SAN ANTONIO – Patients with locally advanced or metastatic breast cancer did not fare better if given eribulin instead of capecitabine in a randomized phase III trial. But results suggest eribulin may have an edge in certain subgroups, including women with triple-negative breast cancer.
Investigators led by Dr. Peter A. Kaufman studied more than 1,000 such patients who had had up to three prior treatments, assigning them evenly to eribulin (Halaven) or capecitabine (Xeloda). Trial results, reported at the San Antonio Breast Cancer Symposium, showed only a nonsignificant trend in overall survival in favor of eribulin, amounting to a gain of about 1.5 months, and essentially no difference in progression-free survival.
Dr. Peter A. Kaufman
However, preplanned subgroup analyses hinted at a potential overall survival benefit among patients having the notoriously hard-to-treat triple-negative disease, as well as those with estrogen receptor–negative or HER2-negative disease.
"Eribulin and capecitabine have similar overall activity in this trial that included patients in the first-, second-, and third-line settings," Dr. Kaufman commented in a press briefing. "Our findings, I want to emphasize, are not definitive but interesting in terms of certain subsets and suggest that the triple-negative subgroup perhaps may be doing better with eribulin."
The reason for the unusual combination of a trend toward overall survival benefit in the absence of a progression-free survival benefit is unknown, he said. "This is not typical of what one sees in metastatic breast cancer, but the data are what the data are."
Given the trial’s findings, what are clinicians to do?
"I think eribulin is a reasonable option to consider therapeutically now for women. It compares reasonably to capecitabine," maintained Dr. Kaufman of the Geisel School of Medicine at Dartmouth and the Norris Cotton Cancer Center in Lebanon, N.H. "Our primary objective was to demonstrate superiority, which we did not meet. But the results are reasonably encouraging."
HER2 Therapy Questioned
In the session where the results were presented, attendee Dr. Daniel F. Hayes of the University of Michigan, Ann Arbor said, "I’m assuming that everyone who was HER2-positive received trastuzumab or some other anti-HER2 therapy, which might affect your subgroup analysis for HER2."
"A large proportion of the study population was enrolled in eastern Europe or South America, parts of the world where trastuzumab and other HER2-targeted therapies were not routinely used at the time of enrollment for this study," Dr. Kaufman replied. "We are sorting out those results and will present them at a further date."
Dr. Hayes also wondered about crossover on the trial and its possible impact on overall survival.
Although there was no planned crossover, about half of patients in the eribulin arm did switch to capecitabine, and that may have affected the results, Dr. Kaufman acknowledged. "We are pulling the data together on post-progression treatments to try to analyze that further."
FDA Requested Bone Scans
Attendee Dr. Steven Vogl of Montefiore Medical Center in the Bronx, New York, took issue with the objective response rates with both eribulin and capecitabine, which were 16% and 20%, respectively, by local investigator assessment but just 11% and 12% on independent review.
"The objective response rate on radiologic review is sort of distressing," he said. "Did someone decide that if the bone scan didn’t get better, that wasn’t a response, even though the node shrank 70%?"
"Correct. So this is one of the reasons why the response rate on the independent analysis is lower," Dr. Kaufman replied. "If the bone scan did not confirm it, those patients were not scored as having responding disease; additionally, if the confirmatory bone scan was not performed, those patients were not scored as having an objective response.
"We are still pulling all this together, but I think that is a large piece of why the independent analysis response rates were lower than the local investigator assessment. But I will highlight also that ... only about 20% of the patients were in the first-line setting, so it’s not a pure first-line trial."
"The 20% response rate in a third-line trial is credible, but the 10% response rate – some of us don’t think that’s worth pursuing," Dr. Vogl added. "I think bone scans shouldn’t be done on studies like this because they take forever to get better."
"Bone scan was actually included at the request of the FDA," Dr. Kaufman noted.
Results Not Statistically Different
Patients were eligible for the trial, formally known as Study 301, if they had locally advanced or metastatic breast cancer and had a received up to three prior regimens (up to two for their advanced disease), which must have included a taxane and an anthracycline.