Nivolumab/Relatlimab Shows PFS/OS Benefit Over Nivolumab in Melanoma

Michael B. Atkins, MD (Moderator)

Deputy Director

Georgetown Lombardi Comprehensive Cancer Center

Scholl Professor and Vice Chair

Department of Medical Oncology

Georgetown University Medical Center

Washington, DC

Targeted Oncology: What data support the combination of nivolumab and relatlimab (Opdualag) for patients with metastatic melanoma?

MICHAEL B. ATKINS: The RELATIVITY-047 trial [NCT03470922]compared nivolumab at 480 mg and relatlimab 160 mg in a fixed-dose combination with nivolumab [Opdivo] monotherapy every 4 weeks, stratified by LAG-3 status, PD-L1 status, BRAF status, and American Joint Committee on Cancer melanoma stage.¹

The median was 63 years of age, which is typical for a melanoma study population. Not much older and certainly not much younger. Sixty percent were male, which is pretty typical.

There were a small number of patients who had M1D [stage] disease, and those were patients with treated brain metastases. [Approximately two-thirds] had an ECOG performance status of 0. The majority of patients had normal LDH [lactate dehydrogenase]. Approximately 45% had an elevated LDH, mostly 2 times above the upper limit of normal. Most patients had LAG-3 expression greater than 1%. Forty percent had PD-L1 expression greater than 1%, and about 40% had BRAF mutations.

Which efficacy points were initially reported for these patients?

The progression-free survival [PFS] was the primary end point by blinded independent central review, and the median PFS’s are much different: 10 months for nivolumab/relatlimab, 4.5 months for nivolumab. I think it's a little bit of an exaggeration because the [Kaplan-Meier] curves are starting to get flat right around the 50% point, and I prefer to look at the differences between the curves at the landmark. It's approximately an 11% difference at 1 year, and 9% difference at 2 years, and although the curves appear to be coming together as you get out to 3 years, the numbers are small. There was a PFS HR of 0.78 [favoring nivolumab/relatlimab], with a median follow-up of 19.3 months.

[What] I think is interesting is that every group seems to have better PFS for the nivolumab/relatlimab arm compared with nivolumab, with exception of the patients with high PD-L1 expression.² One wonders if you have [high] PD-L1 expression on your tumor, maybe you just need to block that pathway to activate the immune system. Or if you don't have PD-L1 expression on your tumor, maybe there are other factors, and one of those could be LAG-3 expression that's preventing the immune system from attacking that tumor.

But once you activate the immune system by blocking LAG-3, you have to keep PD-L1 blocked because PD-1 will become expressed on the T cells…[and] interferon and PD-L1 will be on the tumor cells, and they'll shut off your immune reaction. These data might argue that we want to look at PD-L1 expression on the tumors in some patients in making the decision.

I also want to call attention to the BRAF mutation status…. There is no difference in nivolumab/relatlimab versus nivolumab [based on] the HR between the BRAF wild-type and the BRAF-mutated populations.² That was not the case [in CheckMate 067; NCT01844505] with nivolumab plus ipilimumab [Yervoy].³ The HR compared with nivolumab in the BRAF-mutated population is 0.59 and it's similar to this 0.78 in the BRAF wild-type patients.

What level of PD-L1 expression did the trial use in the high PD-L1 subgroup? What does the difference in efficacy for high PD-L1 and other subgroups suggest?

Over 1% is all we know.¹ That's not really something we typically follow, and if we went to over 5%, there's only about 20% of patients using the assay who have PD-L1 expression at a greater than 5% level.

I think it means that the efficacy is better for nivolumab in that patient population. This is a comparison in these HR plots of nivolumab activity to nivolumab/relatlimab, and we can say relatlimab doesn't add anything to PFS in that population. But it may be hard to sort out, because probably what it means is that in the PD-L1–positive group, that PFS [Kaplan-Meier] curve tends to flatten over the 50% mark where the nivolumab/relatlimab is.

[We] see that 78 of 147 patients have yet to progress on nivolumab, and 80 out of 146 on nivolumab/relatlimab.

The baseline LDH doesn't seem to make a difference.² Patients with [high] LDH tended to benefit, with similar benefit from nivolumab/relatlimab as those who maintained LDH. The melanoma stage categories are small, so we don't know [if that matters]. Performance status doesn't make a difference. Tumor burden doesn't seem to make much of a difference.

How did patients do on the RELATIVITY-047 trial in terms of overall survival (OS) and response rates?

OS data were reported out later.¹ That is a secondary end point, and the HR favored nivolumab/relatlimab for OS at 0.80 compared with nivolumab. But once again, I tend to focus on the margins between the [Kaplan-Meier] curves, particularly the OS curves, and out at 3 years, there was about a 7% difference in those curves, which is a very similar difference that we see between nivolumab/ipilimumab and nivolumab monotherapy in the CheckMate 067 trial.⁴

The response data were also reported out later, and we saw a 10% difference in the response favoring nivolumab/relatlimab over nivolumab.¹ Even though the overall response rate is lower for nivolumab/relatlimab in this study compared with nivolumab [plus ipilimumab], the difference is about the same. It was 13 percentage points difference for nivolumab/ipilimumab versus nivolumab [in CheckMate 067], and it's about 10 percentage points difference [in RELATIVITY-047].^1,4,5 There was a 16% complete response rate [with nivolumab/relatlimab], [whereas] there was around 22% for nivolumab/ipilimumab, but some of those complete responses have been [reported] later.^1,4 The disease control rates were a bit higher with nivolumab/relatlimab compared with nivolumab.¹ Duration of response was not reached [at 19.3 months median follow-up].

What were the outcomes in terms of safety and tolerability?

In the safety summary for this [trial], we can see that the rate of treatment-related adverse events of grade 3 and 4, which is what we tend to focus on, was 21% for nivolumab/relatlimab, and only 11% for nivolumab. It seemed like the toxicity was less for [single-agent] nivolumab in this study than in the CheckMate 067 study.^1,4 Maybe that's because physicians were getting better in terms of managing it and jumping on it quicker, or maybe it's because it was a different patient population.

There is not a lot to look at when you look at the grade 3/4 toxicity types with nivolumab/relatlimab. In particular there's only 1% incidence of [grade 3 or 4] diarrhea, and no grade 3 or 4 hypothyroidism, but patients can [develop] hypothyroidism without getting grade 3 or 4.1 Looking at some more [immune-mediated] toxicities, there was 1% [grade 3 or 4] colitis, and the biggest [grade 3 or 4] toxicity was hepatitis, which was 4%.

REFERENCES:

1. Tawbi HA, Schadendorf D, Lipson EJ, et al. Relatlimab and nivolumab versus nivolumab in untreated advanced melanoma. N Engl J Med. 2022;386(1):24-34. doi:10.1056/NEJMoa2109970

2. Tawbi HA, Hodi SF, Lipson EJ, et al. Nivolumab (NIVO) + relatlimab (RELA) versus NIVO in previously untreated metastatic or unresectable melanoma: OS and ORR by key subgroups from RELATIVITY-047. J Clin Oncol. 2022;40(suppl_16):9505. doi:10.1200/JCO.2022.40.16_suppl.9505

3. Wolchok JD, Chiarion-Sileni V, Gonzalez R, et al. Overall survival with combined nivolumab and ipilimumab in advanced melanoma. N Engl J Med. 2017;377(14):1345-1356. doi:10.1056/NEJMoa1709684

4. Wolchok JD, Chiarion-Sileni V, Gonzalez R, et al. Long-term outcomes with nivolumab plus ipilimumab or nivolumab alone versus ipilimumab in patients with advanced melanoma. J Clin Oncol. 2022;40(2):127-137. doi:10.1200/JCO.21.02229

5. Larkin J, Chiarion-Sileni V, Gonzalez R, et al. Five-year survival with combined nivolumab and ipilimumab in advanced melanoma. N Engl J Med. 2019;381(16):1535-1546. doi:10.1056/NEJMoa1910836