FRIENDSWOOD, Texas, June 25, 2026
Castle Biosciences announced the publication of a prospective multicenter study in Dermatology and Therapy demonstrating that its DecisionDx®-Melanoma i31-SLNB test more accurately predicts sentinel lymph node (SLN) positivity than the Melanoma Institute Australia (MIA) nomogram for patients with cutaneous melanoma (CM). The study evaluated 912 patients enrolled in the multicenter DECIDE trial and showed that integrating 31-gene expression profile (31-GEP) data with clinicopathologic features significantly improves patient risk assessment. The findings strengthen evidence supporting the use of DecisionDx-Melanoma to help clinicians identify patients who may safely avoid sentinel lymph node biopsy (SLNB) while ensuring higher-risk patients receive appropriate surgical evaluation.
DecisionDx-Melanoma Identifies Low-Risk Patients More Accurately
The study found that patients classified as low risk by the i31-SLNB test had an observed SLN positivity rate of only 2.6%, well below the National Comprehensive Cancer Network (NCCN) threshold of 5%, which is commonly used when considering avoiding SLNB. In comparison, patients categorized as low risk by the MIA nomogram had an observed positivity rate of 5.8%, exceeding the NCCN guideline threshold and indicating lower predictive accuracy. Among patients with discordant classifications, those identified as low risk by i31-SLNB but high risk by the MIA nomogram showed an actual positivity rate of just 2.8%, while patients considered low risk by the MIA nomogram but high risk by i31-SLNB demonstrated an 11.5% positivity rate, further highlighting the superior clinical performance of the molecular test.
Gene Expression Profiling Improves Melanoma Risk Stratification
Researchers also reported that DecisionDx-Melanoma achieved significantly stronger overall predictive performance than the clinicopathologic-only MIA nomogram, with an area under the curve (AUC) of 0.74 compared with 0.61 (p=0.001). The test combines tumor biology through a 31-gene expression profile with selected clinicopathologic factors to provide personalized assessments of both SLN positivity risk and the likelihood of melanoma recurrence or metastasis. Because up to 88% of SLNB procedures produce negative results and approximately 15% of patients experience surgery-related complications, improved risk prediction may help reduce unnecessary surgical procedures while maintaining appropriate care for patients at elevated metastatic risk.
Advertisement
Growing Clinical Evidence Supports Precision Melanoma Management
The newly published prospective findings add to an expanding body of evidence supporting DecisionDx-Melanoma, including multiple peer-reviewed studies, prospective clinical trials, and previous multicenter analyses showing the test consistently outperforms conventional risk prediction models. Developed in collaboration with more than 100 leading U.S. institutions, the assay has been clinically validated in over 10,000 patient samples and ordered more than 240,000 times since its commercial launch. By integrating molecular profiling with traditional clinical factors, Castle Biosciences continues to advance precision oncology, providing clinicians with actionable information to guide SLNB decisions, surveillance strategies, imaging recommendations, and personalized treatment planning for patients with stage I–III cutaneous melanoma.
Source: Castle Biosciences, press release
