Differentiating Type 1 From Type 2 Myocardial Infarction Using Growth Differentiation Factor-15

Differentiating type 1 from type 2 myocardial infarction (MI) in emergency department (ED) patients is challenging but key for driving appropriate management decisions. Growth differentiation factor-15 (GDF-15) is a pro-inflammatory cytokine associated with myocardial ischemia. This study aims to determine if GDF-15 alone or in combination with high-sensitivity cardiac troponin (hs-cTnT) can differentiate MI type.

We conducted a secondary analysis of the prospective, multisite STOP-CP U.S. cohort. Adults with chest pain without evidence of STEMI were enrolled from eight U.S. emergency departments (1/2017-9/2018). For this analysis all patients had an adjudicated MI at index and hs-cTnT (Roche, Basel, Switzerland) and GDF-15 measures. Expert adjudicators used the Fourth Universal Definition of MI to determine MI occurrence and type. GDF-15 was compared between patients with type 1 and type 2 MI using a Wilcoxon rank sum test. The area under the receiver operating characteristic curve (AUC) for type 1 MI with 95% confidence intervals (95%CI) was determined for each biomarker alone and in combination. For hs-cTnT, we used the highest measure obtained while in the ED. We determined the optimal cut-point for GDF-15, which we defined as the cut-point that maximizes the type 1 rule-in rate and positive predictive value (PPV). Rule-in rate, PPV, and positive likelihood ratio (+LR) at the optimal GDF-15 cut-point for type 1 MI with 95%CIs were calculated.

Among the 118 STOP-CP patients with an adjudicated diagnosis of index MI, 30.5% (36/118) were female, 60.2% (71/118) were White, and the mean age was 61.4 ± 11.6 years. Among these, 55.9% (66/118) had type 1 and 44.1% (52/118) had type 2 MI. Patients with type 1 MI had higher GDF-15 values than those with type 2 MI (2,710 ± 3,070 ng/L vs 4,400 ± 4,220 ng/L; p=0.007). GDF-15 alone had an AUC of 0.646 (95%CI 0.544-0.748) while hs-cTnT alone had an AUC of 0.480 (95%CI 0.375-0.586) for type 1 MI. In combination, the AUC was 0.664 (95%CI 0.563-0.764). At the optimal GDF-15 cut-point of 1,000 ng/L, 23.3% (95%CI 15.9-32.0%) of patients were ruled-in for type 1 MI with a PPV of 74.1% (95%CI 53.7-88.9%) and a +LR of 2.17 (95%CI 0.99-4.71).

GDF-15 is associated with type 1 MI and demonstrates modest accuracy for differentiating MI type. This suggests that GDF-15 is a promising biomarker, which may play a role in future predictive models for MI type.