INTRODUCTION: Chronic myeloid leukemia (CML), a myeloproliferative neoplasm is primarily treated using tyrosine kinase inhibitors (TKIs). Several next-generation TKIs (dasatinib, nilotinib, bosutinib, ponatinib) have been approved since the first approval of imatinib in 2002. With varying safety profiles, data on non-trial long-term TKI use and associated adverse events (AE) can aid in clinical decision making. Using population based data sources the current study assessed treatment patterns and AEs in a non-trial setting among TKI users with CML.

METHODS: A retrospective cohort analysis was conducted using MarketScan Commercial, and Supplemental Medicare databases (2012-2016). Data includes details on medical (e.g., date of diagnosis, diagnosis codes, procedures) and pharmacy (e.g., drug dose, duration, strength) utilization for over 90 million individuals enrolled in employer-sponsored health-plans in the US. Patients with a prescription claim for TKI along with at least 2 medical claims on separate dates with a diagnosis code for CML (ICD-9-CM: 205.10 - 205.12; ICD-10: C92.10-C92.12) were selected. Patients were aged ≥18 years at TKI initiation and had continuous health plan enrollment ≥6 months before and ≥6 months after TKI initiation. Patients were followed from TKI initiation until health plan disenrollment or end of database, whichever occurred earlier (defines follow-up period). TKI treatment patterns including initial dose, initial and total TKI treatment duration, treatment switching and discontinuation were assessed. Frequency and proportion for incident (i.e. newly observed conditions) AEs including rash, bleeding, gastrointestinal disorders (nausea, vomiting, diarrhea), tumor lysis syndrome and cardiovascular events (CHF, arrhythmia, peripheral arterial occlusive disease, cerebrovascular events) and post-AE outcomes (switching, discontinuation, dose reduction) were assessed. AE list was developed based on commonly reported events in clinical trials involving TKIs for CML and clinical judgement. All analyses were descriptive in nature.

RESULTS: Study included 2,213 CML patients with mean age (standard deviation [SD]) of 55 (15) years, of which 55% were male and 55% had ≥2 co-morbid conditions. Post CML diagnosis, first-line of TKI initiated was imatinib (41%), dasatinib (36%), nilotinib (22%), and bosutinib and ponatinib (1% each). The average (SD) follow-up duration post TKI initiation was 607 (442) days. Mean (SD) duration of first-line TKI prescription was 392 (383) days. 61% of patients continued initial TKI during the follow-up period, 7% discontinued and 32% switched/re-initiated TKI (Figure 1). 39.2% patients had at least 1 incident AE while on index TKI therapy. GI disorders (20.4%) were the most commonly observed AE, followed by cardiovascular events (14.6%), rash (8.4%), bleeding (8.0%), and TLS (0.2%). The average (SD) time to GI disorders, bleeding, cardiovascular event, rash, and TLS was 242 (274) days, 238 (272) days, 219 (261) days, 222 (255) days, and 223 (285) days, respectively. The rates of discontinuation, treatment switch or dose reduction were largely similar across all AEs (Figure 2).

The mean duration on any TKI (all-lines of therapy) was 561 (438) days. The most commonly observed AEs on while of TKI therapy (irrespective of line of therapy) were GI disorders (24.7%), followed by cardiovascular events (17.3%), rash (11.5%), bleeding (10.1%) and TLS (0.5%).

CONCLUSIONS: The study helps address the current literature gap of long-term treatment patterns and AE among patients with CML using TKIs in a non-clinical trial setting. During follow-up 46% patients experienced at least 1 incident AE while on TKI therapy. Majority of patients (61%) continued same TKI therapy following AE. The type and rate of AE were similar following first-line therapy and anytime on TKI (any-line) during the follow-up period. Study results may not be comparable to clinical trial findings due to certain limitations; e.g., this study did not capture all adverse events but assessed incident AE based on a pre-specified list. Also, due to limitations of administrative claims data, assessment of AE grade was not feasible. However, findings from this study can complement clinical trial data in selection of TKIs for long-term use in real-world setting and can also be of value in selecting TKIs for future clinical studies involving novel combinations in CML.


Karve:AbbVie: Employment, Equity Ownership. Huang:ZS Associates: Employment; Novo Nordisk Inc: Equity Ownership; AstraZeneca: Research Funding. Ranade:ZS Associates: Employment. Porwal:ZS Associates: Employment. Desai:AbbVie: Employment, Equity Ownership. Rosenberg:AbbVie: Employment, Equity Ownership.

Author notes


Asterisk with author names denotes non-ASH members.