DS03 <i>Ex vivo</i> confocal laser scanning microscopy of 83 facial basal cell carcinomas: a single-centre prospective 1-year study of real-world Mohs experience

Abstract Ex vivo confocal laser scanning microscopy (CLSM) is a novel integrative noninvasive imaging tool that can be used within Mohs micrographic surgery (MMS) for tumour margin assessment. Studies have reported up to 88% sensitivity and 99% specificity with CLSM in diagnosing basal cell carcinomas (BCC), compared with conventional histology. Its real-world utility has triggered variable opinions among Mohs surgeons. The fourth-generation VivaScope 2500 allows fresh tissue scanning with minimal preparation with acetic acid and acridine orange, creating digitally coloured mosaic images comparable with gold-standard frozen sections stained with haematoxylin and eosin, within minutes. This has the potential to enhance the patient experience, allow multiple stages and repair in a single session, and improve overall efficiency in a Mohs unit. We prospectively analysed our experience with 83 consecutive MMS first-stage specimens over 1 year, involving head and neck BCC, integrating the VivaScope 2500 into one Mohs list. Ethics committee approval was obtained. CLSM is currently used by one surgeon and biomedical scientists trained to use and read the digital images created. We characterized the real-world sensitivity and specificity of CLSM in our cohort. In total 83 head and neck BCCs were processed with the VivaScope 2500 and then cryosectioned. The BCCs included 54 nodular, 12 infiltrative, 12 both nodular and infiltrative, three nodular and micronodular, one micronodular and infiltrative, and one micronodular. There were six cases of noncorrelation, where the suspected tumour seen on CLSM differed from true tissue margins determined by cryosectioning. For all other cases there was concordance in tumour mapping between CLSM and cryosection. Basal budding was appreciated well in both modalities. A diagnostic conundrum was highlighted where CLSM of a thin conchal-bowl specimen showed no tumour but initial cryosection from the microtome revealed a nodular BCC. Was a thin BCC potentially missed? Or has CLSM minimized tissue loss with narrowly clear margins? In other words was this a false negative or a true negative. Our cohort had 78.3% sensitivity and 91.7% specificity for CLSM compared with frozen section and standard histology. The VivaScope 2500 CLSM shows promise in MMS. Training in CLSM image interpretation is important for Mohs surgeons and biomedical scientists. Suboptimal image quality and issues with tissue flattening may lead to false positive results and reduce the reliability of CLSM. Identifying features can be challenging and cause false positive tissue margins, for example follicular structures misinterpreted as tumour. Inflammatory infiltrates can mimic BCC-like features. Our experience highlights important caveats with CLSM when utilized in MMS. Regular pathology reporting, improved flattening techniques and appropriate sample selection can potentially improve image interpretation with CLSM and minimize future errors. Enhanced artificial intelligence algorithms and digital real-time pathologist review could allow for rapid diagnostics.