Introduction: Growth differentiation factor (GDF)-15 has been associated with cardiovascular (CV) outcomes in ACS using research use only assays. We evaluated the prognostic performance of GDF-15 measured with the 1st clinically available assay (Europe).
Hypothesis: GDF-15 levels are associated w/ a gradient of risk for CV death/HF in NSTE-ACS, and this association is consistent across the spectrum of risk according to TIMI risk score (RS).
Methods: We measured baseline GDF-15 (Roche Diagnostics) in 4,330 pts within 48h of NSTE-ACS enrolled in MERLIN-TIMI 36 (Randomized trial of ranolazine). Pts were categorized by a priori thresholds of GDF-15 (<1200, 1200-<1800, ≥1800 pg/mL) & stratified by TIMI RS (0-2, 3-4, ≥ 5). Adj-HRs were adjusted for clinical variables as well as eGFR, NT-proBNP, hs-TnT, and hs-CRP.
Results: There were 2286 (53%), 1104 (25%), and 940 (22%) pts w/ GDF-15 < 1200, 1200-<1800, and ≥1800 respectively. Pts with higher GDF-15 were more likely to be older, female, and have comorbidities. Increasing GDF-15 was associated with a significant graded increase in risk of CV death/HF (HR 1200-<1800: 2.68 [1.93 – 3.72], HR ≥ 1800: 6.29 [4.69 – 8.42], p<0.001 for both, Fig A). Adj-HRs were 1.50 (1.06 – 2.13) and 1.88 (1.30 – 2.72) respectively. When stratified by TIMI RS, pts with a low TIMI RS and high GDF-15 level had a similar event rate to pts with a high TIMI RS and low GDF-15 level (Fig B). Notably, there was an interaction (p=0.004) between TIMI RS and GDF-15, with GDF-15 identifying the greatest incremental risk in those who otherwise appeared to be at lowest risk based on the TIMI RS alone. This risk persisted after adjustment for established biomarkers.
Conclusion: In the 1st evaluation of a clinically available assay, GDF-15 applied using established cut-points strongly predicted risk of CV death/HF, particularly among pts who were seemingly low risk using standard clinical criteria, with potential implications for risk-based therapeutic decision-making.