Normal apical myocardial perfusion in the rat model with Takotsubo syndrome: is subsequent microvascular dysfunction and hypoperfusion an epiphenomenon? Reply

We greatly appreciate the comments offered by Dr Madias in response to our original article ‘Myocardial contrast imaging reveals apparently normal coronary perfusion in stress-induced (takotsubo) cardiomyopathy’.1 Dr Madias, in his letter, poses several important questions that would be worthwhile to address in our model. The first point raised is the presence of tissue oedema in affected myocardial segments in Takotsubo patients.2 The rat model is well suited to study the temporal changes in regional myocardial water content (i.e. oedema). We are currently investigating myocardial water content post-mortem in tissue samples harvested at different time-points after injection of isoproterenol. Our preliminary post-mortem data from tissue harvested 2h after isoproterenol show no significant difference in myocardial water content in apical tissue compared with basal tissue (Figure 1, n = 5, P = 0.5). Although one probably should not be too quick to draw conclusions based on only a few data points, these results indicate that myocardial oedema does not precede development of left ventricular akinesia. However, the best method for the in vivo evaluation of dynamic changes in regional myocardial tissue water is magnetic resonance imaging and we are currently working with this methodology in the rat model. In addition to imaging of tissue oedema, MR techniques can be used to evaluate fluxes in myocardial energy metabolites and may provide important information about myocardial energy metabolism in Takotsubo.3 Myocardial tissue water content. Hearts were excised from five rats that had received 50 mg/kg isoprenaline and that showed typical left ventricular apical ballooning on echocardiography. The hearts were excised as soon as the echocardiographic examination was finished, i.e. 2 h post-isoprenaline, and were immediately frozen in liquid nitrogen. The echocardiograms were carefully studied to identify which areas were clearly akinetic (apex) and hyperkinetic (base), respectively. Tissue samples were then carefully cut from the mid-myocardial portion of each section, excluding the endocardial and epicardial surfaces. Each sample was weighed while still deep frozen and subsequently dried for 24 h at 38° before being weighed again. Dry and wet weights were recorded for each tissue sample and percentage of tissue water content was calculated as: % tissue water = (Wet weight – Dry weight)/Wet weight. The second point relates to electrocardiographic characteristics in this Takotsubo-like animal model. Using standardized limb lead ECG, we have detected apparent ST-T segment deviation, including ST-segment elevation.4 Understanding the mechanisms that underlie these electrocardiographic appearances would be of particular interest. It remains to be established which part, if any, electrophysiological perturbations play in Takotsubo. In other words, whether action potential transmission and ion fluxes are impaired in dysfunctional regions or if loss of function can be explained entirely by alterations in the contractile apparatus (Figure 2). We will be able to come back with a more detailed description of these electrophysiological perturbations and their temporal pattern in the near future. We have previously shown that excess isoprenaline impairs electrophysiological function in isolated cells.5 At present, we are focusing our attention on ex vivo electrophysiological studies of tissue harvested from the rat model, with particular emphasis on depolarization/repolarization patterns and regional differences in action potential duration. Pathomechanisms in Takotsubo. It is our intention to address the issues discussed above as well as to study myocardial perfusion and tissue histology, including morphology of the microvasculature, at several additional time points, including at the very least daily assessment until time of recovery of cardiac function.4 Lastly, we would like to clarify that although we did not observe any apical perfusion defect and no microvascular damage preceded development of cardiac dysfunction in our model, we do not believe that these findings necessarily imply that perfusion defects and/or microvascular dysfunction described in some patients are merely epiphenomena. Instead impaired perfusion in the subacute phase of the disease may instead stem from microvascular obstruction secondary to tissue swelling (i.e. oedema). Alternatively, the akinetic myocardium within the affected areas may be metabolically less active and autoregulatory factors may redistribute blood flow towards unaffected hypercontracting regions with a higher metabolic demand. The points proposed by Dr Madias and discussed above would help to address this issue. Conflict of interest: none declared.