Jason J. Luke, MD, FACP
- In 2011, ipilimumab received regulatory approval for the treatment of advanced melanoma. Due to what was deemed to be inadequate safety follow-up in 10 mg/kg clinical trials, only the dose of 3 mg/kg received approval.
- The phase III CA184-169 trial was launched to address the lingering question surrounding the most efficacious dose for management of advanced disease.
- Phase III clinical trial data for PD-1 antibodies and BRAF-MEK–directed combination therapies support the statement that there is no role for ipilimumab monotherapy (at 3 mg/kg or 10 mg/kg) as a first-line treatment for advanced melanoma. Even so, ipilimumab remains an important therapeutic agent both in clinical practice and for investigation in combination approaches moving forward.
The field of systemic therapeutics for malignant melanoma has changed rapidly over the past 6 years, with 11 regulatory approvals by the U.S. Food and Drug Administration (FDA), including monotherapy agents, as well as combination therapies. These approaches have targeted the molecular signaling nodes of BRAF and MEK, the immune checkpoints CTLA-4 and PD-1 receptors, as well as the first oncolytic virotherapy for cancer.1 Whereas prior to 2011, no therapy was clearly associated with a survival advantage in this disease, since then, the primary endpoint of every major phase III trial has been met. This has led to marked improvement in the overall survival of patients both in the metastatic and adjuvant settings.
The first agent to demonstrate an advantage in overall survival for patients with advanced melanoma was the anti–CTLA-4 antibody ipilimumab. This agent was approved for standard use in 2011 after successful completion of two randomized phase III studies. The first of these, CA184-002, evaluated ipilimumab (3 mg/kg intravenous every 3 weeks for four doses) as a monotherapy versus ipilimumab in combination with a gp100 vaccine versus the gp100 vaccine alone. Although rates of Response Evaluation Criteria in Solid Tumors (RECIST) quality response ranged only from 5.7% to 11.0% in the ipilimumab-containing arms, median survival was improved to 10.0 months relative to 6.4 months for the vaccine alone (HR 0.68; p < 0.001). In the CA184-024 study, ipilimumab was administered in combination with dacarbazine chemotherapy (ipilimumab 10 mg/kg every 3 weeks x four doses and then every 12 weeks and dacarbazine 850 mg/m2 intravenous every 3 weeks for up to 22 weeks) and compared with dacarbazine alone. In this study, the response rate was similar to prior, whereas overall survival was again improved at 11.2 months for ipilimumab plus dacarbazine relative to 9.1 months for dacarbazine alone (HR 0.72; p < 0.001).2,3 Based on these data, in 2011, ipilimumab received regulatory approval for the treatment of advanced melanoma. However, due to what was deemed to be inadequate safety follow-up in 10 mg/kg clinical trials, only the dose of 3 mg/kg received approval. This decision has led to the 3 mg/kg dose being considered “standard”; however, there has continued to be debate in the field about the most appropriate dose of ipilimumab, and clinical trials have continued to investigate the 10 mg/kg dose.
Novel Features of Treatment With Ipilimumab
The development of ipilimumab was novel in multiple ways beyond establishing an overall survival benefit in melanoma. Features of the therapy are in contrast with standard chemotherapy in at least three ways, including an increased potential for long-term benefit, evaluation and management of toxicity, and kinetics of treatment response.
A consistent observation throughout CTLA-4 antibody clinical trials has been that a proportion of patients have had an extended survival independent of RECIST response. In a pooled analysis of 1,861 patients who had participated in the phase II and III clinical trials for ipilimumab, 22% were alive at 3 years, and a plateau on the survival curve suggested the likelihood of long-term survival thereafter.4
The toxicities commonly observed with ipilimumab include inflammatory consequences, such as dermatitis and diarrhea (or colitis), and less commonly hepatitis, uveitis, and hypophysitis. These toxicities have been referred to as immune-related adverse events (AEs), and algorithms have been advanced for their standardized clinical management.5 Rates of AEs with ipilimumab, by Common Terminology Criteria for Adverse Events (CTCAE), appeared to be dose proportional, with 61% (all-grade immune-related AEs) and 15% (grade 3 to 5) at 3 mg/kg in CA184-002, and 78% and 42% at 10 mg/kg in combination with dacarbazine in CA184-024. Table 1 summarizes toxicity rates for ipilimumab across notable clinical trials.
In addition, treatment with ipilimumab is associated with a subset of patients appearing to have radiographic findings consistent with disease progression prior to tumor control or shrinkage. This has been described in approximately 15% of patients, and a formalized set of radiographic criteria have been developed to describe these atypical patterns of response known as the immune-related response criteria (irRC).6 The use of irRC, or immune-related variations on RECIST, has now become common practice for all checkpoint immunotherapy approaches, with the essential consideration being to confirm progression at least 4 weeks from the initial imaging study that indicated progressive disease.
With the success of ipilimumab for advanced disease, evaluation in the adjuvant setting for stage III melanoma following surgery became a priority. The European Organisation for Research and Treatment of Cancer (EORTC) initiated a clinical trial (EORTC 18071) prior to the FDA approval of ipilimumab, randomly selecting patients to receive ipilimumab (10 mg/kg every 3 weeks for four doses and then every 3 months for 2 years) or placebo.7 An improvement was observed in both median relapse-free survival (27.6 vs. 17.1 months; HR 0.76; p < 0.001) and 5-year overall survival (65.4% vs. 54.4%; HR 0.72; p = 0.001). Notable toxicity was observed at 10 mg/kg, including 90% all-grade immune-related AEs and 43% grade 3 to 5 immune-related AEs. Quality of life was significantly lower for the ipilimumab arm relative to placebo, although this did not meet the pre-specified criteria for clinical significance.8 Based on these data, adjuvant use of ipilimumab has become a standard of care, at least for patients with appropriate performance status and perhaps ulcerated primary lesions or multiple involved lymph nodes where the survival benefit was greatest.
The approval of ipilimumab thus followed an unusual route, whereby the dose approved for the treatment of advanced disease (3 mg/kg) was substantially lower than the dose approved for adjuvant therapy (10 mg/kg). A lingering question remained surrounding the most efficacious dose for management of advanced disease, and contingent with regulatory approval, a phase III trial to address this question, CA184-169, was launched.
The results of CA184-169 were presented at the 2016 European Society of Medical Oncology (ESMO) Annual Meeting. In the clinical trial, patients without prior exposure to BRAF or PD-1 inhibitors were randomly selected in a one-to-one fashion to ipilimumab either at 10 mg/kg (365 patients) or 3 mg/kg (362 patients).9 In each arm, treatment was administered as one dose every 3 weeks for four total doses, and no crossover was allowed between arms. A retreatment phase was allowed for any patients who experienced progressive disease after having had an initial response (or stable disease > 3 months) and who had not experienced intolerable toxicity. Stratification factors included stage IV metastasis (M) status (M1a-M1c), including previous treatment for brain metastases; prior versus no prior treatment; and Eastern Cooperative Oncology Group (ECOG) performance status (0 or 1). Key exclusion criteria included brain metastases that were symptomatic or required immediate treatment, as well as history of active autoimmune disease. The primary endpoint of the study was overall survival, with secondary endpoints including progression-free survival, response rate, and disease control rate. Other endpoints included safety during the first four ipilimumab doses as well as quality of life.
The two arms of CA184-169 were overall well matched in terms of study population, including median age, BRAF mutational status, lactate dehydrogenase level, stage IV M stage with and without brain metastases, ECOG performance status, and any prior systemic therapy. The median number of doses of ipilimumab administered was four in each arm, with retreatment being pursued in 6% and 9% of patients in the 10 and 3 mg/kg arms, respectively. The overall survival of each arm is shown in the Figure. With a minimum follow-up of approximately 43 months, the median overall survival was 15.7 versus 11.5 months (HR 0.84; p = 0.04) favoring the 10 mg/kg arm. Landmark overall survival favored the 10 mg/kg arm at all time points after a year including 1 year (54% vs. 48%), 2 years (38% vs. 31%), and 3 years (31% vs. 23%). A plateau on the overall survival curve was observed for both doses of ipilimumab consistent with prior analyses.4 Analysis of pre-specified subgroups favored the higher dose uniformly. In contrast to overall survival, no significant differences were seen between the 10 and 3 mg/kg arms for progression-free survival (2.8 months in each arm), response rate (15% vs. 12%), or disease control rate (32% vs. 28%).
Regarding toxicity, any-grade AEs were observed in 95% and 93% of patients in the 10 and 3 mg/kg arms, respectively, with 59% and 52% CTCAE grade 3 to 5 events, respectively. Regarding treatment-related AEs, the 10 mg/kg arm demonstrated 79% all-grade and 34% grade 3 to 5 AEs, as compared with 54% and 14% in the 3 mg/kg arm. Consistent with prior studies of ipilimumab, the most common treatment-related AEs were diarrhea, rash, and pruritus. In the 10 versus 3 mg/kg dose arms, grade 3 to 4 symptoms of diarrhea were observed in 10% versus 6% of patients, respectively, and colitis in 5% versus 3% of patients, respectively.
There were six treatment-related deaths during the study, including four patients in the 10 mg/kg arm (all related to gastrointestinal complications or multi-organ failure) and two patients in the 3 mg/kg arm (colon perforation and myocardial infarction secondary to complications from diarrhea). Quality of life as measured by the EORTC QLQ-C30 tool was consistently higher for the lower dose of ipilimumab.
The results of CA184-169 are notable in that they open to debate multiple issues regarding the standard of care, as well as clinical investigation in melanoma therapeutics. The combination of ipilimumab at 3 mg/kg with nivolumab at 1 mg/kg has already proceeded to regulatory approval; however, the dose of ipilimumab at 10 mg/kg was never explored in the phase I study of this combination.10 Given that patients already experience relatively high toxicity in order to gain maximal benefit with this combination, it is worth considering whether a higher ipilimumab dose would be even more efficacious. More broadly, an ongoing debate surrounds whether combination therapy with CTLA-4 and PD-1 antibodies up front is preferred to sequential administration of these agents.
Combination Versus Sequential Therapy With CTLA-4 and PD-1 Antibodies
Although no randomized evidence of combination versus sequential approaches exists, some preliminary data are available. In the phase II CheckMate 064 study,11 the sequential administration of PD-1 and CTLA-4 therapy, or vice versa, was compared. Notably, the primary endpoint of this study surrounded safety, and the trial was underpowered for comparisons surrounding efficacy. However, for patients receiving initial nivolumab (3 mg/kg) with a hard stop independent of progression at 3 months and switch to ipilimumab (3 mg/kg) for 3 months followed again by nivolumab thereafter, approximately the same response rate and incidence of toxicity was seen compared with the phase II CheckMate 069 and phase III CheckMate 067 trials of combination ipilimumab plus nivolumab.12,13 It seems reasonable to believe that a higher dose of second-line ipilimumab could perhaps push the efficacy of the PD-1 antibody followed by ipilimumab sequence even higher.
These data suggest that consideration of a sequential approach is reasonable in clinically appropriate patients, while up-front administration of CTLA-4/PD-1 combination therapy might be preferred for those patients whose disease may make sequential administration impractical (e.g., patients with brain metastases, rapid progression, etc.). For patients receiving sequential PD-1 antibody followed by ipilimumab in standard practice, consideration of the 10 mg/kg dose would be reasonable to consider, especially in the population of patients with BRAF wild-type disease, where treatment options are fewer. Such an approach would be clinically justified in that patients would have demonstrated a need to risk the increased toxicity profile of ipilimumab, having already progressed on a PD-1 antibody. This approach might be associated with an increase in cost, although that is not entirely clear, given that the 10 mg/kg dose in the adjuvant setting is currently being charged at the same price as the 3 mg/kg dose in the metastatic setting. An important consideration of “value” between patient and provider would be reasonable as well, given a 16% improvement in the hazard of death with the higher dose, however, also a decrease in quality of life.9
Perhaps more intriguing, these observations suggest that the development of other combination regimens, building either off PD-1 antibody or ipilimumab as a base, have merit. In such a scenario, the 10 mg/kg ipilimumab dose might be prioritized for the reason above. For example, early clinical trials combining PD-1 antibody with an inhibitor of indoleamine-2, 3-dioxygenase,14 or oncolytic virus15 have shown response rates similar to or approaching the ipilimumab plus nivolumab combination. In contrast, however, these have shown no substantial increase in toxicity relative to PD-1 antibody monotherapy. Combinations of these agents with ipilimumab have also been interesting,16,17 potentially justifying sequencing of combination therapies built off PD-1 followed by combinations built off CTLA-4.
Tremendous progress has been made in improving the survival outcomes for patients with advanced melanoma. Although ipilimumab led this revolution, it is worth noting that phase III clinical trial data for PD-1 antibodies and BRAF-MEK–directed combination therapies support the statement that there is no role for ipilimumab monotherapy (at 3 mg/kg or 10 mg/kg) as a first-line treatment for advanced melanoma. Even so, ipilimumab remains an important therapeutic agent both in clinical practice and for investigation in combination approaches moving forward. With the recent results of CA184-169 demonstrating an overall survival advantage for 10 mg/kg as compared with 3 mg/kg of ipilimumab, it is clear that the field continues to move rapidly even regarding standard agents. Although some have described the current time as a golden era in melanoma therapeutics, continued patient participation in clinical trials cannot be stressed enough, both to optimize current treatment approaches and to identify even more efficacious and less toxic regimens moving into the future.
About the Author: Dr. Luke is an assistant professor of medicine in the Melanoma and Developmental Therapeutics Clinics at the University of Chicago.