The cohort was classified based upon morphology of cutaneous lesions, characteristics of the atypical infiltrate on histology, and the presence or absence of cytotoxic cell markers (TIA-1, granzyme B, perforin). Clinical information was obtained from the electronic medical record of each patient when available and is summarized in Supplementary Table 1. Metastatic spread was determined on the basis of positron emission tomography-computed tomography (PET–CT) or computed tomography (CT) imaging highly suspicious for metastatic involvement. One-third of these cases had metastasis proven by biopsy and/or cytology. Antibodies used to determine γδ TCR expression included γ3.20 antibody clone (TCR1153, ThermoFisher Scientific, IL) or H-41 antibody clone (SC-100289, Santa Cruz Biotechnology, TX)60 ,61 (link). Major pathological criteria that were assessed were depth of infiltrate involvement, ulceration, keratinocyte necrosis, and vasculitis. Samples were divided in groups based on depth of skin involvement: (1) epidermal (2) epidermal/dermal, and (3) panniculitic. There was insufficient clinical annotation to distinguish between epidermal vs. epidermal/dermal involvement in five samples. Biopsy specimens were subject to review by expert pathologists (J.G., A.L., and/or A.G.).
Free full text: Click here