Quality-Adjusted Time without Symptoms of Disease or Toxicity (Q-TWiST) of Ponatinib Versus Imatinib, Administered in Combination with Reduced-Intensity Chemotherapy, in Patients with Newly Diagnosed Philadelphia Chromosome–Positive Acute Lymphoblastic Leukemia

Introduction: Ponatinib is a third-generation tyrosine kinase inhibitor that effectively inhibits BCR::ABL1 with or without single-mutation variants, including T315I. In the analysis of the primary endpoint from the PhaLLCON trial, ponatinib showed a higher minimal residual disease-negative complete response rate at the end of induction than imatinib (34.4% vs 16.7%; risk difference, 0.18 [95% CI, 0.06-0.29]; p 0.002), and a significant improvement in progression-free survival (PFS) over imatinib (median PFS; 20.0 vs 7.9 months; HR 0.58 95% CI, 0.41-0.83) in patients with newly diagnosed Ph+ ALL. The safety profile was comparable between treatment arms. This post-hoc analysis evaluated whether the longer PFS with ponatinib vs imatinib was at the expense of patients’ quality of life using a Quality‐adjusted Time Without Symptoms of disease or Toxicity (Q-TWiST) analysis.

Methods: Q-TWiST, an integrated measure of treatment benefit and risk, partitioned patients’ survival time into 3 mutually exclusive health states (HS): TOX (time with toxicity after randomization and before disease progression), TWiST (time in the progression-free period without toxicity; PFS - TOX), and REL (time from disease progression until death or end of follow-up, whichever occurs first; OS - PFS). Restricted mean survival time for each HS was estimated up to the end of follow-up per the current data cutoff (Aug 12, 2022). In the base-case analysis, TOX was defined as time spent with treatment-emergent adverse event (TEAE) grade 3+ and Q-TWiST was estimated by weighing time spent in each HS using the health utility values (TWIST 0.8, TOX 0.6, and REL 0.4) based on data from the literature (utilities range from 0 [death] to 1.0 [“perfect” health]). Sensitivity analyses were performed to assess the robustness of the base-case results by varying the following: utilities for both TOX and REL states, follow-up time, and the definition of TOX using either TEAE grade 2+ or a patient-reported measure of overall treatment tolerability (i.e., the FACT-GP5 response level of “quite a bit” or “very much” bothered by side effects of treatment). A relative Q-TWiST gain of ≥10% over the restricted mean survival of the comparator (imatinib) is deemed to be clinically meaningful.

Results: The analysis included all randomized patients (ponatinib, n=164; imatinib, n=81). As the OS data were still immature at the current data cutoff, the restricted mean survival was similar between arms - 1,082 days for ponatinib and 1,025 days for imatinib. In the base-case analysis, the mean durations of TOX, TWiST, and REL health states were 96, 398, and 588 days for ponatinib, respectively, and 77, 574, 374 days, for imatinib. The difference in mean time in TOX was not statistically different between ponatinib vs. imatinib (18.85 days [95% CI, −11.82, 49.52]; p 0.228). The mean time for the TWiST state was 214.46 days (95% CI, 70.26, 358.67; p 0.004) longer for ponatinib vs. imatinib; and mean time for the REL state was 175.92 days (95% CI, -325.4, -26.45; p 0.021) shorter with ponatinib. The estimated difference in Q-TWiST was significantly higher for ponatinib vs. imatinib (112.51 days [95% CI, 21.91, 203.11]; p 0.015), and the relative Q-TWiST gain was 10.98%, exceeding the pre-specified clinical meaningfulness threshold. Sensitivity analysis results consistently supported the robustness of the base-case findings.

Conclusion: This Q-TWiST analysis demonstrates that treatment with ponatinib, compared to imatinib, resulted in a significantly and meaningfully longer quality-adjusted survival, further supporting the clinical benefit-risk profile of ponatinib for patients with newly diagnosed Ph+ ALL. These analyses should be revisited upon maturity of the OS data.