DecisionDx-Melanoma Outperforms MSKCC Nomogram in Predicting Sentinel Lymph Node Positivity in Melanoma

DecisionDx-Melanoma test identifies more patients who can safely forego SLNB than using current guidelines alone or a clinicopathologic-only nomogram.

Castle Biosciences, Inc.’s DecisionDx-Melanoma outperformed a nomogram developed at the Memorial Sloan Kettering Cancer Center (MSKCC) in predicting the risk of sentinel lymph node (SLN) positivity in patients with cutaneous melanoma (CM), according to a new study in Anticancer Research.

The study evaluated the performance of two tools used to identify patients at low and high risk of SLN positivity:

1. Castle’s DecisionDx-Melanoma test, which uses advanced algorithms to integrate a patient’s clinical and pathologic factors with his/her tumor biology to provide a personalized prediction of the risk of SLN positivity, and melanoma recurrence and metastasis.
2. The MSKCC nomogram, which uses logistic regression and only clinical and pathologic factors to predict a patient’s SLN positivity risk.

Patients from previously published multicenter cohorts with T1-T2 tumors who had undergone the SLNB procedure (n=465) were analyzed using both DecisionDx-Melanoma and the MSKCC nomogram. Following current National Comprehensive Cancer Network (NCCN) guidelines, a risk prediction of less than 5% in the study was considered low risk for SLN positivity, where patients could safely forego the SLNB procedure. A true-to-false-negative ratio was evaluated using this 5% risk threshold. A 5% threshold represents 19 truly negative nodes for one positive SLN missed (19:1 ratio; 1/20 [5%]), meaning that for every 100 patients who avoided SLNB based on NCCN criteria, 5 would have had a missed positive SLN.

The DecisionDx-Melanoma test resulted in a 36:1 true-to-false-negative ratio (108/3), meaning that for every 100 patients who avoided SLNB based on the test’s results, only 2.7 would have had a positive SLN. This is well below the 5% low-risk threshold established by NCCN. DecisionDx-Melanoma’s performance was better than that of the MSKCC nomogram, which resulted in a 9:1 true-to-false-negative ratio, suggesting that for every 100 patients avoiding SLNB using the MSKCC nomogram, 10 would have had a positive SLN. DecisionDx-Melanoma also demonstrated better accuracy in predicting SLN positivity, including higher sensitivity (95% vs. 81%) and negative predictive value (97% vs. 90%) than the MSKCC nomogram.

Importantly, in patients with T1 tumors, for whom the decision to perform SLNB is least clear, using the DecisionDx-Melanoma test to help guide decision-making could have reduced the number of SLNBs by 43.7%, compared with standard NCCN SLNB guidance using American Joint Committee on Cancer staging, while maintaining a low false-negative rate.

 “By providing an accurate prediction of a patient’s likelihood of having a positive SLN, DecisionDx-Melanoma test results can inform important conversations between clinicians and patients and provide added confidence in decisions to proceed without the surgery if a patient’s risk is low,” says Michael Tassavor, M.D., study author, board-certified dermatologist and fellowship-trained Mohs surgeon currently practicing in New Jersey, in a news release.

DecisionDx-Melanoma has been validated to identify patients who have less than a 5% risk of a positive SLN, indicating that these patients may consider avoiding the SLNB surgical procedure.