Conversations in Gene Expression–Based Testing: A Clinician’s Roadmap to Multidisciplinary Melanoma Care

So let's review the evidence for GEP tests that assesses the risk of disease progression in cutaneous melanoma.

Let's start with a cpgEP test, which includes clinical pathologic features along with eight genes associated with sentinel node metastases in cutaneous melanoma.

Now, an exciting and important recent study that was published in late 2025 evaluated in a prospective fashion the ability of this test to predict sentinel lymph node biopsy positivity.

In this study, the Merlin001 study enrolled just under 2,000 patients.

The total number of enrolled patients who had a sentinel lymph node biopsy was about 1,800 patients, and the assay was successfully completed within approximately 1,700 tests.

So Dr. Zager, this was a very ambitious study and a very well-run and designed study.

Can you tell us a little bit about the results?

In this study, the investigators looked at the T1 to T3 population and saw that there was about an 18% sentinel lymph node positivity rate.

But when dichotomized between high-risk and low-risk cpgEP, we saw a threefold increase of sentinel lymph node positivity.

So we can use this test to identify a patient who are high-risk for sentinel lymph node positivity, and we should absolutely do the procedure versus low risk of sentinel lymph node positivity, now between 5% and 10%, allowing us to now have that surgeon-patient conversation of whether or not we should be proceeding with sentinel lymph node biopsy.

These outcomes are another data point that lets us have that conversation between patient and surgeon, but not to necessarily get bogged down just with the exact percentage, that 5% threshold.

We've arbitrarily decided is where the risk-benefit is low enough to not necessarily recommend sentinel lymph node biopsy or in that gray zone between 5% and 10%, but those are numbers that we've created.

Now, cpgEP is not the only gene expression profiling test available.

There's also a 31 GP test where there's 28 discriminate genes and three control genes tests validated to evaluate the risk of metastases as a prognostic endpoint.

This test looks at recurrence-free survival or distant metastasis-free survival, different than the endpoint of sentinel lymph node biopsy.

It's important to know how they were developed and with which endpoint.

There has been a recent study looking at the 31 GP test as it relates to sentinel lymph node biopsy, the DECIDE study, a fairly different study than the Merlin study.

This was a observational study that was not blinding of that test and the therapeutic approach.

So there is some caution when we compare these and make conclusions.

I definitely agree with you in that it's hard to compare a prospective blinded study to an observational study.

And certainly 31 gene expression profile was really not created to predict sentinel lymph node positivity.

So this is kind of a byproduct of their data.

Then lastly, there's the NIVO-MELA trial, which is a randomized controlled prospective multicenter phase III trial with biomarker-based risk stratification.

These patients are stratified by high risk and low risk to see if they need adjuvant immunotherapy or just observation, and then the low-risk patients will be just followed up for five years.

This study's specifically gonna look at overall response rates, progression-free survival, and overall survival, and obviously any safety signals.