Biomarker Guided Therapy For Heart Failure With Mid-Range EF

Document Type

Conference Proceeding

Publication Date


Publication Title

Journal of Cardiac Failure


Background: Heart failure (HF) with mid-range ejection fraction (HFmrEF, EF 40-49%) accounts for ∼15% of HF patients and these are a heterogeneous group with a phenotypic character that reflects aspects of both preserved and reduced EF HF. There are no proven treatments for this group but identifying subsets of patients that will respond to conventional therapies such as beta blockers (BB) would be desirable. We hypothesized that HFmrEF patients with elevated Nt-proBNP or ST2 would have more favorable response to BB therapy.

Methods: A prospective registry was conducted that enrolled 1760 HF patients who met Framingham HF criterion and who received care through our health system and insurance product. Participants donated blood for biomarker measurements at enrollment and their medication exposure was quantified over time using pharmacy claims data. The current study utilized all participants who had EF 40-49% and baseline biomarker measurement (n=338). The primary endpoint was all-cause mortality. Biomarker categories were dichotomized at Nt-BNP >1000ng/L and ST2 ≥35ng/mL. We quantified the HR for BB exposure overall and divided by biomarker level using Cox models adjusted for MAGGIC score and BB propensity score.

Results: Vital status and Nt-BNP measurements were available for all 338 patients, while 279 had ST2 measured, and 201 had both markers (Table 1). For Nt-BNP, 200 (58%) had elevated levels while 143 (42%) had low values. BB use in patients with elevated Nt-BNP levels had a hazard ratio (HR) of 0.43 (p=0.1) compared to HR 0.97 (p=0.97) in the low Nt-BNP group. For ST2, 195 patients (70%) had levels <35 ng/mL while 84 (30%) had elevated levels. BB use in the elevated ST2 cohort had a HR 0.08 (p=0.71) vs. 0.62 (p=0.72) in the low ST2 group, a statistically significant interaction. (p int= 0.089). Both low marker groups had very low event rates (3.5% and 3.1% 1 year death, respectively). The combination of the two markers together (either elevated vs. both low) identified an especially low risk group with 1.5% dying and predicted the BB benefit association; in the low marker group BB HR =0.99 (p >0.99) while in patients with either ST2 or Nt-BNP elevated the BB HR was 0.06 (p=0.03).

Conclusions: In this observational study of HFmrEF patients, those with low ST2 and Nt-BNP have low risk of death and little benefit of BB therapy whereas among the approximately two-thirds of patients with either marker elevated a substantial BB benefit association was seen.





First Page


This document is currently not available here.