Shah RV; Vanderbilt University Medical Center Nashville TN
Hwang SJ; Murthy VL; Zhao S; Tanriverdi K; Gajjar P; Duarte K;
Schoenike M; Farrell R; Brooks LC; Gopal DM; Ho JE; Girerd N; Vasan RS;
Levy D; Freedman JE; Lewis GD; Nayor M

Journal of the American Heart Association. 12(21):e029980, 2023 Nov 07.

BACKGROUND: While exercise impairments are central to symptoms and
diagnosis of heart failure with preserved ejection fraction (HFpEF), prior
studies of HFpEF biomarkers have mostly focused on resting phenotypes. We
combined precise exercise phenotypes with cardiovascular proteomics to
identify protein signatures of HFpEF exercise responses and new potential
therapeutic targets.
METHODS AND RESULTS: We analyzed 277 proteins (Olink) in 151 individuals
(N=103 HFpEF, 48 controls; 62+/-11 years; 56% women) with cardiopulmonary
exercise testing with invasive monitoring. Using ridge regression adjusted
for age/sex, we defined proteomic signatures of 5 physiological variables
involved in HFpEF: peak oxygen uptake, peak cardiac output, pulmonary
capillary wedge pressure/cardiac output slope, peak pulmonary vascular
resistance, and peak peripheral O2 extraction. Multiprotein signatures of
each of the exercise phenotypes captured a significant proportion of
variance in respective exercise phenotypes. Interrogating the importance
(ridge coefficient magnitude) of specific proteins in each signature
highlighted proteins with putative links to HFpEF pathophysiology (eg,
inflammatory, profibrotic proteins), and novel proteins linked to distinct
physiologies (eg, proteins involved in multiorgan [kidney, liver, muscle,
adipose] health) were implicated in impaired O2 extraction. In a separate
sample (N=522, 261 HF events), proteomic signatures of peak oxygen uptake
and pulmonary capillary wedge pressure/cardiac output slope were
associated with incident HFpEF (odds ratios, 0.67 [95% CI, 0.50-0.90] and
1.43 [95% CI, 1.11-1.85], respectively) with adjustment for clinical
factors and B-type natriuretic peptides.
CONCLUSIONS: The cardiovascular proteome is associated with precision
exercise phenotypes in HFpEF, suggesting novel mechanistic targets and
potential methods for risk stratification to prevent HFpEF early in its
pathogenesis.