Researchers have identified several key diagnostic criteria for the diagnosis of melanoma and basal cell carcinoma with reflectance confocal microscopy.
Researchers have identified several key diagnostic criteria for reflectance confocal microscopy (RCM) diagnosis of melanoma and basal cell carcinoma (BCC), according to a small study published in JAMA Dermatology.
According to the study, RCM is a noninvasive method of examining the epidermis and papillary dermis at cell-level resolution, and use of the method has been shown in studies to improve the diagnostic accuracy for melanoma and BCC. However, use of the method is also operator dependent.
“Because the knowledge base of RCM is still being formed and because formal training programs have been launched only recently, heterogeneity in criteria recognition and in diagnostic accuracy is expected among different RCM evaluators,” wrote study author Francesca Farnetani, MD, of the department of dermatology at the University of Modena and Reggio Emilia, Modena, Italy, and colleagues.
Therefore, in this study, Farnetani and colleagues tested interobserver reproducibility of RCM descriptors among evaluators with different levels of experience. The Web-based study looked at nine dermatologists-six experienced users and three recent users-from six countries who were asked to evaluate a set of RCM images that included 100 biopsy-proven lesions for the presence of predefined RCM descriptors. The main outcome of the study was descriptors with fair to good interrater agreement.
For melanoma, the identified fair to good descriptors were presence of pagetoid cells (good), atypical cells at the dermal-epidermal junction (good), and irregular epidermal architecture (fair). For BCC, the descriptors were aspecific junctional pattern (good), basaloid cords (good), and ulceration (good).
The researchers calculated the overall diagnostic accuracy of the evaluators and found that they attained a mean accuracy of 82.7%. They noted that this accuracy is lower than that seen in other studies looking at RCM images but added that this was “likely due to the inclusion of lesions that were dermoscopically challenging for diagnosis by dermoscopy experts.”
The mean sensitivity of the evaluators was 88.9% with a mean specificity of 79.3%. Based on RCM experience, those evaluators with more experience had higher sensitivity but similar specificity compared with recent RCM users.
“Taken together, these findings confirm that there is consistent recognition of several key diagnostic criteria for melanoma and BCC,” the researchers wrote. “However, few of the published diagnostic criteria for melanoma and BCC were inconsistently recognized: agreement on their identification should be further tested, or RCM users could be better trained in their recognition.”