P R E D I C T I O N O F I N S I G N I F I C A N T P R O S T A T E C A N C E R I N M E N W I T H S T A G E T I C D I S E A S E
Kam il Çam, M .D . / Levent Tü rk e ri, M .D . / Atıf Akdaş, M .D .
D e p a rtm e n t o f U rology, S c h o o l o f M e d icin e , M arm ara U n ive rsity, Is ta n b u l, Turkey.
ABSTRACT
Objective: In this retrospective study, we evaluated the ability of preoperative parameters to predict tumor significance in men with stage T1c disease who underwent radical retropubic prostatectomy.
Patients and Methods: A total of 26 consecutive patients who underwent radical retropubic prostatectomy owing to clinically localized disease fulfilled the criteria for insignificant cancer namely PSA density lower or equal to 0.10, presence of Gleason grade <4 (Gleason sum <6), number of cores (<3) involved with cancer, and tumor percentage less than 50% for insignificant prostate cancer. Additionally, we only included the patients with a PSA level of less than 10 ng/ml.
Results: According to pathological examination, 6 (23%) of these patients had bilateral (pT2c) disease, 5 (19.2%) had extracapsular disease (pT3a-b), 2 (7.7%) had seminal vesicle (pT3c) involvement and 2 (7.7%) had positive surgical margins. Consequently, at least 50% of these cases were significant cancer.
Conclusion: In conclusion, our data demonstrated that the patient and the physician must be aware of the variability and insufficiency of these preoperative algorithms for predicting insignificant cancers in discussing treatment alternatives.
K ey W o rd s: Prostate cancer, Stage T1c disease, Insignificant tumor.
IN TRODUCTION
There is an ongoing considerable amount of debate on prostate cancer screening and treatment of localized disease (1). Many physicians and organizations suggest early diagnosis because prostate cancer is the cause of more than 40.000 deaths in USA annually (2), and it is second to lung cancer as the leading cause of male cancer deaths. On the other hand, autopsy studies have shown a relatively high rate of latent cancer, ranging from 10% at the age of 50 years to 80% by the age of 80, raising concern over detection of insignificant cancers and risk of overtreatment (3). In other words, 30 to 40% of men over 50 years of age have prostate cancer, but only 8% of these become clinically significant (4). Nevertheless, effective treatment is only available for the localized disease and the significant number of deaths claimed by prostate cancer. Therefore, the use of prostate specific antigen (PSA) for early diagnosis has gained popularity throughout the world. This widespread use of PSA has resulted in earlier diagnosis and treatment in the natural course of the disease (5). However, some of these tumors may be unimportant clinically, needing no treatment at all and they are denoted as insignificant tumors which indicates that they will not cause any disease specific symptoms. In summary, the prevalence of prostate cancer far exceeds its
Kamil Qam, et al
morbidity and mortality. Therefore methods are needed to distinguish insignificant manageable cancers from clinically important cases that require aggressive treatment. Several studies defined clinically insignificant prostate cancers based on the final pathologic stage, cancer volume, surgical margin status, and Gleason grade (2, 6). It is generally accepted that these are the ones with tumor volume less than 0.5 cm3 and confined to the prostate with no primary or secondary Gleason grade 4 to 5 (7). Consequently, up to 20% of stage T1c tumors (nonpalpable tumors diagnosed on the basis of an elevated PSA) were proposed to be insignificant (8). Then several models were tested in the prediction of insignificant cancers (9-10). Epstein's definition for predicting an insignificant tumor was PSA density 0.10 or less, no Gleason grade of 4 or 5 in the biopsy specimen, fewer than 3 biopsy cores involved, and no core with over 50% cancer involvement. In this retrospective study, we evaluated the ability of these preoperative parameters to perdict tumor significance in men with stage T1c disease who had undergone retropubic prostatectomy.
PATIENTS A N D M ETHODS
A total of 26 consecutive patients who had undergone radical retropubic prostatectomy owing to clinically localized disease, fulfilled the Epstein's insignificant cancer criteria, namely PSA density lower or equal to 0.10, presence of Gleason grade <4 (Gleason sum <6), number of cores (<3) involved with cancer, and tumor percentage less than 50% (9). In addition, we only included patients with a PSA level between 4 to 10 ng/ml. All patients had nonpalpable tumors. Transrectal ultrasound (TRUS) guided biopsies were done as an outpatient procedure under antibiotic prophylaxis. Ultrasound guided transrectal systematic biopsies consisting of 4 cores (base, midportion, apex of peripheral zone and transitional zone) from each half of the prostate were performed. On the other hand, if any hypoechoic lesion was present, lesion directed biopsies were also performed in addition to the systematic biopsies. After the histological diagnosis of prostate cancer, metastatic disease was ruled out by bone
scan, chest-X-ray, and abdominopelvic computerized tomography. Local staging was performed with digital rectal examination and TRUS.
All patients then underwent bilateral pelvic lymphadenectomy and radical retropubic prostatectomy. The histopathological examination of surgical specimens was performed according to the step-sectioning technique by the same pathololgy department. The pathological stage was compared with preoperative parameters. Then, we assessed these cases based on parameters, namely tumor volume less than 0.5cm3 and confined to the prostate with no primary or secondary Gleason grade 4 to 5 that describe insignificant cancer after radical prostatectomy (7).
RESULTS
The mean age of the patients was 64 (57-70). Preoperative mean PSA value was 6.1 ng/ml (4- 9.87) and median Gleason score was 4 (2-6). The ability of pretreatment criteria based on a combination of PSA density (<0.10), Gleason grade (<4), number of cores (<3) involved with cancer and percentage (<50%) of cancer within the core to predict the insignificant cancer was assessed. All of these cases fulfilled these preoperative parameters. The pathological data for the 26 patients who had undergone radical retropubic prostatectomy are shown in table I.
Table I. Pathological d a t a of 2 6 radical p r o sta te c to m y c a s e s . Number of patients (%) Pathological stage pT2a NO 10(38.4% ) pT2b NO 3 (11.5%) pT2c NO 6 (23%) pT3a NO 3(11.5% ) pT3b NO 2 (7.7%) pT3c NO 2 (7.7%) Surgical margin Positive 2 (7.7%) Gleason Score 3-4 9 (34.6%) 5-6 11 (42.3%) 7 4 (15.3%) 8-9 2 (7.7%)
Prediction of insignificant prostate cancer
We reassessed these cases based on just basic parameters, namely pathological stage (tumors confined to the prostate) and no primary or secondary Gleason grade 4 to 5 that describe insignificant cancer after radical prostatectomy (7). First of all, there was a remarkable discrepancy bétween grades of TRUS biopsy and radical prostatectomies in these patients. While the concordance rate was only 30.6%, most cases (67.7%) were upgraded after the surgery. Most importantly, 6 (23%) of these patients had bilateral (pT2c) disease, 5 (19.2%) had extracapsular disease (pT3a-b) 2 (7.7%) had seminal vesicle (pT3c) involvement and 2 (7.7%) had positive surgical margins. Consequently, at least 50% of the cases were significant cancer, by definition.
DISCUSSION
Prostate cancer is the leading cancer in men, and also it is the second cause of death from cancer. There is no effective treatment for metastatic disease, while definitive treatment of organ confined disease is available. Therefore, there is a significant need for methods to either reduce the incidence or provide early detection when definitive treatment is possible in order to prevent deaths from prostate cancer. Of these, only the latter is currently available at hand through radical prostatectomy. Consequently, a significant increase has been observed in the detection of clinically localized prostate cancer through extensive use of PSA (5). This in turn has resulted in a remarkable increase in the number of radical prostatectomies throughout the world. Meanwhile, an ongoing debate has been raised about whether the PSA screening causes the detection of insignificant cancers or not? It is known that stage T1c cancers diagnosed by moderate elevations of PSA are the most frequent cancers detected by screening programs (11). Most studies clearly demonstrated that the majority of stage T1c cancers were significant. Epstein et al reported 49% of extracapsular extension with a 17% of positive surgical margin for stage T1c disease (9). Similarly, Oesterling et al revealed the comparable properties between stage T1c (nonpalpable) and stage T2 (palpable) tumors in terms of Gleason grade, tumor volume, and DNA polidy (12). However, a nonnegligible proportion
(7 to 20%) of stage T1c tumors were reported to be clinically insignificant and had been potentially overtreated. To overcome this drawback several models based on preoperative parameters have been proposed in the prediction of insignificant cancers. Epstein and Goto specifically constructed their models by comparing systematic biopsies and corresponding radical prostatectomy specimens (9,10). They reported that the preoperative parameters were able to predict 73% and 75% of insignificant cancers, respectively. In the present study, we fonud that preoperative parameters including PSA level less than 10 ng/ml as an additional parameter were insufficient for the prediction of insignificant cancer based on our data. Categorically, 6 (23%) of these patients had bilateral (pT2c) disease, 5 (19.2%) had extracapsular disease (pT3a-b), 2 (7.7%) had seminal vesicle (pT3c) involvement and 2 (7.7%) had positive surgical margins. At least 50% of cases were misidentified by these criteria and treated inappropriately. If we reviewed the results of the previous studies, although this algorithm will miss only 3% of important cancers, there is a significant risk of misidentification of insignificant cancers (9). The authors also stated that their algorithm was based on their data and may not be transferable to a different population. Weldon et al reported that 94% of 33 patients who seemed to be insignificant based on preoperative parameters, in fact harbored significant tumors (stage pT3 in 52% and pT2 in 30%) (13). Similarly, Kojima et al reported the insufficiency of these parameters in predicting insignificant cancer (14). Gardner et al reported that of the 83 men with microfocal prostate cancer on biopsy, 75 (90%) had clinically significant disease after radical prostatectomy (15). They stated that preoperative variables were of no use in identifying patients with insignificant cancers. On the other hand, Epstein also suggested repeat biopsies (including transition zone biopsies) and serial PSA measurements for those cancers predicted to be insignificant based on these preoperative variables in the case of conservative management (16). Aditionally, he advised considering the age, comorbidity and treatment preferences of the patient. Most of the stage T1c tumors are undoubtedly significant. In one study, 40% of stage T1c cancers and PSA 4 to 10ng/ml as a comparable population with our study group were pathologically stage T3 (17). In
Kamil Çam, et al
order to exclude the small but not negligible proportion of insignificant stage T1 diseases, according to our results the current preoperative models are insufficient. First of all the insignificance of a cancer is related to complex and clearly undetermined factors which should include host features (age, general status and so forth) and tumor characteristics (doubling time and most probably distinct biological behaviour of the particular tumor). On the other hand, current models rely on essentially TRUS biopsy findings. But, it is obvious that in TRUS biopsy, very limited amount of tissue which may not reflect the actual properties of the definitive specimen can be obtained. Fernandes et al have shown 20.7% of correlation and 74.6% of upgrading in Gleason grades on TRUS biopsy after prostatectomy for well differentiated tumors in a review of 466 radical prostatectomy cases (18). Similarly, we found the highest discrepancy between grades of TRUS biopsy and radical prostatectomies in patients with well differentiated tumors (Gleason 2-4) in a total of 75 radical prostatectomies (unpublished data). While the concordance rate was only 28.6% with upgrading after the surgery in 64.3% in the well differentiated group, 70.7% of patients in the intermediate group and 83.3% in the poorly differentiated group remained within the same groups after the surgery. In the intermediate group (Gleason 5-7) the rate of upgrading was 21.9% and downgrading was 7.3%. In patients with poorly differentiated tumors (Gleason 8-10) on TRUS biopsy only 16.6% of the cases were downgraded in comparison to grades on radical prostatectomy specimens. In this particular study group, the upgrading rate was 67.7%, and with respect to the upgrading alone, 23% of the men actually had significant (Gleason sum >6) tumors based on the final pathology. This inconsistency between the grades of the TRUS biopsy and prostatectomy specimens is firstly thought to be due to the fact that histological assessment may be difficult in the limited amount of tissue obtained by the biopsy. The other explanation is that this limited amount of tissue may not reflect the final grading on a prostatectomy specimen. Prostate cancer has intense tumor heterogeneity and more than two grades can be found in the radical prostatectomy specimen. It is shown that more than 50% of cancers consist of two or more
patterns and the average of different Gleason grade per case was postulated to be 2.7 (19-20). Therefore, a sampling insufficiency in which the core needle may miss the tissue of higher grade found in the prostatectomy specimen is possible. Finally, interpretation of the specimen needs experience and the reproducibility of the Gleason system between or within observers considering the difficulty in the assessment of tumors that bridge two grades may account for some of the dissimilarity. In conclusion, when deciding upon different treatment alternatives, the Gleason score on TRUS biopsy should be interpreted with caution. It is especially important to know that a well differentiated tumor on a TRUS biopsy is a poor predictor of a well differentiated tumor after surgical treatment. The treatment algorithms for prostate cancer should not be concluded upon biopsy histology alone and the restrictions of Gleason grading based on a TRUS biopsy should be considered when discussing treatment options. Similarly, a sampling insufficiency may also be possible in the determination of tumor percentage. In addition to this, it has been suggested that measurement of the volume of the cancer in the TRUS biopsy specimen cannot reflect the actual tumor volume found in the radical prostatectomy specimen (21).
In conclusion, our results showed the insufficiency and unreliability -or at least non transferability to our population- of these preoperative parameters for the prediction of insignificant cancers. Therefore, the patient and physician must be aware of the variability and insufficiency of these algorithms for predicting insignificant cancers in discussing treatment alternatives. According to our results, the decision to observe especially young patients with prostate cancer depending on these parameters carries a considerable risk of progression and would be potentially life threatening. Therefore, we need more sensitive ways that would enhance our ability to identify insignificant cancers. These tests should be primarily based on the differentiation of distinct biological behaviour of the particular tumors and in this regard molecular studies and genetic markers may play an important role in the near future.
Prediction of insignificant prostate cancer
REFERENCES
1. Kram er BS, Brown ML, Prorok PC, Potosky AL, Gohogan JK. Prostate cancer screening: what we know and what we do not know. Ann Intern Med 1993,1 19:914-923.
2. Wingo PA, Tong T, Boiden S. Cancer statistics. CA Cancer J Clin 1995;45:8-30.
3. Bostwick DG, Cooner Wli, Denis L, et al. The association o f benign prostatic hyperplasia
and cancer o f the prostate. Cancer
1992;70:291-301.
4. Stamey TA, Preiha PS, MclTeal JE, et al. Localized prostate cancer: relationship o f tu m o r volum e to clinical significance fo r the tre a tm e n t o f p ro sta te cancer. Cancer 1993;71:933-938.
5. Catalona WJ, Smith DS, R atliff TL, Basler JW. Detection o f organ confined prostate cancer is increased through prostate specific antigen based screening. JAMA 1993;270:948-954. 6. Epstein JL Walsh PC, Carmicheal M, Brendler
CB. Pathologic and clinic findings to predict tu m o r extent o f nonpalpable (stage Ti c) prostate cancer. JAMA 1 9 94;2 71 .-368-374 . 7. O hori M, Wheeler TM, Dunn JK, et al. The
p a th o lo g ic a l features and prognosis o f p ro sta te cancer detectab le with current diagnostic tests. J Urol 1994; 152:1714-1720. 8. G ibod LB. E ditorial: S ignificant versus insignificant prostate cancer-Can we identify the tigers from pussy cats? J Urol 1996; 156:
1069-1070.
9. Epstein JL Walsh PC, Brendler CB. Radical p ro s ta te c to m y fo r im palpable prostate cancer. The John Hopkin s experience with tum ours found on transurethral resection (stages T la and Ti b) and on needle biopsy (stage Tic). J Urol 1994,152:1721-1729. 10. Goto Y, O h o ri M, Arakawa A, et al.
D istinguishing c lin ic a lly im p o rta n t from u n im p o rta n t pro sta te cancers before treatment. Value o f systematic biopsies. J Urol 1996,156:1059-1063.
11. H um phrey PA, Keetch DW, Sm ith DS, Shepherland DL, Catalona WJ. Prospective
characterization o f pathological features o f prostatic carcinomas detected via prostate sp ecific antigen based screening. J Urol
1996;155:816-820.
12. Oesterling JE, Su man WJ, Zincke H, Bostwick DG. PSA detected (clinical stage T ic o r BO) prostate cancer. Urol Clin n o rth Am er
1993;20:687-693.
13. Weldon VE, Travel PR, Meuwirth H, Cohen R. Failure o f focal prostate cancer on biopsy to predict. Focal prostate cancer the importance o f prevalence. J Urol 1995;154:1074-1077. 14. Kojim a M, Troncoso P, Babaian RJ. Use o f
prostate specific antigen and tum or volume in p re d ictin g needle biopsy grading error.
Urology 19 9 5 ;4 5 :8 0 7 -8 12.
15. G ardner TA, Lem er ML, Schlegel PH, Waldbaum ED Jr, et al. Microfocal prostate cancer biopsy cancer volume does not predict actual tum or volume. Br J Urol 1998;98:839- 843.
16. Epstein JL Can insignificant prostate cancer be predicted preoperatively in men stage T1 disease? Semin Urol O ncol 1996; 14:165-1 73. 17. Lerner SE, Seay TM, Blute ML, Bergstralh EJ,
Barrett D, Zincke H. Prostate specific antigen detected prostate cancer (clinical stage Tic): an interim analysis. J Urol 1996; 155:821-826. 18. Fernandes ET, Sundaram CP, Long R, Sol tan i
M, Ercole CJ. Biopsy Gleason score: how does it correlate with the fin a l patholo gical diagnosis in prostate cancer? Brit J Urol
1997;79:615-617.
19. Gleason DF. Histologic grading o f prostate
cancer: a perspective. Hum Pathol
1992;23:273-279.
20. Aihara M, Wheeler TM, O hori M, et al. Heterogeneity o f prostate cancer in radical
p ro sta te cto m y specim ens. Urology
1994;43:60-66.
21. Cupp MR, Bostwick DG, Myers RP, Oesterling JE. The volume o f prostate cancer in the biopsy specim en cannot reliably predict the quantity o f cancer in the radical prostatectoy specim en on an in d ivid u a l basis. J Urol
