Surgical results
In his ASCO presentation, Dr. Spicer reported that definitive surgery was canceled in 16% of patients in the nivolumab/chemotherapy arm, and 21% of the chemotherapy arm. Reasons for surgery cancellation generally included patients declining surgery, unresectable disease, and poor lung function. “Cancellation of surgery due to neoadjuvant therapy toxicity was rare,” Dr. Spicer said in his presentation.
Among patients who did proceed to surgery, the median duration of the procedure was 184 minutes in the nivolumab/chemotherapy arm and 217 minutes in the chemotherapy arm. That half-hour difference in favor of the combination arm suggests that the complexity of surgery was not increased by the addition of nivolumab, Dr. Spicer said.
Median time to surgery was about 5 weeks in both arms, which was “well within accepted standards for a neoadjuvant therapeutic approach,” Dr. Spicer said. Most delays beyond 6 weeks were due to administrative issues, and occurred in similar proportions (21% of the nivolumab/chemotherapy arm and 18% of the chemotherapy arm).
The addition of nivolumab to chemotherapy improved pCR rates regardless of baseline stage of disease, according to Dr. Spicer. Furthermore, the depth of pathological regression in the primary tumor was “dramatically different” across stage groupings, he said. Median residual viable tumor percentage in stage IB/II patients was 28% for nivolumab/chemotherapy and 79% for chemotherapy, and in stage IIIA patients, it was 8% for nivolumab/chemotherapy and 70% for chemotherapy.
Overall, thoracotomy was the most frequent surgical approach in this international phase 3 trial, Dr. Spicer said. However, among patients with stage IIIA disease, minimally invasive approaches were used 30% of the time in the nivolumab/chemotherapy arm and 19% in the chemotherapy arm. Conversely, the rate of conversion from a minimally invasive to open approach in patients with stage IIIA disease was 11% for nivolumab/chemotherapy and 20% for chemotherapy alone.
Lobectomy was more frequent in the nivolumab/chemotherapy arm (77%) compared to the chemotherapy arm (61%), a difference that Dr. Spicer described as clinically important. He said the difference appears to be attributable to a lower rate of pneumonectomy in the nivolumab/chemotherapy arm (17%) than in the chemotherapy arm (25%).
Despite less extensive lung resection being required, the rate of R0 resection was numerically higher in the nivolumab/chemotherapy arm (83%) than in the chemotherapy arm (78%), said Dr. Spicer.
Length of hospital stay was “within expected ranges” from geographic regions represented in the trial, Dr. Spicer said. Median length of stay was 4.0 and 6.0 days, respectively, for nivolumab/chemotherapy and chemotherapy alone in North America, 9.5 and 13.0 days in Europe, and 11.0 and 13.0 days in Asia.
Likewise, 90-day surgical complications were well within expected ranges, according to the investigator, with anemia, pain, and wound complications being the most commonly reported. Rates were generally similar between study arms, other than a twofold higher rate of pain in the chemotherapy arm, possibly due to the lower rate of minimally invasive surgery or higher rate of conversion to an open procedure, compared to the nivolumab/chemotherapy arm, he said.
Awaiting survival
Rates of 30- and 90-day mortality are expected to be evaluated when survival endpoints are available, according to Dr. Spicer. Beyond pCR rate, event-free survival is also a primary endpoint of the study, while overall survival is a secondary endpoint.
The study was supported by Bristol Myers Squibb. Dr. Spicer reported disclosures related to AstraZeneca, Bristol Myers Squibb Foundation, Merck, and Roche. Dr. Rusch reported research funding with Genelux and Genentech, and travel expenses from Intuitive Surgical.