DNA ploidy pattern in papillary renal cell carcinoma. Correlation with clinicopathological parameters and survival

PATHOLOGY RESEARCH AND PRACTICE

© Gustav Fischer Verlag

DNA Ploidy Pattern in Papillary Renal Cell Carcinoma. Correlation with Clinicopathological Parameters and Survival* Maria T. del Vecchio 1, Stefano Lazzi 1, Alessandra Bruni 1, Paola Mangiavacchi 1, Gabriele Cevenini 2 and Pietro Luzi 1 ,Institute of Pathological Anatomy and Histology, University of Siena, Siena, Italy 21nstitute of Thoracic and Cardiovascular Surgery and Biomedical Technology, University of Siena, Siena, Italy

Summary Papillary renal cell carcinoma (PRCC) is a less frequent histomorphologic variant of renal cortical carcinoma (RCC). Morphologically, PRCC differs from other forms of RCC in that it is associated with frequent tumor infiltration by macrophages and lymphocytes, and a tendency for central necrosis and cystic change. Follow-up data revealed that survival rates are higher among patients with PRCC than among patients with other forms of RCe. The authors explore the DNA content in a series of PRCC and correlate the findings with nuclear grade, pathological stage and survival. Using Flow Cytometry, we analysed the DNA ploidy pattern of 37 paraffin-embedded PRCe. At least 3 tumor fragments were analysed in each case. To obtain the reference diploid standard, the non-tumor renal tissue from the same case was added to the solution. Tumor ploidy was classified as diploid and aneuploid. The degree of DNA content abnormalities was given by the DNA Index (DI). An aneuploid DNA profile was found in 65% of the tumors. 25% of the aneuploid tumors presented near diploid peaks (1.10 < 01 < 1.30; low degree aneuploidy), 25% were hyperdiploid, while 22% had a hypodiploid profile (01 < 0.90). A homogeneous DNA ploidy pattern was observed in 25 tumors (68%), while there was intratumoral heterogeneity in 12 tumors (32%). Patients with aneuploid DNA patterns had high grade/stage tumors and died at the end of the follow-up

* Study funded by Ministero dell'Universita e della Ricerca Scientifica e Tecnologica (MURST 60%). Pathol. Res. Pract. 194: 325-333 (1998)

period, while patients with diploid/near diploid profiles had low grade/stage tumors and survived. However, the multi-way analysis of variance performed in order to investigate the prognostic significance of ploidy pattern against tumor stage and grade showed a highly significant main effect of ploidy pattern. Moreover, the patients with hypodiploid DNA profile presented the worst prognosis. These results suggest that the DNA profile of PRCC is a highly significant prognostic index. Key words: Papillary renal cell carcinoma - DNA ploidy - Flow cytometry - Prognosis

Introduction The clinical behaviour of renal cell carcinoma (RCC) varies considerably. It is therefore important to identify parameters determining its clinical outcome. Throughout the years, a variety of features including pathologic stage at diagnosis, histologic or nuclear grade, nuclear area and shape, nucleolar organizer regions, expression of ~2 microglobulin, kinetic cell and nuclear DNA content have been suggested to carry prognostic information [9, 11, 14, 15, 17,22, 31, 32, 37,40,42,46, 50}. However, RCC presents many histomorphological variants which have not been considered in most previous studies. Address for correspondence: Prof. Pietro Luzi, Institute of Pathological Anatomy and Histology, Uiliversity of Siena, Via delle Scotte 6, 53100 Siena, Italy. Tel.: ++39/5 77-26 32 37, Fax: ++39/5 77-26 32 35 0344-0338/98/0194-0325$5.00/0

326 . M. T. del Vecchio et al.

Papillary renal cell carcinoma (PRCC) is a less frequent variant of renal cortical carcinoma (RCC), representing approximately 14% of all renal parenchymal neoplasms [39]. Clinical presentation, sex distribution and age range of patients with PRCC are similar to those studied with other forms of RCC [28]. Morphologically, PRCC differs from other forms ofRCC in that it is associated with frequent tumor infiltration by macrophages and lymphocytes, and a tendency for central necrosis and cystic change. The morphological criteria for differentiating PRCC from other forms of RCC have been established, but overlapping histologic features sometimes do exist. Recently, cytogenetic and molecular genetic studies have identified characteristics that further distinguish these tumors from conventional RCC [7, 23, 24, 27, 28,45]. For the most part, followup data have revealed that survival rates are higher among patients with PRCC than among patients with otherforms ofRCC [3, 4, 21, 29, 35,44]. Since paraffin-embedded material has been important for retrospective studies that have been useful in establishing the predictive value of DNA content measurement [18], the DNA ploidy pattern of RCC has recently been investigated [8, 14, 19, 33, 41, 43, 49]. Although some PRCCs may have been included in a number of these studies, analysis of DNA ploidy pattern of this category has been exclusively performed by EI-Naggar et al. [12]. The aim of this study was to investigate the DNA ploidy pattern in an archival series of formalin-fixed, paraffin-embedded samples of PRCC examined by Flow Cytometry (FCM) and to correlate the findings with nuclear grade, pathological stage and survival.

Material and Methods From the surgical pathology files of the Institute of Pathological Anatomy, University of Siena, all cases of RCC curatively resected between 1984 and 1990 were identified. The histological slides (average 7) of these neoplasms were reviewed by a pathologist (M.T.d.Y.). Neoplasms were classified as papillary if more than 75% of the neoplasm had a papillary configuration [27, 30, 35]. Low grade tubulo-papillary tumors <0,5 cm with no clear cells were classified as adenomas [38J, and were therefore excluded. Forty-eight neoplasms (13% of all RCC) were classified as PRCC. Cases were rejected if insufficient tissue was available for FCM. Follow-up data were obtained through the general practitioner and/or relatives of the patient. These data were reviewed in order to determine if the death was tumor-related or the result of other causes. Forty-three cases were included in the study. Nuclear grading according to Fuhrman et al. [l5J was used. Each tumor was graded by the most malignant or highest grade exhibited. The pathological stage was assigned according to the system 'of Robson et al. [46J. The area in the block containing the tumor was marked (tumor-enriched area). The tumor-enriched and non-tumor

areas were isolated by perpendicular and horizontal cutting of the block, put into separate vials and sent to the flowcytometry laboratory. On the basis of the histologic pictures, the tumor-enriched area was subdivided into fragments (at least 3) representing different nuclear patterns, if present, to study the heterogeneity of the DNA content. Non-tumor kidney tissue was carefully examined to exclude small tumor or tumorlets. All analysis was performed without knowledge of the tumor characteristics. Nuclei were isolated according to the method of Headley et al. [18J, making some modifications in the procedure. The fragments were placed in a glass tube and dewaxed overnight in 10 rnl of xilene at room temperature. The fragments were then rehydrated in a sequential series of ethanol solutions: 100%, 95%, 70% and 50%, taking 50 minutes for each step. The fragments were left in distilled water overnight. The fully rehydrated tissue was then minced with surgical scissors and incubated in 2 ml 0.5% pepsin (SIGMA P 6887; SIGMA, St. Louis, MO) in saline solution at pH 1.5 in a 37°C water bath for 60 minutes with frequent vortexing. The enzymatic reaction was controlled under phase-contrast microscopy and stopped by the addition of 2 ml ice cold PBS buffer (phosphate buffer saline). After washing, the pellet was resuspended in 2 ml of PBS and filtered through 50 11m mesh to remove aggregates. The nuclei were simultaneously counted in a hemocytometer, adjusted to lxlO6/ml and stained with 500 11m of LysislDNA solution (Propidium iodide, SIGMA, 50 I1g/ml; Rnase, SIGMA R-4875, 0,2 mg/ml; Nonidet P40, SIGMA, 0.5% v/v; EDTA, SIGMA, 0.5 mM in PBS calcium and magnesium free, at pH 7.2). The presence of tumor cells in the material used was verified immediately after its disaggregation on a May-Grunwald Giemsa-stained cytospin. A cytologist performed the cytospin reading. In each case, at least 10,000 nuclei were measured. FCM was performed on an FACStar plus flowcytometer (Becton & Dickinson; San Jose, CA) using a 488 nm argon laser (100 mW). The flowcytometer was connected to a HewlettPackard 300 microcomputer (Hewlett Packard, Fort Collins, CO) using an FACStar plus software for instrument control and data acquisition. To obtain a reference diploid standard, the non-tumor renal tissue treated as described above was added to the solution. Peak samples with a coefficient of variation (CV) of more than 8% diploid were excluded. Thirtyseven cases were analyzed. An analysis of the histograms generated by FCM was carried out by means of the Modfit DNA modelling system (Verity Software House, Tofahour, MC). The mathematical model was fitted to data by means of the Marquardt [36J non-linear least-square optimization algorithms. Debris filtering was performed through debris subtraction routines based on a multicut for the beginning and end of the debris components. The software was implemented on a Compaq 386120 (Compaq Computer Corporation, Houston, TX). According to the current literature, tumor ploidy was classified, for practical purposes, as diploid (0.90 < DI < 1.10) and aneuploid [34, 47 J. The definition of DNA aneuploidy included the requirement that at least a third peak different from GO/G1 and G2M peaks be present in the histogram. The DNA aneuploid population should comprise at least 10% of the nuclei, after correction for aggregates. The degree of DNA content abnormalities was given by the DNA index (DI). Tumors with DI > 1.30 were considered hyperdiploid. Tumors with DI

DNA Ploidy in Papillary Renal Carcinoma . 327

< 0.90 were considered hypodiploid. Tumors with DI between 1.10 and 1.30 were considered near diploid aneuploid (low degree aneuploidy). Statistical analysis A multi-way analysis of variance was performed in order to investigate the prognostic significance of ploidy pattern against tumor stage and grade. According to the DI, as previously described, the tumors were grouped in diploid, near diploid and aneuploid. With regard to the stage and grade, the tumors were grouped as low stage/grade (stage I; grades 1-2) and high stage/grade (stage II, III and IV; grades 3-4). However before performing the Anova analysis, the covariates age and sex were introduced as adjusting factors and their possible influence on survival was statistically evaluated by the MannWhitney rank test for sex and a Pearson correlation analysis for age. Moreover, the Wilcoxon Rank two-sided test was used to evaluate the differences of survival between patients with hyperdiploid and hypodiploid tumors. All statistical computations were performed with the BMDPpackage [l0}. A univariate survival analysis based on the Kaplan-Meier product-limit was also performed to estimate the survival distribution [20} for ploidy pattern.

Table 1. Correlation between nuclear grade and tumor stage of 37 PRCC Stage Nuclear Grade

I

II

III

IV

1 2 3

9 6 0 0

1 1 2 2

0 2 6 1

0 0

4

4

3

Table 2. Correlation between DI, nuclear grade and tumor stage of 37 PRCC Nuclear grade DI

0.90-1.10 1.11-1.30 >1.30 <0.90 M

6 4 0 0 0

Stage

2

3

4

7 2 0 0 0

0 0 8 4 0

0 0 4 1 1

11 4 0 0 0

II

III

IV

0 2 2 1 1

2 0 4 3 0

0 0 6 I

0

Results

M = multiploid tumor

The patients were comprised of 24 men and 13 women. The age ranged from 52 to 78 years (mean 66.3). Tumor size ranged from 2 to 7 cm (mean 3.5). The tumors were characterized by a papillary pattern of growth; in only 2 cases associated tubular areas were observed, even if they occupied less than 25% of the total examined areas. In 12 cases, necrotic and/or hemorrhagic areas were present. The papillary fronds, with thin fibrovascular core, were lined by small cuboidal eosinophilic cells (14 cases) and/or clear cells (Fig. 1). Clear cells were prevalent in 4 cases. We found large macrophages in the papillary stalks in only 2 cases. Psammoma bodies were present in 2 cases. Intratumoral lymphoid infiltrate was present in 15 cases. Follow-up of patients ranged from 60 to 144 months. During this period, 18 (49%) patients were alive and 19 (51 %) died of disease. Among the 37 patients, 15 (41 %) had a tumor in stage I, 6 (16%) in stage II, 9 (24%) in stage III and 7 (18%) in stage IV. Ten tumors (27%) were grade 1,9 (24%) were grade 2, 12 (32.4%) were grade 3 and 6 (16%) were grade 4. In Table 1, cross correlation between stage and grade is presented. 100% of stage I tumors were of low grade. Overall, high grade tumors were of higher stages. In an overall analysis, 13 (35%) tumors showed a diploid pattern and 24 (65%) were aneuploid. Of the 24 aneuploid tumors, 6 were near diploid aneuploid (25%), 12 (50%) were hyperdiploid (01) 1.30),5 (22%) were hypodiploid (01 < 0.90) and 1 (3%) tumor was multiploid, with both hypodiploid and abnormal hyper-

Table 3. Survival of patients with PRCC by tumor stage, nuclear grade and DI Survival (nO of patients) Stage I II III IV

A 15 2 1 0

D 0 4 8 7

4

10 8 0 0

0 1 12 6

DI 0.90-1.10 1.11-1.30 >1.30 <0.90 M

12 6 0 0 0

0 12 5 1

Nuclear Grade

1 2 3

I

A = alive; D = dead; M = multiploid tumor diploid stem lines. In Table 2, the cross correlation between DNA ploidy status and tumor grade/stage is shown. Stage I and low grade tumors were found to be diploid while aneuploid pattern was prevalent in tumors with higher stages and grades. Table 3 presents survival of patients with PRCC in relation to tumor stage, nuclear grade and DNA ploidy

328 . M. T. del Vecchio et aI.

Fig.la-d

DNA Ploidy in Papillary Renal Carcinoma . 329

pattern. None of the nineteen patients who died of disease were stage I, four were stage II, eight were stage III and seven were stage IV. Eighteen of the 19 patients who died had high grade tumors and 100% of the survived patients had low grade tumors. 100% of the patients with DNA aneuploid tumors died because of their disease and 18 of 19 patients with diploid/near diploid

100

I

90

~

70

~ 60 ~

z

w 50

1f ~ w

40

~

30

256

a DI=O.61

())

OJ

()

2

::J

Diploid

~

~ 80

5

ill

Near diploid

DNA content lived disease-free. All five patients with DNA hypodiploid tumors died. In Fig. 2 the survival data of patients with PRCC in relation to ploidy pattern are also given by the KaplanMeier curve. Follow-up of patients with hyperdiploid tumors ranged from 22 to 84 months (mean 46.3), while that of patients with hypodiploid tumors ranged from 23 to 58 months (mean 33.2; p < 0.001). The one-way analysis of variance showed that all three considered variables had a highly significant in-

rJ)

'--

::;;

G 20

Hypodiploid

I

~

V)

C

::J

o

Hyperdiploid

o

10

o

o

12

24

36

48

60 72 84 96 SURVIVAL (months)

108

120

132

144

o

Fig. 2. Survival data by the Kaplan-Meier curves for ploidy pattern.

256 ...,

----,

200

400

600

800

1000

Fluorescence Three Height

256

b

())

OJ

()

rJ)

~

~

rJ)

c

V)

~

5

o

V)

§ o

o

o o

200

400

600

800

1000

Ruorescence Three Height

Fig. 3. Histogram of PRCC showing a hyperdiploid peak.

....

200

400

600

800

1000

Fluorescence Three Height

Fig. 4. Histogram of PRCC showing a hypodiploid peak (a). Diploid peak increased while hypodiploid peak proportionally decreased after non-tumor kidney tissue was added (b).

Fig. 1. a) Fuhrman nuclear grade in PRCC. Grade 1: regular, uniform, round nuclei; DI = 1.00. b) Grade 2: nuclei larger than nuclei of grade I with more variable size and small nucleoli; DI 1.09. c) Grade 3: large nuclei with irregular size and shape and conspicuous nuclei; Dr = 0.77. d) Grade 4: multilobated nuclei with chromatin clumping; Dr = 1.52. Hematoxylin and eosin, x188 and x375.

330 . M. T. del Vecchio et al.

Table 4. Median values of surviyal (months) of 37 patients with PRCC subdivided for stage, grade and ploidy pattern Stage

Survival n° of patients D

Grade

Ploidy Pattern

Low

High

Low

High

D

ND

A

127 15

51 22

122

40

119

130 6

40

19

13

18

18

=diploid tumor; ND =near diploid tumor; A =aneuploid tumor

Table 5. DNA ploidy in 37 case of PRCC Homogeneous DNA-diploid DNA-aneuploid Heterogeneous DNA-diploid/aneuploid DNA-aneuploid (different abnormal stem lines) DNA multiploid

13 12

3 8

fluence on survival. The two-way analysis of variance performed with ploidy pattern and stage/grade, respectively, showed a highly significant main effect of ploidy pattern on survival (p < 0.001; P < 0.00l). No main effect was found for either stage or grade even if the stage had an almost significant p-value (p = 0.056). No twoway interactions were found between ploidy pattern and stage on survival. Due to empty cells, the two-way interactions between ploidy pattern and survival could not be performed nor could a three-way analysis of variance. The effects of the covariates sex and age on survival were highly significant ( p < 0.00l). In Table 4 median values of survival in months of patients subdivided for stage, grade and ploidy pattern are presented. Heterogeneity of DNA content was observed in 12 high grade/stage tumors (32%; Table 5). Areas of different nuclear appearance sometimes produced different

er, Fuhrman et al. [15J reported the contrary. In the study of Lager et al. [30J, an aggressive form of papillary renal carcinoma was strictly correlated with high nuclear grading. Moreover, the papillary renal carcinoma showed characteristics akin to those of other vari-

128

a

01=0.77

01=1.00

o

200

400

600

800

1000

Fluorescence Three Height

256

-r-----------------, b

01=1.00

DJ.

Patterns of DNA aneuploid histograms obtained in this study are shown in Figs. 3 through 5. 01=0.77

Discussion Papillary renal carcinoma is morphologically and genetically distinct from the typical renal cell carcinoma [39 J. In terms of the clinical behaviour, conflicting data have been reported. Their less aggressive potential, asserted by Mancilla Jimenez et al. [35 J, has been sustained in subsequent studies [3, 4, 21, 29, 44J. Howev-

o

200

400

600

800

1000

Fluorescence Three Height

Fig. 5. Histogram of PRCC showing two aneuploid peaks (a). Diploid peak increased while both hyperdiploid and hypodiploid peaks decreased after tumor kidney tissue was added (b).

DNA Ploidy in Papillary Renal Carcinoma . 331

ants in the series presented by El-Naggar et al. [12]. A strong correlation was observed between DNA ploidy status and clinical outcome. We have analyzed the DNA ploidy pattern and clinicopathological parameters in a series of 37 papillary renal carcinoma. For the most part, the demographic data in the present study were similar to those in previous investigations of PRCC. The only exception was the mortality rate, which was higher in our results than in previous reports [12,30]. An aneuploid DNA profile was found in 65% of the tumors, which is higher than that reported by El-Naggar et al. [12]. However, they did not report the DI, while in our series 25% of the aneuploid tumors were near diploid aneuploid (low-degree aneuploidy). In our series, a strong correlation between nuclear grade and stage, as well as between nuclear grade/stage and ploidy pattern was found. Tumors of low nuclear grade were of low stage and diploid/near diploid aneuploid, while high grade tumors were of advanced stages and aneuploid. No correlation between nuclear grade and DNA ploidy was found in the previous study [12]. We have also found a strong correlation between nuclear grade/stage, ploidy pattern and clinical outcome. Patients with high grade/stage tumors and an aneuploid DNA pattern died at the end of the follow-up period, while patients with low grade/stage and diploid/near diploid tumors survived. In our study, it appears that PRCCs can be satisfactorily grouped as diploid/near diploid versus aneuploid. Only one patient with a diploid tumor had a stage III tumor with lymph node metastasis. Highly aneuploid tumors are most aggressive when histologic and cytologic criteria are applied. However, the hypodiploid neoplasms presented a significantly shorter survival than other aneuploid tumors. Interesting results have been found with the multiway analysis of variance though the sample data were limited. The results of this study showed that the ploidy pattern evaluated together with grade and stage had a very high significative influence on survival, while less influence was attributed to grade and stage. Furthermore, the ploidy pattern seems to have no interactions with grade and stage. On the other hand, the one-way analysis of variance showed that, when considered alone, all the three variables had a significative influence on survival. This finding suggests that the ploidy pattern in PRCC seems to be a highly sensitive prognostic index especially in correlation with stage. A homogeneous DNA ploidy pattern was observed in 25 (68%) tumors, while there was intratumoral heterogeneity in 12 (32%). Three tumors presented distinct DNA diploid and DNA aneuploid areas while various abnormal stem lines were observed in 8 aneuploid tumors. Moreover, 1 tumor showed 2 abnormal stem lines with both hypodiploid and hyperdiploid peak. These findings

vary somewhat with the reported data of El-Naggar et al. [12J in PRCC, but are in accordance with the results of Ljungberg et al. [33J and Baretton et al. [2J in RCC. Most aneuploid tumors showed a hyperdiploid pattern with DI > 1.30. Furthermore, 6 (25%) tumors in this study had a DI between 1.10 and 1.30. The latter value has been used to define lower risk patients [39]. The recent application of cytogenetic analysis has shown that malignant papillary tumors exhibit trisomies of chromosomes 16, 12 or 20 in addition to more typical alterations of benign papillary tumors. The latter include trisomies of chromosomes 17 and 7 and loss of the Y chromosome [25, 28]. It is noteworthy that we found a very high percentage of hypodiploid tumors, since this unusual finding has not been previously reported in papillary renal cell carcinoma. Unfortunately, molecular cytogenetic studies were not performed in our study and we cannot correlate the hypodiploidism with specific chromosomal abnormalities. However, it could be suspected that loss of chromosomes might sometimes occur in the late stage of neoplasia [39J, as in our cases. Consistent cytogenetic alterations, namely monosomies of chromosomes 1,6, 14, 15 and 22, were also found in collecting duct renal carcinoma, a tumor with a more aggressive course [16]. In our cases, neither morphological nor cytochemical aspects were consistent with this type of neoplasia. A hypodiploid number of chromosomes also characterizes the chromophobe renal cell carcinoma, generally showing an excellent prognosis [1, 6, 26, 48]. In our series, all patients with a hypodiploid DNA pattern showed a worse prognosis. For the most part, however, a hypodiploid DNA pattern has been said to be associated with a poorer prognosis than other aneuploid types [5,13J. The identification of tumors with a DI < 0.90 is an especially distinct problem in paraffin-embedded tissue. The lack of a known DNA diploid reference population in nuclei from paraffin-embedded tissue makes the identification of these populations difficult. However, efforts should be made to identify such populations since they may be clinically relevant at some tumor sites [47]. It has been recommended that the first peak represents the DNA diploid population in paraffin-embedded material. However, acceptance of the first peak as a normal diploid GO/G 1 may have led to a misinterpretation of the second peak as a hyperdiploid or near diploid GO/G 1 peak, ignoring the hypodiploid population. This may explain why these tumors are quite rare overall. To avoid this particular problem, we always add normal renal tissue obtained from the same block, in order to augment the diploid population after analyzing the tumor-enriched tissue. In fact, the normal tissue component representing the normal counterpart of neoplastic cells is considered the best DNA content stan-

332 . M. T. del Vecchio et al. dard [47]. The technical problems in DNA content assessment of older paraffin-embedded specimens are well known [47]. There is a significant increase in background noise from broken nuclei and cellular debris, and such problems in DNA content analysis become more probable as the sample age increases. Since a large amount of background noise may cause more widespread CV, we rejected the cases in which the CV of the diploid peak was >8%. A routine debris filtering and subtraction was further implemented in the software, allowing us to more correctly classify the DNA histograms. In conclusion, we analyzed the DNA pattern in a series of PRCC. Ploidy pattern was strictly correlated with nuclear grade, pathological stage and prognosis. The multi-way analysis of variance performed to investigate the prognostic significance of ploidy pattern against tumor grade and stage showed the main effect of ploidy pattern. Tumors with diploid/near diploid pattern presented a good prognosis. A high percentage of aneuploid neoplasms had a DI < 0.90, and showed a shorter survival than other aneuploid tumors.

References 1. Akhtar M, Kardar H, Linjawi T, McClintock J, Ali MA (1995) Chromophobe cell carcinoma of the kidney: a clinicopathologic study of 21 cases. Am J Surg Pathol19: 1245-1256 2. Baretton G, Kuhlmann B, Krech R, Lohrs U (1991) Intratumoral heterogeneity of nuclear DNA-content and proliferation in clear cell type carcinomas of the kidney. Virchows Archiv B Cell Pathol61: 57-63 3. Blath RA, Mancilla-Jimenez R, Stanely RJ (1976) Clinical comparison between vascular and avascular renal cell carcinoma. J Uro11l5: 514-519 4. Boczko S, Fromowitz FB, Bard RH (1979) Papillary adenocarcinoma of kidney: a new prospective. Urology 14: 491-495 5. Coulson PB, Thornthwaite JT, Woolley TW, Sugarbaker EV, Seckinger D (1987) Prognostic indicators including DNA histogram type, receptor content, and staging related to human breast cancer survival. Cancer Res 44: 4187-4196 6. Crotty TB, Farrow GM, Lieber MM (1995) Chromophobe cell renal carcinoma: clinicopathological features of 50 cases. J Urology 154: 964-967 7. Dal Cin P, Gaeta J, Huben R, Li F, Prout G, Sandberg A (1989) Renal cortical tumors: cytogenetic characterization. Am J Clin Pathol92: 408-414 8. DeKernion JB, Mukamel E, Ritchie AWS, Blyth B, Hannah J, Bohman R (1989) Prognostic significance of the DNA content of renal carcinoma. Cancer 64: 1669-1673 9. Delahunt B, Ribas n, Nacey IN, Bethwaite PB (1991) Nucleolar organizer regions and prognosis in renal cell carcinoma. J Path 163: 31-37 10. Dixon WJ (1981) BMDP Statistical Software, pp 576-594. University of California Press, California

II. Elfving P, Mandahl N, Lundgren R, Limon J, Bak-Jensen E, Ferno M, Olsson H, Mitelman F (1997) Prognostic implication cytogenetic findings in kidney cancer. Br J Urol

80:698-706

12. EI-Naggar AK, Ro JY, Ensign LG (1993) Papillary renal cell carcinoma: Clinical implication of DNA content analysis. Hum Pathol24: 316-321 13. Ferno M, Baldetorp B, Borg A, Olsson H, Sigurdsson H, Killander D (1992) Flow cytometric DNA index and S-phase fraction in breast cancer in relation to other prognostic variables and to clinical outcome. Acta Oncologica 31: 157-165 14. Flint A, Grossman HB, Liebert M, lloyd RV, Bromberg J (1995) DNA and PCNA content of renal cell carcinoma and prognosis. Am J C1in Pathol103: 14-19 15. Fuhrman SA, Lasky LC, Limas C (1982) Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol6: 655-663 16. Fuzesi L, Cober M, Mittermayer C (1992) Collecting duct carcinoma: cytogenetic characterization. Histopathology 21: 155-160 17. Green LK, Ayala AG, Ro JY, Swanson DA, Grignon DJ, Giacco GG et al (1989) Role of nuclear grading in stage I renal cell carcinoma. Urology 34: 310-315 18. Hedley DW, Friedlander ML, Taylor IW (1985) Application of DNA flow cytometry to paraffin-embedded archivial material for the study of aneuploidy and its clinical significance. Cytometry 6: 327-333 19. Jochum W, Schroder S, AI-Taha R, August C, Gross AJ, Berger J, Padberg BC (1996) Prognostic significance of nuclear DNA content and proliferative activity in renal cell carcinomas. Cancer 77: 514-521 20. Kaplan EL, Meier P (1958) Non parametric estimation for incomplete observation. JAm Stat Assoc 53: 457-481 21. Kesari D, Sibi Y, Yoffe B, Bernstein-Lipshitz L, Segal S (1991) Misdiagnosed papillary renal adenocarcinoma. Br J Uro167: 253-256 22. Kloppel G, Knoffel WT, Raisch H, Otto U (1986) Prognosis of renal cell carcinoma related to nuclear grade, DNA content and Robson stage. Eur Uroll2: 426-431 23. Kovacs G (1989) Papillary renal cell carcinoma: a morphologic and cytogenetic study of 11 cases. Am J Pathol 134:27-34 24. Kovacs G (1990) Application of molecular cytogenetic techniques to the evaluation of renal parenchymal tumors. J Cancer Res Clin Onco1116: 318-323 25. Kovacs G (1993) Molecular differential pathology of renal cell tumors. Histopathology 22: 1-8 26. Kovacs A, Kovacs G (1992) Low chromosome number in chromophobe renal cell carcinomas. Genes Chromosom Cancer 4: 267-268 27. Kovacs G, Wilkens L, Papp T, de Riese W (1989) Differentiation between papillary and nonpapillary renal cell carcinomas by DNA analysis. J Natl Cancer Inst. 81: 527-530 28. Kovacs G, Fuzesi L, Emanuel A, Kung H (1991) Cytogenetics of papillary renal cell tumors. Genes Chromosom Cancer 3: 249-255 29. Krieger IN, Sniderman KW, Seligson GR, Sos TA (1979) Calcified renal cell carcinoma: a clinical, radiographic and pathology study. J Uroll21: 575-580 30. Lager DJ, Huston BJ, Timmerman TG, Bonsib SM (1995) Papillary renal tumors. Cancer 76: 669-673

DNA Ploidy in Papillary Renal Carcinoma . 333 31. Larsson P, Roos G, Stenling R, Ljungberg B (1993) Tumor-cell proliferation and prognosis in renal-cell carcinoma. Int J Cancer 55: 566-570 32. Levin I, Kuperman 0, Goldstein J, Neumann L, Segal S, Kaneti J, Lismer L, Machpud A, Klein B (1991) Cellular ~2 microglobulin expression as a prognostic indicator in renal cell carcinoma. Acta Oncologica 30: 941-945 33. Ljungberg B, Larsson P, Stenling R, Roos G (1991) Flow cytometric deoxyribonucleic acid analysis in stage I renal cell carcinoma. J Uro1146: 697-699 34. Luzi P, Bruni A, Mangavacchi P, Cevenini G, Marini D, Tosi P (1994) Ploidy pattern and cell cycle in breast cancer as detected by image analysis and Flow Cytometry. Cytometry 18: 79-87 35. Mancilla-Jimenez R, Stanley RJ, Blath RA (1976) Papillary renal cell carcinoma: a clinical, radiologic and pathologic study of 34 cases. Cancer 38: 2469-2480 36. Marquardt DI (1963) An algorythm for least squares estimation of non-linear parameters. J Soc Ind Appl Math 2: 431-441 37. Medeiros LJ, Gelb AB, Weiss LM (1988) Renal cell carcinoma. Prognostic significance of morphologic parametersin 121 cases. Cancer 61: 1639-1651 38. Medeiros LJ, Jones EC, Aizawa S, Aldape HC, Cheville JC, Goldstein NS, Lubensky lA, Ro J, Shanks J, Pacelli A, Jung S-H (1997) Grading of renal cell carcinoma Workgroup No 2. Cancer 80: 990-991 39. Motzer RJ, Bander NH, Nanus DM (1996) Renal cell carcinoma (review article). New Engl J Med 335: 865-875 40. Murphy GF, Partin AW, Maygarden SJ, Mohler JL (1990) Nuclear shape analysis for assessment of prognosis in renal cell carcinoma. J Uro1143: 1103-1107 41. Nakano E, Kondoh M, Okatani K, Seguchi T, Sugao H (1993) Flow cytometric analysis of nuclear DNA content of renal cell carcinoma correlated with histologic and clinical features. Cancer 72: 1319-1323 42. Nativ 0, Sabo E, Raviv G, Medalia 0, Moskovitz B, Goldwasser B (1995) The role of nuclear morphometry

43.

44. 45.

46. 47.

48.

49.

50.

for predicting disease outcome in patients with localized renal cell carcinoma. Cancer 76: 1440-1444 Novelli MR, Mac Iver G (1992) Renal cell carcinoma: comparison of morphological and flow cytometric parameters of primary tumour and invasive tumour lying within the renal vein. J Pathol167: 229-233 Orain I, Buzelin F, Ferry N (1987) Tubulopapillary tumors of the kidney. A proposal of 20 new cases and a review of the literature. J Uro193: 1-9 Presti JC, Rao PH, Chen Q, Reuter VE, Li FP, Fair WR, Jhanwar SC (1991) Histopathological, cytogenetic, and molecular characterization of renal cortical tumors. CancerRes51: 1544-1552 Robson C, Churchill B, Anderson W (1969) The result of radical nephrectomy for renal cell carcinoma. J UrollOl: 297-301 Shankey TV, Rabinovitch PS, Bagwell B, Bauer KD, Duque RE, Hedley DW, Mayall BH, Wheeless L (1993) Guidelines for implementation of clinical DNA Cytometry. Cytometry 14: 472-477 Speicher MR, Schoell B, du Manoir S, Schrock E, Ried T, Cremer T, Storkel S, Kovacs A, Kovacs G (1994) Specificlossofchromosomes 1,2,6, 10, 13, 17,and21 inchromophobe renal cell carcinomas revealed by comparative genomic hybridization. Am J Pathol145: 356-364 Tannapfel A, Hahn HA, Katalinic A, Fietkav RJ, Kuhn R, Wittekind CW (1996) Prognostic value of ploidy and proliferation markers in renal cell carcinoma. Cancer 77: 164-171 Tosi P, Luzi P, Baak JPA, Miracco C, Santopietro R, Vindigni C, Mattei FM, Acconcia A, Massay MR (1986) Nuclear morphometry as an important prognostic factor in Stage 1 renal cell carcinoma. Cancer 58: 2512-2518

Received: June 27, 1997 Accepted in revised form: March 16, 1998