(BSPOT34) Characteristics of Patients with ATTR Amyloidosis and Methodological Approaches for Establishing Clinical Subtype in the Netherlands

Background: Lack of international consensus on transthyretin amyloidosis (ATTR), classification, coupled with limited specificity of ICD-10 codes, poses a challenge in conducting real world research on this understudied population.

Objectives: 1. To characterize amyloidosis patients; 2. To establish the feasibility of determining relevant ATTR subtypes of cardiomyopathy, peripheral neuropathy, hereditary, and/or wildtype

Methods: We selected patients from the hospital data of the PHARMO Data Network based on hospital admissions or ambulatory consultations of amyloidosis (ICD-10 E85) between 2012-2022. Patients were descriptively characterized in terms of demographics, comorbidities, and high-cost medicine use.

Results: We identified 11,128 patients using broad ICD-10 codes. Applying different algorithmic approaches for cohort selection yielded other cohorts of n=6,460, n=8,658, and n=1,220 (ATTR-CM only). Of the 11,128 patients, 56% were male with a mean age (±SD) of 61 (±23). The most common hospital-based comorbidities were cardiovascular related, including: hypertension (16%), atrial fibrillation (9%), presence of cardiac and vascular implants and grafts (6%), heart failure (6%), chronic ischemic heart disease (6%), lipoprotein metabolism disorder or other lipidaemia (6%), and Type 2 diabetes (6%). The most frequently prescribed high-cost medication classes were antineoplastic agents (3%), selective immunosuppressants (2%), and alkylating agents (2%).

Conclusions: Identified comorbidities point to a higher incidence of the cardiomyopathy subtype. To establish and validate subtypes, linked data from the PHARMO Data Network can provide clinical insights into biopsy results and gene sequences (from the Pathology Registry) as well as bloods (e.g., troponin T, NT pro-BNP) and immunochemistry (e.g., immunoglobulins) (from clinical labs). Unstructured free text notes (from general practitioner data) can be used to complement coded diagnoses. For patients without an amyloidosis classification, algorithms could be developed to predict subtype.