Biomarker-Based Risk Score May Predict Heart Failure in Type 2 Diabetes

man clenching chest in pain
Severe heartache, man suffering from chest pain, having heart attack or painful cramps, pressing on chest with painful expression.
People with type 2 diabetes are at an increased risk of several health complications, including heart failure. Researchers analyzed data from two large multinational trials to see if certain biomarkers could predict those at greatest risk for being hospitalized due to a serious cardiovascular event.

A biomarker-based risk score to predict hospitalization for heart failure (HHF) among patients with type 2 diabetes (T2D) was able to identify patients at an increased risk and those who may benefit from dapagliflozin therapy. These findings were published in Diabetes Care.

Data were sourced from the Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus–Thrombolysis in Myocardial Infarction 53 (SAVOR-TIMI 53) and Dapagliflozin Effect on CardiovascuLAR Events–Thrombolysis in Myocardial Infarction 58 (DECLARE-TIMI 58) clinical trials. The HHF risk score was formulated using the placebo arm of the SAVOR-TIMI 53 study (n=6106), and the placebo arm of the DECLARE-TIMI 58 study (n=7251) was used as the validation cohort.

The SAVOR and DECLARE cohorts had median baseline high-sensitivity troponin T (hsTnT) levels of 11.9 (interquartile range [IQR], 8.1-18.5) and 10.2 (IQR, 6.9-15.4) ng/L and N-terminal prohormone of B-type natriuretic peptide (NT-proBNP) levels of 140 (IQR, 63-329) and 75 (IQR, 36-168) pg/mL, respectively.

Hospitalization for heart failure occurred among 168 of the SAVOR and 251 of the DECLARE placebo cohorts.

In the multivariate model, the strongest predictors of HHF were NT-proBNP, prior heart failure (HF), and hsTnT (all P <.001), accounting for 95.1% of the variance.

The maximum possible TIMI biomarker score for heart failure among people with T2D in the derivation cohort was 11 points, broken down as follows (using one score from each category):

  • 2 points for prior HF
  • 1 point for hsTnT 6-<10 ng/L
  • 2 points for hsTnT 10-<14 ng/L
  • 3 points for hsTnT ³14 ng/L
  • 2 points for NT-proBNP 50-<125 ng/L
  • points for NT-proBNP 125-<450 ng/L
  • 6 points for NT-proBNP ³450 ng/L

In the validation cohort, individuals who were considered low risk (n=3655; 0-3 points) had a HHF incidence rate of 1.1 per 1000 patient-years (py), intermediate risk (n=2115; 4-6 points) of 7.7 per 1000 py, high risk (n=884; 7-8 points) of 18.5 per 1000 py, and very high risk (n=597; 9-11 points) of 53.1 per 1000 py (P <.001).

Among the full DECLARE trial, dapagliflozin decreased the risk of HHF compared with placebo among the intermediate- (HR, 0.64; 95% CI, 0.43-0.95), high- (HR, 0.63; 95% CI, 0.43-0.94), and very high- (HR, 0.72; 95% CI, 0.54-0.96) risk subgroups, but had no effect among those at low risk vs placebo (HR, 0.98; 95% CI, 0.50-1.92).

Researchers noted several limitations of the study and acknowledged the findings may not be generalizable, as 2 clinical trials from which the study data came from had very strict inclusion and exclusion criteria.

The study authors concluded their biomarker-based risk score was useful in predicting risk for HHF in patients with T2D and could potentially help identify patients at higher risk of HHF who would benefit from sodium-glucose cotransporter-2 (SGLT2) inhibitors such as dapagliflozin, the drug studied in the DECLARE-TIMI 58 trial.

Disclosure: Multiple authors declared affiliations with pharmaceutical companies, including AstraZeneca, which markets dapagliflozin under the brand name Farxiga.  Please refer to the original article for a full list of disclosures.


Berg DD, Wiviott SD, Scirica BM, et al. A biomarker-based score for risk of hospitalization for heart failure in patients with diabetes. Diabetes Care. Published online September 17, 2021. doi:10.2337/dc21-1170. Additional supplemental material at