Prediction of prostate cancer recurrence using quantitative phase imaging: Validation on a general population.

Shamira Sridharan,Emily Dube,Virgilia Macias,Krishnarao Tangella,Max Xiangtian Kong,Jonathan Melamed,André Kajdacsy-Balla,Gabriel Popescu

doi:10.1038/srep33818

Abstract

Prediction of biochemical recurrence risk of prostate cancer following radical prostatectomy is critical for determining whether the patient would benefit from adjuvant treatments. Various nomograms exist today for identifying individuals at higher risk for recurrence; however, an optimistic under-estimation of recurrence risk is a common problem associated with these methods. We previously showed that anisotropy of light scattering measured using quantitative phase imaging, in the stromal layer adjacent to cancerous glands, is predictive of recurrence. That nested-case controlled study consisted of specimens specifically chosen such that the current prognostic methods fail. Here we report on validating the utility of optical anisotropy for prediction of prostate cancer recurrence in a general population of 192 patients, with 17% probability of recurrence. Our results show that our method can identify recurrent cases with 73% sensitivity and 72% specificity, which is comparable to that of CAPRA-S, a current state of the art method, in the same population. However, our results show that optical anisotropy outperforms CAPRA-S for patients with Gleason grades 7–10. In essence, we demonstrate that anisotropy is a better biomarker for identifying high-risk cases, while Gleason grade is better suited for selecting non-recurrence. Therefore, we propose that anisotropy and current techniques be used together to maximize prediction accuracy.

Highlights

They rely on prostate specific antigen (PSA) levels and Gleason score (GS) reports which are prone to errors from assay sensitivity and inter-observer variability respectively[12]
Anisotropy (g) of light scattering was measured using the scattering phase theorem in the unstained prostatectomy samples imaged using spatial light interference microscopy (SLIM), a quantitative phase imaging (QPI) method[21,22]
Anisotropy was calculated in the single stromal layer adjoining 6–18 glands from each of the 33 patients with post-prostatectomy biochemical recurrence of prostate cancer and 159 patients who did not have recurrence

Summary

Introduction

They rely on PSA levels and Gleason score (GS) reports which are prone to errors from assay sensitivity and inter-observer variability respectively[12]. Biomarker-based approaches have been developed as both stand-alone predictors of prostate cancer recurrence and in combination with nomogram-based approaches[13,14,15,16,17,18,19]. The combination of those biomarkers with existing nomograms resulted in improved performance, but they are still subject to the vulnerabilities associated with subjective parameters in nomograms. We demonstrated that anisotropy had the ability to identify recurrence with 77% sensitivity and 62% specificity, while CAPRA-S showed poor discriminatory ability as multiple CAPRA-S parameters were used as matching criterion. Our results showed that anisotropy has added value over CAPRA-S at GS ≥ 7; and at CAPRA-S ≥ 3 which corresponds to the intermediate and high risk groups

Objectives

Methods

Results

Conclusion