Questions Over Adjuvant Ipilimumab in Melanoma

Alexander M. Castellino, PhD

February 16, 2017

Adjuvant ipilimumab (Yervoy, Bristol-Myers Squibb) was approved in October 2015 for treatment of stage III melanoma on the basis of efficacy data from the European Organization for Research and Treatment of Cancer (EORTC) 18071 trial. However, there has been concern over the toxicity that was seen, as ipilimumab was used at a high dose (10 mg/kg, compared with the more commonly used dose of 3 mg/kg).

Now, investigators report secondary outcomes on health-related quality of life (HRQoL) from the same study, published online February 2 in Lancet Oncology. The researchers report that despite the toxicity seen with high-dose ipilimumab and the high rate of treatment discontinuation due to an adverse event, the overall HRQoL with ipilimumab was similar to that in patients who received placebo: "no clinically relevant differences (10 points or more) in global health scores were observed during or after induction."

The authors write, "It is remarkable that these side-effects did not translate into worse patient-reported global health status compared with that in the placebo group. One possible explanation is that because this trial is placebo controlled, patients might experience side-effects as a confirmation of treatment efficacy," they add.

However, an editorial and an expert not related to the study disagree. They suggest that the HRQoL of patients receiving ipilimumab is significantly and clinically worse than that seen with placebo. It all depends on the bar one sets. With the study protocol defining "clinically meaningful" as a 10-point difference from baseline or between treatment groups, the bar may have been set too low (high threshold) to pick up clinical differences between the two.

The conclusion the study authors make about there being no effect on quality of life "requires careful interpretation" because the "the estimate of the true difference might be underestimated," comment Michael Brundage, MSc, MD, and Timothy Hanna, MD, MSc, from Cancer Care and Epidemiology at Queen's University, Kingston, Ontario, Canada. Writing in an accompanying commentary, they add that "the per-protocol outcome assessments might have diminished the apparent effect of treatment by averaging global scores across time."

In addition, the lack of adherence with completing the HRQoL questionnaire during and at the end of induction may have contributed to the effect, they propose.

An Advance in Treatment, but Toxic

Adjuvant ipilimumab for melanoma was recently hailed as one of the "clinical cancer advances" of last year by the American Society of Clinical Oncology in its annual report. It pointed out that long-term data from EORTC 18071 show that at 5 years, more patients in the ipilimumab group had not experienced a recurrence (41% vs 30%) or metastasis (48% vs 39%).

"These findings suggest that high-dose ipilimumab may help patients with stage III melanoma live longer after surgery. However, the adverse effects of this treatment are common and can be life threatening," the report noted.

Melanoma experts were approached for comments by Medscape Medical News on HRQoL conclusions and toxicity seen with high-dose ipilimumab.

In the study, patients received ipilimumab, 10 mg/kg, every 3 weeks for four doses (12 weeks of induction) and then every 3 months for a maximum of 3 years (maintenance).

"This is an important analysis for understanding the use of ipilimumab in the adjuvant setting," Michael A. Postow, MD, a medical oncologist in the Melanoma and Immunotherapeutics Services, Memorial Sloan Kettering Cancer Center, New York, New York, said.

"In the adjuvant setting, the bar is set high because we are treating patients who may be cured with surgery alone, and understanding what these toxicities mean for these patients is important," he added.

Dr Postow told Medscape Medical News that he reserves adjuvant ipilimumab for highly motivated patients who may be able to better withstand its side effects or when there is a perception of a higher-risk melanoma scenario, such as in patients with 4 or more positive lymph nodes.

He does not use adjuvant ipilimumab for patients with stage IIIA melanoma.

Nikhil I. Khushalani, MD, from the Department of Cutaneous Oncology at the H. Lee Moffitt Cancer Center, Tampa, Florida, was more emphatic. "We are adding an agent to decrease the risk of relapse, and how it impacts patients' quality of life is an important question to answer," he added.

"Ipilimumab 10 mg/kg is too toxic for routine clinical use," Dr Khushalani told Medscape Medical News.

In the study, the toxicity of ipilimumab was "substantial," note the editorialists — 42% of patients had grade 3/4 immune-related toxicities, and adverse events caused ipilimumab discontinuation in over half (52%) of the patients.

Grade 3/4 adverse events in patients who received ipilimumab included gastrointestinal (16% vs <1% for placebo), hepatic (11% vs <1% for placebo), and endocrine (8% vs 0% for placebo) events; in addition, 5 treatment-related deaths were reported among patients receiving ipilimumab (vs 0 for placebo).

Dr Khushalani said that he prefers to use interferon in the adjuvant setting because it has similar efficacy and "the toxicities with interferon are self-limited," he said. Once the drug is stopped, the toxicities resolve, he explained.

To support his comment that efficacy is similar, Dr Khushalani cited two studies. Prospective data from European Cooperative Oncology Group (ECOG) 684 (J Clin Oncol. 1996;14:7-17) showed 5-year relapse-free survival with high-dose interferon of 37%, which is "not far off from the 40% seen with adjuvant ipilimumab," he said. Also, retrospective data presented at the Society for Melanoma Research 2016 Congress show that the distant metastasis–free survival with interferon was 48%, which is similar to that of ipilimumab, Dr Khushalani noted. However, he also emphasized that comparison across studies is difficult and must be interpreted with caution.

"I provide adjuvant ipilimumab only in the context of a clinical trial or for those patients who may not want to go on interferon," he added.

Both experts noted that ongoing studies are looking at PD-1 inhibitors as adjuvant treatment for melanoma. For example, the Southwestern Oncology Group 1404 trial (NCT02506153) is comparing the current standard of care — high-dose interferon or high-dose ipilimumab — with pembrolizumab (Keytruda, Merck & Co).

In the meantime, the fully accrued ECOG 1609 trial (NCT01274338) will likely establish the new standard of care for adjuvant treatment of melanoma, commented Dr Khushalani. This trial is comparing high-dose interferon with ipilimumab, 10 mg/kg, and ipilimumab, 3 mg/kg. Results are expected in May 2018.

Underreporting in the Study?

HRQoL, a prespecified secondary endpoint of EORTC 1807, was measured by using EORTC QLQ-C30 — a 30-item questionnaire for cancer patients across nine items, including global health status, fatigue, pain, and nausea and vomiting. It also contained single-item measures of dyspnea, insomnia, appetite loss, constipation, diarrhea, and financial impact. Questionnaires were completed at baseline and at weeks 4, 7, 10, and 24 and every 12 weeks after that for 2 years.

The EORTC 18071 investigators point out that this HRQoL questionnaire, although routinely used as a validated instrument in oncology practice, has not been validated in a trial evaluating an immunotherapy and lacks symptom scales for immune-related adverse events (eg, endocrine effects, rash).

The investigators report that average global health scores during induction and during maintenance differed by fewer than 5 points between patients receiving placebo and those receiving ipilimumab (in favor of placebo). Although the differences between ipilimumab and placebo were statistically significant in favor of patients receiving placebo, they were not considered clinically relevant — defined per protocol as a difference of 10 points or more. When analyzed at each assessment after baseline, the maximum difference of approximately 8 points was seen at week 10, with a lower score in the ipilimumab group.

However, at week 10 (when induction was complete), diarrhea and fatigue scores were 10 points worse than baseline scores for patients treated with ipilimumab. Between-treatment differences for diarrhea and insomnia were also more than 10 points worse for patients treated with ipilimumab.

"It is clear that the global health score was better for patients on placebo versus those on ipilimumab," Dr Khushalani said. He indicated being at odds with how the authors defined "clinical significance" — as a difference of greater than 10 points (from baseline or between treatment differences).

Dr Brundage and Dr Hanna also pointed out that "the 10-point threshold may be too high, since others assessing ipilimumab in the setting of advanced disease have deemed smaller thresholds (5-10 points) important, and lower thresholds may be appropriate in the context of a randomised comparison."

Acknowledging that a lower threshold may be relevant, the EORTC 18071 investigators write, "[The] observed deteriorations in the ipilimumab group cannot be dismissed as trivial and should be interpreted with care."

Dr Khushalani concurred. "There were five deaths associated with ipilimumab in this study," he said. "These are five deaths too many for a disease state where you're looking for a cure," he added.

The main reason cited for underestimation of HRQoL for patients receiving ipilimumab may be attributed to missing data, Dr Brundage and Dr Hanna point out. At week 10, patient-reported data were missing for 9% of patients receiving placebo and 21% of patients receiving ipilimumab, they note.

"Patients with worse global HRQoL might have been less likely to complete the instrument," they write. "Although the authors appropriately conducted sensitivity analyses addressing this problem, they also acknowledge that fully correcting for potential bias cannot be achieved using statistical techniques alone," they add.

Dr Khushalani also noted the increased lack of adherence over time for patients receiving ipilimumab. He noted that at week 24, after induction was complete, only 69% of patients taking ipilimumab completed their questionnaire compared with 81% for patients taking placebo. He pointed out that the lack of adherence may well be attributed to the high rate of discontinuation for an adverse event during induction. "For patients who discontinue therapy, there is no incentive to complete the questionnaire," he said. "If all patients had completed their questionnaire, we may have seen a bigger treatment effect," he added.

The study was funded by Bristol-Myers Squibb. Dr Brundage and Dr Hanna have disclosed no relevant financial relationships. Disclosures of all study authors are available in the publication. Dr Postow has received honoraria from Bristol-Myers Squibb and Merck & Co; has had an advisory role for Bristol-Myers Squibb, Amgen, and Novartis; and has received research funding from Bristol-Myers Squibb. Dr Khushalani has served on advisory boards for Bristol-Myers Squibb, Castle Biosciences, and Genentech; has received research funding from Merck, Bristol-Myers Squibb, Novartis, Amgen, GlaxoSmithKline, and the National Comprehensive Cancer Network; and has served on the data safety monitoring board for AstraZeneca.

Lancet Oncol. Published online February 2, 2017. Abstract, Comment

Follow Medscape Oncology on Twitter: @MedscapeOnc


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.