The use of the PD1 inhibitor nivolumab appeared to be feasible and safe in a small study of patients with untreated, stage I–IIIa non–small-cell lung cancer.
The use of the PD1 inhibitor nivolumab appeared to be feasible and safe in a small study of patients with untreated, stage I–IIIa non–small-cell lung cancer (NSCLC) presented at the European Society for Medical Oncology (ESMO) 2016 Congress, held October 7–11 in Copenhagen, Denmark (abstract LBA41_PR). Eighty percent of patients in the study had pathologic evidence of tumor regression.
“Until now nivolumab and the other anti–PD-1 and anti–PD-L1 drug studies have only been reported in metastatic or advanced lung cancer,” said lead author Patrick Forde, MD, assistant professor of oncology at Sidney Kimmel Comprehensive Cancer Center in Baltimore, Maryland, in a press release. “This was the first study of neoadjuvant PD-1 blockade in early-stage lung cancer.”
Forde and colleagues enrolled 20 patients with untreated, resectable early NSCLC. All patients underwent pretreatment tumor biopsy, and then received two doses of 3 mg/kg nivolumab at 4 and 2 weeks prior to resection. The primary endpoints were safety and feasibility of preoperative nivolumab administration. Forde presented results from the first 16 patients.
According to the researchers, there were no significant safety concerns and no delays to surgery with nivolumab. Twelve of 15 patients (80%) had pathologic evidence of tumor regression and 40% had major pathologic responses. All of those tumors had dense infiltration of immune cells and either a complete pathologic response or isolated remaining tumor cells. An additional 5 patients had some regression of their tumor noted and evidence of immune infiltration.
According to Forde, the researchers plan to expand the study with one cohort receiving a third dose of nivolumab preoperatively and another cohort receiving a combination of nivolumab and ipilimumab preoperatively.
“There is a potential for bias when comparing a small biopsy, which might not represent the whole tumor, with the resected tumor,” said Pieter Postmus, MD, PhD, chair of thoracic oncology at the University of Liverpool, United Kingdom, commenting on these results in a press release. “This is not a validated way to measure response to a treatment. It describes a biological effect but whether that has any clinical impact on survival is unproven.”
“Although we do not know for the time being if a major pathologic response is correlated with improved survival, this method could first be validated in a cohort of patients with advanced disease by comparing the percentages of viable tumor cells in tumor biopsies taken before and 4 to 8 weeks after immunotherapy,” Postmus continued. “If in this way regression-as defined in the preoperative study-correlates with survival in patients with advanced cancer, it is likely to hold true in less advanced or resectable patients. Long-term survival data will be the ultimate test for these neoadjuvant immunotherapy strategies.”