Abstract
Background:
There is a lack of published estimates of the occurrence of major symptoms and toxicities (SxTox events) and associated costs during AML episodes of care in real-world data. This analysis evaluated the frequency and cost implication of SxTox events in a US commercially insured population with AML during these episodes of care: high intensity chemotherapy (HIC), low intensity chemotherapy (LIC), hematopoietic stem cell transplant (HSCT) and relapsed-refractory (R/R) treatment.
Methods:
A large US healthcare claims database (PharMetrics Plus™) was linked to charge detail master (CDM) hospital data, to identify incident adult patients (≥ 2 outpatient or ≥ 1 inpatient claim with an AML diagnosis) between 1/1/2008 and 3/31/2016. Patients had continuous health plan enrolment for ≥ 6 months pre and ≥ 3 months post the first diagnosis date (index) to assure collection of representative data. Episodes evaluated included HIC induction (inpatient high dose cytarabine+anthracycline use within 3 months post-index), HIC consolidation (cytarabine +/- anthracycline use within 2 months following a prior HIC), LIC (receipt of low-intensity or less toxic chemotherapy in the outpatient setting within 3 months post-index), HSCT (a record of transplant specific diagnosis/procedure codes) and R/R patients (ICD-9 diagnosis code of 205.02 for relapsed AML, or a new line of therapy (new agent or re-initiation of treatment after 90-day gap), after a prior treatment of HIC, LIC, or HSCT). SxTox events of interest observed during episodes were identified via diagnostic and treatment codes and reported as frequencies. A generalized linear model (GLM) was used to assess the contribution of events to total episode costs while controlling for age, gender and Charlson Comorbidity Index (CCI) score.
Results:
The study sample consisted of 1,542 HIC induction (mean age 47.0 years; mean follow-up 2.1 months), 591 consolidation (mean age 47.0 years; mean follow-up 1.5 months), 628 LIC (mean age 64.9 years; mean follow-up 2.0 months), 1,000 HSCT (mean age 51.4 years; mean follow-up 6.4 months) and 910 R/R patients (mean age 52.3 years; mean follow-up 12.6 months). The SxTox events most frequently observed during HIC induction episodes were blood and lymphatic system disorders (98.8%), infections (91.1%), gastrointestinal (GI) disorders (69.1%), nervous system disorders (56.2%), cardiovascular (CV) disorders (44.6%), bleeding (39.0%), and skin and subcutaneous tissue disorders (31.2%). Most HIC consolidation patients experienced blood and lymphatic system disorders (96.4%) and infections (89.5%); while other SxTox events were less frequent except bleeding event rates were higher (73.6%) than during induction (Table 1). LIC patients experienced lower rates of SxTox events compared to HIC induction; frequently observed events were blood and lymphatic system disorders (84.1%), infections (64.3%), bleeding (54.0%), GI (47.5%), CV (36.5%), and renal disorders (27.9%). HSCT patients had similar SxTox event rates as those occurring during HIC induction, while the following SxTox events were more frequent: GI disorders, bleeding events, renal, and CV disorders. R/R patients had a higher occurrence and wider range of high frequency SxTox events, including blood and lymphatic system disorders (96.8%), infections (96.4%), GI disorders (83.2%) bleeding (68.9%), CV disorders (67.5%), nervous system disorders (62.7%), skin and subcutaneous tissue disorders (36.6%), renal disorders (35.6%), and vascular disorders (31.5%). Multivariate GLM results revealed that R/R episodes were sensitive to the occurrence of any SxTox events, which significantly increased costs during R/R episodes, while the occurrence of SxTox events had a mixed impact on increasing total costs in other episodes. CV disorders consistently increased total costs in the following episodes: HIC induction (22% higher; p<0.001), consolidation (23% higher; p=0.005), LIC (36% higher; p<0.001), and R/R episodes (32% higher; p<0.001).
Conclusions:
This analysis observed that major symptoms and toxicities are associated with a significant clinical and economic burden during AML treatment episodes in the US. These SxTox events had the broadest impact on R/R AML patients which may suggest R/R AML patients are less tolerant of currently available treatments. Therapies with lower toxicity would fill an unmet need, especially in R/R AML.
Pandya: Astellas Pharma Global Development: Employment. Chen: QuintilesIMS: Employment. Wilson: Astellas Pharma Global Development: Employment. Groves: QuintilesIMS: Employment. Horvath: Astellas Pharma Global Development: Employment. Wade: QuintilesIMS: Employment.
Author notes
Asterisk with author names denotes non-ASH members.