Quality-Adjusted Time Without Symptoms of Disease or Toxicity (Q-TWiST) in Patients With Newly Diagnosed Philadelphia Chromosome-Positive Acute Lymphoblastic Leukemia: A Comparison of Ponatinib Versus Imatinib

Abstract

Background: In the phase 3 ponatinib-3001 trial (PhALLCON, NCT03589326), ponatinib demonstrated superior efficacy over imatinib with comparable safety in patients with newly diagnosed Philadelphia-positive acute lymphoblastic leukemia (Ph+ ALL). This post hoc analysis evaluated the net benefits of ponatinib using a quality-adjusted time without symptoms of disease or toxicity (Q-TWiST) approach.

Methods: Overall survival (OS) time for patients from PhALLCON was partitioned into three health states: TOX (time with grade 3+ treatment-emergent adverse events [TEAEs] before disease progression), TWiST (time without toxicity before progression), and REL (time from progression until death or end of follow-up). Q-TWiST was calculated as the sum of health utility-weighted restricted mean durations of the three states. A relative Q-TWiST gain of ≥ 10% was considered clinically important. Sensitivity analyses were conducted by varying TOX and REL utilities, follow-up time, and the TOX definition (using grade 2+ TEAEs or patient-perceived treatment tolerability assessed by the FACT-GP5).

Results: Among all randomized patients (ponatinib n = 164, imatinib n = 81), restricted mean OS was similar between arms (1082.2 vs. 1024.8 days; p = 0.373). In the base-case analysis, mean TWiST was 214.5 days longer with ponatinib versus imatinib (95% CI 70.3-358.7; p = 0.004), REL was shorter by 175.9 days (325.4-26.5; p = 0.021), and TOX was not significantly different between arms (p = 0.228). The relative Q-TWiST gain (10.98%) was clinically important. Sensitivity analyses consistently supported the robustness of the base-case findings.

Conclusion: Ponatinib may prolong quality-adjusted survival compared with imatinib, supporting the benefit-risk profile of ponatinib as a front-line treatment for Ph+ ALL.

Trial registration: NCT03589326.

Keywords: Philadelphia chromosome‐positive; Q‐TWiST analysis; acute lymphoblastic leukemia; imatinib; ponatinib; quality of life; tyrosine kinase inhibitor.

FIGURE 1

Partitioned survival curves in the base‐case analysis. (A) Ponatinib. (B) Imatinib. Abbreviations: REL, relapse (period from disease progression until end of follow‐up/death); TEAE, treatment‐emergent adverse event; TOX, toxicity (sum of all periods in which patients experienced grade 3+ TEAEs during the progression‐free period); TWiST, time without symptoms or toxicities.

FIGURE 2

Threshold utility plot of absolute Q‐TWiST gain (in days) with ponatinib relative to imatinib at the maximum follow‐up of 1247 days. The TOX and REL utility values varied from 0 to 1, while that of TWiST was fixed at 0.8. Red lines represent the average Q‐TWiST gain (in days) with ponatinib compared to imatinib, and numbers associated with the red lines are the absolute Q‐TWiST gain for the corresponding utility values. The color of the shaded horizontal lines represents the significance of the comparison, with yellow indicating a significantly better effect (p < 0.05) with ponatinib than imatinib and light gray indicating no significant difference in treatment effect. Abbreviations: Q‐TWiST, quality‐adjusted TWiST; REL, relapse (period from disease progression until end of follow‐up/death); TEAE, treatment‐emergent adverse event; TOX, toxicity (sum of all periods in which patients experienced grade 3+ TEAEs during the progression‐free period); TWiST, time without symptoms or toxicities.

FIGURE 3

Absolute Q‐TWiST gain (in days) with ponatinib relative to imatinib at various follow‐up times. The Q‐TWiST gain was estimated using the utility values in the base‐case analysis. Error bars represent the 95% CI. Abbreviations: CI, confidence interval; Q‐TWiST, quality‐adjusted TWiST; TWiST, time without symptoms or toxicities.