DOI: 10.1161/circ.148.suppl_1.12926 ISSN: 0009-7322

Abstract 12926: Characteristics of Tafamidis-Treated Patients With Mixed Phenotype Transthyretin Amyloidosis: A Real-World Study in the Medicare Fee-for-Service Population

Haechung Chung, Sapna Prasad, Sameer Swarup, Darrin Benjumea, Aaron Crowley, Jose M Alvir, Franca S Angeli, Cindi Sounthonevat
  • Physiology (medical)
  • Cardiology and Cardiovascular Medicine

Introduction: Transthyretin amyloidosis (ATTR amyloidosis) is highly heterogeneous, characterized by predominantly cardiac, predominantly neurologic or mixed phenotypes. Diagnosis of mixed phenotype is often delayed, due in part to fragmented care. Research based on administrative claims data has improved knowledge of the disease, but investigation of the mixed phenotype has been limited by deficiencies in ICD diagnosis codes.

Hypothesis: Proxy algorithms can be used to identify and characterize mixed phenotype ATTR amyloidosis in Medicare patients.

Methods: This noninterventional, retrospective, observational cohort study was conducted using de-identified patient data from the US Medicare database (Jan 1, 2016-Dec 31, 2021). Based on literature review and clinician consultation, 3 distinct proxy algorithms were developed to select 3 cohorts of tafamidis-treated patients with possible mixed phenotype ATTR amyloidosis ( Figure ). Baseline patient characteristics were analyzed descriptively.

Results: In 1,243 patients, mean (SD) age ranged from 76.0 (6.3)-79.5 (6.5) y, and 68-83% were men. Comorbidity burden was similar across cohorts (Charlson comorbidity indices: 1.6 [0.9]-1.7 [1.0]). The most common cardiovascular comorbidities were hypertensive diseases, arrhythmia, conduction disorders and LVH; the most common noncardiovascular comorbidities were cataracts, chronic and acute kidney disease, constipation and diabetes ( Table ).

Conclusions: In this real-world study, proxy algorithms helped identify and characterize mixed phenotype ATTR amyloidosis in Medicare patients. Additional research is needed to further validate these algorithms.

More from our Archive