Prostate health index predicts biochemical recurrence in patients with organ-confined prostate cancer treated with radical prostatectomy

INTRODUCTION & OBJECTIVES: A non-negligible proportion of patients with organ-confined PCa may experience biochemical recurrence
(BCR) after definitive treatment. Identification of patients at high risk of BCR may help to determine those individuals who may benefit of a
strict follow-up schedule or adjuvant treatments. We tested the hypothesis that prostate health index (PHI) could predict BCR in a
contemporary population of patients with organ-confined PCa treated with robot-assisted radical prostatectomy (RARP).
MATERIAL & METHODS: The study population consisted of 228 patients with organ-confined PCa (defined as patients with pT2cR0 pN0
disease at final pathology) treated with RARP and pelvic lymph node dissection for PCa at a single high-volume institution between 2010
and 2011. Patients subjected to neoadjuvant or adjuvant therapies were excluded. Pathological characteristics of PCa and PHI ([-
2]proPSA/free PSA)*tPSA) were available in all patients. According to EAU guidelines, BCR was defined as a postoperative value of tPSA
³0.2 ng/mL and rising after RARP. The Kaplan-Meier method was used to determine BCR-free survival rates and Cox regression models
were fitted to determine the predictors of BCR. Finally, the predictive accuracy (AUC) of each predictor was determined with the Harrell's
concordance index and the minimum p-value method was used to determine the most informative PHI cut-off value for BCR.
RESULTS: Mean total PSA and mean PHI were respectively 6.8 ng/mL (range: 1-46) and 50 (16-302). Pathologic Gleason score was 6 in
83 (36.4%), 7 in 139 (61.0%), and 8 in 6 (2.6%) patients. Mean prostate weight and mean tumour volume were 56.3 gr (16-401) and 11.7%
(0.3-65%), respectively. Overall, 16 (7.0%) patients experienced BCR after RARP. The 2- and 3-year BCR-free survival rates were 96.7 and
84.8%, respectively. At Cox regression analyses, PHI was significantly associated with BCR (HR: 1.017; p<0.001). In accuracy analyses,
PHI emerged as the most accurate predictor of BCR (AUC: 65.2%), and significantly outperformed tPSA (61.9%), tumor volume (58.0%),
and pathologic Gleason score (55.3%) (p<0.001). Specifically, the 3-year BCR-free survival rate was 88.7% in patients with a preoperative
PHI <82 relative to 44.4% in patients with a PHI ³82 (log-rank test: p=0.005).
CONCLUSIONS: Prostate health index levels may help to identify patients at high risk of having BCR despite having organ-confined
disease. External validation of our findings is required to further evaluate the role of PHI within this subgroup of patients.