Cell Proliferation in Prostate Cancer Patients with Lymph Node Metastasis: a Marker for Progression

2820 Vol. 5, 2820–2823, October 1999 Clinical Cancer Research

Cell Proliferation in Prostate Cancer Patients with Lymph Node Metastasis: A Marker for Progression

Liang Cheng,1 Thomas M. Pisansky, expression in the primary cancer and lymph node metastasis Thomas J. Sebo, Bradley C. Leibovich, was associated with systemic progression-free survival in ؍ Dharamdas M. Ramnani, Amy L. Weaver, univariate analysis (P 0.027 and 0.048, respectively) but was not significant after adjusting for nodal cancer volume and 0.17, respectively). Our data indicate that 0.52 ؍ Beth G. Scherer, Michael L. Blute, Horst Zincke, (P 2 and David G. Bostwick assessment of cell proliferation in nodal metastasis is pre- Departments of Pathology [L. C.] and Urology [L. C.], Indiana dictive of clinical outcome in prostate cancer patients with University School of Medicine, Indianapolis, Indiana 46202 and the regional lymph node metastasis. Division of Radiation Oncology [T. M. P.], the Departments of Pathology [T. J. S., D. M. R., D. G. B.], Urology [B. C. L., M. L. B., H. Z., D. G. B.] and the Section of Biostatistics [A. L. W., B. G. S.], INTRODUCTION Mayo Clinic, Rochester, Minnesota Prostate cancer is the most common noncutaneous malig- nancy and second leading cause of cancer-related mortality in the American male, accounting for 179,300 new cases and ABSTRACT 37,000 deaths in 1999 (1). About 10% of patients have pelvic The biological aggressiveness of lymph node-positive lymph node metastases at the time of diagnosis (2, 3), and many prostate cancer is closely linked to cancer volume in nodal received androgen deprivation (4) with radiotherapy (5) or sur- metastases. We evaluated MIB-1 (Ki-67) labeling index and gery (6–8). Despite the adverse prognostic significance of nodal bcl-2 expression in primary cancer and matched nodal me- involvement, the clinical outcome for these patients is variable tastases from 138 node-positive patients treated with radical (2). Additional factors are needed to more accurately determine prostatectomy and bilateral pelvic lymphadenectomy be- outcome in these patients. In this study, we sought to determine tween 1987 and 1992 at the Mayo Clinic. One hundred whether cell proliferation was associated with patient outcome. twenty-eight patients (93%) received androgen deprivation therapy within 90 days after radical prostatectomy. Mean patient age was 66 years (range, 51–78). The median fol- MATERIALS AND METHODS low-up was 6.7 years (range, 0.03–11). MIB-1 (Ki-67) label- Patients. The study group consisted initially of 269 pa- ing index was determined by digital image analysis, and tients who underwent radical prostatectomy and bilateral pelvic nodal cancer volume was determined by the grid method. lymphadenectomy between January 1987 and December 1992 at Systemic progression, defined as the presence of distant Mayo Clinic (2). All patients had regional lymph node metas- metastasis documented by biopsy or radiographic examina- tasis, and none had clinical evidence of distant metastases at the tion, was used as an outcome end point in the Cox propor- time of surgery. The final study population was limited to 138 tional hazard models. MIB-1 labeling index in nodal metas- patients with sufficient tissue in the primary cancer and metastases was predictive of systemic progression-free survival tasis for immunostaining. One hundred twenty-eight patients The 8-year systemic progression-free survival (93%) received androgen deprivation therapy within 90 days .(0.001 ؍ P) was 100% for those with MIB-1 labeling index <3.5% com- after prostatectomy, and this consisted of orchiectomy or lutein- pared with 78% for those with MIB-1 labeling index >7.8%. izing hormone-releasing hormone agonist with or without an MIB-1 labeling index correlated with Gleason score, DNA antiandrogenic agent. After surgery, patients were evaluated ploidy, and nodal cancer volume (P < 0.001, 0.04, and quarterly for the first 2 years, semiannually for 2 additional 0.001, respectively). After controlling for nodal cancer years, and annually thereafter. Follow-up examinations included < 3 volume, MIB-1 labeling index remained significant in pre- physical examination, serum PSA measurement, chest radiog- bcl-2 raphy, and computerized tomography of the abdomen and pelvis .(0.047 ؍ dicting systemic progression-free survival (P as clinically indicated. A radionuclide bone scan was conducted at least annually or as clinically indicated. Serum PSA was measured using the Hybritech Tandem-R assay (Hybrityech, Inc., San Diego, CA) in all patients. In those followed after Received 4/22/99; revised 7/12/99; accepted 7/15/99. The costs of publication of this article were defrayed in part by the surgery at another institution, PSA determination was performed payment of page charges. This article must therefore be hereby marked at Mayo Clinic by means of a mailed blood specimen; alterna- advertisement in accordance with 18 U.S.C. Section 1734 solely to tively, patients were contacted annually, and additional medical indicate this fact. Supported in part by Grant #IRG-84-002-16 from the information was obtained from the local physician, as necessary American Cancer Society (to. L. C.). 1 To whom requests for reprints should be addressed, at Department of (2, 6–10) Pathology, Indiana University School of Medicine, UH 3465, 550 North University Boulevard, Indianapolis, IN 46202. E-mail: lcheng@ iupui.edu. 2 Present address: Bostwick Laboratories, 6722 Patterson Avenue, Rich- mond, VA 23226. 3 The abbreviation used is: PSA, prostate-specific antigen.

Specimens. The radical prostatectomy and bilateral pel- Table 1 Characteristics of 138 prostate cancer patients with regional vic lymphadenectomy specimens were examined by frozen sec- lymph node metastasis tion at operation and subsequently by permanent sections, as No. of patients (%) described previously (2, 11–13) All histological evaluations Age (years) were performed without knowledge of the clinical outcome. Ͻ60 27 (20) Briefly, the apex and base of the prostate were amputated or 60–69 67 (48) submitted as en face (shave) margins, and the prostate was Ն70 44 (32) serially sectioned perpendicular to the long axis of the gland Preoperative PSA (ng/ml) Ͻ10 26 (20) from the apex to the tip of the seminal vesicles. After gross 10–19 29 (22) examination of the whole prostate slices, frozen sections were Ն20 76 (58) selected to encompass the cancer; the length, width, and height Pathologic stagea T 4 (3) were determined by microscopic examination of frozen sections. 2a T 9 (7) The number of cancer sections submitted for frozen examina- 2b T3a 18 (13) tions from the radical prostatectomy specimens varied from 8 to T3b 107 (77) 20 in this series, depending on cancer volume, prostate volume, Surgical margins and the preference of the pathologist. Approximately 14 prostate Positive 92 (67) blocks were examined per case, and the method of sampling Negative 45 (33) Gleason grade remained constant during the study period (2). Յ6 8 (6) The lymph nodes from the pelvic lymph node dissection 7 74 (54) were totally embedded. The median number of lymph nodes Ն8 56 (41) sampled was 14 (range, 4–33). Nodal cancer volume (size) was DNA ploidy Diploid 68 (50) determined in permanent sections by the grid method (2), and Tetraploid 48 (36) the total cancer volume of all positive nodes (nodal cancer Aneuploid 19 (14) volume) was used for statistical analysis. Number of positive nodes The 1997 Tumor-Node-Metastasis system was used for 1 61 (44) pathological staging (14). Grading of the primary cancer was 2–4 61 (44) Ն5 16 (12) performed according to the Gleason system (15). Prostatectomy Nodal cancer volume (ml)b specimens were examined for DNA ploidy in all patients by Յ0.02 19 (14) flow cytometry with the Hedley technique (16), and DNA 0.02–0.04 38 (28) histograms were classified as diploid, tetraploid, or aneuploid. 0.05–0.19 42 (30) Ն0.20 38 (28) (17) a Immunohistochemistry. For inclusion in the study, ad- Pathological staging according to the 1997 American Joint Com- mittee on Cancer (AJCC) cancer staging (14). equate tissue from both the primary cancer and matched nodal b Nodal cancer volume was measured by the grid method (2). metastases were required for immunostaining. Representative blocks of the primary cancer and paired lymph node metastases were selected from the same patient for immunostaining. Stain- ing was performed on 6-␮m, formalin-fixed, paraffin-embedded for univariate associations of variables with systemic progres- sections using the avidin-biotin complex technique. Primary sion-free survival. Analysis of association of continuous vari- monoclonal antibodies were used for evaluation of MIB-1 la- ables with survival was performed using single degree of free- beling index (Ki-67; Immunotech, Westbrook, ME; dilution, dom (linear) terms in the Cox model. Nodal cancer volume was 1:50) and bcl-2 expression (Dako, Carpinteria, CA; dilution analyzed on the log base-2 scale because it was positively 1:20). 3,3-Diaminobenzidine was used as the chromogen, and skewed. To avoid bias, no arbitrary cutoff points were chosen 0.2% methyl green was used as the counterstain. Quantification for continuous variables in the Cox regression analysis (18). of MIB-1 (Ki-67) labeling was performed using the CAS 200 Continuous variables were split into three even groups (MIB-1) digital image analyzer and proliferation index software pro- for the purpose of illustration and estimating survival rates at 8 grams (Becton Dickinson, Cellular Imaging Systems, San Jose, years. The associations of MIB-1 labeling index and bcl-2 CA). At least 1000 cells were analyzed in each case. Immuno- expression with clinical and pathological findings were assessed Ͻ reactivity for bcl-2 was evaluated semiquantitatively on a 5% using the Spearman rank correlation coefficient. P 0.05 was incremental scale ranging from 0 to 90%. All immunostains considered significant, and all Ps were two-tailed. were evaluated without knowledge of clinical outcome. Positive and negative controls were run in parallel and gave appropriate RESULTS results. Patient characteristics are shown in Table 1. The mean Statistical Analysis. The number of cancer-specific patient age was 66 years (range, 47–79), and the median fol- deaths (8 events) was low; therefore, we focused on clinical low-up was 6.7 years (range, 0.03–11). Twelve patients had systemic progression as a surrogate outcome variable. Progres- distant metastasis (systemic progression), 8 died of prostate sion was defined as the presence of distant metastasis docu- cancer, and 21 died of other causes. There was no significant mented by biopsy or radiographic examination as the end point difference in age, preoperative PSA concentration, surgical mar- for analysis. Survival was estimated using the Kaplan-Meier gin status, DNA ploidy, nodal cancer volume, systemic progres- method. The Cox proportional hazards model was used to test sion-free survival, and the length of follow-up between patients

Fig. 2 Kaplan-Meier curves (systemic progression-free survival) ac- Fig. 1 MIB-1 immunostaining in the lymph node metastasis from cording to MIB-1 labeling index in the lymph node metastasis. To avoid prostate cancer. bias, no arbitrary cutoff points were chosen for continuous variables in the Cox regression analysis, and patients were divided into three even groups according to MIB-1 labeling index for the purpose of illustration. Numbers within parentheses represent number of patients still under observation at 3, 5, and 7 years. in the study group and those excluded from study because of lack of adequate tissue for immunostaining (data not shown). However, the study group had higher mean Gleason grade and more positive nodes than those excluded (P Ͻ 0.01). DISCUSSION The median MIB-1 labeling index was 8.4% (range, In this report, we found that cell proliferation of nodal 0–67%) in primary cancer and 5.3% (range, 0–30%) in nodal metastases determined by digital image analysis of MIB-1 la- metastases. To avoid bias, no arbitrary cutoff points were cho- beling index was significantly associated with systemic progres- sen for continuous variables in the Cox regression analysis (18). sion-free survival in prostate cancer patients with regional Continuous variables were split into three even groups (MIB-1) lymph node metastasis, and this finding was independent of for the purpose of illustration and estimating survival rates at 8 nodal cancer volume. Patients with higher MIB-1 labeling in- years. MIB-1 labeling index in nodal metastases was predictive dices had worse prognosis than those with a lower index. Fur- of systemic progression-free survival (P ϭ 0.001); the 8-year thermore, MIB-1 labeling index in nodal metastases was correlated with Gleason score, DNA ploidy, and nodal cancer systemic progression-free survival was 100% for those with volume. Our data indicate that cell proliferation in nodal me- MIB-1 labeling index Ͻ3.5% compared with 78% for those with tastasis is closely associated with the biological behavior of MIB-1 labeling index Ͼ7.8% (Fig. 1–2). MIB-1 labeling index prostate cancer and that digital image analysis of MIB-1 label- in nodal metastases was correlated with Gleason score (P Ͻ ing index in nodal metastases provides useful prognostic infor- 0.001), DNA ploidy (P ϭ 0.04), and nodal cancer volume (P Ͻ mation regarding patient outcome. 0.001). After controlling for nodal cancer volume, MIB-1 label- In a previous study, we found that nodal cancer volume ing index in node metastases remained significant in predicting was the best predictor of systemic progression-free survival in systemic progression-free survival (P ϭ 0.047). MIB-1 labeling prostate cancer patients among several classical clinical and ϭ index in primary cancer correlated with Gleason score (P pathological factors (2). The risk of distant metastasis in patients 0.008). MIB-1 labeling index in primary cancer was also mar- with regional metastasis increased proportionally with increas- ϭ ginally associated with systemic progression-free survival (P ing nodal cancer volume when treated by radical prostatectomy 0.060); however, this difference was not significant after adjust- and immediate androgen deprivation (2). Cher et al. (19) dem- ϭ ing for nodal cancer volume (P 0.178). onstrated that lymph node-positive patients with lower cellular The median bcl-2 expression was 5.0% (range, 0–90%) in proliferative fraction in lymph node metastases had significant primary cancer and 5.0% (range, 0–90%) in nodal metastases. survival advantage over those with higher proliferative fraction, bcl-2 expression in nodal metastasis correlated with nodal can- and our results are similar. However, in their study, histological cer volume (P ϭ 0.02) and was associated with systemic pro- grade and pathological stage of the primary cancer were not gression-free survival in univariate analysis (P ϭ 0.048) but not significant in predicting outcome, suggesting that the behavior after adjusting for nodal cancer volume (P ϭ 0.176). Among of lymph node-positive prostate cancer was more closely linked patients with elevated bcl-2 levels (Ͼ5%) in nodal metastasis, to the biological aggressiveness of the metastatic cancer rather 83% of patients were free of systemic progression at 7 years, than the primary cancer (19). They hypothesized that patients whereas 92% with lower bcl-2 levels (Յ5%) were free of with a high proliferative fraction in lymph node metastases have systemic progression (P ϭ 0.048). bcl-2 expression in the pri- increased nodal cancer volume and were more likely to have mary cancer was associated with systemic progression-free sur- distant metastasis with decreased survival than those whose vival in univariate analysis (P ϭ 0.027) but not in multivariate cancer had a low proliferative rate. analysis (P ϭ 0.52). There was an association between bcl-2 In addition to cellular proliferation, other factors may be expression in primary cancer and DNA ploidy (P Ͻ 0.001). important in cell cycle regulation, tumor progression, and ther-

apeutic response. For example, the bcl-2 oncoprotein inhibits 8. Cheng, C. W., Bergstralh, E. J., and Zincke, H. Stage D1 prostate apoptosis and prolongs the duration of cell survival. Its overex- cancer. A nonrandomized comparison of conservative treatment options pression may be a predictor of clinical outcome in patients with versus radical prostatectomy. Cancer, 71: 996–1004, 1993. localized prostate cancer (20, 21), and it may be associated with 9. Ghavamian, R., Bergstralh, E. J., Blute, M. L., Slezak, J., and Zincke, H. Radical retropubic prostatectomy plus orchiectomy versus orchiec- emergence of the androgen-independent state (22–24). In the tomy alone for pTxNϩ prostate cancer: a matched comparison. J. Urol., present study, bcl-2 expression was associated with other pre- 161: 1223–1228, 1999. viously established prognostic factors (e.g., nodal cancer vol- 10. Seay, T. M., Blute, M. L., and Zincke, H. Long-term outcome in ume), but an independent association with systemic progression- patients with pTxNϩ adenocarcinoma of prostate treated with radical free survival was not demonstrated conclusively. However, the prostatectomy and early androgen ablation. J. Urol., 159: 357–364, 1998. statistical test of associations was of borderline significance 11. Cheng, L., Slezak, J., Bergstralh, E. J., Cheville, J. C., Sweat, S., ϭ Zincke, H., and Bostwick, D. G. Dedifferentiation in metastatic pro- (P 0.17) for patient outcome, which suggested that further gression of prostate cancer, Cancer, 86: 657–663, 1999. study of this factor may be of value. 12. Cheng, L., Leibovich, B. C., Bergstralh, E. J., Scherer, B. G., The present study may have potential limitations. Although Pacelli, A., Ramnani, D. M., Zincke, H., and Bostwick, D. G. p53 this relatively large population of node-positive patients had alteration in regional lymph node metastases from prostate carcinoma: a extended follow-up, the number of outcome events (i.e., distant marker for progression? Cancer, 85: 2455–2459, 1999. metastases) were limited. Consequently, some variables that 13. Cheng, L., Bostwick, D. G., Li, G., Wang, Q., Hu, N., Vortmeyer, were inconclusive because of limited statistical power may A. O., and Zhuang, Z. Allelic imbalance in the clonal evolution of prostate cancer. Cancer, 85: 2017–2022, 1999. attain statistical significance if the sample size is increased. We 14. Fleming, I. D., Cooper, J. S., Henson, D. E., Hutter, R. V. P., used clinical (systemic) progression as an end point in this Kennedy, B. J., Murphy, G. P., O’Sullivan, B., Sobin, L. H., and Yarbro, study; however, most studies use biochemical progression as an J. W. AJCC Cancer Staging Manual, pp. 241–246. Philadelphia: Raven end point. The clinical utility of biochemical progression for and Lippincott, 1997. evaluation of treatment outcome is yet to be established (25, 26). 15. Bostwick, D. G. Neoplasms of the prostate. In: D. G. Bostwick and At the Mayo Clinic, most lymph node-positive patients treated J. N. Eble (eds.), Urologic Surgical Pathology, pp. 343–422. St. Louis: Mosby, 1997. with radical prostatectomy also underwent immediate adjuvant 16. Hedley, D. W., Friedlander, M. L., Taylor, I. W., Rugg, C. A., and hormonal therapy (2, 6–10). Therefore, our findings strictly Musgrove, E. A. Method for analysis of cellular content of paraffin- apply to this treatment strategy, and the results reported herein embedded pathological material using flow cytometry. J. Histochem. should be interpreted cautiously when compared with other Cytochem., 31: 1333–1335, 1983. studies with different therapeutic regimens. 17. Cheng, L., Sebo, T. J., Slezak, J., Pisansky, T. M., Bergstralh, E. J., In summary, the biological aggressiveness of lymph node- Neumann, R. M., Iczkowski, K. A., Zincke, H., and Bostwick, D. G. positive prostate cancer is closely linked to cancer volume in the Predictors of survival in prostate cancer patients treated with salvage radical prostatectomy after radiation therapy. Cancer, 83: 2164–2171, 1998. nodal metastases. However, the assessment of cellular proliferation 18. Altman, D. G., Lausen, B., Sauerbrei, W., and Schumacher, M. in nodal metastasis provides significant prognostic information in Dangers of using “optimal” cutpoints in the evaluation of prognostic prostate cancer patients treated by radical prostatectomy, bilateral factors. J. Natl. Cancer Inst., 86: 829–835, 1994. pelvic lymphadenectomy, and androgen deprivation. 19. Cher, M. L., Stephenson, R. A., James, B. C., and Carroll, P. R. Cellular proliferative fraction of metastatic lymph nodes predicts survival in stage D1 (TxNϩM0) prostate cancer. J. Urol., 155: 1674–1677, 1996. REFERENCES 20. Bauer, J. J., Sesterhenn, I. A., Mostofi, F. K., McLeod, D. G., Srivas- 1. Landis, S. H., Murray, T., Bolden, S., and Wingo, P. A. Cancer tava, S., and Moul, J. W. Elevated levels of apoptosis regulator proteins p53 statistics, 1999. CA Cancer J. Clin., 49: 8–31, 1999. and bcl-2 are independent prognostic biomarkers in surgically treated clin- 2. Cheng, L., Bergstralh, E. H., Cheville, J. C., Slezak, J., Corica, F. A., ically localized prostate cancer. J. Urol., 156: 1511–1516, 1996. Zincke, H., Blute, M. L., and Bostwick, D. G. Cancer volume of lymph 21. Bubendorf, L., Sauter, G., Moch, H., Jordan, P., Blochlinger, A., node metastasis predicts progression in prostate cancer. Am. J. Surg. Gasser, T. C., and Mihatsch, M. J. Prognostic significance of Bcl-2 in Pathol., 22: 1491–1500, 1998. clinically localized prostate cancer. Am. J. Pathol., 148: 1557–1565, 1996. 3. Zincke, H., Bergstralh, E. J., Larson-Keller, J. J., Farrow, G. M., 22. McDonnell, T. J., Troncoso, P., Brisbay, S. M., Logothetis, C., Myers, R. P., Lieber, M. M., Barrett, D. M., Rife, C. C., and Gonchoroff, Chung, L. W., Hsieh, J. T., Tu, S. M., and Campbell, M. L. Expression N. J. Stage D1 prostate cancer treated by radical prostatectomy and of the protooncogene bcl-2 in the prostate and its association with adjuvant hormonal treatment. Evidence for favorable survival in patients emergence of androgen-independent prostate cancer. Cancer Res., 52: with DNA diploid tumors. Cancer (Phila.), 70: 311–323, 1992. 6940–6944, 1992. 4. Millikan, R., and Logothetis, C. Update of the NCCN guidelines for 23. Colombel, M., Olsson, C. A., Ng, P. Y., and Buttyan, R. Hormone- treatment of prostate cancer. Oncology, 11: 180–193, 1997. regulated apoptosis results from reentry of differentiated prostate cells 5. Lawton, C. A., Winter, K., and Byhardt, R. Androgen suppression onto a defective cell cycle. Cancer Res., 52: 4313–4319, 1992. plus radiation versus radiation alone for patients with D1 (pNϩ) ade- 24. Colombel, M., Symmans, F., Gil, S., O’Toole, K. M., Chopin, D., nocarcinoma of the prostate: results based on a national prospective Benson, M., Olsson, C. A., Korsmeyer, S., and Buttyan, R. Detection of randomized trials, RTOG 85–31. Int. J. Radiol. Oncol. Biol. Phys., 38: the apoptosis-suppressing oncoprotein bc1–2 in hormone-refractory hu- 931–939, 1997. man prostate cancers. Am. J. Pathol., 143: 390–400, 1993. 6. Myers, R. P., Larson-Keller, J. J., Bergstralh, E. J., Zincke, H., 25. Eisenberger, M. A., and Nelson, W. G. How much can we rely on the Oesterling, J. E., and Lieber, M. M. Hormonal treatment at time of level of prostate-specific antigen as an end point for evaluation of clinical radical retropubic prostatectomy for stage D1 prostate cancer: results of trials? A word of caution! J. Natl. Cancer Inst., 88: 779–781, 1996. long-term followup. J. Urol., 147: 910–915, 1992. 26. Schellhammer, P., Cockett, A., Boccon-Gibod, L., Gospodarowicz, 7. Zincke, H. Extended experience with surgical treatment of stage D1 M., Krongrad, A., Thompson, I. M., Scardino, P., Soloway, M., and adenocarcinoma of prostate. Significant influences of immediate adjuvant Adolfsson, J. Assessment of endpoints for clinical trials for localized hormonal treatment (orchiectomy) on outcome. Urology, 33: 27–36, 1989. prostate cancer. Urology, 49: 27–38, 1997.

Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 1999 American Association for Cancer Research. Cell Proliferation in Prostate Cancer Patients with Lymph Node Metastasis: A Marker for Progression

Liang Cheng, Thomas M. Pisansky, Thomas J. Sebo, et al.

Clin Cancer Res 1999;5:2820-2823.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/5/10/2820

Cited articles This article cites 23 articles, 2 of which you can access for free at: http://clincancerres.aacrjournals.org/content/5/10/2820.full#ref-list-1

Citing articles This article has been cited by 4 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/5/10/2820.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/5/10/2820. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.