Clinical Phenomapping Of Obese Heart Failure With Preserved Ejection Fraction

Vivek Jani,Soumya Vungarala,Edwin Yoo,Steven Hsu,Virginia Hahn,David Kass,Kavita Sharma
DOI: https://doi.org/10.1016/j.cardfail.2023.10.010
IF: 6.592
2024-01-01
Journal of Cardiac Failure
Abstract:Background Heart failure with preserved ejection fraction (HFpEF) encompasses >50% of all HF and is a heterogeneous syndrome with few proven treatments. Phenomapping is a machine-learning based technique for sub-group identification which we sought to perform in an obese HFpEF cohort, a predominant form of HFpEF in which phenomapping to date is limited. Methods We prospectively studied 378 patients with HFpEF and performed detailed clinical, laboratory, echocardiographic, and hemodynamic phenotyping. A total of 207/378 (55%) and 104/378 (28%) underwent right heart catheterization and pulmonary function testing, respectively. We then assembled 104 clinical, laboratory, echocardiographic, and hemodynamic features for phenomapping using machine learning cluster analysis. Variables were compared between group clusters using a Kruskal Wallis test. Results Cluster analysis (Figure 1A) identified three phenogroups: Obese/Metabolic (Ob/Mt) Occult HFpEF (n=186 patients), Ob/Mt Manifest HFpEF (n=81), and Elderly/Restrictive HFpEF (n=65). The mean age of the elderly group was 75±8 years, significantly higher than both obese groups (p=0.0001). Both Ob/Mt groups had higher BMI (Occult 38±11, Manifest 38±12) than Elderly/Restrictive HFpEF (p=0.0001) and had evidence of diabetes and dyslipidemia. Ob/Mt Manifest HFpEF patients had small left ventricles (p=0.006), higher E/A ratio (p<0.0001), elevated filling pressures (pulmonary capillary wedge pressure, PCWP; Manifest 22±6 vs. Occult 13±4 mmHg, p<0.0001), and elevated mean pulmonary pressure (Manifest 37±8 vs. Occult 13±4 mmHg). Elderly/Restrictive patients showed the smallest left ventricles and highest E/A ratio. Further, Elderly/Restrictive patients had markedly depressed cardiac output (p=0.049) and cardiac index (p=0.04). Kaplan-Meier survival analysis for a combined endpoint of mortality and heart failure hospitalization showed worst outcomes in Ob/Mt Manifest (p<0.0001) compared to the other subgroups (Figure 1B). Stepwise regression found differences in survival between these groups was predominantly driven by RV stroke work index, mean pulmonary artery pressure, and PCWP. Conclusion Clinical phenomapping reveals a new sub-group of obese HFpEF patients that are volume-overloaded with much worse survival and more heart failure hospitalizations than other subgroups.
cardiac & cardiovascular systems
What problem does this paper attempt to address?