Diagnostic accuracy of ex vivo fluorescence confocal microscopy in Mohs surgery of basal cell carcinomas: a prospective study on 753 margins

Background Frozen histological sections are used for intraoperative margin assessment during Mohs surgery. Fluorescence confocal microscopy (FCM) is a new tool that offers a promising and faster alternative to frozen histology. Objectives To evaluate prospectively in a clinical setting the accuracy of FCM vs. frozen sections in margin assessment of basal cell carcinoma (BCC). Methods Patients with BCC scheduled for Mohs surgery were prospectively enrolled. Freshly excised surgical specimens were examined by FCM and then frozen sections were evaluated. Permanent sections were obtained, in order to validate the sample technique. A blind re-evaluation was also performed for discordant cases. Sensitivity and specificity levels, as well as positive and negative predictive values (PPV and NPV, respectively), were calculated and receiver-operating characteristic curves generated. Results We enrolled 127 BCCs in as many patients (40 center dot 2% females). Seven hundred and fifty-three sections were examined. All BCCs were located in the head and neck area. In evaluating the performance of FCM vs. frozen sections, sensitivity was 79 center dot 8%, specificity was 95 center dot 8%, PPV was 80 center dot 5% and NPV was 95 center dot 7% [area under the curve 0 center dot 88, 95% confidence interval 0 center dot 84-0 center dot 92 (P < 0 center dot 001)]. Forty-nine discordant cases were re-evaluated; 24 were false positive and 25 false negative. The performance of FCM and frozen sections was also evaluated according to the final histopathological assessment. Conclusions We found high levels of accuracy for FCM vs. frozen section evaluation in intraoperative BCC margin assessment during Mohs surgery. Some technical issues prevent the wide use of this technique, but new devices promise to overcome these limitations.