Introduction

Eligibility criteria for active surveillance (AS) for low-risk prostate cancer have been defined by multiple series reported in the literature. In clinical practice occasional patients who do not meet all criteria may choose AS. We investigated outcomes in borderline cases.

Methods

We investigated our institutional database of 990 men on AS between 1997-2014. Our guidelines for AS eligibility, formalized in 2008, include Gleason ≤6, stage ≤cT2a, PSA ≤10 ng/mL, ≤3 of 12 cores positive at diagnosis, and ≤20% of any core involved at diagnosis. For this analysis, we defined borderline cases for AS as those patients with one or more of either Gleason score 7, PSA >10, stage cT2a, >33% of cores positive at diagnosis, or >20% of any core involved at diagnosis. Survival analyses were conducted using Kaplan-Meier and Cox proportional hazards.

Results

In the entire cohort (n=990), mean age at diagnosis was 66.9 years (±7.9) and median PSA 5.1 (IQR 4-6.87). While the majority met all AS criteria, 310 patients (31.3%) met at least one of the borderline AS criteria; 2.4% of patients had Gleason 7, 7.6% had PSA >10, 8.0% were cT2a, 3.9% (37/943) had >33% of cores positive at diagnosis, and 18.4% (156/848) had >20% of any core involved. With mean follow-up 4.5 years, univariate survival analysis demonstrated no difference in freedom from treatment (FFT) between patients with Gleason 7 vs. ≤6, >33% vs. ≤33% cores involved, or PSA >10 vs. ≤10. Lower FFT was noted among patients with cT2a vs. ≤cT1c disease (62.0% vs. 70.8%, P=0.04), patients with >20% vs. ≤20% of any core involved (61.5% vs. 71.8%, P=0.009), as well as those with PSA density ≥0.15 vs. <0.15 (61.1% vs. 72.0%, P = 0.0006). In multivariate analysis, >20% core involvement and PSA density ≥0.15 remained a significant predictor for treatment (P=0.003), adjusting for PSA >10, Gleason >6, >33% cores involved, and stage. Among the 310 borderline AS cases, there were only 6 (1.9%) cases of metastasis and 1 (0.3%) prostate cancer-specific death. These adverse outcomes were equivalent to the remainder of the cohort meeting strict AS criteria, which included 10 (1.5%) cases of metastasis and 2 (0.3%) prostate cancer-specific deaths.

Conclusions

Active surveillance remains a viable option for select patients who are borderline cases per current AS criteria. However, patients with higher volume disease and higher PSA density may be more likely to progress to treatment. Long-term clinical outcomes in these patients should continue to be investigated.

Funding

None