Patterns of positive surgical margins after surgery in prostate cancer patients and their association with clinical recurrence.

Introduction and objective: We aimed to report the different patterns of positive surgical margins (PSMs) in Prostate cancer (PCa) patients treated with radical prostatectomy (RP) in and their association with various scenario of clinical recurrence (CR).

Material and methods: We retrospective evaluated 644 consecutive patients submitted to open RP between 1995 and 2016 with PSMs at final pathology and complete clinical and pathological data. Our analyses include only patients with minimum follow up of 3 years (n= 494). PSMs patterns were categorized as burden (focal vs multifocal), site (apical-anterior vs posterolateral vs base-bladder neck vs multiple) and side (unilateral vs bilateral). CR was categorized as local, nodal and systemic (including skeletal, visceral and multiple). Kaplan-Meier analyses assessed CR-free survival (CR-FS) and cancer specific mortality free survival (CSM-FS). Uni and Multivariate Cox regression analyses were performed to assess the predictors of CR and various scenario of CR.

Results: PSMs were categorized as focal and multifocal in 56.3% and 43.7%, as apical-anterior, posterolateral, base-bladder neck and multiple in 21.9%, 26.1%, 3.8% and 48.2%, as unilateral and bilateral in 75.5% and 24.5%, respectively. At median follow up of 7.8 years, the scenario of CR was local, nodal and systemic in 50.6%, 9.4% and 40%, respectively. At 10-year follow-up, the CR-FS rates were 80.8% and 46.1% in overall population and after biochemical recurrence (Figure 1a and 1b). Patients with unilateral PSMs experienced higher CR-FS rates as referred to those with bilateral PSMs (87.1% vs 71.3%, p<0.001, Figure 1d). The CSM-FS rates after CR was 59.6% (Figure 1e) and patients with systemic CR experienced lower CSM-FS rates as compared to those with local recurrence (54.1% vs 68.8%, p=0.02; Figure 1f). At univariate analysis, PSMs side was able to predict CR (HR: 2.53, p=0.001) and only systemic recurrence (HR: 5.8, p<0.001) after stratifying according to site of CR (Table 1). At multivariable analyses, pathologic Gleason score 8-10, pathologic stage ≥T3b and adjuvant therapy status represent independent predictors of CR (Table 2, all p≤0.04), while patterns of PSMs were not found to be independent predictors of CR.

Conclusions: PSMs patterns, namely unilateral vs. bilateral, could be a useful tool to predict CR and the different scenario of recurrence in patients with PSMs at RP in order to select patients to proper adjuvant treatments.