ABSTRACT
Summary: The Gleason score obtained on prostatic biopsies is a key parameter in the management of localized prostate cancer.
Objectives: We conducted this study to evaluate the correlation between Gleason scores (GS) obtained on prostate biopsies and radical prostatectomy to establish the accuracy of biopsy grading in the prediction of final grades.
Materials and Methods: Forty patients with localized prostate cancer diagnosed between 2000 and 2010, and treated with radical prostatectomy, were included in this study. All patients underwent transrectal needle biopsies (TRNB) with at least 12 cores. Gleason scores on biopsies and radical prostatectomy specimens were determined and the concordance between the 2 scores was then evaluated. Histological grading using the conventional Gleason grading system (3 groups) and the modified Gleason grading system (5 groups) was also performed. The correlation between grades obtained on biopsies and radical prostatectomy specimens was also evaluated.
Results: The mean age of patients was 61.1 years, with a mean PSA value of 10 ng/ml. In 32.5% of cases, the biopsy’s Gleason score correlated with the one obtained on radical prostatectomy. Using the conventional Gleason grading system, the correlation highly improved, with 62.5% of patients remaining in the same group after radical prostatectomy. However, using modified Gleason grading, the correlation was slightly improved and estimated at 37%.
Conclusion: In this study we have noticed that the accuracy of Gleason scores determined by transrectal needle biopsy in patients with prostate cancer seems unreliable. The classification of patients into 3 distinct groups (well, moderately, and poorly differentiated tumors) increases the concordance between the biopsy GS and the definitive GS, but the modified Gleason grading system seems to be more precise and better reflects the Gleason score.
Sallami Satáa, Ines Chelly, Amira Ben Salem, Hanen Chorfi, Haifa Nfoussi, Nidhameddine Kchir
Submitted June 11, 2012 - Accepted for Publication July 9, 2012
KEYWORDS: Prostate carcinoma, prostate biopsy, radical prostatectomy, pathology, Gleason score
CORRESPONDENCE: Sallami Satáa, MD, La Rabta Hospital-University, Tunis, Tunis, Tunisia ()
CITATION: UroToday Int J. 2012 December;5(6):art 57. http://dx.doi.org/10.3834/uij.1944-5784.2012.12.02
INTRODUCTION
Prostate cancer is an increasingly frequent pathology, and it is considered the second most frequent urinary cancer in Tunisian men [1]. After the clinical or biochemical suspicion of prostate cancer, the diagnosis can be histologically confirmed on transrectal needle biopsy. Moreover, the degree of tumor differentiation is determined by establishing the Gleason score (GS). This score is obtained by combining the primary and secondary patterns for a number score from 2 to 10 [2,3]. It is one of the most powerful predictors of biological behavior, and it is one of the most influential factors used to determine treatment for prostate cancer and a choice of external radiotherapy, brachytherapy, cryotherapy, or radical prostatectomy [5]. However, in some cases, the GS established on transrectal needle biopsy could be different from the one for radical prostatectomy, which has a very important impact on prognosis and therapeutic options [3,4,5,6].
In this retrospective study, we evaluate the correlation between Gleason scores obtained on prostate biopsies and radical prostatectomy to establish the accuracy of biopsy grading in the prediction of final grades.
MATERIALS AND METHODS
We retrospectively analyzed the medical records of 40 patients who had undergone radical prostatectomy for prostate cancer, from January 2000 to February 2010.
Exclusion criteria included patients who underwent prostate biopsies who showed a high or low PIN grade, patients using 5-alpha-reductase inhibitors, and patients with a positive biopsy for prostate cancer that were not candidates for/refused radical prostatectomy. Inclusion criteria included patients who underwent transrectal needle biopsy with at least 12 cores and were candidates for radical prostatectomy.
The diagnosis was made on transrectal needle biopsy (TRNB), which was performed on patients with elevated serum PSA and/or abnormal digital rectal examination findings. Patients underwent at least 12 core biopsies (from the apex, the middle, and the base of the prostate, with 4 cores for each site). Additional target biopsies were performed when needed (prostatic nodes; hypoechogenic lesions). The same senior pathologist examined all the biopsies. Various preoperative biopsy findings were recorded, including the primary and secondary Gleason pattern grade from each positive biopsy site. The Gleason score was then established by combining the highest Gleason score among all biopsy sites and the Gleason score from the site with the highest tumor volume on the needle biopsy. The tumor was then classified using the conventional Gleason grading system with 3 categories: Well-differentiated tumors (score: 2 to 4), moderately differentiated tumors (score: 5 to 7), and poorly differentiated tumors (score: 8 to 10) [2]. The modified Gleason grading system included 5 categories: Well-differentiated cancer (score: 2 to 6), moderately differentiated (score: 3 + 4 = 7), Moderately/poorly differentiated (score: 4 + 3 = 7), poorly differentiated (score: 8), and undifferentiated (score: 9 to 10) [7]. The 2002 TNM staging system of the American Joint Committee on Cancer (AJCC) was used for clinical staging.
All these patients underwent an open, retropubic radical prostatectomy. For each patient, we evaluated these parameters: Serum PSA level, clinical stages, biopsy data, and postoperative criteria (pathological TNM stage 2002, Gleason score). The highest GS established on the TRNB and radical prostatectomy specimens were compared. A “downgrade” was defined as the Gleason score for the prostatectomy specimen greater than that of the biopsy specimen, whereas an “upgrade” was defined as the opposite.
Statistical analysis was performed using Student t test for continuous variables, Pearson's chi-square test, and Fisher's exact test for categorical variables. The software used for statistical analysis was SPSS version 12.0 (SPSS Inc., Chicago, IL, USA). A P value of less than 0.05 was considered statistically significant.
RESULTS
There were 40 patients with a mean age of 61.1 ± 4.4 years (52 to 69) and a mean serum PSA level of 10 ± 6.3 ng/ml (1.9 to 26.3). For the TRNB, the mean GS established was 6.52 (4 to 9). Among the 40 patients, 1 (2.5%) had a GS of 4 (2 + 2), 1 (2.5%) had a GS of 5 (2 + 3), 26 (65%) had a GS of 6 (3 + 3), 1 (2.5%) had a GS of 7 (3 + 4), 2 (5%) had a GS of 7 (4 + 3), 6 (15%) had a GS of 8 (4 + 4), 1 (2.5%) had a GS of 9 (5 + 4), and 2 (5%) had a GS of 9 (4 + 5) (Table 1).
For the radical prostatectomy specimens, the mean GS was 6.87 (5 to 9). One case (2.5%) had a GS estimated at 5 (2 + 3), 16 cases (40%) had a GS estimated at 6 (3 + 3), 13 cases (32.5%) had a GS estimated at 7 (9 patients exhibited a pattern of 3 + 4, and 4 patients exhibited a pattern of 4 + 3), 7 cases (17.5%) had a GS estimated at 8 (2 cases showed a pattern of 3 + 5 and 5 cases a pattern of 4 + 4), and 3 cases (7.5%) had a GS estimated at 9 (2 cases exhibited a pattern of 4 + 5 and 1 case had a pattern of 5 + 4) (Table 1).
For the group of 26 patients with TRNB GS estimated at 6, 11 cases (42.3%) had identical Gleason scores on the needle biopsy and prostatectomy specimens, whereas 14 cases (53.8%) were under-graded (9 cases with radical prostatectomy GS of 7, 4 cases with a GS of 8, and 1 case with a GS of 9), and 1 case (3.8%) was over-graded (GS on radical prostatectomy estimated at 5) (Table 2). For the group of 7 patients with TRNB GS estimated at 7, concordance with radical prostatectomy specimen GS was observed in only 1 case. There was an under-grade in 1 case (GS estimated at 9 after radical prostatectomy) and an over-grade in 1 case (GS 6 after radical prostatectomy). For the group of 6 patients with TRNB GS estimated at 8, there was concordance with radical prostatectomy specimen GS in only 1 case, an over-grade in 4 cases (1 case with GS estimated at 7 and 3 cases with GS estimated at 6 after radical prostatectomy), and an under-grade in 1 case (GS estimated at 9 after radical prostatectomy). For the 3 patients with TRNB GS estimated at 9, there was an over-grade in all cases (2 case with GS estimated at 6 and 1 case with GS estimated at 8 after radical prostatectomy) (Table 2).
Overall, the preoperative biopsy predicted the prostatectomy Gleason score accurately in 13 cases (32.5%), with the highest accuracy rates obtained for Gleason scores of 6. The Gleason scores were discordant in 27 cases (67.5%), with an under-grade in 45% of cases and an over-grade in 22.5% of cases. The difference between the 2 GS in these cases was not significant (P = 10.6).
Using the conventional Gleason grading system to classify the results obtained on TRNB, this series accounted for 1 well-differentiated tumor, 30 moderately differentiated tumors, and 9 poorly differentiated tumors (Table 3). Comparing the results of TRNB using this classification to the results of radical prostatectomy specimens showed that the concordance highly improved, estimated at 62.5% (N = 25). There was an under-grade in 20% (N = 8) and an over-grade in 17.5% (N = 7) (Table 4). The difference between the 2 GS in different groups was also not significant (P = 1.0).
The majority of tumors were classified as moderately differentiated (score: 5 to 7) on TRNB (75%) as well as on radical prostatectomy specimens (72.5%). The highest accuracy rate was noted in this group (73.3%). Using the modified Gleason grading system for classification, the series of biopsies accounted for 28 well-differentiated, 3 moderately differentiated, 6 poorly differentiated, and 3 undifferentiated tumors (Table 5). The concordance according to the 5 groups was 35% (N = 14), under-staging in 20% (N = 8), and over-staging in 45% (N = 18) (Table 6). The difference between the 2 GS in the 5 groups was not significant (P = 0.40). The majority of patients using this new classification also had well-differentiated tumors (score: 2 to 6) on biopsy (70%) and prostatectomy specimens (45%). The highest accuracy rate was also noted in this group (46.4%) (Table 6).
DISCUSSION
In clinical practice, treatment decisions for prostate cancer depend on many factors, including prostatic biopsy GS, preoperative PSA, and clinical stages [8]. The Gleason score is a system based on tumor architecture. It is determined by adding the scores of the 2 most represented volumetric quotas. The course of action is based on the hypothetical correlation between Gleason score of the biopsy and that of the prostatectomy specimen. Gleason score represents the histological assessment of the different cellular differentiation degrees in prostate cancer. The most frequent tumor patterns are regarded as primary grade while those that are next in frequency are regarded as secondary grade. The sum of both patterns enables classification on the basis of differentiation and a good prognosis of patterns (a good prognosis and good cellular differentiation yields a GS of 2 to 5, a medium prognosis and moderate cellular differentiation a GS of 6 to 7, and a bad prognosis and bad cellular differentiation a GS of 8 to 10) [2,9]. Recent studies have focused on the discordance in GS between TRNB and radical prostatectomy specimens. Global discordance varies from 31 to 72% [3,5,6,9,10,11,12,13,14,15,16,17,18] (Table 5). Under-staging ranges from 26 to 46% [3,5,6,9,10,11,12,13,16,19,20]. Our results accord with those reported in medical literature.
Several factors could explain this discordance and must be taken into consideration when the treatment decision is based on Gleason score:
- The prostatic biopsies are heterogeneous and do not fully reflect the real tumor architecture.
- The reproducibility of the GS is low during iterative analysis by the same or by other pathologists [2,21].
- Other factors, including PSA level, prostate volume, size of the tumor mass, and preoperative clinical stage, are known to be associated with discordance of the GS between TRNB and radical prostatectomy [22,23].
The clinical implication of this discordance is very important. Indeed, an under-grade on biopsy should not have any impact on treatment decisions, especially if the patient does not change groups. However, in some cases, the therapeutic attitude may be different between 2 very close GS. For example, a patient with a GS of 8 will not be a candidate for prostatectomy while those with a GS of 7 may benefit from surgery. Therefore, many studies focused on different factors causing this discordance between TRNB GS and RP GS and the ways to overcome these obstacles.
Because of the heterogeneity of prostatic biopsies, some series studied the impact of increasing the number of biopsy cores to decrease the discrepancy of the 2 GS. Kahl et al. found that an extended biopsy scheme with cores numbering more than 12 can help to acquire more accurate GS than a biopsy scheme with cores numbering less than 12 [24]. Moon [25] suggests that at least 13 cores should be taken during TRNB and more aggressive biopsies should be performed at the apex and of hypoechogenic lesions for more precise GS, preoperatively. San Francisco et al. [26] compared 2 groups of patients with a different number of biopsy cores who underwent radical prostatectomy: A group of 340 patients (biopsies with 9 cylinders or less) and a group of 126 patients (biopsies with 10 cylinders or more). They found a concordance with 67 and 76% of cases, respectively. Thus, they concluded that increasing the number of biopsy cylinders better defines the final GS after radical prostatectomy. Divrik et al. [27] also showed that an extended biopsy scheme beyond its superior diagnostic capability also improves the concordance of Gleason scores of needle biopsies and radical prostatectomy specimens. However, Thickman found that there is no interest in increasing the number of carrots to improve the prediction of the GS [15].
Arrabal-Polo et al. [28], comparing 2 groups of patients who underwent radical prostatectomy after sextant prostate biopsy in 98 patients (64% correlation) and 12 cylinder in 30 patients (60% correlation), also didn’t find a significant difference in GS correlation.
For the other factors causing the GS discordance, Nayyar et al. reported that upgrading the GS was not associated with a high PSA level [29], whereas Dong et al. reported that a PSA level greater than 5.0 ng/ml is associated with upgrading the GS [30]. Freedland et al. and Tilki et al. also reported that a high PSA level is associated with upgrading the GS [31,32]. In our study, however, we did not determine the PSA level to be a predictive parameter of upgrading.
Another reported factor involved in GS discordance is the prostate volume. Moussa et al. reported that smaller prostate volume is associated with upgrading the GS [33]. Furthermore, Turley et al. reported that patients with a prostate volume < 20 cm³ had more than 5 times the risk of upgrading compared with patients with a prostate volume > 60 cm³ [34]. In Moon’s study [25], patients with a prostate volume < 36.5 cm³ had a higher risk of upgrading after radical prostatectomy. Different results have been found by different studies, and until now the accuracy of GS established on TRNB is still problematic, especially since therapeutic attitude is often based on this GS.
The largest discrepancies in GS were found for well-differentiated tumors [14,21,35,36], as in our study. It is worth mentioning that patients with a GS of 7 (78%) show the highest percentage of concordance, patients with a GS of 6 or less show the highest percentage of under-staging, and cases with a GS of 8 to 10 (35%) show the highest percentage of over-staging [6]. Garnett found a concordance of 30% and noted that the prediction of the GS is more accurate when the scores are high [37]. Prost et al. [18] reported the same conclusions; the correlation was 54, 73, and 100%, respectively, for groups of tumors well, moderately, and poorly differentiated. The same conclusions were reported by Köksal and Rodríguez Faba [9,38].
In our study, the concordance between TRNB GS and RP GS was 32.5% for a GS of 6 and 7, and the correlation was around 70%, a similar value to that seen in Algaba’s and Bostwick’s studies [10,16,35]. In contrast to the other studies [13,18,37], in our series, the over-grading was more frequent for a higher GS, estimated at 66.6% for a GS of 8 and 100% for a GS of 9.
The classification of patients into 3 distinct groups (well, moderately, and poorly differentiated tumors) allows better concordance between the TRNB results and RP results. Using this conventional Gleason grading system, the correlation in our series was highly improved with 62.5% of patients remaining in the same group after radical prostatectomy. Similar results were found by Cookson et al. [17] who reported a high correlation of 80% between biopsy and final specimen using this classification. This correlation was of 72.6% in the Prost [18] series. Table 7 resumes the correlation of Gleason scores between prostatic biopsy and radical prostatectomy specimens in the literature. In our series we also studied the correlation between the TRNB results and RP results using the modified Gleason grading system with 5 groups of tumor differentiation. We found that the correlation did not significantly improve (in only 35% of cases) with this new classification.
There are many limitations in our study. First, this is a single institute, retrospective study dealing with a relatively small population. Second, this study did not identify the relation between “no concordance” and clinical and biochemical factors such as the PSA level, the free PSA, and prostate volume. Third, because the biopsy strategy was not consistent and because of the number of urologists, we are not sure that the biopsy specimens were taken at exactly the same location in every patient.
And finally, the biopsies and the radical prostatectomy specimens were examined by only 1 pathologist, which allows a uniform interpretation of the results but depends also on the pathologist’s experience. Our result implies that more than half of the patients diagnosed with a GS of 6 by TRNB had more aggressive cancer than they seemed to have. Therefore, even if patients are diagnosed with moderate-risk prostate cancer after TRNB, physicians should always consider parameters that can be predictive of over-grading. This awareness affects treatment policy, particularly the watchful waiting criteria.
CONCLUSIONS
In this study, we have noticed that the accuracy of Gleason scores determined by transrectal needle biopsy in patients with prostate cancer seems to be very low. The greatest correlation percentage is found in tumors showing a moderated degree of cellular differentiation (GS of 5 to 7), whereas the over-grading was more frequent for a higher GS.
The classification of patients into 3 distinct groups (tumors well, moderately, and poorly differentiated) increases the correlation between the biopsy GS and final score. The modified Gleason grading system didn’t reflect the definitive GS. In practice, physicians must be aware of the limits of the GS established on biopsy and clinical and biochemical parameters must be used with the biopsy GS to provide better information concerning prognosis, and the most adapted therapeutic option for the patient and his prostatic carcinoma.
REFERENCES
- Ben-Abdallah, M., S. Zehani, et al. (2009). “1994-2003 evolution and projections by 2024.” North Tunisia Cancer Registry: 1999-2003: 25-31.
- Gleason, D. F. and G. T. Mellinger (1974). “Prediction of prognosis for prostatic adenocarcinoma by combined histological grading and clinical staging.” J Urol 111(1): 58-64.
PubMed - Montesino Semper, M., J. Jimenez Aristu, et al. (2004). “Correlation between Gleason score on prostate biopsies diagnostic of adenocarcinoma and radical prostatectomy specimens.” Arch Esp Urol 57(5): 519-523.
PubMed - Partin, A. W., G. D. Steinberg, et al. (1992). “Use of nuclear morphometry, gleason histologic scoring, clinical stage, and age to predict disease-free survival among patients with prostate cancer.” Cancer 70(1): 161-168.
PubMed ; CrossRef - Kvale, R., B. Moller, et al. (2009). “Concordance between Gleason scores of needle biopsies and radical prostatectomy specimens: a population-based study.” BJU Int 103(12): 1647-1654.
PubMed ; CrossRef - Rajinikanth, A., M. Manoharan, et al. (2008). “Trends in Gleason score: concordance between biopsy and prostatectomy over 15 years.” Urology 72(1): 177-182.
PubMed ; CrossRef - Epstein, J. I. (2011). “Update on the Gleason grading system.” Ann Pathol 31(5 Suppl): S20-26.
PubMed ; CrossRef - D’Amico, A. V., R. Whittington, et al. (1998). “The combination of preoperative prostate specific antigen and postoperative pathological findings to predict prostate specific antigen outcome in clinically localized prostate cancer.” J Urol 160(6 Pt 1): 2096-2101.
PubMed - Rodriguez Faba, O., J. M. Fernandez Gomez, et al. (2003). “Assessment of the Gleason score in biopsies and specimens of radical prostatectomy.” Arch Esp Urol 56(7): 781-784.
PubMed - Algaba Arrea, F., A. Chivite de Leon, et al. (2004). “Evidence of the radical prostatectomy Gleason score in the biopsy Gleason score.” Actas Urol Esp 28(1): 21-26.
PubMed - Altay, B., A. Kefi, et al. (2001). “Comparison of Gleason scores from sextant prostate biopsies and radical prostatectomy specimens.” Urol Int 67(1): 14-18.
PubMed ; CrossRef - Noguchi, M., T. A. Stamey, et al. (2001). “Relationship between systematic biopsies and histological features of 222 radical prostatectomy specimens: lack of prediction of tumor significance for men with nonpalpable prostate cancer.” J Urol 166(1): 104-109; discussion 109-110.
PubMed ; CrossRef - Salomon, L., A. Hoznek, et al. (1998). “Value of biopsies in staging of prostatic cancer before radical prostatectomy.” Prog Urol 8(6): 969-976.
PubMed - Spires, S. E., M. L. Cibull, et al. (1994). “Gleason histologic grading in prostatic carcinoma. Correlation of 18-gauge core biopsy with prostatectomy.” Arch Pathol Lab Med 118(7): 705-708.
PubMed - Thickman, D., W. C. Speers, et al. (1996). “Effect of the number of core biopsies of the prostate on predicting Gleason score of prostate cancer.” J Urol 156(1): 110-113.
PubMed ; CrossRef - Bostwick, D. G. (1994). “Gleason grading of prostatic needle biopsies. Correlation with grade in 316 matched prostatectomies.” Am J Surg Pathol 18(8): 796-803.
PubMed ; CrossRef - Cookson, M. S., N. E. Fleshner, et al. (1997). “Correlation between Gleason score of needle biopsy and radical prostatectomy specimen: accuracy and clinical implications.” J Urol 157(2): 559-562.
PubMed ; CrossRef - Prost, J., N. Gros, et al. (2001). “Correlation between Gleason score of prostatic biopsies and the one of the radical prostatectomy specimen.” Prog Urol 11(1): 45-48.
PubMed - Fleshner, N. E., M. O’Sullivan, et al. (1997). “Prevalence and predictors of a positive repeat transrectal ultrasound guided needle biopsy of the prostate.” J Urol 158(2): 505-508; discussion 508-509.
PubMed ; CrossRef - Lange, P. H. and P. Narayan (1983). “Understaging and undergrading of prostate cancer. Argument for postoperative radiation as adjuvant therapy.” Urology 21(2): 113-118.
PubMed ; CrossRef - Yang, X. J., K. Lecksell, et al. (1999). “Significance of small foci of Gleason score 7 or greater prostate cancer on needle biopsy.” Urology 54(3): 528-532.
PubMed ; CrossRef - Kim, S. C., J. H. Hong, et al. (2009). “Predictive factors for upgrading or upstaging in biopsy Gleason score 6 prostate cancer.” Korean J Urol 50: 836-842.
CrossRef - Montironi, R., R. Mazzucchelli, et al. (2006). “Prostate carcinoma II: prognostic factors in prostate needle biopsies.” BJU Int 97(3): 492-497.
PubMed ; CrossRef - Kahl, P., S. Wolf, et al. (2009). “Saturation biopsy improves preoperative Gleason scoring of prostate cancer.” Pathol Res Pract 205(4): 259-264.
PubMed ; CrossRef - Moon, S. J., S. Y. Park, et al. (2010). “Predictive factors of Gleason score upgrading in localized and locally advanced prostate cancer diagnosed by prostate biopsy.” Korean J Urol 51(10): 677-682.
CrossRef - San Francisco, I. F., W. C. DeWolf, et al. (2003). “Extended prostate needle biopsy improves concordance of Gleason grading between prostate needle biopsy and radical prostatectomy.” J Urol 169(1): 136-140.
PubMed - Divrik, R. T., A. Eroglu, et al. (2007). “Increasing the number of biopsies increases the concordance of Gleason scores of needle biopsies and prostatectomy specimens.” Urol Oncol 25(5): 376-382.
PubMed ; CrossRef - Arrabal-Polo, M. A., A. Jimenez-Pacheco, et al. (2010). “Relationship between biopsy Gleason score and radical prostatectomy specimen Gleason score in patients undergoing sextant vs 12 core biopsies.” Arch Esp Urol 63(9): 791-796.
PubMed - Nayyar, R., P. Singh, et al. (2010). “Upgrading of Gleason score on radical prostatectomy specimen compared to the pre-operative needle core biopsy: an Indian experience.” Indian J Urol 26(1): 56-59.
CrossRef - Dong, F., J. S. Jones, et al. (2008). “Prostate cancer volume at biopsy predicts clinically significant upgrading.” J Urol 179(3): 896-900; discussion 900.
PubMed ; CrossRef - Freedland, S. J., C. J. Kane, et al. (2007). “Upgrading and downgrading of prostate needle biopsy specimens: risk factors and clinical implications.” Urology 69: 495-499.
PubMed ; CrossRef - Tilki, D., B. Schlenker, et al. (2011). “Clinical and pathologic predictors of Gleason sum upgrading in patients after radical prostatectomy: results from a single institution series.” Urol Oncol 29(5): 508-514.
PubMed ; CrossRef - Moussa, A. S., J. Li, et al. (2009). “Prostate biopsy clinical and pathological variables that predict significant grading changes in patients with intermediate and high grade prostate cancer.” BJU Int 103(1): 43-48.
PubMed ; CrossRef - Turley, R. S., R. J. Hamilton, et al. (2008). “Small transrectal ultrasound volume predicts clinically significant Gleason score upgrading after radical prostatectomy: results from the SEARCH database.” J Urol 179(2): 523-527; discussion 527-528.
PubMed ; CrossRef - Catalona, W. J., A. J. Stein, et al. (1982). “Grading errors in prostatic needle biopsies: relation to the accuracy of tumor grade in predicting pelvic lymph node metastases.” J Urol 127(5): 919-922.
PubMed - Mills, S. E. and J. E. Fowler, Jr. (1986). “Gleason histologic grading of prostatic carcinoma. Correlations between biopsy and prostatectomy specimens.” Cancer 57(2): 346-349.
PubMed ; CrossRef - Garnett, J. E., R. Oyasu, et al. (1984). “The accuracy of diagnostic biopsy specimens in predicting tumor grades by Gleason’s classification of radical prostatectomy specimens.” J Urol 131(4): 690-693.
PubMed - Koksal, I. T., F. Ozcan, et al. (2000). “Discrepancy between Gleason scores of biopsy and radical prostatectomy specimens.” Eur Urol 37(6): 670-674.
PubMed ; CrossRef