Progression on active surveillance for prostate cancer in Black men: a systematic review and meta-analysis

Abstract

Background: Several studies evaluated prostate cancer (PCa) outcomes in Black men on active surveillance (AS); most studies contained few Black men and results were conflicting. We performed a systematic review and meta-analyze of race and outcomes on AS.

Methods: A systematic search was performed for articles of men with Grade Group 1 or 2 (GG1 or GG2) PCa on AS. All studies required race-specific comparative progression data. Progression to treatment, PSA, or biopsy progression were considered and relative risk (RR) estimates of Black men progressing were extracted and pooled using random-effects models. Differences by study-level characteristics were evaluated using subgroup and a cumulative meta-analysis by time.

Results: In total, 12 studies were included (3137 Black and 12,206 non-Black men); eight prospective (27%, n = 4210) and four retrospectives (73%, n = 11,133) cohorts. The overall RR of progression for Black men was 1.62 (95%CI, 1.21-2.17), I² = 64% (95% CI, 32-80%), (χ² = 30.23; P = 0.001; τ² = 0.16). Black men with GG1 PCa alone had a higher pooled progression: RR = 1.81 (95% CI, 1.23-2.68). Including only studies with clinical progression (excluding progression to treatment), potentiated results: RR = 1.82 (95%CI, 1.27-2.60). However, a cumulative meta-analysis demonstrated decreasing pooled effect over time, with contemporary studies after 2019 showing a tempered effect (RR: 1.29, 95% CI: 1.20-1.39).

Conclusions: Many studies attribute racial disparity in PCa to delayed presentation of disease, however, AS is unique since all AS eligible men have a low grade and stage PCa. Our findings suggest Black men may have an increased risk of progression during AS, but the association is not so strong that Black men should be discouraged from undergoing AS. Indeed, contemporary evidence suggests stricter inclusion, better confirmatory testing or better access to care may temper these findings. Importantly, these results utilize self-reported race, a social construct that has many limitations.