Diagnostic performance of prostate health index (PHI) in predicting prostate cancer on prostate biopsy.

INTRODUCTION: Prostate health index (PHI) has been shown to have better diagnostic accuracy in predicting prostate cancer (PCa) in men with total prostate-specific antigen (PSA) levels between 4-10ng/ml. However, little is known of its value in men with elevated PSA beyond this range. This study aimed to evaluate the diagnostic performance of PHI in Malaysian men with elevated PSA values ≤ 20ng/ml. MATERIALS AND METHODS: From March 2015 to August 2016, all men consecutively undergoing transrectal ultrasound (TRUS)-guided prostate biopsy with total PSA values ≤ 20ng/ ml were recruited. Blood samples were taken immediately before undergoing prostate biopsy. The performance of total PSA, %fPSA, %p2PSA and PHI in determining the presence of PCa on prostate biopsy were compared. RESULTS: PCa was diagnosed in 25 of 84 patients (29.7%). %p2PSA and PHI values were significantly higher (p<0.05) in patients with PCa than those without PCa. The areas under the receiver operating characteristic curves for total PSA, %fPSA, %p2PSA and PHI were 0.558, 0.560, 0.734 and 0.746, respectively. At 90% sensitivity, the specificity of PHI (42.4%) was five times better than total PSA (8.5%) and two times better than %fPSA (20.3%). By utilising PHI cut-off >22.52, 27 of 84 (32.1%) patients could have avoided undergoing biopsy. CONCLUSION: Findings of our study support the potential clinical effectiveness of PHI in predicting PCa in a wider concentration range of total PSA up to 20ng/ml.