Gleason Score at Surgical Margin Is Not an Independent Predictor of Biochemical Recurrence after Radical Prostatectomy
Swetha Paluru, Viacheslav Iremashvili, Saleem A Umar, Soum Lokeshwar, Murugesan Manoharan, Ramgopal Satyanarayana, Mark S Soloway, Merce Jorda. University of Miami , Jackson Memorial Hospital, Miami; University of Miami, Jackson Memorial Hospital, Miami
Background: A positive surgical margin is known to increase the risk of prostatic adenocarcinoma (PA) recurrence after surgery. Gleason score at surgical resection margin (GSM) may independently affect postoperative outcome; this association however has not been confirmed. The objective of this study is to retrospectively analyze the association between GSM and biochemical recurrence (BCR) after radical prostatectomy (RP).
Design: We identified 154 RPs with positive surgical resection margin (SRM) and different Gleason patterns amounting to scores of 7 (3+4, 4+3) and 8 (3+5 and 5+3). Since GSM may impact BCR, these cases were selected for their disparate Gleason patterns. All slides with positive margins were reviewed and the GSM was recorded for each case. The BCR-free survival in patients with same overall Gleason score (GS) and different GSM was estimated using the Kaplan-Meier method and results were compared with the log-rank test. To determine if the GSM has independent predictive value with regards to biochemical outcome, we also compared the predictive performance of two multivariate Cox regression models. One model comprised traditional pathologic and clinical variables, including pathologic stage and GS, lymph nodes status, visually estimated percent of carcinoma, length of surgical margin and preoperative PSA. The other model added GSM to the first model variables. The predictive performance of these two models was quantified using the Harrell's c-index.
Results: Thirteen (12%) of 109 patients with PA GS 7 (3+4) and GS 8 (3+5) had GSM of 7 (4+3) or 8 (5+3). Twenty-eight (62%) of 45 patients with PA GS 7 (4+3) and GS 8 (5+3) had GSM of 7 (3+4). Over a median follow-up of 4.2 years, 52 (34%) patients had BCR. No difference in BCR-free survival was found between the two models. GSM was not significantly associated with BCR-free survival in the multivariate analysis (p=0.285). A marginal increase in predictive performance of the multivariate model was noted after inclusion of GSM (c-index changed from 0.689 to 0.691).
Conclusions: GSM is not associated with biochemical outcome after radical prostatectomy in both univariate and multivariate analyses. GSM does not add predictive accuracy to the standard model containing established prognostic factors. We found no evidence to support the inclusion of GSM in routine pathologic reporting of radical prostatectomy specimens.
Category: Genitourinary (including renal tumors)
Wednesday, March 21, 2012 9:30 AM
Poster Session V # 114, Wednesday Morning