Prognostic Gene Expression Profiling and Sentinel Lymph Node Risk Prediction in Melanoma

Conclusions

• Gene expression profiling in melanoma is promising for refining recurrence and sentinel lymph node risk, but current evidence does not show that it improves patient outcomes beyond standard staging or nomogram-based tools.
• The 31-GEP, 11-GEP, and CP-GEP tests can stratify patients into different risk groups, yet their real-world value remains uncertain because management changes based on test results have not been proven to produce better clinical outcomes.
• For T1A melanoma, GEP testing is generally not recommended because the baseline risk of metastasis is low and false reassurance or false positives could lead to inappropriate management changes.
• Predictive performance for sentinel lymph node positivity is reasonably good for some GEP-based approaches, but free clinical-pathologic nomograms appear to perform similarly in many studies and may be more cost-effective.
• The MERLIN/CP-GEP data suggest a low-risk group with a relatively low rate of sentinel node positivity, but the threshold has not been low enough to clearly replace standard clinical decision-making.
• The main unresolved question is not whether GEP tests can generate risk scores, but whether those scores should change treatment, surveillance, or biopsy decisions in ways that truly improve outcomes.
• Prospective randomized trials and head-to-head comparisons across tests and against nomograms are needed before routine widespread use can be justified.
• At present, GEP testing may be useful in selected cases, but its clinical utility, optimal patient population, and value relative to cost remain unproven.