Prevalence, associated causes, and explanatory variables of non-adherence of generic imatinib in chronic myeloid leukemia

Abstract

Introduction: Imatinib has become the standard first-line therapy CML in chronic accelerated and blastic phases and has demonstrated hematologic, cytogenetic, and molecular response rates of over 90%. The non-adherence represents one of the main reasons for not achieving therapeutic success, increased disease progression, risk of resistance, and higher mortality rates. Although drug adherence to imatinib has been analyzed in several studies, its prevalence and associated causes are widely variable. Objective: To determine the prevalence of non-adherence of generic imatinib in chronic myeloid leukemia, associated causes, the proportion of patients who improve this non-compliance, and its explanatory variables. Methods: A retrospective observational cross-sectional analytical study was performed on patients with Chronic Myeloid Leukemia regarding treatment with imatinib generic brand, dispensed by a health institution between August 2018 and December 2022. Drug adherence status was defined as the dependent variable and sociodemographic, clinical, and pharmacological variables as independent variables. The information was extracted from a database of pharmacotherapeutic follow-ups performed on patients automatically identified through algorithms as non-adherent. Pharmacists contacted the patient to intervene and improve this compliance. Univariate, bivariate, and multivariate analysis was performed. Results: A total of 315 patients were analyzed, with a mean age of 52±17 years, median treatment time of 250 days [RIC 97-362], 11% were polymedicated and 16.8% were multimorbid; 30.5% were non-adherent. The associated causes were lack of specialist appointment (39.6%), delay in authorization by the insurer (31.9%) and not having a current medical prescription (5.5%). Of the non-adherent patients, 86% had a history of pharmacological non-adherence and 61% of non-timely claims. After the pharmacist’s intervention, 80% improved this inadequate compliance. During the bivariate analysis, the variables age, time on treatment, and having a history of non-adherence were found to be significant, which, in the multivariate model, behaved as risks with statistical significance. When adjusting the model for the other variables, although they continued to behave as risks, only the variable of a history of non-adherence remained statistically significant. Conclusion: The prevalence of non-adherence is higher than that reported in the literature, the associated causes are mainly administrative, and the persons at greater risk of presenting this health outcome are those with greater age, less time in treatment, and those with a history of non-adherence to other medications.