Dr. Singh’s Review Favors DecisionDx in Melanoma GEP Testing

By Sophie Lindqvist

In a large SEER database study spanning thousands of cutaneous melanoma patients across multiple stages, DecisionDx-Melanoma testing showed an absolute mortality difference of greater than 1%, according to Dr. Singh’s review.

That result anchors the central difference between the two gene expression profiling (GEP) assays currently available for cutaneous melanoma. Dr. Singh, speaking in the Expert Perspectives episode “GEP Assays in Melanoma: Comparing the Evidence and NCCN Risk Thresholds,” said the assays do not carry the same evidentiary weight.

Two assays, unequal evidence

Dr. Singh identified two GEP tools currently available for cutaneous melanoma:

• DecisionDx-Melanoma: Validation data support accurate stratification into low-risk and high-risk cohorts. The assay provides an exact percentage risk of recurrence, metastasis and nodal involvement, and Dr. Singh said those estimates have been shown to be accurate.
• Merlin: The test is based on MERLIN_001, which did not meet its primary endpoint. Dr. Singh said Merlin has not demonstrated that patients classified as low risk are truly low risk.

DecisionDx therefore supplies both a risk category and numerical estimates for three clinically distinct outcomes. Merlin, as presented in this review, does not yet have the same validated low-risk claim.

The NCCN 5% benchmark

Dr. Singh anchors his interpretation in NCCN guidance, which defines low risk as a less than 5% likelihood of melanoma in the lymph nodes. He uses that definition as the benchmark for judging whether a test can identify a genuinely low-risk group.

The NCCN threshold gives the nodal estimate particular importance. A melanoma GEP assay must provide more than a reassuring label if its low-risk classification is to align with the guideline’s definition.

Dr. Singh also said he always follows NCCN, while noting that clinicians sometimes move ahead of guidelines. He compared that situation with breast cancer prognostic assays, which took years to enter guidelines before becoming standard of care.

The comparison illustrates his view that guideline entry can lag the development and use of prognostic testing. It does not erase the present difference in validation between the two melanoma assays.

Survival evidence behind the preference

Beyond risk stratification, DecisionDx-Melanoma has demonstrated a survival benefit in the large SEER database study. The analysis included thousands of patients across multiple stages, and tested patients showed an absolute mortality difference of more than 1%.

Dr. Singh also cited prospective evidence from the DECIDE trial. In that study, management changes based on GEP results were associated with improved survival. The prospective evidence connects test-informed decisions with patient outcomes rather than validation alone.

He framed his standard plainly: “A survival benefit is always something he looks for in a test.” For Dr. Singh, that combination of a large database finding and prospective trial evidence carries substantial weight when comparing assays.

What does this change for melanoma testing?

What does this change for a clinician choosing a melanoma GEP assay? Within Dr. Singh’s current evidence review, DecisionDx-Melanoma has the stronger evidentiary support for cutaneous melanoma, while Merlin still requires additional validation data.

That conclusion rests on three findings: accurate low- and high-risk stratification, quantified risks for recurrence, metastasis and nodal involvement, and survival signals from both the large SEER analysis and prospective DECIDE evidence.

By contrast, MERLIN_001 missed its primary endpoint, leaving Merlin’s central low-risk classification unproven in Dr. Singh’s assessment. He described the field as quickly evolving and said he would like to see additional data on Merlin in the future.

The next episode, “Using GEP Results to Guide SLNB and Imaging in Stage I-II Melanoma,” will move into the clinic and examine how GEP results change sentinel lymph node biopsy, imaging and surveillance decisions for stage I and II patients.

Dr. Singh’s broader aim was direct: “Offering cutting-edge precision medicine tools helps ensure patients receive the best possible outcomes.” As additional Merlin data emerge, the practical question is whether its low-risk classification can gain comparable validation footing.