Second-generation tyrosine kinase inhibitors vs imatinib in adult ph+ ALL: Target-trial emulation using a propensity-matched multinational cohort

Abstract Background: Philadelphia chromosome–positive acute lymphoblastic leukemia (Ph + ALL) accounts for 20–30 % of adult ALL and carries the worst genetic prognosis despite imatinib-based regimens. Second-generation BCR-ABL1 tyrosine-kinase inhibitors (2G-TKIs; dasatinib, nilotinib, bosutinib) achieve deeper kinase inhibition and superior central-nervous-system penetration, yet their incremental benefit over imatinib in modern adult practice remains uncertain because head-to-head phase-3 trials are lacking. Thus, we emulated a randomized trial to compare overall survival (OS) and toxicity of frontline 2G-TKI versus imatinib in newly diagnosed adult Ph + ALL. Secondary objectives included detailed characterization of adverse-event profiles and validation of findings with falsification (negative-control) outcomes. Methods: TriNetX Research Network (157 M patients) was queried for adults (≥18 y) with incident Ph + ALL (ICD-10 C91.0 plus BCR-ABL1 code) from January 2014–February 2025. Patients initiating a TKI within 90 d of diagnosis were eligible; those with prior TKI use, pre-TKI transplantation, missing chemotherapy, or <30 d follow-up were excluded. A prespecified target-trial framework was applied. Propensity scores derived from demographics, comorbidities, BMI, and baseline laboratories enabled 1:1 nearest-neighbour matching (caliper 0.1) of imatinib and 2G-TKI recipients. Primary endpoint was 5-year OS (robust Cox models). Secondary endpoints including grade ≥3 cytopenias, thromboembolism, neuropathy, metabolic and gastrointestinal toxicities, and transplantation were compared with modified Poisson regression. Sensitivity analyses used an on-treatment cohort (≥12 administrations in 6 mo) and a U.S.-only subset. Gout, nephrolithiasis, and cholelithiasis served as negative-control outcomes. Results: After 1:1 matching (n=599 per group; median follow-up 1,140 vs 1,129 days), second-generation TKIs improved five-year overall survival to 77.1% versus 67.1% (HR 0.677, 95% CI 0.544 to 0.842; P<0.001). This translates to about 10 more survivors per 100 treated at five years. Findings were consistent in sensitivity analyses: on-treatment HR 0.508 (95% CI 0.361 to 0.716; P<0.001) and U.S.-only HR 0.753 (95% CI 0.594 to 0.955; P=0.022). The benefit was most evident in patients older than 35 years (HR 0.56, 95% CI 0.438 to 0.718); estimates in 18–35 years did not reach significance (HR 1.29, 95% CI 0.699 to 1.899). By agent, dasatinib accounted for the aggregate advantage (HR 0.70, 95% CI 0.572 to 0.865), whereas nilotinib and bosutinib were neutral in small cohorts. Safety trade-offs were expected: per 100 patients treated with a second-generation TKI rather than imatinib, approximately 14 more experienced peripheral neuropathy (RR 2.06, 95% CI 1.62 to 2.63; P<0.001), 8 more had grade 3 or higher thrombocytopenia (RR 1.14, 95% CI 1.04 to 1.25; P=0.006), and 5 more had venous thrombosis (RR 1.23, 95% CI 1.01 to 1.51; P=0.045); nausea or vomiting and hyperglycemia were also more frequent (absolute differences ~7 and ~6 per 100, respectively). Allogeneic transplantation use and central nervous system relapse were similar between groups (transplant RR 0.88, 95% CI 0.74 to 1.04; P=0.121; CNS relapse RR 1.53, 95% CI 0.84 to 2.79; P=0.165). Negative control outcomes were comparable, supporting minimal residual confounding Conclusion: Frontline second-generation BCR-ABL1 TKIs, predominantly dasatinib, deliver a clinically meaningful five-year survival advantage over imatinib in real-world adult Ph+ ALL without increasing treatment-related mortality. The benefit is concentrated in adults older than 35 years, whereas estimates in 18–35 years are not significant, supporting age-informed TKI selection. Transplant use and CNS relapse are similar between groups, and falsification outcomes are neutral, strengthening causal confidence. In practice, second-generation TKIs should be the preferred initial therapy, with agent choice often favoring dasatinib when appropriate and proactive monitoring for peripheral neuropathy, thrombocytopenia, venous thrombosis, gastrointestinal intolerance, and hyperglycemia. These data provide a benchmark for MRD-adapted, chemotherapy-sparing strategies that incorporate blinatumomab and inform integration of third-generation TKIs in modern sequencing.