A model for characterizing anatomical locations of merkel cell carcinoma

Summary: Abstract Body: Merkel cell carcinoma (MCC) and melanoma are two of the most common skin cancers, with ultraviolet radiation (UVR) exposure playing a role in the pathogenesis of each. Nearly 50 times more new cases of melanoma are diagnosed annually in the United States compared with MCC, allowing for easier collection and analysis of melanoma data. We sought to examine applicability of a model used for classification of melanoma primary tumor sites to a population of MCC cases. Gordon et al. (2017) characterized primary tumor sites of cutaneous melanoma (n=5,973) using a modified UVR exposure model and a model describing visibility of the lesion upon self-examination. The UVR exposure model consisted of five categories: chronic, moderately intermittent, highly intermittent, rare, or other UVR exposure. Additionally, the primary site of the lesion was classified as either easily or poorly visible upon self-examination. We implemented these models in a smaller MCC cohort (n=105). The highest quantity of MCCs were present in chronically UVR exposed areas (n=45, 42.9%). Most of the MCC lesions were also present in easily visible sites upon self-examination (n=88, 83.8%), emphasizing the importance of awareness and self-examination for earlier detection of this aggressive malignancy. Application of the melanoma model to the MCC cohort did not allow for adequate power to perform survival analysis. This confirmed our expectations that the MCC sample size was too limited for analysis through this model, with the melanoma instrument unable to be directly applied to the small MCC registry size. Dichotomization of variables may be necessary to assess UVR exposure for MCC. A more manageable analysis may result from establishing two distinct groups for UVR exposure of primary sites, rather than several categories. Next steps will explore anatomic determinants through application of the UVR exposure variables in a binary method. Although the adaptation of the instrument from melanoma to MCC was restricted by the rarity of MCC, a binary model may assist with further analysis. Claire Reynolds¹, Kelsey Ouyang¹, Bryan Carroll¹ 1. Dermatology, University Hospitals, Cleveland, OH, United States. Clinical Research: Epidemiology and Observational Research