Quality-adjusted time without symptoms or toxicities (Q-TWiST) analysis of ZUMA-7, a randomized controlled trial of axicabtagene ciloleucel versus standard of care for second-line large B-cell lymphoma.

Abstract

7555 Background: Quality of life (QoL) post cancer therapy is increasingly important. We conducted a Q-TWiST analysis to compare the quality-adjusted survival of axicabtagene ciloleucel (axi-cel, a chimeric antigen receptor T-cell therapy) versus standard of care (SOC) among relapsed/refractory (R/R) large B-cell lymphoma (LBCL) patients enrolled in ZUMA-7 (NCT03391466), a Phase 3, randomized, open-label, multicenter study. ZUMA-7 met its primary endpoint of event-free survival; events were defined as death, progression, or new lymphoma therapy. Methods: Overall survival (OS) in the intention-to-treat (ITT) cohort was partitioned into 3 mutually exclusive health states as measured through a median follow-up duration of 23.5 months (m): time with grade 3/4 adverse events (AE) before an event (TOX); time without symptoms from events or toxicity (TWiST); and time after event (REL). Q-TWiST was calculated as the average time spent in each state, weighted by state-specific QoL utility (u, ranging from 0 to 1), assuming a base case u(TOX)=0.5, u(REL)=0.5, u(TWiST)=1. Threshold analyses assessed the mean Q-TWiST differences between axi-cel and SOC by varying u(TOX) and u(REL) from 0 to 1. Relative gains for axi-cel (axi-cel Q-TWIST gain divided by mean SOC OS) ≥15% were considered “clearly clinically important” using published norms. Nonparametric bootstrap 95% confidence intervals (CI) were computed. Subgroup analyses were performed for relapse status and age. Sensitivity analyses varying the follow-up duration from 3 to 37.7m were explored. Results: For the ITT cohort (n=359), mean time spent in TOX and TWiST was significantly longer for the axi-cel cohort compared to SOC, and mean time spent in REL was significantly shorter for axi-cel (Table). Using the base case, quality-adjusted survival was significantly longer for axi-cel by 3.7m, representing a 21.9% relative gain. In threshold analyses, the difference in Q-TWiST ranged from 1.2m (u(TOX)=0, u(REL)=1) to 6.2m (u(TOX)=1, u(REL)=0) in favor of axi-cel. Q-TWiST gains favored axi-cel across all subgroup analyses. Q-TWiST gains from axi-cel increased with longer follow-up. Conclusions: Axi-cel was associated with statistically significant and “clearly clinically important” gains in quality-adjusted OS vs. SOC in R/R LBCL, regardless of the relative decline in QoL associated with treatment toxicity, disease progression, or additional cancer treatment.[Table: see text]