Speckle tracking-vs conventional echocardiography for the detection of myocardial injury-A study on patients with subarachnoid haemorrhage.

Abstract

Myocardial injury with regional wall motion abnormalities (RWMA) is common in subarachnoid haemorrhage (SAH). We hypothesized that the diagnostic performance of left ventricular (LV) global and regional longitudinal strain (GLS and RLS, respectively), assessed with speckle tracking echocardiography is superior to standard echocardiography for the detection of myocardial injury in SAH. Seventy-one unselected patients with verified SAH were included. Echocardiography was performed within 48 hours after admission. hsTnT was followed daily up to 3 days post-admission. RWMA, LV ejection fraction (LVEF), GLS and RLS were analysed by two experienced echocardiographists, blinded to the information on plasma hsTnT. A reduced GLS was defined as >-15%. Two cut-off levels were used for the definition of RLS, ie when segmental strain was >-15% (liberal) or >-11% (conservative) in ≥2 adjacent segments. Myocardial injury was defined as a peak hsTnT ≥90 ng/L. The incidence of myocardial injury was 25%. The hsTnT (median, 25% and 75% percentile) in patients with (a) reduced LV ejection fraction (LVEF <50%, n = 10) was 502 (175-718), (b) RWMA (n = 12) was 648 (337-750), (c) reduced GLS (n = 12) was 502 (132-750) and (d) reduced RLS (n = 42) was 40 (10-216), respectively. The specificity/sensitivity for LVEF, RWMA, GLS and RLS to detect myocardial injury 98%/50%, 100%/67%, 96%/56% and 54%/94%, respectively. The intra- and inter-observer variability for assessment of RLS was high. The diagnostic performance of GLS by strain imaging is not superior to standard echocardiography for the detection of myocardial injury in SAH. RLS could not reliably detect regional myocardial injury.