629 Performance monitoring of a streamlined and scalable non-invasive gene expression assay for pigmented lesions

Abstract

The DermTech Pigmented Lesion Assay (PLA) is a qPCR-based assay that aids in the detection of melanoma by assessing up-regulation of two melanoma-associated genes, LINC00518 and PRAME, in skin samples collected by non-invasive adhesive patches. This study is to establish assay performance characteristics of the PLA post optimization, and to investigate the prognostic potential of the assay by comparison of qPCR amplification cycle in PLA positive samples with the histopathologic diagnoses of cutaneous pigmented lesions. RNA isolated from non-invasively collected skin cells was analyzed from seventy-five pigmented lesions. The PLA was optimized to simplify the laboratory workflow and increase scalability of the assay through a series of process improvements including RNA quantification by 1-step qRT-PCR (vs. reverse transcription followed by qPCR) and assay miniaturization. PLA result was determined by evaluating qPCR amplification of LINC00518 and PRAME multiplexed with housekeeping genes β-Actin and PPIA respectively. Samples that demonstrated over-expression of LINC00518 and PRAME relative to the housekeeping genes were considered PLA positive. Blinded to the PLA results, biopsy specimens from the pigmented lesions were examined and classified by a dermatopathologist as invasive melanoma (n=22); melanoma in situ (MIS) (n=26); benign nevus (24); or other (n=3). All twenty-two lesions classified as invasive melanoma by histopathology were PLA-positive, with an average LINC00518 Ct of 32.2 and PRAME Ct of 30.9. Seventeen of the twenty-six samples classified as MIS were PLA-positive, with an average LINC00518 Ct of 33.8 and PRAME Ct of 33.2. Of the twenty-seven samples not classified as melanoma, five were PLA-positive, with an average LINC00518 Ct of 34.4 and PRAME Ct of 35.8. The results indicate that PLA Ct values are highly correlated with histopathologic diagnoses. The optimized PLA assay demonstrated a sensitivity of 100% for invasive melanoma and 65% for MIS, with an overall specificity of 81%.