Virtual vs Traditional Microscopy in Dermatopathology: Which Is Better?

Diagnosis from whole-slide images (WSI) on a digital microscope is noninferior to diagnosis from glass slides using traditional microscopy (TM), according to the results of a large retrospective dermatopathology cohort study published in JAMA Dermatology.

Intraobserver concordance of diagnosis with WSI vs TM was evaluated in a large cohort of 499 cases that represented the spectrum of diagnoses observed in a high-volume dermatopathology laboratory. The focus was on image resolution, and specifically eosinophils in inflammatory cases and mitotic figures within melanomas. In addition, the investigators examined the workflow efficiency of WSI vs TM. Interobserver ground truth consensus (GTC) diagnosis of the 499 cases was determined by 3 dermatopathologists.

The mean intraobserver concordance between WSI and TM diagnosis was 94% (471 of 499). Interobserver concordance between WSI and GTC diagnosis and TM and GTC diagnosis was also 94%. Moreover, concordance among all observations (GTC, WSI, and TM) was 91%. Overall, melanocytic lesions comprised 30% of total cases, which included 15 malignant melanomas, and inflammatory cases represented 7%.

Diagnoses from WSI were shown to be statistically significantly noninferior to those from TM compared with GTC diagnosis, in most cases. The WSI technique approached, but did not attain, significant noninferiority for the subset of melanocytic lesions (P =.053). With discordance in TM diagnosis of this type of lesion being widely recognized, further examination of melanocytic neoplasms is recommended.

The investigators concluded that diagnosis from WSI is equivalent to diagnosis from glass slides using the TM method, thus supporting the viability of WSI for primary diagnosis of cutaneous lesions in the clinical setting.

Reference

Kent MN, Olsen TG, Feeser TA, et al. Diagnostic accuracy of virtual pathology vs traditional microscopy in a large dermatopathology study [published online October 11, 2017]. JAMA Dermatol. doi: 10.1001/jamadermatol.2017.3284