Breaking Down the Benefit of Neoadjuvant Nivolumab/Ipilimumab in Stage III Macroscopic Melanoma

Neoadjuvant treatment with nivolumab (Opdivo) plus ipilimumab (Yervoy) elicits high pathologic response rates in patients with stage III macroscopic melanoma, according to Christian U. Blank, MD, PhD, who added that it is important to consider this approach prior to surgery, as this could potentially induce a broader immune response than might be achieved in the adjuvant setting.

“In the neoadjuvant setting, you can really see whether the patient has benefited from immunotherapy, and if not, you can even adjust the adjuvant therapy,” said Blank. “For example, if [a patient has a] BRAF mutation, you can switch them to a BRAF or MEK inhibitor; you give the patient a second chance. If you don't consider neoadjuvant therapy [prior to surgery], you take away that patient's chance [of potentially experiencing a better outcome]; this is very important.”

In the phase 1b OpACIN trial, investigators examined both adjuvant and neoadjuvant ipilimumab plus nivolumab in patients with stage III B/IIIC macroscopic melanoma; this was the first trial in the field to evaluate checkpoint inhibition in the neoadjuvant setting, according to Blank.

Updated data from the trial, which were presented during the 2020 AACR Virtual Annual Meeting II, showed that after 3-years of follow-up, patients who were given the combination in the neoadjuvant setting experienced a relapse-free survival (RFS) of 80% versus 60% of those who received it in the adjuvant setting.

Data from the subsequent OpACIN-neo trial were also released at the meeting. Here, the median RFS had not yet been reached in either arm.

After a median follow-up of 36 months for OpACIN and 18 months for OpACIN-neo, only 1.4% of patients with a pathologic response on neoadjuvant therapy relapsed compared with 65.2% of those without a response. “We have shown that the [patients with a] pathologic response rate are the ones who do not relapse,” stressed Blank. “This is important.”

In an interview with OncLive, Blank, a medical oncologist within the Division of Immunology at The Netherlands Cancer Institute, discussed the promise of neoadjuvant ipilimumab/nivolumab and research efforts examining interferon-gamma signature and tumor mutational burden (TMB) as potential predictors of response.

OncLive: Could you provide some background on the efficacy of adjuvant ipilimumab in patients with high-risk stage III melanoma? What is the rationale for neoadjuvant immunotherapy in this patient population?

Blank: First of all, we have nice adjuvant therapies with targeted treatments such as dabrafenib (Tafinlar)and trametinib (Mekinist), and with immunotherapies, such as pembrolizumab (Keytruda) and nivolumab. The RFS for patients with stage III melanoma in the macroscopic setting ranges from 60%-70%. However, the problem is that all the trials [conducted in this setting] have excluded early relapses. Patients who relapse following surgery, before they started adjuvant therapy at week 12 were excluded from these trials. The RFS in the intention-to-treat population, or what you call event-free survival, is about 15%-25% lower than what we have achieved for these patients [who relapse]. This is very important if you want to compare [trials in the] adjuvant [setting] with these new trials in the neoadjuvant setting. 

With neoadjuvant therapy, there is the advantage that [when] you start the immunotherapy, you still have the whole tumor in the patient; that means you can induce a broader and more intense immune response than if you remove everything and [are left with] the smoldering small tumor cells that are sometimes not immunogenic, [which could result] in a lower immune response. We have shown this before in the OpACIN trial, where [in the neoadjuvant arm], we compared ipilimumab plus nivolumab [2 doses] before surgery and then analyzed [2 doses of the combination] after surgery.

When you compare the tumor-specific T-cell clones that were expanded in the periphery, you saw that with the neoadjuvant approach, more clones were expanded. We also found that more clones that were initially not detectable in the periphery, became detectable; this is very important because if you don't expand these T-cell [clones] very well, then [the patient] will relapse.

Could you discuss the OpACIN and OpACIN-neo trials and the updated data presented during the 2020 AACR Virtual Annual Meeting II?

OpACIN was the first trial in melanoma that tested checkpoint inhibition. [This approach] was previously tested in the neoadjuvant setting for non–small cell lung cancer. Now, with the 3-year update, we see that none of the patients who had a pathologic response relapsed. As such, the pathologic response at week 6 after these 2 courses of neoadjuvant therapy was a very good prognosticator for long-term outcomes; that’s important because we hope that the FDA will soon accept such early outcome markers as a long-term outcome because then patients can get earlier access to this very promising treatment. This study looked at a very small patient population; therefore, we [conducted] the OpACIN-neo study.

In OpACIN, we [gave] ipilimumab [at 3 mg/kg] and nivolumab [at 1 mg/kg] and this was pretty toxic; the rate of grade 3/4 toxicities was 90%. Therefore, we tested different [dosing] schemes in OpACIN-neo: 1 was sequential and 1 gave only 2 courses instead of the 4 courses [of therapy] that we had given in OpACIN. We also lowered the dosing of ipilimumab and increased the dosing of nivolumab, so ipilimumab was given at 1 mg/kg and nivolumab was given at 3 mg/kg. By doing this, we found that we could lower the [rate of] grade 3/4 toxicity to 20%, while preserving the pathologic response rate, which was 77% in this trial, compared with the 80% [observed with the previous dosing]; it was equally effective.

Only 2% of responders relapsed per the updated data—while 62% of non-responders relapsed. As such, at week 6, you can tell your patients, “You have a very good response so you can plan your life: Get your house, get married, and [have] children.” Unfortunately, with your patients who do not respond, you have to follow up very strictly; they have a big chance of relapse.

We are currently working on [identifying] baseline markers that, in the future, [will tell us not to] treat [certain] patients anymore with ipilimumab/nivolumab, [that we should use an] alternative [approach]. In the future, we [hope to] identify favorable patients [at baseline] who receive the treatments we are testing at the moment. For the patients who have unfavorable tumors, [we want provide them with] access to alternative new trials where we can give them triplet or quadruplet therapies in the hope that we can increase their response rates, as well.

Could you elaborate on the findings from the translational analyses examining TMB and interferon-gamma gene expression score at baseline? How might these factors serve as predictors of the response?

With our translational research, we [set out to] find baseline biomarkers to predict which patients are favorable and which will respond [to treatment]. If [a patient] has a high interferon signature and a high TMB, we found that these are independent markers. I expected that the high TMB would also be high in the interferon signature; however, this was not the case. These are independent biomarkers. However, if both are favorable, then you have a 100% response rate; this is beautiful. If [1 of the 2 biomarkers] is favorable, either high interferon signature and low TMB or high TMB and low interferon signature, then you see a response rate of 80% to 90%, which is still [amazing]. If both are low, then you see a response rate of 37%; this is the group of patients for which you’ll want to extend treatment, [by giving a] triplet therapy or another alternative strategy in an attempt to bring [the response rate] into the range of 80%-90%.

We are working on that already with a personalized trial [called] the DONIMI trial. Based on the interferon signature, we will extend treatment for patients [with unfavorable biomarkers]. [We will also look at] de-escalated [therapy] for patients who have very high interferon signature. For them, we would try just an anti–PD-1 agent only; this elicits a good response, but the percentage of response is low. [Therefore], if you can identify the patients [for whom this is appropriate], then they can benefit.

This is the future [of melanoma, where] we will be able to personalize [treatment] based on [certain] signatures. You will be able to say to your patient, “You have this signature, and therefore, you need this treatment.” Then, we will hopefully no longer have so many patients with stage IV disease anymore.

What are the next steps for this research?

In the melanoma field, under the lead of the International Neoadjuvant Melanoma Consortium, different trials with different combinations are already ongoing. For example, [some are agents] are being examined in [triplet regimens] with targeted therapies or with an HDAC inhibitor.

We already have small trials that have the same design. The next step will be to clearly define the responders and non-responders in these different trials. For example, by looking at RNA signature, [we will try to] see where there is overlap and where there are differences. When we have differences, then we can say, “This is a patient group that should be treated preferentially by this combination,” or if we have another patient for which there is overlap, then you want to learn which of the 2 [approaches is] better.

This is the future; we will really work to define which patient needs which treatment up front. Melanoma is [leading the way with this]; I cannot think of any other malignancy where we are on the [verge] of personalized neoadjuvant therapy.

What is your take-home message regarding this research?

First, please consider neoadjuvant [therapy as an option]. Mainly in the United States, patients are still operated on first and then go onto the adjuvant trials, as they are referred to the centers. This puts patients at a disadvantage because in the adjuvant setting, you can't see whether the patient benefits [from immunotherapy] or not; you have to blindly give [immunotherapy] and hope that there is a benefit.

The second thing is—and this is mainly to the FDA: Please consider pathologic response [as a surrogate for] long-term [outcomes] and approve these combinations [earlier]. The quality from immunotherapy [with regard to] pathologic response is way better for patients than [it is] from targeted therapy or chemotherapy; the correlation is much better. In that way, this is a really excellent surrogate marker.

Reference

1. Blank CU, Versluis JM, Rozerman EA, et al. 36-months and 18-months relapse-free survival after (neo)adjuvant ipilimumab plus nivolumab in macroscopic stage III melanoma patients—update of the OpACIN and OPACIN-neo trials. Presented at: 2020 AACR Virtual Annual Meeting; June 22-24, 2020; Virtual. Abstract 3412.