Racial Disparities in Active Surveillance for Prostate Cancer

Purpose Active surveillance protocols track low risk prostate cancer progression over time. However, given the lack of uniform criteria for managing low risk prostate cancer, men who qualify for active surveillance might have less intensive surveillance and, thus, experience poorer outcomes. In this study we examined racial disparities in the frequency and intensity of active surveillance between African-American and Caucasian men. Materials and Methods Using the linked SEER-Medicare data set we identified 13,374 men with low risk prostate cancer (defined by the D'Amico criteria) diagnosed from 2004 to 2009 and then followed through 2011. A total of 2,916 men did not receive any treatment (radiation, hormonal therapy or surgery) within 1 year after diagnosis. Men were considered to be on active surveillance if they had at least 1 of the following 3 surveillance strategies within 2 years after diagnosis, namely 1 or more prostate biopsies, 4 or more prostate specific antigen tests, and/or 4 or more visits to the doctor with prostate cancer listed as the diagnosis. To compare the frequency of active surveillance between the groups (African-American vs Caucasian) we used the chi-square test. To estimate the odds ratio of active surveillance we used multivariable logistic regression after adjusting for possible confounders such as year of diagnosis, age at diagnosis, socioeconomic status and Charlson score. Results Of the 2,916 untreated men 1,141 (39%), including 963 (37%) Caucasian men and 178 (58%) African-American men (p <0.0001), did not undergo any of the 3 surveillance strategies but instead were essentially on watchful waiting. Caucasian men (vs African-American) were more likely to be on active surveillance, with 1,646 (63.1%) vs 129 (42.0%) opting for 1 surveillance strategy (p <0.0001), 783 (30.0%) vs 50 (16.3%) opting for any 2 strategies (p <0.0001) and 193 (7.4%) vs 11 (3.6%) going through all 3 (p=0.01). On multivariable analysis African-American men had significantly lower odds of being on active surveillance than Caucasian men (OR 0.52, 95% CI 0.40–0.67). Men with more comorbidities (Charlson score 1 or greater) had significantly higher odds of being placed on active surveillance than watchful waiting (OR 1.7, 95% CI 1.46–2.12). Conclusions Among those not treated for low risk prostate cancer, Caucasian men were placed on active surveillance more frequently than African-American men, who often defaulted to de facto watchful waiting after an initial period of active surveillance. This discrepancy raises questions about the factors favoring watchful waiting over active surveillance. Moreover, given the lack of consensus regarding the most efficient active surveillance strategy, we anticipate that racial disparities in the use of active surveillance will persist, especially in African-American patients.