Dr Weiss Reviews Key Melanoma Updates From ASCO 2025

At the 2025 American Society of Clinical Oncology (ASCO) Annual Meeting, a number of melanoma-focused studies were presented, offering updates on current approaches and ongoing challenges in the field. In an interview with Targeted Oncology, Sarah Weiss, MD, Director of the Melanoma/Cutaneous Oncology Program at Rutgers Cancer Institute and RWJBarnabas Health, reviewed several key abstracts.

Here, we focus on findings from the RELATIVITY-098 trial (NCT05002569),¹ the SWOG S2000 study (NCT04511013),² and the DANTE trial (ISRCTN15837212),³ highlighting clinical implications, safety considerations, and areas where further research is needed.

Targeted Oncology: Starting with RELATIVITY-098, can you give some background about this study and what was being looked at?
Sarah Weiss, MD: In high-risk melanoma that’s been completely resected, the current standard of care is 1 year of adjuvant therapy with a PD-1 inhibitor to try to reduce the risk of recurrence.Despite this treatment, some of our patients will still develop melanoma recurrences, so there is a need to further optimize adjuvant therapy strategies. Some prior adjuvant studies have tried adding a second immune-based therapy, such as ipilimumab [Yervoy], to a PD-1 inhibitor, but those results were negative. RELATIVITY-098 is the first study to report on whether the addition of a LAG-3 inhibitor to a PD-1 inhibitor improves recurrence-free survival compared toa PD-1 inhibitor alone in resected stage III to IV melanoma.

We know from RELATIVITY-047 that in the metastatic setting, the combination improves outcomes compared to nivolumab [Opdivo] alone. But in this study, the addition of relatlimab [Opdualag] in the adjuvant setting actually did not improve recurrence-free survival. It is not entirely clear yet why that is, but biomarker studies are ongoing. There is a similar Regeneron adjuvant study using their LAG-3/PD-1 combination of fianlimab and cemiplimab [Libtayo] compared to PD-1 monotherapy. The accrual is complete, but the results are still pending. It will be interesting to see if that study confirms or differs from the RELATIVITY-098 findings.

Were there any safety considerations that are particularly notable?

The safety profile was relatively consistent with what we typically expect in each arm. However, in the combination arm [PD-1 and LAG-3], there was a higher incidence of endocrinopathies, particularly adrenal insufficiency requiring steroids and thyroid dysfunction. These are often long-lasting or permanent adverse events. Also, there were slightly more cases of neurologic or muscle-related toxicities, like myositis or myocarditis, with the combination. These are particularly concerning, as they can be difficult to manage and can lead to significant morbidity.

How might these findings influence future research for patients with completely resected high-risk melanoma?

We really need to understand what we’re doing with adjuvant therapy. We do not have effective biomarkers to help guide patient selection for who will benefit most from adjuvant therapy, as not all do. Future research should focus on blood-based or tumor-based biomarkers—for example, from the sentinel lymph node, circulating tumor cells, or immune cell populations or other novel markers—to help predict which patients need adjuvant therapy most. This study did suggest that resected melanoma patients may have lower levels of circulating LAG-3–positive T cells compared with metastatic patients—so maybe the target just was not there. That’s an interesting point and needs to be explored further.

Moving onto the SWOG S2000 study, can you provide some background and discuss the key takeaways that were presented at ASCO regarding this trial?

This was a small, randomized phase 2 study for patients with symptomatic melanoma brain metastases—a population typically excluded from clinical trials. Historically, outcomes with immunotherapy have been poor in these patients. Even with ipilimumab/nivolumab, progression-free survival is short. This study included patients with BRAF-mutant melanoma and symptomatic brain metastases, and steroids were allowed.Patients were randomized to receive either ipilimumab/nivolumab or the triplet combination of nivolumab plus targeted therapy with encorafenib [Braftovi] and binimetinib [Mektovi]. Although small, the study suggested that targeted therapy combined with a PD-1 inhibitor may improve progression-free survival compared to ipi[limumab]/nivo[lumab] in this population. This is a difficult group to treat, and we’re trying to understand how to sequence or combine targeted therapy with immunotherapy.

Were there any surprising toxicity differences between the arms, and how do you think these findings can affect real-world manageability of each regimen?
Toxicity was quite high in the triplet arm, which isn’t surprising based on past studies with similar regimens. These patients are already symptomatic and dealing with metastatic disease, so adding a high rate of grade 3 toxicity on top of that is challenging. Many patients needed dose modifications of targeted therapy due to toxicity. So, while the results are promising, we still need to optimize these regimens. Future studies should assess whether upfront targeted therapy for perhaps 4 to 8 weeks can achieve enough disease control to allow a switch immunotherapy at that point.

The treatment of brain metastases really requires a multidisciplinary approach involving neurosurgeons, radiation oncologists, and medical oncologists. This is a population that’s often excluded from studies, so this is a really commendable effort. We need more trials focused on this group that represent real world practice.

Looking at the DANTE trial, would provide some background at what this study looked at?

This was a study done in the [United Kingdom] to explore the optimal duration of immune checkpoint inhibitors in advanced melanoma. The study had many challenges, but essentially, once patients had been on a PD-1-based immunotherapy for 1 year, they were randomized to either continue therapy up to 2 years or stop after 1 year and be monitored. The study was significantly underpowered to draw definitive conclusions, but from the data available, 1 year of therapy didn’t appear inferior to 2 years. However, patients were on different treatment regimens, and the study group wasn’t uniform. COVID-19 impacted accrual and logistics. Many patients tolerating therapy didn’t want to stop. That said, it’s an important study because the question of duration affects not only patient outcomes, but also health care costs, resource use, and long-term toxicities.

Are there any clinical or biological markers that might help identify patients who can safely discontinue therapy early?
There are several things being looked at, and this area definitely needs more research. PET scans have been studied to see if they can help predict when patients can stop therapy. Circulating tumor DNA has been retrospectively analyzed and is being looked at in clinical trials. These tools, possibly combined, might help us define who can safely come off treatment. What we’ve learned from earlier trials is that achieving a confirmed complete response is key. If a patient has consistent imaging showing complete response, the chance of recurrence is very low. For those patients, it’s likely safe to stop therapy early. The harder question is around patients who still have stable residual disease—when is it okay to stop therapy for them? Also, we don’t fully understand if certain patients like those with brain metastases need longer therapy or if their treatment duration should be different from those with extracranial disease.

Is there anything else about this trial that you think is important for a community oncologist to know?

I think the takeaway is that the optimal duration of immunotherapy in advanced melanoma is still unclear. Traditionally, treatment has been continued for 2 years, but that is not necessary for all patients. We just don’t have the data yet to say definitively when to stop, although other studies such as the PET-Stop trial [NCT04462406] are ongoing. So, decisions should be individualized, taking into account tolerability, patient comorbidities, and treatment response. PET scans and biopsies can be helpful to confirm disease status. Circulating tumor DNA could also be considered, though it’s not yet standard. But overall, we need more data to guide how we act on these findings.

What else from ASCO did you find to be particularly interesting or practice-changing?

One key takeaway is the growing and continued focus on neoadjuvant therapy in melanoma. Treating patients while the tumor is still in place seems to yield better outcomes. There were many small neoadjuvant trials this year, each with various therapies and small patient numbers. While these studies show promise for some patients, we still don’t know the optimal regimen or how to match therapies to specific patients. The one consistent thing is that PD-1–based immunotherapy is appropriate. Another takeaway is that cell therapy, especially tumor-infiltrating lymphocytes, is being optimized in early-phase trials and remains a promising strategy. Lastly, across melanoma and other tumor types, there seems to be a renewed interest in VEGF inhibition, which is something to watch closely.

REFERENCES:

1. Long GV, Ascierto PA, Guo J, et al. Nivolumab plus relatlimab vs nivolumab alone for the adjuvant treatment of completely resected stage III-IV melanoma: primary results from RELATIVITY-098. J Clin Oncol. 2025;43(suppl 17):LBA9500. doi:10.1200/JCO.2025.43.17_suppl.LBA9500

2. Eroglu Z, Moon J, Najjar YG, et al. A randomized phase 2 trial of encorafenib + binimetinib + nivolumab vs ipilimumab + nivolumab in BRAFV600-mutant melanoma brain metastases: SWOG S2000 (NCT04511013). J Clin Oncol. 2025;43(suppl 17):LBA9507. doi:10.1200/JCO.2025.43.17_suppl.LBA9500

3. Danson S, Collinson M, Plummer ER, et al. Comparison of 1 year versus minimum 2 years of anti-PD1-based immunotherapy as first-line treatment for metastatic melanoma: Results of the DANTE phase III trial. J Clin Oncol. 2025;43(suppl 17):LBA9508. doi:10.1200/JCO.2025.43.16_suppl.1070