Variation in positive surgical margin status following radical prostatectomy for pT2 prostate cancer
Recommended Citation
Tan WS, Krimphove M, Cole A, Berg S, Marchese M, Lipsitz S, Loppenberg B, Nabi J, Abdollah F, Choueiri T, Kibel A, Sooriakumaran P, and Trinh QD. Variation in positive surgical margin status following radical prostatectomy for pT2 prostate cancer. Eur Urol Suppl 2019; 18(1):e294-e295.
Document Type
Conference Proceeding
Publication Date
2019
Publication Title
Eur Urol Suppl
Abstract
Introduction & Objectives: Positive surgical margin (PSM)following radical prostatectomy for pT2 prostate cancer is considered a surgical quality metric. We evaluated patient, institutional, surgical approach and cancer-specific factors associated with PSM variability. Materials & Methods: A total of 45,426 men from 1,152 institutions with pT2 prostate cancer and known margin status following radical prostatectomy were identified using the National Cancer Database (2010-2015). Patient demographics and comorbidity, socioeconomic status, geographical and institutional information, cancer-specific variables and type of surgical approach were extracted. Multilevel hierarchical mixed effects logistic regression model was performed to determine the factors associated with a risk of PSM and their contribution to a PSM status. Results: Median PSM rate of 8.5% (IQR: 5.2-13.0%, range: 0-100%). Robotic (OR: 0.90, 95% CI: 0.83-0.99)and laparoscopic (OR: 0.74, 95% CI: 0.64-0.90)surgical approach, academic institution (OR: 0.87, 95% CI: 0.76-1.00), high institution surgical volume (>297 cases [OR: 0.83, 95% CI: 0.70-0.99)and East North Central USA (OR: 0.71, 95% CI: 0.52-0.96)were independently associated with a lower PSM. Black men (OR: 1.13, 95% CI: 1.01-1.26)and adverse cancer specific features (PSA 10-20, PSA >20, cT3 stage, Gleason 7, 8, 9-10; all p>0.01)were independently associated with a higher PSM. The overall multilevel hierarchical logistic regression model accounts for 24.9% of PSM variation. Patient-specific, institution-specific and cancer-specific factors accounted for 9.1%, 15.6% and 61.1% of the variation within the overall regression model respectively. [Figure presented]Conclusions: Cancer-specific factors account for 15.2% of PSM variation with the remaining 84.8% of PSM variation due to patient, institution and other factors not accounted for in the model. Non cancer-specific factors represent potentially addressable factors which are important for policy makers in their efforts to improve patient outcome.
Volume
18
Issue
1
First Page
e294
Last Page
e295
