Effects of hyperoxia on systolic and diastolic function assessed by strain analysis in patients with coronary artery disease during general anaesthesia

Fig. 1. Assessment of left ventricular diastolic function in a sample patient. Traditional assessment uses the ratio of peak early diastolic trans-mitral inflow (E, middle panel) to peak early diastolic annular mitral tissue velocity assessed by tissue Doppler imaging (TDI, e’, left panel) and results in a global value for diastolic function (E/e’). A higher E/e’ represents a worsening of diastolic function. Diastolic strain curves show the same features as TDI assessment. Replacing TDI e’ by segmental early diastolic peak strain rate derived from myocardial deformation analysis (eDSR, right panel) allows for global, but also regional quantification of diastolic function, when assessing the diastolic strain curve of each individual myocardial segment (AHA segments 7–12 of the midventricular short axis plane depicted in different colours). eDSR, early diastolic strain rate.

Fig. 2. Global circumferential strain (LV GCS) traces during normoxaemia and hyperoxia. Displayed are traces of global left ventricular circumferential strain (LV GCS) in a sample patient both during normoxaemic (blue) and hyperoxic (red) FiO2 showing circumferential shortening over time. Systolic peak strain (dots) is attenuated (less negative peak strain) under hyperoxia. More negative values denote better systolic function. GCS, global circumferential strain; GLS, global longitudinal strain.

Fig. 3. Relationship between baseline myocardial function and strain changes during hyperoxia: Change in GCS and GLS as a function of baseline (normoxaemia) strain pooled across the two cross-over groups (denoted by the shape). A positive change in the GCS and GLS rate, respectively, denote a worsening systolic function, whereas a negative change denotes an improvement. The solid line and associated grey bands mark the unadjusted linear association as computed by a linear regression. The purple shaded area represent changes in strain rates that are deemed clinically irrelevant (see Methods). GCS, global circumferential strain; GLS, global longitudinal strain.

Fig. 4. Regional diastolic strain rate changes under hyperoxia in the mid-ventricular short-axis view. The six segments of a LV midventricular short axis image according to the American Heart Association 16-Segment Model are shown. In the upper panel all segments show similar segmental strain rates during normoxaemia. During hyperoxia (lower panel) the segmental analysis shows a decoherence with slowing predominant in the early diastolic strain rate (eDSR) peak: Only the segments subtended to the left anterior descending artery (LAD) (segments 7 and 8) remain normal, while the other segments show a depression in eDSR, i.e., segments 11 (inferolateral wall) and 12 (anterolateral wall) subtended to the left circumflex artery (LCx), as well as segments 9 (inferoseptal wall) and 10 (inferior wall) downstream of the right coronary artery (RCA). This results in an increase in E/eDSR that nearly doubles demonstrated in segment 12 (red). This reflects significant depression of regional diastolic function in this patient with three-vessel disease (visual assessment in preceding coronary angiography: left main stenosis 70% to 89%, mid LAD 70% to 89%, proximal LCx 90% to 99%, proximal and mid RCA…