A new diagnostic test showed 90 percent sensitivity and 91 percent specificity for differentiating between melanomas and benign melanocytic nevi in a validation study.
Chicago - A new genomic diagnostic test (myPath Melanoma, Myriad Genetics) is a highly accurate tool for discriminating between malignant melanoma and benign nevi, according to results of a validation study presented at the annual meeting of the American Society of Clinical Oncology.
The test is performed using formalin-fixed paraffin-embedded tissue sections of melanocytic lesions and analyzes the expression of 23 genes. The gene signature was created in a development study that included 254 melanomas and 210 benign melanocytic nevi. The validation study was conducted using 211 melanomas and 226 nevi and showed the test had 90 percent sensitivity and 91 percent specificity for differentiating between the two lesion types.
In November 2013 an early access program was launched that brought the test into the hands of a select group of dermatopathologists. The experience of these initial users shows that the gene signature test provides diagnostic information independent of histopathology and that it has led to modifications in diagnosis and physician recommendations for management.
“One might assume the test would have its greatest application for evaluating difficult to diagnose lesions. However, the early users have found it also can generate a clinically useful result in seemingly straightforward cases, causing them to take a second look at the histopathology that may lead to a new observation,” says Colleen Rock, Pharm.D., Ph.D., medical science liaison, Myriad Genetic Laboratories, Salt Lake City.
“The genomic assay is not intended to replace histopathology, but we think it is a valuable addition to the dermatopathologist’s toolbox for diagnosing melanocytic lesions.”
Loren E. Clarke, M.D., is a dermatopathologist and vice president, medical affairs, Myriad Genetic Laboratories. He tells Dermatology Times he believes the new assay can have a significant impact on patient care.
“Its potential to reduce uncertainty in differentiating between benign and malignant melanocytic lesions is important considering that about 1.5 million skin biopsies are done annually in the U.S. to exclude melanoma. If histopathology is indeterminate in just 15 percent of those cases, that means there are 200,000 patients per year who are without a definitive diagnosis.
“Based on our validation testing and initial clinical experience, we believe incorporation of this new test into pathology practice can provide a more informed diagnosis and potentially more appropriate management for a significant number of patients,” he says.
Development of the test began with a review of the published literature that identified 79 candidate genes with the potential to distinguish malignant melanoma from benign nevi. The list was refined to 40 genes based on their differential expression (measured by quantitative RT-PCR) in a set of 31 malignant and 52 benign lesions, and taking into account the technical reliability of the measurements.
Final refinement of the gene signature leading to the 23 gene assay was performed using the training cohort of 464 melanocytic lesions and a forward selection process in a series of logistic regression models. The 23 genes in the final signature represent one gene involved in cellular differentiation, eight genes involved in immune signaling, five genes with multiple functions including immune regulation, and nine housekeeper genes.
“The excellent diagnostic performance of the test can be understood by the fact that the genes it analyzes regulate melanocyte differentiation and immune responses that represent critical differences between benign and malignant melanocytic lesions,” Dr. Rock says.
Through an analytic algorithm, the test generates a single numeric score that can be used as an adjunctive tool in the diagnosis of melanocytic lesions.
“I think the simplicity of the scoring system, combined with the objectivity of the quantitative RT-PCR based assay, has the potential to improve the reliability of melanoma diagnoses made in pathology labs using traditional methods,” Sancy Leachman, M.D., Ph.D., chairwoman, department of dermatology, Oregon Health & Science University, Portland, Oregon, tells Dermatology Times. She was principal investigator for the validation trial.
In the clinical validation study, the 437 melanocytic lesions analyzed had diagnostic scores ranging from -16.7 to +11.1. These scores distributed in a bimodal fashion around a pre-defined lesion classification cut-off of “0” such that the vast majority of malignant lesions had a score >0 and the vast majority of benign lesions had a score <0. The predictive accuracy of the test was also demonstrated by receiver operating characteristic curve analysis in which the area under the curve was 0.96.
Dr. Leachman has served on the medical and scientific advisory board for Myriad Genetics and was involved in the design of the validation study. She reports no other relevant financial interests.