Related Experiment Video
Updated: Sep 29, 2026

Retzius-Sparing Robot-Assisted Radical Prostatectomy
Published on: May 19, 2022
Does Positive Margin in Secondary Tumors at Radical Prostatectomy Predict Biochemical Recurrence in Multifocal
Nikunj B Shah1, Angelica Menendez1, Sanoj Punnen2,3
1Departments of Pathology and Laboratory Medicine.
Abstract:
Positive surgical margin (M+), higher-grade group (GG), extraprostatic extension (EPE), seminal vesicle invasion (SV+), and tumor volume (TV) are among the factors at radical prostatectomy (RP) demonstrating strong independent correlations with subsequent biochemical recurrence (BCR). However, the prognostic significance of M+ in secondary tumors on BCR when a dominant tumor is excised with a clear margin has not been studied. A review of 1,687 RPs with multifocal prostate cancer was performed to assess GG, margin status, EPE, SV+, and TV for each individual tumor. The final analytical cohort consisted of 1,449 RPs, as patients without follow-up were excluded. The highest GG determined the dominant tumor. In tumors with equal GG, higher stage tumors were considered dominant. In tumors with equal GG and stage, a tumor with a larger TV was considered dominant. In all cases, the dominant tumor was excised with clear margins. Associations between secondary tumor margin status and BCR were evaluated using multivariable logistic regression analysis. The dominant tumors were GG1 (n=195 [13.5%]), GG2 (n=551 [38.0%]), GG3 (n=308 [21.2%]), GG4 (n=72 [5.0%]), and GG5 (n=323 [22.3%]). The median TV of the dominant tumor was 0.73 mL (range: 0.03 to 22). EPE was seen in 481 (33.2%) cases and SV+ in 77 (5.3%). In 74 (5.1%) RPs, a secondary tumor showed M+, composed of GG1 32 (43.2%), GG2 33 (44.6%), GG3 4 (5.4%), GG4 3 (4.1%), and GG5 2 (2.7%). BCR was observed in 137 (9.4%) patients, and 62 (4.3%) had persistent PSA. The median follow-up for patients with BCR was 855 days (range: 395 to 1,575) and for those without, 729 (range: 236 to 1,582), P=0.045. In the Cox proportional hazards model, after adjusting for GG, TV, EPE, and SV+, M+ in secondary tumors predicted BCR (HR: 3.14; P=0.0002). Except for GG2 and EPE, all other RP findings correlated statistically significantly with BCR: GG3 (HR: 9.53; P=0.0002), GG4 (HR: 16.4; P<0.0001), GG5 (HR: 14.2; P<0.0001), TV (HR: 1.16; P<0.0001), and SV+ (HR: 3.03; P=0.0001). While published literature indicates that M+ in the dominant tumor is one of the strongest predictors of BCR, M+ in secondary tumors appears to have a lesser effect on BCR than dominant tumor characteristics.
