Induction Exposure Dose of Ipilimumab and Failure of Adjuvant Nivolumab Plus Ipilimumab in Melanoma

The approval of ipilimumab (anti–cytotoxic T-cell lymphocyte [CTLA]4 Ab) in 2011 initiated the era of modern immunotherapy in cancer and has subsequently led to a dramatic improvement in outcomes for patients with melanoma. With the eventual success of other checkpoint blocking antibodies against programmed cell death protein 1 (PD-1; nivolumab and pembrolizumab) and more recently LAG3 (relatlimab), combination immune checkpoint blockade has become a standard for most patients. Long-term outcomes from randomized phase III trials suggest median survival of patients with metastatic melanoma at more than six and a half years after treatment. The success of immunotherapy for metastatic disease has also led to exploration of these agents as adjuvant therapies where anti–PD-1 and BRAF therapies are now standard-of-care considerations for stage III, with pembrolizumab also approved for stage IIB/C.¹ The report of CheckMate (CM) 915 that this editorial accompanies suggests no improvement in outcomes for adjuvant nivolumab plus ipilimumab versus nivolumab alone,² but with the caveat that ipilimumab was dosed as 1 mg/kg once every 6 weeks, a regimen not previously studied in melanoma. The negative result of this study suggests a need for review and reconsideration of ipilimumab dosing strategies broadly in oncology.

THE TAKEAWAY

• In the article that accompanies this editorial, the authors of CheckMate 915² investigate the role of combined anti–programmed cell death protein 1 (nivolumab) plus anti–cytotoxic T-cell lymphocyte 4 (ipilimumab) with nivolumab maintenance versus nivolumab alone in patients with resected, stage IIIB-D or stage IV melanoma and determine that low-dose ipilimumab did not improve relapse-free survival but did add toxicity. Given the low cumulative induction dose of ipilimumab used in CheckMate 915, these results do not clarify whether the addition of ipilimumab could improve relapse-free survival and potentially call into question infrequent and low-dose ipilimumab combination regimens being used in other tumor types.

Despite the substantial progress for patient outcomes delivered by immune checkpoint blockade, an area that has been understudied is the optimal dose and frequency of ipilimumab administration. A series of trials assessed ipilimumab monotherapy, exploring different doses but maintaining the administration schedule of every 3 weeks for four doses. The phase III MDX010-20 and CA184-024 trials demonstrated the superiority of ipilimumab at 3 mg/kg over gp100 vaccine or dacarbazine, respectively, leading to US Food and Drug Administration approval. Dose comparison studies explored 0.1-10 mg/kg, and the randomized phase III CA184-169 study demonstrated an improvement in overall survival for 10 versus 3 mg/kg, albeit with a doubling of treatment-related adverse events (TRAEs). These data suggest that higher doses of ipilimumab monotherapy are associated with greater efficacy and toxicity, but as a point of emphasis, all explored once every 3-week dosing schedules.³

With the rise of PD-1 inhibitors in advanced melanoma, efforts shifted to the assessment of combination regimens of anti–PD-1 plus CTLA4. Although higher-dose ipilimumab as monotherapy appears superior, lower doses combined with anti–PD-1 potentially maintain efficacy and are associated with lower toxicity. The seminal phase III CM 067 study established standard dose combination of nivolumab and ipilimumab (3 mg/kg) given once every 3 weeks during the induction period. However, a subsequent randomized phase III study (CM 511) called into question the optimal dose of ipilimumab when combined with nivolumab. Efficacy outcomes were similar in this study, with lower toxicity, using 1 mg/kg of ipilimumab. These results suggest that lower ipilimumab dosing (albeit still given once every 3 weeks) alongside PD-1 blockade could limit toxicity but maintain efficacy.⁴

At approximately the same time that lower-dose ipilimumab was being explored in randomized clinical trials, translational evidence emerged, suggesting that the immune-modulatory effects of immune checkpoint blockade may be an early on-treatment phenomenon. More specifically, a proliferative burst in CD8+ T cells seen over the first approximately 2 weeks is associated with improved long-term outcomes. This raised a question then whether fewer doses of nivolumab and ipilimumab may be sufficient to maintain efficacy and limit toxicity. In a randomized phase II trial called ADAPT-IT, one cohort of patients stopped standard-dose nivolumab plus ipilimumab after only two doses if they were demonstrating tumor reduction.⁵ In this study, overall response rate and TRAEs were similar between the early stop versus standard regimens. These results again emphasize that the antitumor immune response associated with ipilimumab is driven by an opening burst or maintaining area under the curve (AUC) for exposure early during treatment.⁵ In the PD-1 failure setting, it is also notable that anti–PD-1 plus ipilimumab given every 3 weeks generates a 30% response rate, whereas anti–PD-1 plus quavonlimab (anti-CTLA4) 0.25 mg once every 6 weeks demonstrated a response rate of only 9%.^6,7 Summarizing these studies of ipilimumab combinations, lower doses of ipilimumab might maintain efficacy with anti–PD-1⁴; however, too low and/or infrequent dosing can sacrifice the benefit.

With the success of anti–PD-1 and anti-CTLA4 for metastatic disease, these approaches have become standard considerations in the adjuvant setting for patients with stage IIB through resected stage IV melanoma. Specifically, the phase III CM 238 trial established superior relapse-free survival (RFS) for nivolumab versus ipilimumab, whereas KEYNOTE-054 and SWOG-S1404 established the RFS advantage for pembrolizumab versus placebo or ipilimumab, respectively. For adjuvant nivolumab plus ipilimumab combination, the only robust data to date have come from the IMMUNED randomized phase II study. This trial demonstrated a major advantage to the combination (given at standard dose) compared with nivolumab alone in patients with resected stage IV disease. A few other pilot studies had also treated small numbers of patients although only one had begun to assess different nivolumab plus ipilimumab dosages (Weber IIT, 2016).⁸ In that study, there appeared to be a minimal difference between ipilimumab at 3 or 1 mg/kg once every 3 weeks and nivolumab, both demonstrating impressive RFS rates. Thus, available data appear to favor dual immune checkpoint blockade in the adjuvant setting, but questions remain over the optimal ipilimumab dose.

On this background, the adjuvant phase III CM 915 study has now been observed to show no improvement in outcomes for nivolumab versus nivolumab plus ipilimumab, when given as 1 mg/kg once every 6 weeks. To inform their dosing strategy, the authors relied on data from CM 012, a phase I study in advanced non–small-cell lung cancer (NSCLC) that assessed nivolumab at 1 or 3 mg/kg every 2 weeks combined with ipilimumab at 1 mg/kg once every 6 weeks or 12 weeks.⁹ CM 012 suggested favorable efficacy and safety profiles but similar response rates across the modestly sized treatment groups. Given what was perceived to be retained efficacy with lower and infrequent ipilimumab dosing in NSCLC, the investigators of CM 915 selected a dose of ipilimumab 1 mg/kg once every 6 weeks to be administered with nivolumab. This was a substantially lower cumulative ipilimumab dose than previous studies in melanoma. As shown in Figure 1, CM 915 administered a relatively meager AUC induction of ipilimumab as compared with the larger doses seen in IMMUNED and the Weber IIT (both using every 3-week dosing). These two later studies notably demonstrated not only higher 2-year RFS but also greater TRAEs. A debate surrounding ipilimumab since early in clinical development has centered on whether toxicity is linked to efficacy. Multiple analyses have investigated this with a consistent suggestion that minimal efficacy is observed if there is no immune-related toxicity. Although patients can benefit without experiencing high-grade TRAEs, those who do experience them have outcomes that are at least as good as those who do not.

FIG 1.

Cumulative dosing of ipi and nivo (or pembrolizumab) during the 12-week induction phase across six studies in patients receiving adjuvant treatment for resected (stage III/IV) melanoma. Two-year RFS (teal line) and grade 3/4 TRAEs (gray line) are also shown for each study. ^aPembro. ^bCumulative induction phase dosing (mg/kg) calculated on the basis of a 70-kg patient. ICB, immune checkpoint blockade; ipi, ipilimumab; nivo, nivolumab; RFS, relapse-free survival; TRAE, treatment-related adverse event.

Returning to CM 915, adjuvant nivolumab plus ipilimumab versus nivolumab alone demonstrated a 2-year RFS of 64.6% versus 63.2%, respectively, with 53.6% versus 52.4% in the programmed death-ligand 1 < 1% cohort. The latter stratification assessed the role of anti-CTLA4 augmentation in poorly immunogenic tumors. This was pursued given that the benefit to ipilimumab combination was most obviously observed in the advanced disease study CM 067 for programmed death-ligand 1 < 1%. Grade 3-4 TRAEs were reported in 32.6% versus 12.8% of the nivolumab plus ipilimumab versus nivolumab cohorts, respectively, although the authors highlight similar rates of early discontinuation because of toxicity. The latter results prompt the authors to dismiss early cessation as a confounding variable that could compromise survival benefit in the nivolumab plus ipilimumab group. An alternative view, however, could be that the similar rates of discontinuation may reflect an inadequate AUC during the priming phase of treatment and not enough toxicity from ipilimumab. For example, the RFS benefit seen in the IMMUNED trial was also accompanied by a greater proportion of early treatment cessation in the standard-dosed ipilimumab combination cohort. In the IMMUNED trial, patients received a median of two doses, notably consistent with the efficacy window described in ADAPT-IT.

The results of CM 915 are informative to the field of adjuvant melanoma and potentially other cancer types. These data emphasize the need for optimization of ipilimumab dose and schedule as monotherapy, but particularly in combination with anti–PD-1. Considering the totality of data from studies in melanoma, the cumulative ipilimumab induction dose used in CM 915 was too low and seemingly negated the potential benefits of combination therapy. More broadly, these results call into question ipilimumab dosing regimens in other cancers pursuing administration at intervals longer than 3 weeks (ie, NSCLC and mesothelioma).

AUTHOR CONTRIBUTIONS

Conception and design: All authors

Financial support: Jason J. Luke

Administrative support: Jason J. Luke

Provision of study materials or patients: Jason J. Luke

Collection and assembly of data: All authors

Data analysis and interpretation: All authors

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Induction Exposure Dose of Ipilimumab and Failure of Adjuvant Nivolumab Plus Ipilimumab in Melanoma

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jco/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).