Castle’s DecisionDx-Melanoma gene expression profiling (GEP) test can help predict mortality risk among patients with stage I–III cutaneous melanoma (CM), according to research presented at the 2025 American Society of Clinical Oncology (ASCO) Annual Meeting in Chicago, IL.

The 31-gene expression profile test (31-GEP) identifies patients at a low (Class 1A), intermediate (Class 1B/2A), or high risk (Class 2B) of tumor recurrence and metastasis.

Higher 5-Year MSS

In the study, patients with a Class 1A 31-GEP result had a significantly higher five-year melanoma-specific survival (MSS) than those with Class 1B/2A or Class 2B results. Multivariable analysis revealed that a Class 2B result, a Class 1B/2A result, a positive lymph node status, Breslow thickness, ulceration, age, and mitotic rate were significant predictors of melanoma-specific mortality (MSM).

Among all stages, 31-GEP-tested patients had a lower median survival time (MSM) than untested propensity score-matched patients.

Further, testing with DecisionDx-Melanoma was associated with a 32% reduction in mortality risk compared to untested patients, the study showed.

“Management decisions for melanoma patients, such as referrals for sentinel lymph node biopsy and surveillance intensity, are guided by a patient’s risk of dying from their disease,” says study author Merve Hasanov, MD, Oncologist and Director of the Division of Medical Oncology at The Ohio State University Comprehensive Cancer Center in Columbus, OH, in a news release. “The findings being presented at ASCO demonstrate that the DecisionDx-Melanoma test can provide more precise mortality risk predictions based on a patient’s unique tumor biology and can improve upon the population-based risk assessment provided by the American Joint Committee on Cancer (AJCC) staging. This can, in turn, help clinicians tailor treatment approaches to individual patient needs.”

For the study, researchers compiled data from the SEER registry for patients with stage I–III CM (2013–2019) who were linked to patients with 31-GEP test results provided by Castle Biosciences (N = 13,560) using an independent third-party data source. Survival differences between 31-GEP-tested and untested patients were assessed by matching the two groups according to clinicopathological factors, diagnosis year, ethnicity, and socioeconomic status.