Q-TWiST analysis of pembrolizumab or placebo plus chemotherapy for patients with advanced HER2-negative gastric or gastroesophageal junction (G/GEJ) cancer in KEYNOTE-859.
Abstract
376 Background: In the phase 3 KEYNOTE-859 study (NCT03675737; N = 1579), patients (pts) with locally advanced or metastatic HER2-negative G/GEJ cancer treated with first-line pembrolizumab (pembro) plus chemotherapy (chemo) had a significant and clinically meaningful improvement in OS with manageable toxicity vs placebo (pbo) plus chemo while maintaining HRQoL in the intention-to-treat and prespecified PD-L1 combined positive score (CPS) populations. We report data from an analysis of quality-adjusted time without symptoms of disease progression or toxicity (Q-TWiST). Methods: Q-TWiST is a measure of survival weighted by health states utility values. The analysis categorized survival time into 3 health states: time with grade ≥3 AEs (toxicity [TOX]) and time without symptoms or toxicities (TWiST)—both before disease progression or death—and relapse (REL), defined as time from disease progression to death. Q-TWiST was calculated as the restricted mean survival time (RMST) spent in each state, with a weighted health state utility for that state. Average utility weights were derived based on pooled EQ-5D-5L scores using the US mapping algorithm and standardized base case utility weights. Relative gains in Q-TWiST of ≥10% and ≥15% were “clinically important” and “clearly clinically important,” respectively, as reported in the literature. Treatment difference 95% CIs were generated using the nonparametric bootstrapping method. Data are reported for all randomly assigned pts and pts with PD-L1 CPS ≥1 and CPS ≥10 tumors. The data cutoff date was August 22, 2023. Results: At the maximum follow-up of 56 mo for pembro plus chemo vs pbo plus chemo, RMST was 5.16 vs 2.86 mo for TOX, 7.77 vs 5.87 mo for TWiST, and 6.85 vs 7.12 mo for REL. Between-treatment differences are shown in the table. The difference in restricted mean Q-TWiST and relative gain favored pembro plus chemo. Conclusions: A clearly clinically important Q-TWiST gain was observed for pembro plus chemo vs pbo plus chemo in all pts with advanced HER2-negative G/GEJ cancer and pts with PD-L1 CPS ≥1 and CPS ≥10 tumors. Clinical trial information: NCT03675737 . All ptsN = 1579 PD-L1 CPS ≥1n = 1235 PD-L1 CPS ≥10n = 553 RMST in TOX, mo (95% CI) 2.30(1.28 to 3.44) 2.65(1.46 to 3.99) 4.28(2.08 to 6.41) RMST in TWiST, mo (95% CI) 1.90(−0.02 to 3.79) 2.28(0.07 to 4.49) 3.72(0.43 to 7.19) RMST in REL, mo (95% CI) −0.28(−2.43 to 2.01) −0.22(−2.88 to 2.39) −0.45(−4.52 to 3.71) Restricted meanQ-TWiST, mo (95% CI) a 3.31(2.10 to 4.59) 3.98(2.60 to 5.43) 6.45(4.26 to 8.74) Relative Q-TWiST gain, % (95% CI) a 20.90(12.49 to 30.56) 25.34(16.04 to 36.26) 38.05(23.21 to 56.59) Restricted meanQ-TWiST, mo (95% CI) b 2.91(1.58 to 4.14) 3.50(2.02 to 4.88) 5.63(3.31 to 7.99) Relative Q-TWiST gain, % (95% CI) b 18.38(9.78 to 27.46) 22.27(12.65 to 32.52) 33.19(18.76 to 49.97) a Based on US mapping algorithm. b Based on standardized base case utility weights.
Article Details
Journal Info
Journal of Clinical Oncology
Lippincott Williams & Wilkins
Authors (5)
Lucjan Wyrwicz
Department of Oncology and Radiotherapy, Maria Sklodowska-Curie National Research Institute of Oncology, Warsaw, Poland
Vasiliki Kalampoki
Merck & Co., Inc., Rahway, NJ
Adriana Valderrama
Merck & Co, Inc, Rahway, NJ
Karthik Ramakrishnan
Merck & Co., Inc., Rahway, NJ
Kate Young