New biomarker-driven model predicts clinical outcomes in HFpEF
Click Here to Manage Email Alerts
Researchers have developed a risk prediction model for clinical outcomes in HF with preserved ejection fraction, now including N-terminal pro-B-type natriuretic peptide and high-sensitivity cardiac troponin.
The new risk calculator, which was developed using patient data from the EMPEROR-Preserved trial, utilized readily available clinical information to predict a composite endpoint of HF hospitalization or CV death, all-cause death, CV death and HF hospitalization.
The model was externally validated using the patient cohort from PARAGON-HF, according to the study published in the European Journal of Heart Failure.
“The biomarker-driven risk models that we present here provide effective risk discrimination for patients with HFpEF both for HF-related morbidity and mortality. ... Their practical value is enhanced by the fact that these models require only two widely available biomarkers and a handful of clinical variables,” Stuart John Pocock, PhD, professor of medical statistics at the London School of Hygiene and Tropical Medicine, U.K., and colleagues wrote. “The present work demonstrates that these biomarkers are similarly important in the prediction of risk of patients with HFpEF. There is a dearth of risk models for HFpEF patients, particularly contemporary models that incorporate biomarkers with strong prognostic value.”
Pocock and colleagues utilized patient data from the EMPEROR-Preserved randomized controlled trial to synthesize a biomarker-driven prognostic tool to predict the primary composite outcome of HF hospitalization or CV death, all-cause death, CV death and HF hospitalization in patients with chronic HFpEF.
In the EMPEROR-Preserved trial, researchers evaluated the use of the empagliflozin (Jardiance, Boehringer Ingelheim/Eli Lilly) to lower risk for the primary outcome of CV death or hospitalization for HF compared with placebo in patients with HFpEF.
As Healio previously reported, the researchers demonstrated for the first time that the SGLT2 inhibitor empagliflozin improved the primary outcome in patients with HFpEF compared with placebo.
Predictive ability of the new model
Using this data, the researchers who undertook the present analysis developed a risk calculator that effectively estimated a patient’s 2-year risk for the primary composite endpoint with a C statistic of 0.75.
The model included NT-proBNP and high-sensitivity troponin, which were the strongest predictors of the primary composite outcome, as well as time from last hospitalization, NYHA class III or IV, history of chronic obstructive pulmonary disease (COPD), insulin-treated diabetes, low hemoglobin and time since HF diagnosis (P for all < .001).
Researchers reported that patients in the top 10th of risk had an approximately 22 times greater primary composite outcome event rate compared with those in the lowest 10th of estimated risk.
A model for HF hospitalization alone had even better discrimination, according to the study, with a C statistic of 0.79.
“To the best of our knowledge, only four risk models have been developed for HFpEF,” the researchers wrote. “The CHARM and MAGGIC risk scores did not include BNP or NT-proBNP. ... Still, the levels of BNP were added to the MAGGIC score a posteriori showing a strong association with outcomes. I-Preserve included NT-proBNP, but the models were rather complex with 12 variables, and they included quality of life scores not routinely performed in clinical practice. The 3A3B score was tested in Japanese patients with left ventricular ejection fraction 50% from the CHART-2. ... Its risk scores did not include cardiac troponins or the combination of natriuretic peptides with cardiac troponins.”
In order of predictive strength, Pocock and colleagues ranked NT-proBNP and high-sensitivity troponin at the top, followed by history of COPD, low albumin, older age, LVEF of 50% or more, NYHA class III or IV and insulin-treated diabetes (P for all < .001).
External validation of the new model
The researchers conducted an external validation of the new risk estimation model using the PARAGON-HF cohort.
In the PARAGON-HF trial, researchers assessed sacubitril/valsartan (Entresto, Novartis) for the reduction of the primary composite outcome of CV death or HF hospitalization compared with valsartan in patients with HFpEF.
As Healio previously reported, sacubitril/valsartan did not reduce the rate of the primary outcome in patients with HFpEF and misses its primary endpoint.
External validation using the PARAGON-HF cohort yielded C statistics of 0.71 for the primary composite outcome; 0.71 for HF hospitalization; 0.72 for all-cause mortality; and 0.72 for CV death.
“Importantly, our biomarker-driven risk models are simple to use since they rely on only a few readily available clinical variables plus NT-proBNP and high-sensitivity cardiac troponin T, which can be easily obtained in most clinical settings,” the researchers wrote. “Our findings support the joint assessment of NT-proBNP and high-sensitivity cardiac troponin T in the comprehensive risk assessment of all heart failure patients, both HFpEF (as reported here) and HFrEF (as previously reported).”
References:
- Anker SD, et al. N Engl J Med. 2021;doi:10.1056/NEJMoa2107038.
- Solomon SD, et al. N Engl J Med. 2019;doi:10.1056/NEJMoa1908655.
Collapse
Click Here to Manage Email Alerts