With an increasing number of effective targeted therapies and immunotherapies for the treatment of melanoma, the possibility of individualized therapy is getting closer to reality. Researchers presented data on different potential pathways for individualized therapy in melanoma at the recent 2024 ASCO Annual Meeting.

Following these presentations, featured expert John M. Kirkwood, MD, was interviewed by Conference Reporter Editor-in-Chief Tom Iarocci, MD. Dr Kirkwood’s clinical perspectives on these findings are presented here.

I think that performing risk stratification and risk quantification is a huge part of our job in the clinic. For melanoma, looking at the pathology is important. For example, if there are T cells in a patient’s primary tumor, that individual may be predisposed to benefit from available immunotherapies. Clinicians should review this with their patients, along with the event curves that are discussed in the 8th edition of the American Joint Committee on Cancer (AJCC) melanoma staging system. However, one needs to realize that the 8th edition only gives us data on a life horizon of 10 years.

I recently saw some patients in the clinic who had early disease; several of them were in their 40s and 1 was in their teens. If they did not have melanoma, their life horizon would be in the 70s and 80s. So, we are talking about a more than 50-year horizon for teenaged patients and considerably longer than 10 years for nearly all of our patients if they did not have melanoma. We should not be thinking about a 10-year horizon. It is important for patients to know that if the data at 10 years show a 23% relapse risk for a stage IIIB melanoma, it is just the beginning of the curve. If we follow it for 10 more years or longer, the curve will continue to fall because melanoma has a long risk arc. We should be thinking about 20- to 30-year data, not 10-year data.

Hopefully, gene expression profiling and immunohistochemistry-driven profiling will continue to be of prognostic value and will give us a better risk-benefit profile. Circulating tumor DNA (ctDNA) can be measured with a probe that captures the most aberrant genes in a patient’s tumor, providing us with a “footprint of the tumor” in the blood. This allows the patient to not have to wait for their positron emission tomography or computed tomography scans, which can only be done at wider 6- or 12-month intervals; they can have assessments of their blood ctDNA every 3 months or even more frequently. While some of us are using ctDNA today, more of us will have the opportunity to use it in the near future.

I think that biomarkers for assessing risk and toxicity are 2 ends of a spectrum that we should be focusing on, not just biomarkers that can identify sensitivity to treatment. For example, as discussed in abstract 9567 from ASCO 2024, if there are no tumor-infiltrating lymphocytes in the tumor, we know that they are in a group of patients who may need more than ICIs. However, if we can develop better, more robust profiles for predicting benefit, and then also identify patients with biomarkers of sensitivity to toxicities, I think that we will make progress exponentially faster by deploying therapies in patients who are most likely to benefit and avoiding treatment in those who are most likely to experience toxicity.

At ASCO 2024, Jeffrey S. Weber, MD, PhD, gave a lovely summary of the 3-year update from the KEYNOTE-942 trial, in which the mRNA-4157 vaccine was added to pembrolizumab in patients with resected melanoma (abstract LBA9512). So-called individualized neoantigen vaccines are of huge interest and build on what we learned from the COVID-19 vaccines. We hope to see this impact the treatment of stage III disease or maybe even a broader swath of the disease.

Although these phase 2 data are limited and involve fewer patients than we will need to really make a firm conclusion, researchers have already launched a phase 3 trial in a much broader group of the melanoma population (abstract TPS9616) comparing pembrolizumab with mRNA-4157 against pembrolizumab with placebo. This trial is anticipated to complete accrual by the third quarter of 2024, and it may take another year after that before we see results. But if the hazard ratio is what it was reported to be in the pilot study (ie, 0.51), it will not take long to see an impact in the phase 3 trial. So, we are talking about a relative difference of, say, 16.6% at 24 months and 19.2% at 30 months. Those are smaller gains seen with the mRNA-4157 vaccine, the neoantigen-specific approach. But I have no reason to think that it will not have an incremental impact over anti–PD-1 therapy alone in adjuvant treatment.