Outcomes of men on active surveillance for low-risk prostate cancer at a safety-net hospital

Purpose To characterize demographic, disease, and cancer outcomes of men on active surveillance (AS) at a safety-net hospital and characterize those who were lost to follow-up (LTFU). Methods From January 2004 to November 2014, 104 men with low-risk prostate cancer (PCa) were followed with AS at Zuckerberg San Francisco General Hospital (ZSFG). Criteria for AS have evolved over time; however, patients with diagnostic prostate-specific antigen (PSA) 10 ng/mL or less, clinical stage T1/2, biopsy Gleason score 3 + 3 or 3 + 4, 33% or fewer positive cores, and 50% or less tumor in any single core were potentially eligible for AS. Men were longitudinally followed with a PSA or digital rectal examination or both every 3 to 6 months, and repeat prostate biopsy every 1 to 2 years. Clinical staging and grading were based on a physical examination and at least a 12-core biopsy, respectively. LTFU was defined as failure to successfully contact patients with 3 phone calls or any urology visit recorded within 18 months from a prior visit or biopsy. A secondary chart review was performed using the electronic medical record at ZSFG as well as EPIC Systems CareEverywhere which allows access to select non-ZSFG institutions to confirm that patients were truly LTFU. Results Among the 104 men on AS at ZSFG, the median age at diagnosis of PCa was 61.5 years (range: 44–81). The median follow-up period was 29 months (range: 0–186 months) during which 18 (17.3%) men were LTFU and 48 (46%) remained on surveillance. Men underwent a median of 7 (1–21) serum PSA measurements and an average of 2 prostate biopsies (1–5). In total, 22 (20.6%) men had definitive treatment with the median time from diagnosis to active treatment being 26 (range: 2–87) months. Radiation therapy was more common than radical prostatectomy (12.5% vs. 7.7%). There was 1 PCa–related death and 3 noncancer deaths. Initial adherence to AS was poor; however, men committed to AS initially were ultimately more compliant over time. Conclusion AS for low-risk PCa is challenging among a vulnerable population receiving care in a safety-net hospital, as rates of LTFU were high. Our findings suggest the need for AS support programs to improve adherence and follow-up among vulnerable and underserved populations.