Penalized Model-Based Unsupervised Phenomapping Unravels Distinctive HFrEF Phenotypes With Improved Outcomes Discrimination From Sacubitril/Valsartan Treatment Independent of MAGGIC Score

J Am Heart Assoc. 2023 Sep 19;12(18):e028860. doi: 10.1161/JAHA.122.028860. Epub 2023 Sep 8.

Abstract

Background The angiotensin receptor-neprilysin inhibitor (LCZ696) has emerged as a promising pharmacological intervention against renin-angiotensin system inhibitor in reduced ejection fraction heart failure (HFrEF). Whether the therapeutic benefits may vary among heterogeneous HFrEF subgroups remains unknown. Methods and Results This study comprised a pooled 2-center analysis including 1103 patients with symptomatic HFrEF with LCZ696 use and another 1103 independent HFrEF control cohort (with renin-angiotensin system inhibitor use) matched for age, sex, left ventricular ejection fraction, and comorbidity conditions. Three main distinct phenogroup clusterings were identified from unsupervised machine learning using 29 clinical variables: phenogroup 1 (youngest, relatively lower diabetes prevalence, highest glomerular filtration rate with largest left ventricular size and left ventricular wall stress); phenogroup 2 (oldest, lean, highest diabetes and vascular diseases prevalence, lowest highest glomerular filtration rate with smallest left ventricular size and mass), and phenogroup 3 (lowest clinical comorbidity with largest left ventricular mass and highest hypertrophy prevalence). During the median 1.74-year follow-up, phenogroup assignment provided improved prognostic discrimination beyond Meta-Analysis Global Group in Chronic Heart Failure risk score risk score (all net reclassification index P<0.05) with overall good calibrations. While phenogroup 1 showed overall best clinical outcomes, phenogroup 2 demonstrated highest cardiovascular death and worst renal end point, with phenogroup 3 having the highest all-cause death rate and HF hospitalization among groups, respectively. These findings were broadly consistent when compared with the renin-angiotensin system inhibitor control as reference group. Conclusions Phenomapping provided novel insights on unique characteristics and cardiac features among patients with HFrEF with sacubitril/valsartan treatment. These findings further showed potentiality in identifying potential sacubitril/valsartan responders and nonresponders with improved outcome discrimination among patients with HFrEF beyond clinical scoring.

Keywords: MAGGIC score; angiotensin receptor–neprilysin inhibitor (LCZ696); guideline‐directed medical therapy; heart failure with reduced ejection fraction (HFrEF); phenomapping.

Publication types

  • Multicenter Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Antihypertensive Agents
  • Female
  • Heart Failure* / drug therapy
  • Humans
  • Male
  • Stroke Volume
  • Valsartan / therapeutic use
  • Ventricular Function, Left

Substances

  • Antihypertensive Agents
  • sacubitril
  • sacubitril and valsartan sodium hydrate drug combination
  • Valsartan