KEYFORM-007 Results: Co-Formulated Favezelimab and Pembrolizumab in Metastatic Colorectal Cancer

At the 2025 American Society of Clinical Oncology Gastrointestinal Cancers Symposium, Neil Segal, MD, PhD, of Memorial Sloan Kettering Cancer Center, presented the results of the phase III KEYFORM-007 trial, which evaluated the efficacy and safety of co-formulated favezelimab (an anti–lymphocyte activation gene-3 protein antibody) and pembrolizumab (a programmed cell death 1 [PD-1] inhibitor) versus standard-of-care (SOC) treatments in patients with PD ligand 1 (PD-L1)–positive microsatellite stable/mismatch repair proficient (MSS/pMMR) metastatic colorectal cancer (mCRC).

Study Design and Objectives

The trial enrolled 441 patients with a PD-L1 combined positive score ≥1 MSS/pMMR mCRC whose disease had progressed during standard treatment or who could not tolerate standard treatment. Participants were randomized 1:1 to receive either co-formulated favezelimab/pembrolizumab or SOC (regorafenib or TAS-102 [trifluridine and tipiracil hydrochloride]). The primary endpoint was overall survival (OS), with secondary endpoints including progression-free survival (PFS), objective response rate (ORR), and duration of response (DOR).

Key Results

After a median follow-up of 28 months, the study revealed that favezelimab/pembrolizumab did not demonstrate a survival benefit over SOC. Median OS was 7.3 months with favezelimab/pembrolizumab compared with 8.5 months with SOC (hazard ratio, 0.98, P=.4183). Similarly, median PFS was 2.1 months versus 2.6 months, respectively, with no statistical significance.

Despite the lack of OS and PFS benefits, the ORR was notably higher with favezelimab/pembrolizumab, with 6.8% of patients achieving a response compared with 0.9% in the SOC arm. Responses in the favezelimab/pembrolizumab arm were durable, with the median DOR not reached among responders, compared with 6.5+ months in the SOC arm.

The co-formulated therapy exhibited a manageable safety profile, with fewer grade ≥3 treatment-related adverse events than SOC (20% vs 36%). However, adverse events of special interest, such as immune-related events, were more frequent with favezelimab/pembrolizumab (38% vs 6%).

The KEYFORM-007 trial authors concluded that although the safety profile was manageable, co-formulated favezelimab/pembrolizumab did not improve OS or PFS compared with SOC for patients with PD-L1–positive MSS/pMMR mCRC.