Presentation Authors: Dong Hoon Koh*, Jongsoo Lee, Hyeok Jun Goh, Hyun Ho Han, Ji Eun Heo, Seoul, Korea, Republic of, Jae Won Park, Young Sig Kim, Goyang, Korea, Republic of, Dong Hoon Lim, Gwangju, Korea, Republic of, Joong Shik Lee, Seoul, Korea, Republic of, Rae Cho In, Goyang, Korea, Republic of, Won Sik Jang, Won Sik Ham, Young Deuk Choi, Seoul, Korea, Republic of
Introduction: Men with positive surgical margin (PSM) at radical prostatectomy (RP) have higher rates of biochemical recurrence (BCR). The role of adjuvant radiotherapy (ART) for patients with PSM after radical prostatectomy has been shown to prevent BCR, but overtreatment remains a significant concern. The present study, therefore, sought to identify men with PSM at higher risk for BCR after RP in the absence of ART. We developed a postoperative nomogram with these identified risk factors
Methods: We retrospectively reviewed 2301 men who underwent RP for prostate cancer at our institution between 2001 and 2010. The patients who received neoadjuvant or adjuvant therapy and patients without adequate medical records were excluded. Among these patients, 977 patients with positive surgical margins were analyzed. The Kaplan-Meier method was used to estimate and compare BCR-free survival (BCRFS). Cox regression models were used to investigate predictors of BCR. Nomogram predictor variables included preoperative prostate specific antigen (PSA), pathologic T stage, RP Gleason score (GS) and tumor volume. The concordance index was used as an accuracy measure, with bootstrapping to correct for optimistic bias. Calibration plots were constructed.
Results: The median age was 66.0 (interquartile range [IQR]: 62âˆ’70), the median time of follow-up was 78.0 months (IQR: 64âˆ’95). In the univariate and multivariate analysis, preoperative PSA (hazard ratio [HR] 1.465, p < 0.001), pathologic T stage â‰¥ T3 (HR 1.393, p = 0.010), RP GS7 (HR 1.961, p < 0.001), RP GS â‰¥ 8 (HR 4.554, p < 0.001), and tumor volume â‰¥ 5cc (HR 1.469, p = 0.001) were significant predictors for BCR after RP. A nomogram was constructed on the basis of a Cox regression model. The bootstrap-corrected concordance index was 0.713.
Conclusions: Our study demonstrated the heterogeneity of oncologic outcomes in patients with PSM. Stratifying men with PSM by prognostic factors such as preoperative PSA, pathological T stage, RP GS, and tumor volume will improve identification of those at higher risk for BCR that should be strongly considered for ART. A nomogram was developed to predict 5-year BCRFS after RP in PSM patients. This tool should be useful for determining whether to perform ART in PSM patients.