Impact of surgical margin length on biochemical recurrence after radical prostatectomy: Results from a multicenter cohort

Background: Although positive surgical margins predict biochemical recurrence (BCR) after radical prostatectomy, the prognostic value of surgical margin length (SML) remains unclear. We aimed to quantify the SML’s impact on BCR, characterize temporal recurrence patterns, and assess the incremental clinical value beyond established risk factors. Materials and methods: We retrospectively analyzed 4659 patients with pT2N0 status undergoing radical prostatectomy at 3 tertiary centers. The SMLs were categorized as negative, <1 mm, or ≥1 mm. Multivariable Cox regression, time-dependent receiver operating characteristic curve, decision curve analysis, logistic regression for early (≤24 months) versus late recurrence, and landmark analyses were performed. Subgroup analyses according to International Society of Urological Pathology grade, age, and prostate-specific antigen level were used to assess effect modification. Results: Positive surgical margins occurred in 436 patients (9.4%), including 282 (64.7%) and 154 (35.3%) with SML <1 mm and ≥1 mm, respectively. By multivariable analysis, SML <1 mm (hazard ratio [HR], 2.64; 95% confidence interval, 1.92–3.62) and ≥1 mm (HR, 5.49; 95% confidence interval, 3.94–7.65) independently predicted BCR. The addition of SML improved the C-index from 0.672 to 0.739 and the 5-year area under the curve from 0.680 to 0.724. Decision curve analysis confirmed net clinical benefits. An SML ≥1 mm predicted early recurrence ≤24 months (odds ratio, 9.98; p < 0.001) but lost significance in 24-month landmark analysis (HR, 1.21; p = 0.68). Conversely, an SML <1 mm remained an independently prognostic factor beyond 24 months (HR, 3.43; p < 0.001). A significant interaction was observed between SML <1 mm and the International Society of Urological Pathology grade ( p = 0.0004), indicating an attenuated prognostic impact on high-grade disease. Conclusions: The SML independently predicted BCR with a clear dose-response relationship in pT2N0 prostate cancer. An SML ≥1 mm identified a high-risk phenotype for early recurrence warranting intensified short-term surveillance and earlier salvage consideration, while an SML <1 mm predicted late recurrence requiring sustained long-term monitoring. The 1-mm threshold is clinically pragmatic and reproducible. After external validation, these findings could support routine SML reporting in pathology reports to enable time-stratified, personalized postoperative surveillance strategies.