LONGITUDINAL ADHERENCE PATTERNS OF ASCIMINIB VERSUS OTHER TYROSINE KINASE INHIBITORS IN CHRONIC MYELOID LEUKEMIA: A GROUP-BASED TRAJECTORY ANALYSIS
Author(s)
Gabriel Marquez, MS1, Dominick Latremouille-Viau, MSc2, Frédéric Kinkead, MSc2, Carmine Rossi, PhD2, Luo Li, PhD3, Andrea Damon, PhD3, Annie Guerin, PhD2, David Wei, PhD3.
1Novartis Pharmaceuticals Corporation, Dublin, Ireland, 2Analysis Group Inc., Montréal, QC, Canada, 3Novartis Pharmaceuticals Corporation, East Hanover, NJ, USA.
1Novartis Pharmaceuticals Corporation, Dublin, Ireland, 2Analysis Group Inc., Montréal, QC, Canada, 3Novartis Pharmaceuticals Corporation, East Hanover, NJ, USA.
OBJECTIVES: Group-based trajectory modeling (GBTM) captures dynamic adherence patterns over time, representing a novel application in oncology. This approach is particularly relevant in chronic myeloid leukemia (CML), where adherence to tyrosine kinase inhibitors (TKIs) is a key indicator of treatment tolerability and determinant of long-term efficacy. Longitudinal adherence trajectories were compared between asciminib and other TKIs using GBTM, and their clinical relevance was evaluated with guideline-recommended response milestones and treatment modifications in patients with CML.
METHODS: Adults with CML initiating second-line (2L) TKIs (asciminib or other TKIs; index date) in the US HealthVerity database (2016-2025) were included. GBTM with a cubic polynomial was used to identify adherence patterns based on monthly proportion of days covered (PDC) measures during the 12-month post-index period. Model selection was based on statistical fit, clinical interpretability, and minimum group size. Composition of TKIs within each trajectory group was reviewed and compared.
RESULTS: Among 1,702 patients, 90 received asciminib and 1,612 other TKIs (both cohorts: mean age, 55 years; 56% male). Three trajectories were identified: highly adherent (N=825; stable mean monthly PDC>90%), slow decline (N=387; early TKI discontinuation between months 3 and 12 post-index, corresponding to key assessment milestones), and rapid decline (N=490; poor adherence and discontinuation within the first 3 months post-index, preceding initial milestones and suggesting early tolerability issues). Model fit was robust (APPA>0.7 and OCC>5 across all groups). After entropy balancing, asciminib-treated patients were more likely to be highly adherent (65.6% vs 47.5%; RR=1.38, p<0.001) and less likely to be rapid decliners (12.2% vs 29.7%; RR=0.41, p<0.001) than those receiving other TKIs.
CONCLUSIONS: GBTM provides a clinically meaningful framework to assess adherence dynamics in CML, aligning with key guideline-recommended response milestones and treatment decisions. Asciminib was associated with higher sustained adherence than other TKIs, which is consistent with its superior efficacy and favorable safety/tolerability.
METHODS: Adults with CML initiating second-line (2L) TKIs (asciminib or other TKIs; index date) in the US HealthVerity database (2016-2025) were included. GBTM with a cubic polynomial was used to identify adherence patterns based on monthly proportion of days covered (PDC) measures during the 12-month post-index period. Model selection was based on statistical fit, clinical interpretability, and minimum group size. Composition of TKIs within each trajectory group was reviewed and compared.
RESULTS: Among 1,702 patients, 90 received asciminib and 1,612 other TKIs (both cohorts: mean age, 55 years; 56% male). Three trajectories were identified: highly adherent (N=825; stable mean monthly PDC>90%), slow decline (N=387; early TKI discontinuation between months 3 and 12 post-index, corresponding to key assessment milestones), and rapid decline (N=490; poor adherence and discontinuation within the first 3 months post-index, preceding initial milestones and suggesting early tolerability issues). Model fit was robust (APPA>0.7 and OCC>5 across all groups). After entropy balancing, asciminib-treated patients were more likely to be highly adherent (65.6% vs 47.5%; RR=1.38, p<0.001) and less likely to be rapid decliners (12.2% vs 29.7%; RR=0.41, p<0.001) than those receiving other TKIs.
CONCLUSIONS: GBTM provides a clinically meaningful framework to assess adherence dynamics in CML, aligning with key guideline-recommended response milestones and treatment decisions. Asciminib was associated with higher sustained adherence than other TKIs, which is consistent with its superior efficacy and favorable safety/tolerability.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
MSR143
Topic
Clinical Outcomes, Methodological & Statistical Research
Disease
No Additional Disease & Conditions/Specialized Treatment Areas, Oncology