erlandsson, R., Boldog, F., Ingvarsson, S., Muller-Brechlin, R., Klein, G. and Sumegi, J.: Consistent chromosome 3p deletion and loss of heterozygosity in renal cell carcinoma. Proc. Natl. Acad. Sci. USA, 8 5 1571, 1988.

7. van der Hout, A. H., Kok, K., van den Berg, A,, Oosterhuis, J. W., Canitt , B. and Buys, C. H. C. M.: Direct molecular analysis of a deletion of 3p in tumors from patients with sporadic renal cell carcinoma. Cancer Genet. Cytogenet., 32: 281, 1988.

8. Anglard, P., Tory, K., Brauch, H., Weiss, G. H., Latif, F., Merino, M. J., Lerman, M. I., Zbar, B. and Linehan, W. M.: Molecular analysis of genetic changes in the origin and development of renal cell carcinoma. Cancer Res., 51: 1071, 1991.

9. Morita, R., Ishikawa, J., Tsutsumi, M., Hikiji, K., Tsukada, Y., Kamidono, S., Maeda, S. and Nakamura, Y.: Allelotype of renal cell carcinoma. Cancer Res., 51: 820, 1991.

10. Presti, J. C. Jr., Rao, P. H., Chen, Q., Reuter, V. E., Li, F. P., Fair, W. F. and Jhanwar, S. C.: Histopathological, cytogenetic, and molecular characterization of renal cortical tumors. Cancer Res., 51: 1544, 1991.

11. Presti, J . C. Jr., Reuter, V. E., Cordon-Cardo, C., Mazumdar, M., Fair, W. R. and Jhanwar, S. C.: Allelic deletions in renal tumors: histopathological correlations. Cancer Res., 53: 5780, 1993.

12. Bergerheim, U., Nordenskjold, M. and Collins, P.: Deletion map- ping in human renal cell carcinoma. Cancer Res., 4 9 1390, 1989.

13. Cairns, P., Tokino, K., Eby, Y. and Sidransky, D.: Localization of tumor suppressor loci on chromosome 9 in primary human renal cell carcinomas. Cancer Res., 5 5 224, 1995.

14. Kovacs, G., Wilkens, T. P. and deRiese, W.: Differentiation between papillary and non-papillary renal cell carcinomas by DNA analysis. J. Natl. Cancer Inst., 81: 527, 1989.

15. Kovacs, G.: Papillary renal cell carcinoma - a morphologic and cytogenetic study of 11 cases. Am. J. Pathol., 134: 27, 1989.

16. Fuhrman, S. A., Lasky, L. C. and Limas, C.: Prognostic significance of morphologic parameters in renal cell carcinoma. Am. J. Surg. Pathol., 6 655, 1982.

17. Beahrs, 0. H., Henson, D. E., Hutter, R. V. P. and Myers, M. H. (eds.), Manual for Staging of Cancer, 3rd Edition. American Joint Committee on Cancer, pp. 199-201. Philadelphia: J . B. Lippincott Co., 1988.

18. Isola, J., DeVries, S., Chu, L., Ghazvini, S. and Waldman, F.: Analysis of changes in DNA sequence copy number by comparative genomic hybridization in archival paraffin-embedded tumor samples. Am. J. Pathol., 145 1301, 1994.

19. Kallioniemi, A., Kallioniemi, O., Sudar, D., Rutovitz, D., Smith, H., Gray, J., Waldman, F., Pinkel, D.: Comparative genomic hybridization: A powerful new method for cytogenetic analysis of solid tumors. Science, 258 818, 1992.

20. Piper, J., Rutovitz, D., Sudar, D., Kallioniemi, A,, Kallioniemi, 0.. Waldman, F. W., Gray, J. W. and Pinkel, D.: Computer image analysis of comparative genomic hybridization. Cytom- etry, 19: 10, 1995.

21. Guan, X. Y., Trent, J . M. and Meltzer, P. S.: Generation of

INITIATING GENETIC EVENTS IN SMALL RENAL NEOPLASMS 1561

band-specific painting probes from a single microdissected chromosome. Human Molecular Genetics, 2 1117, 1993.

22. Pierre, R. V. and Hoagland, H. C.: Age-associated aneuploidy - loss of Y chromosome from human bone marrow cells with aging. Cancer, 3 0 889, 1972.

23. Guttenbach, M., Koschorz, B., Bernthaler, U., Grimm, T. and Schmid, M.: Sex chromosome loss and aging: in situ hybridization studies on human interphase nuclei. Am. J. Hum. Genet., 57: 1143, 1995.

24. Presti, J. C., Jr., Moch, H., Reuter, V. E., Cordon-Cardo, C. and Waldman, F. M.: Renal cell carcinoma genetic analysis by

comparative genomic hybridization and restriction fragment length polymorphism analysis. J . Urol., 156 281, 1996.

25. Linehan, M., Miller, E., Anglard, P., Merino, M. and Zbar, B.: Improved detection of allele loss in renal cell carcinomas after removal of leukocytes by immunologic selection. J. Natl. Can- cer Inst., 81: 287, 1989.

26. Moch, H., Presti, J . C. Jr., Sauter, G., Buchholz, N., Jordan, P., Mihatsch, M. J . and Waldman, F.: Genetic aberrations detected by comparative genomic hybridization are associated with clinical outcome in renal cell carcinoma. Cancer Res., 5 6 27, 1996.

Documents

INITIATING GENETIC EVENTS IN SMALL RENAL NEOPLASMS DETECTED BY COMPARATIVE GENOMIC HYBRIDIZATION