The addition of sargramostim to ipilimumab significantly extended OS and decreased toxicity in patients with advanced melanoma, according to results of a phase 2 study.
However, the combination of sargramostim (Leukine, Sanofi-Aventis) — a granulocyte–macrophage colony stimulating factor — and the CTLA-4 blockade ipilimumab (Yervoy, Bristol-Myers Squibb) was not associated with prolonged PFS, results showed.
F. Stephen Hodi
“The addition of an immune signaler to ipilimumab both improved patient outcomes regarding better survival as well as decreased the severe side effects,” F. Stephen Hodi, MD, director of the Melanoma Treatment Center and Center for Immuno-Oncology at Dana-Farber Cancer Institute, told HemOnc Today. “Specifically, severe lung and gastrointestinal side effects were decreased. These are the organs we are concerned about the most with regard to treatment with immune checkpoint agents.”
The analysis included 245 patients with unresectable stage III or IV melanoma who had received at least one prior therapy. Researchers randomly assigned 123 patients 10 mg/kg ipilimumab on day 1 plus 250 micrograms subcutaneous sargramostim on days 1 to 14 of a 21-day cycle. The other 122 patients received ipilimumab alone.
Median follow-up was 13.3 months (range, 0.03-19.9).
Patients who received ipilimumab plus sargramostim demonstrated a median OS of 17.5 months (95% CI, 14.9-not reached), whereas patients who received ipilimumab alone demonstrated a median OS of 12.7 months (95% CI, 10-not reached).
Significantly more patients in the combination arm achieved 1-year OS (68.9% vs. 52.9%; P=.01), and the combination also was associated with a significant reduction in the risk for death (HR=0.64; 95% CI, not applicable-0.9).
Although the combination extended OS, median PFS was 3.1 months in both arms (P=.37).
Current data do not elucidate why a difference existed with OS but not PFS, Hodi said.
“It could be that the treatment is causing inflammation that looked like early disease progression, but we won’t know without further studies,” Hodi said in a press release.
Researchers evaluated safety data from 118 patients assigned the combination and 120 patients assigned ipilimumab alone.
Overall, significantly fewer patients in the combination arm experienced a grade 3 to grade 5 adverse event (44.9% vs. 58.3%; P=.04). Significantly more patients who received ipilimumab alone experienced grade 3 to 5 gastrointestinal toxicities (26.7% vs. 16.1%; P=.05) and pulmonary toxicities (7.5% vs. 0%; P=.003).
Seven patients assigned ipilimumab alone died due to a treatment-related adverse event, whereas two treatment-related deaths occurred in the combination arm.
“[The combination] reveals the possibilities of combining an immune signaling molecule with taking the brakes off at the same time,” Hodi said in the release.
Disclosure: Hodi reports research funding, nonfinancial support and personal fees from and patents licensed to Bristol-Myers Squibb, the Cancer Therapy Evaluation Program/NCI, Eastern Cooperative Oncology Group and Sanofi/Genzyme. See the study for a list of the remaining researchers’ relevant financial disclosures.
Sanjiv S. Agarwala
Hodi and colleagues report the results of a trial conducted by the Eastern Cooperative Oncology Group that included 245 patients with unresectable stage III or stage IV melanoma who had received at least one prior therapy. Researchers randomly assigned patients to 10-mg/kg ipilimumab every 3 weeks for four doses, followed by maintenance at the same dose once every fourth cycle with or without 250 micrograms of sargramostim subcutaneously on days 1 through 14 of a 21-day cycle. The analysis showed that patients who received the combination experienced significantly better survival that those who received ipilimumab alone.
Although the results are intriguing and provocative, they must be interpreted with caveats. Surprisingly, there was no improvement in either response rates or PFS for the combination. Although this phenomenon may occur in the setting of immunotherapy, this does provide food for thought and some skepticism. Also, although this was a randomized trial, this was performed as a phase 2 randomized study and needs a phase 3 trial for confirmation. As the currently approved dose of ipilimumab is 3 mg/kg every 3 weeks for four doses without maintenance, the results of this trial cannot be extrapolated to current clinical practice. Until we have confirmatory data, ipilimumab should continue to be used as a single agent at the currently approved dose and schedule. Further data on this combination are awaited with interest.
The authors investigated the combination of ipilimumab and GM-CSF vs. ipilimumab alone in patients with metastatic melanoma, with the primary objective to compare OS between study arms. This trial was designed to detect an improvement in median OS from 10.4 to 14.8 months with 80% power. This was accomplished with a relatively small sample size (245 patients) by utilizing a one-sided alpha=.10 significance level. There was a pre-specified interim analysis allowing for the possibility of achieving the primary objective early, provided interim results met a more stringent significance level: alpha=.05. As the authors indicate, the interim results did achieve statistical significance, with P
=.01 for the primary analysis. However, these interim results were associated with a high censoring rate (58% of study participants censored in OS analysis). This resulted in the inability to estimate the upper limit of the confidence interval for median OS in either arm; therefore, these interim results should be interpreted with caution. The high censoring rate is somewhat surprising considering enrollment was completed between Dec. 28, 2010, and July 28, 2011 and data cutoff for the interim OS analysis did not occur until December 2012. This would suggest that all surviving patients on this study could have been followed for a time interval ranging from approximately 16 months to 23 months. Even considering the survival follow-up schedule of every 3 months, this seems inconsistent with the reported overall median follow-up time of only 13.3 months and overall 58% censoring rate. Another surprising result was the apparent treatment effect modification associated with gender. The combination therapy was associated with a 56% decreased risk for death in men (HR=0.44) and numerically associated with a 32% increased risk for death in women (HR=1.32). Moreover, the OS censoring rate differed markedly, as 65% of men were censored but only 43% of women were censored. The authors appropriately conclude that these study results require confirmation in larger studies. However, it would be prudent to consider larger studies only after conducting an updated OS analysis of the current study.