Relationship between intramyocardial haemorrhage and left ventricular remodelling in patients with left ventricular systolic dysfunction following myocardial infarction and the effect of empagliflozin

Background

Intramyocardial haemorrhage (IMH) following myocardial infarction (MI) is an independent predictor of adverse left ventricular (LV) remodelling. Patients with reduced LV ejection fraction (LVEF) after MI are considered at high risk of remodelling and therefore likely to benefit from anti-remodelling therapies. However, in contemporary trials many patients do not develop progressive adverse remodelling, and these trials have reported neutral effects of anti-remodelling therapies. The presence of IMH may identify those at highest risk of adverse remodelling post-MI and most likely to benefit from anti-remodelling therapy.

Purpose

We examined remodelling trajectories according to IMH status in patients with LVEF<45% following acute MI. We assessed the effect of empagliflozin on cardiovascular magnetic resonance (CMR) measures of remodelling according to the presence or not of IMH.

Methods

Patients within 12 hours and 14 days of a type 1 MI and an LVEF<45% were randomised to empagliflozin 10mg once daily or placebo. IMH was defined as a region of T2* signal intensity <20ms within the infarct area. The primary outcome was the change in LV end-systolic volume index (LVESVI) from baseline to 24 weeks.

Results

Of 104 randomised patients, 93 had IMH data at baseline and follow-up. 45 (48%) had IMH at baseline. Patients with IMH were more likely to be male, had a higher heart rate and higher peak troponin. Baseline LVEF was lower, high-sensitivity troponin I was higher, and infarct size was larger in patients with IMH. Remodelling patterns in patients with and without IMH were divergent (Figure). In patients with IMH, LVESVI did not change between baseline and 24 weeks (-0.9±11.4mL/m2) whereas in patients without IMH, LVESVI decreased (-14.7±14.7mL/m2; between-group P<0.001). In patients with IMH, LV end-diastolic volume index (LVEDVI) increased (9.1±12.7mL/m2), whereas in patients without IMH LVEDVI decreased (-7.8±16.8mL/m2; P<0.001). LVEF improved in patients with and without IMH but to a lesser degree in those with IMH (7.4±7.1% vs. 10.7±7.7%; P=0.004). There was no difference in the change in left atrial volume index, LV mass index, NT-proBNP, high-sensitivity troponin I or infarct size in patients with and without IMH. Empagliflozin had no effect on the change in LVESVI in patients with and without IMH (P for interaction=0.61).

Conclusion

In a contemporary group of patients with LV systolic dysfunction after an acute MI, the presence of IMH identified those most likely to undergo progressive adverse remodelling. Patients with IMH represent a high-risk subgroup who may benefit from targeted therapeutic interventions. Whilst there was no evidence of benefit from empagliflozin in patients with IMH, these subgroup analyses are exploratory and should be considered hypothesis-generating.

For image description, please refer to the figure legend and surrounding text.