A randomized phase III trial of adjuvant gefitinib in patients with completely resected non–small-cell lung cancer (NSCLC) was closed early, but available results suggest that the drug is unlikely to benefit this group of patients.
The study (NCIC CTG BR19) was undertaken due to the generally poor survival of patients with completely resected NSCLC along with some evidence of activity with gefitinib in advanced NSCLC. Once accrual of the study was ongoing, however, results from two other trials showed no benefit with gefitinib in a similar disease setting; because of those results, this trial was closed early and final results were analyzed after all patients already included were followed for at least 4 years. Led by Glenwood D. Goss, MD, of the Ottawa Hospital Cancer Center, results of the study were published online ahead of print in the Journal of Clinical Oncology.
In total, 503 patients out of a planned 1,242 were included. All patients had completely resected NSCLC and were randomized to either placebo (252 patients) or gefitinib (251 patients) 250 mg/day for 2 years. The median follow-up was 4.7 years; three patients were lost to follow-up.
There was no difference with regard to overall survival; the median overall survival on gefitinib was 5.1 years, and it had not been reached for placebo patients, yielding a hazard ratio (HR) of 1.24 (95% CI, 0.94–1.64; P = 0.14). A Cox regression model found age of at least 65 years and tumor size of at least 4 cm to be associated with shorter overall survival, but gefitinib was not significantly associated with survival (though it was “potentially harmful”).
Similarly, median disease-free survival was 4.2 years in gefitinib patients and it was not reached for placebo patients; the HR for recurrence was 1.22 (95% CI, 0.93–1.61; P = 0.15). Only larger tumor size was significantly associated with shorter disease-free survival, while gefitinib had no significant associations and could have been harmful.
EGFR mutation status was available in 359 tumors, and only 15 patients were found to have mutations. There were still no differences for disease-free survival and overall survival in patients stratified by EGFR mutation status.
Toxicity, as expected, was greater in patients who received gefitinib. Those patients had a higher incidence of various adverse events including rash, diarrhea, and dry skin, though they had lower incidence of chest pain, muscle pain, and dyspnea. The patient refusal or withdrawal rate after the start of therapy was 24% in the gefitinib group and 7% in the placebo patients.
“Unfortunately early termination of the study does not allow for statistically robust conclusions,” the authors wrote. Still, the relatively large sample size and balanced study arms led them to note that gefitinib is unlikely to be of benefit in this population of patients.