Prostate health index significantly reduced unnecessary prostate biopsies in patients with PSA 2-10 ng/mL and PSA >10 ng/mL: Results from a Multicenter Study in China.

Rong Na,Dingwei Ye,Jun Qi,Fang Liu,Brian T Helfand,Charles B Brendler,Carly A Conran,Vignesh Packiam,Jian Gong,Yishuo Wu,Siqun L Zheng,Zengnan Mo,Qiang Ding,Yinghao Sun,Jianfeng Xu
DOI: https://doi.org/10.1002/pros.23382
2017-01-01
Abstract:Background: The performance of prostate health index (phi) in predicting prostate biopsy outcomes has been well established for patients with prostate-specific antigen (PSA) values between 2 and 10 ng/mL. However, the performance of phi remains unknown in patients with PSA >10 ng/mL, the vast majority in Chinese biopsy patients. We aimed to assess the ability of phi to predict prostate cancer (PCa) and high-grade disease (Gleason Score 7) on biopsy in a Chinese population. Methods: This is a prospective, observational, multi-center study of consecutive patients who underwent a transrectal ultrasound guided prostate biopsy at four hospitals in Shanghai, China from August 2013 to December 2014. Results: In the cohort of 1538 patients, the detection rate of PCa was 40.2%. phi had a significantly better predictive performance for PCa than total PSA (tPSA). The areas under the receiver operating characteristic curve (AUC) were 0.90 and 0.79 for phi and tPSA, respectively, P < 0.0001. A considerable proportion of patients in the cohort had PSAs >10 ng/mL (N=838, 54.5%). The detection rates of PCa were 35.9% and 57.7% in patients with tPSA 10.1-20 and 20.1-50 ng/mL, respectively. The AUCs of phi (0.79 and 0.89, for these two groups, respectively) were also significantly higher than tPSA (0.57 and 0.63, respectively), both P < 0.0001. If a phi <= 35 was used as the cutoff, 599/1538 (39%) biopsies could have been avoided at a cost of missing small numbers of PCa patients: 49 (7.93%) PCa patients, including 18 (3.69%) high-grade tumors. Conclusions: Results from this study suggest that phi can be used to predict PCa and high-grade disease in Chinese men with high PSA levels (>10 ng/mL).
What problem does this paper attempt to address?