In:
Circulation, Ovid Technologies (Wolters Kluwer Health), Vol. 146, No. Suppl_1 ( 2022-11-08)
Abstract:
Introduction: Subphenotyping of HF may lead to pathophysiological insights and treatment opportunities. We aimed to identify novel circulating-protein based subphenotypes (SP) in patients with HFrEF. We used repeated proteomic assessments to account for the dynamic nature of HF. Methods: In 382 patients, we performed trimonthly blood sampling during a median follow-up of 2.1 years. We selected all baseline samples and two samples closest to the primary endpoint (PEP; composite of cardiovascular mortality, HF hospitalization, LVAD implantation, and heart transplantation) or censoring and applied an aptamer-based multiplex proteomic approach. We used unsupervised machine learning to derive subphenotypes from 4210 protein trajectories per patient represented by linear mixed effect intercept and slope coefficients. Results: We identified 4 SPs. SP1 was older with higher SBP, EF and history of hypertension. SP2 was older with lower SBP, lower EF, more comorbidities, and worse prognosis (adjHR(95% CI): 4.06 (2.07-7.96)) compared to SP1. SP3 was younger, had lower SBP and EF, more frequently cardiomyopathy (in particular dilated), and a worse prognosis (2.70 (1.33-5.48)). SP4 contained few patients and thus was harder to interpret. It differed from SP3 via higher SBP, higher EF, and seemingly better prognosis. SP allocation was driven by subsets of proteins associated with various biological functions and diseases, e.g. myocardial stress, inflammation, and renal dysfunction. Clinical characteristics of the SPs were aligned with these associations. The prognostic model using SPs based on repeated measurements had a better fit than using SPs based on baseline values (c-index: 0.78 vs 0.72). Conclusions: We found 4 circulating-protein based HFrEF subphenotypes driven by varying protein subsets, with different clinical characteristics and prognosis. Repeated protein measurements improved discriminative ability of the subphenotypes for clinical outcomes.
Type of Medium:
Online Resource
ISSN:
0009-7322
,
1524-4539
DOI:
10.1161/circ.146.suppl_1.12012
Language:
English
Publisher:
Ovid Technologies (Wolters Kluwer Health)
Publication Date:
2022
detail.hit.zdb_id:
1466401-X
Permalink