Confocal microscopy for immediate analysis of prostate biopsies.

Background: Diagnosis of prostate cancer (PC) is based on haematoxylin-eosin (HE) staining, time-consuming, histopathological evaluation. Fluorescent confocal microscopy (FCM) allows tissue analysis at the same time as sample collection. The aim of this paper was to evaluate concordance between HE and FCM, and the additional value of a second pathologist.

Methods: Two hundred ninety-seven magnetic resonance imaging-targeted prostate biopsy cores from 95 region of interest (ROI) were intraoperatively analyzed using FCM (VivaScope), and subsequently sent for conventional HE examination. A positive or negative diagnose was made by two pathologists on FCM. Additionally, the agreement of the International Society of Uropathology (ISUP) grade was evaluated. Agreement between FCM and HE analysis in terms of the presence/absence of PC was analyzed at ROI level. Concordance between two different pathologists was evaluated. If one pathologist considered the ROI with suspected PC and the other disagreed, it was categorized as absence of PC.

Results: FCM allowed intraoperative assessment with strong histopathological evaluation agreement. Cohen's Kappa to detecting PC was 0.79 for pathologist 1 (P1), 0.81 for pathologist 2 (P2), and 0.85 for combined pathologists (P1-P2). Positive predictive values for P1, P2 and P1-P2 were 89%, 86% and 100%, respectively. Negative predictive values for P1, P2 and P1-P2 were 90.7%, 95.5 and 86.5%, respectively. Cohen's Kappa concordance for ISUP 1-2 detection was 0.7 for P1, 0.8 for P2 and 0.85 for P1-P2, and for ISUP 3-5, it was 0.69 for P1, 1 for P2 and 1 for P1-P2.

Conclusions: FCM allowed rapid and accurate evaluation of prostate biopsy with a good correlation between two pathologists.