Validating the total cancer location density metric for stratifying patients with low-risk localized prostate cancer at higher risk of grade group reclassification while on active surveillance.

Abstract

To validate a previously proposed prognostic metric, Total Cancer Location (TCLo) density, in a contemporary cohort of men with grade group (GG) 1 prostate cancer (PCa) on active surveillance (AS). We evaluated 123 patients who entered AS with maximum GG1 PCa at diagnostic and/or confirmatory biopsy. TCLo was defined as the total number of PCa locations identified on both biopsy sessions. TCLo density was calculated as TCLo / prostate volume [ml]. Primary endpoint was progression-free survival (PFS), defined as time from confirmatory biopsy to grade group reclassification (GGR) on repeat biopsy or prostatectomy. Optimal cut-point for TCLo density was predefined in a previously reported cohort and applied to this contemporary cohort. Kaplan-Meier and multivariable Cox regression analysis were used to estimate the association of predictors with PFS. During median follow-up of 7.8 years, (IQR 7.3-8.2) 34 men had GGR. Using previously defined cut-points, PFS at 5-years was 60% (95% CI: 44%-81%) vs. 89% (95% CI: 83%-96%) in men with high (≥0.06 ml-1) vs. low (<0.06 ml-1) TCLo density, and 63% (95% CI: 48%-82%) vs. 90% (95% CI: 83%-96%) in men with high (≥3) vs. low (≤2) TCLo (log-rank test: P < 0.0001, respectively). Adjusting for age, prostate volume, percent of positive cores and PSA, both higher TCLo density (HR [per 0.01 ml-1 increase]: 1.18, 95% CI: 1.05-1.33, P = 0.005) and TCLo (HR: 1.69, 95% CI: 1.20-2.38, P = 0.002) were associated with shorter PFS. The previously suggested prognostic value of TCLo density was confirmed in this validation cohort. TCLo alone performed similarly well. Patients with high TCLo density (≥0.06 ml-1) or TCLo (>2) were at greater risk of GGR while on AS. With external validation, these metric may help guide risk-adapted surveillance protocols.