CHICAGO – Pembrolizumab with and without chemotherapy proved superior for overall survival compared with the EXTREME regimen when used first line in certain subgroups of patients with recurrent/metastatic head and neck squamous cell carcinoma (R/M HNSCC), according to “practice-changing” final results from the randomized phase 3 KEYNOTE-048 study.
Compared with 300 patients randomized to receive the EXTREME regimen (a certuximab loading dose followed by carboplatin or cisplatin and 5-fluorouracil), 281 who received pembrolizumab plus chemotherapy (P+C) had superior overall survival (OS) with comparable safety–including both those with programmed death-Ligand 1 (PD-L1) combined positive score (CPS) of 20 or greater (median 14.7 vs 11.0 months; hazard ratio, 0.60) and with CPS of 1 or greater (median, 13.6 vs. 10.4 months; HR, 0.65), Danny Rischin, MD, reported at the annual meeting of the American Society of Clinical Oncology.
The differences were highly statistically significant, said Dr. Rischin, a professor and director of the Division of Cancer Medicine and head of the Department of Medical Oncology at Peter MacCallum Cancer Centre, Melbourne, Australia.
“And this benefit in overall survival in CPS greater than or equal to 20 and greater than or equal to 1 appeared to be present across all the subgroups that we looked at,” he added.
The response rates did not differ between P+C and EXTREME groups, but the median duration of response was significantly greater with P+C vs. EXTREME in both the CPS of 20 or greater and 1 or greater (7.1 vs. 4.2 months and 6.7 vs. 4.3 months, respectively).
Additionally, the final results of the study showed an OS benefit with P+C vs. EXTREME in the total population (13.0 vs. 10.7 months; HR, 0.72), Dr. Rischin said.
The difference between the groups with respect to progression-free survival (PFS), however, was not statistically significant and did not reach the superiority threshold, he noted.
In the 301 patients who received pembrolizumab alone, OS was superior in the CPS 20 or greater and 1 or greater populations (median, 14.8 vs. 10.7 months; HR, 0.58 and 12.4 vs. 10.3 months; HR, 0.74, respectively), compared with EXTREME, but was noninferior in the total population (median 11.5 vs. 10.7 months; HR, 0.83), and safety was favorable .
Again, PFS did not differ between the groups (median, 2.3 vs. 5.2 months; HR, 1.34), and while the overall response rates did not differ significantly, the median duration of response was substantially longer with pembrolizumab at 22.6 vs. 4.5 months with EXTREME, he said.
Study participants had locally incurable R/M HNSCC and no prior systemic therapy in the R/M setting. Those in he P+C arm received pembrolizumab at 200 mg plus 6 cycles of cisplatin at 100 mg/m2 or carboplatin AUC 5, and 5-fluorouracil at a dose of 1000 mg/m2/day for 4 days every 3 weeks; those in the pembrolizumab alone arm received 200 mg every 3 weeks for up to 35 cycles, and those in the EXTREME arm received certuximab at a 400 mg/m2 loading dose followed by 250 mg/m2 weekly with carboplatin AUC 5 or cisplatin at 100 mg/m2, and 5-FU at 1000 mg/m2/day for 4 days for 6 cycles.