ACCURATE outcomes have been better predicted in ST-segment elevation myocardial infarction (STEMI) patients having undergone percutaneous coronary intervention (PCI) using a novel echocardiographic risk stratification model. These recently published results could lead to massively streamlined and improved efficiency of current risk stratification pipelines in the clinic.
Three-hundred and seventy-three patients who had undergone PCI after exhibiting STEMI were enrolled in this single-centre prospective study (derivation cohort), all of whom were given echocardiograms a median of 2 days later. A validation cohort of 298 additional patients from a clinical registry over a set period was used as a control, with primary outcome being a composite of cardiovascular mortality and heart failure.
A mean left ventricle ejection fraction of 46% was found in the derivation cohort, with 75% of patients exhibiting prevalent diastolic dysfunction. In the derivation cohort, 80 patients (21.4%) experienced the primary endpoint, comprising 70 cases of heart failure, 13 cardiovascular deaths, and three patients who had heart failure and subsequently died (median follow-up: 5.4 years). This was compared to 127 patients (42.65) in the validation cohort (median follow-up: 3.5 years).
Three key variables for risk stratification were identified using a classification and regression tree analysis: wall motion score index (WMSI), global longitudinal strain rate e (GLSRe), and estimates of left ventricle filling pressure (E/e’). Patients were categorised into three groups based on heart failure and/or cardiovascular death: high risk (WMSI ≥2.22), intermediate risk (WMSI <2.22; E/e’ ≥7.6; and GLSRe <0.82s), and low risk (WMSI <2.22 and GLSRe ≥0.82s or E/e’ <7.6 and GLSRe <0.82s). Incremental risk was detected when comparing patients in the low- versus intermediate-risk group (hazard ratio: 2.52; 95% confidence interval: 1.24–5.11; p=0.011) and versus the high-risk group (hazard ratio: 4.37; 95% confidence interval: 1.40–13.66; p=0.011).
“The model suggests that by only considering three measures (WMSI, GLSRe and E/e’) clinicians can risk-stratify patients in a simple and efficient manner without having to consider a large variety of other proposed predictors of outcome,” noted the authors.
