Quality-adjusted time without symptoms of disease or toxicity (Q-TWiST) analysis of pembrolizumab (pembro) versus chemotherapy (chemo) in microsatellite instability–high (MSI-H)/mismatch repair–deficient (dMMR) metastatic colorectal cancer (mCRC) in the KEYNOTE-177 trial.

Abstract

3631 Background: In KEYNOTE-177 (NCT02563002),first-line pembro provided significantly longer progression-free survival and a trend toward improvement in overall survival (OS) compared with chemo in participants (pts) with MSI-H/dMMR mCRC after > 5 years of median follow-up. The objective of this analysis was to assess Q-TWiST for pts treated in KEYNOTE-177. Methods: Q-TWiST combines efficacy, safety, and quality of life in a single measure. The analysis categorized survival time into 3 health states: time with grade ≥3adverse events before disease progression or death (toxicity [TOX]), time without symptoms or toxicity before disease progression (TWiST), and time from disease progression to death (relapse [REL]). In all randomly assigned pts, the restricted mean survival time (RMST) in each state was first weighted by a quality-of-life utility value, measured as treatment-specific EQ-5D-3L scores for each health state using the US value set, and then summed to calculate the Q-TWiST value. Relative gains (defined as the Q-TWiST differences between pembro and chemo divided by the RMST of chemo) of ≥15% were defined as “clearly clinically important.” Treatment difference 95% CIs were generated using the nonparametric bootstrapping method. The data cutoff date was July 17, 2023. Results: At a maximum follow-up of 84 months, pts in the pembro arm had a 13.8-mo (95% CI, 4.8-21.7) longer RMST in TWiST (24.3 vs 10.5 mo), a 4.8-mo (95% CI, −0.5 to 10.2) longer RMST in TOX (10.7 vs 5.9 mo), and an 11.1-mo (95% CI, −21.2 to −0.8) shorter RMST in REL (18.1 vs 29.3 mo) compared with chemo. As a proportion of overall preprogression time, pts in the pembro arm spent less time in TOX than the chemo arm (31% vs 36%). For the analysis of restricted mean Q-TWiST based on treatment-specific utility weights, the difference between pembro and chemo favored pembro by 9.1 mo (95% CI, 2.3-15.5), a 20.0% (95% CI, 4.8-36.8) relative Q-TWiST gain. When TOX definition included grade ≥2 adverse events, the relative Q-TWiST gain was 19.7% (95% CI, 4.7-36.5). When OS was adjusted for crossover to anti–PD-(L)1 therapy as second-line treatment, relative Q-TWiST gain was 39.7% (95% CI, 20.1-63.7). Conclusions: Pembro provided clearly clinically important improvement in quality-adjusted survival time based on Q-TWiST analyses compared with chemo as first-line treatment in pts with MSI-H/dMMR mCRC. The magnitude of results related to established thresholds for clinically important Q-TWiST gain suggests that results from this analysis provide additional evidence for the use of pembro in this population. Clinical trial information: NCT02563002 .