OUP user menu

Ventricular repolarization gradients in a patient with takotsubo cardiomyopathy

Hiroshi Furushima, Masaomi Chinushi, Akiko Sanada, Yoshifusa Aizawa
DOI: http://dx.doi.org/10.1093/europace/eun166 1112-1115 First published online: 19 June 2008

Abstract

A 61-year-old woman had recurrent syncopal attacks caused by torsades de pointes associated with remarkable QT prolongation (QTc = 740 ms). Left ventriculography showed apical akinesis (ballooning) and basal hyperkinesis, but coronary angiography was normal. This was compatible with takotsubo cardiomyopathy. The wall motion of the left ventricle (LV) normalized within 2 months, and the remarkable QT prolongation and negative T-waves gradually normalized. However, polymorphic ventricular tachycardia recurred at 2.5 months after its initial onset, and we measured repolarization gradients using activation recovery intervals (ARIs) in an electrophysiological study. During atrial pacing at a cycle length of 1000 ms, the negative T-waves were observed in leads II, III, aVF, and V2–6 with QT prolongation, and the ARIs in both the epicardium and the endocardium increased from the basal site to the apical site. Moreover, the ARI tended to be longer in the epicardium than the endocardium at each level of the LV. In contrast, atrial extrastimulation changed the T-wave morphology (from negative to biphasic) in leads II, III, aVF, and V2–6 and changed the ARI gradients both from the LV basal site to the apical site and from the epicardium to the endocardium. These results suggest that the T-wave abnormalities seen in takotsubo cardiomyopathy during sinus rhythm are due to abnormal LV repolarization gradients.

Takotsubo cardiomyopathy is a cardiac disease characterized by transient left ventricular (LV) dysfunction in the absence of obstructive coronary artery disease1 and often has the prolongation of the QT interval with a deep negative T-wave. The electrophysiological characteristics of the QT prolongation with a negative T-wave have not well been studied.

Case report

A 61-year-old woman was admitted before 2 months because of a syncopal attack. She had no history of head trauma of syncopal episodes before this episode, and brain computed tomography was normal. An electrocardiogram (ECG) displayed deep and diffuse T-wave inversion with the prolongation of the QTc interval to 740 ms (Figure 1A). One day after admission, she had syncopal attacks and ECG monitoring showed torsades de pointes (TdP) (Figure 1D). Overdrive ventricular pacing at 90 b.p.m and mexiletine administration were effective in preventing premature ventricular contraction (PVCs) and TdP. Echocardiography showed akinesis of the LV apex, but did not show dilatation and abnormal wall motion of the right ventricle. Late potential on signal-averaged ECG was negative. At cardiac catheterization, coronary angiography was normal; however, left ventriculography showed akinesis of the LV apex with systolic ballooning, with normal contraction of other LV regions (Figure 2). The LV wall motion and QTc interval gradually improved (QTc = 590 ms) (Figure 1B) within 3 weeks.

Figure 1

(A) Electrocardiogram on admission to the first hospital shows diffuse negative T-waves with prolongation of QTc interval to 740 ms (HR = 62 b.p.m.). (B) Electrocardiogram on admission to our hospital 2 months after onset. The QT interval and negative T-wave gradually improve when compared with those in (A) (HR = 48 b.p.m., QTc = 590 ms). (C) Electrocardiogram 6 months after onset. The QT interval and negative T-wave remarkably improved (HR = 58 b.p.m., QTc = 470 ms). (D) Electrocardiogram monitoring on the day after admission to the first hospital. Torsades de pointes occurs in association with remarkable QT prolongation.

Figure 2

Left ventriculography in a right anterior oblique view in diastole (left panel) and systole (right panel). There is apical ballooning during systole, and the ejection fraction of the left ventricle was 29%.

In an electrophysiological study (EPS), intra-cardiac repolarization gradients were measured between the apical and basal sites in both the epicardium and the endocardium of the LV. The LV epicardial recordings were made using a six-electrode catheter, which consist of three electrode pairs. The catheter was introduced into the coronary sinus and was advanced down the great cardiac vein towards the apical site. The LV endocardial recordings were made using a four-electrode catheter. It was positioned along the endocardial surface of the anteroseptal segment of the LV in close transmural proximity to the epicardial catheter. To estimate local refractoriness, the activation recovery interval (ARI), which was defined as the interval between the time of the minimal first derivative of the intrinsic deflection of the QRS complex and the maximal first derivative of the T-wave on the unipolar electrogram,2 was calculated.

Figures 3 and 4 show the surface ECG and ARIs measured in six epicardial electrodes (Epi 1, apical site; Epi 6, basal site) and four endocardial electrodes (Endo 1, apical site; Endo 4, basal site), with atrial extrastimuli of 620 and 580 ms at a basic paced cycle length of 1000 ms. At a basic cycle length, the ARIs in both the epicardium and the endocardium increased from the basal site to the apical site. The ARIs were usually longer in the epicardium than in the endocardium at the same level of the LV. At an atrial extrastimulus of 620 ms, the difference in ARIs between the basal and apical sites became smaller and T-waves became less negative in leads II, III, aVF, and V2–6 (Figure 3). When the coupling interval of the atrial extrastimulus was shortened to 580 ms, the ARI gradients were changed when compared with the basic cycle length, that is, the ARIs in both the epicardium and the endocardium increased from the apical site to the basal site (Figure 4A), and the ARIs were almost identical in the epicardium and endocardium (e.g. Epi 1 vs. Endo 3 or Endo 4 in Figure 4A). Furthermore, T-waves changed from negative to biphasic in leads II, III, aVF, and V2–6 (Figure 4B).

Figure 3

Simultaneous recording of the surface electrocardiogram in III, V3, and V5, and unipolar electrograms (A) and 12-lead electrocardiogram (B) during atrial extrastimulation of 620 ms at a basic cycle length of 1000 ms. Endo 1 to Endo 4 (5 mm electrode distance) and Epi 1 to Epi 6 (2 mm electrode distance between Epi 1 and Epi 2, Epi 3 and Epi 4, and Epi 5 and Epi 6, respectively, and 5 mm electrode distance between Epi 2 and Epi 3 and Epi 4 and Epi 5, respectively) means electrode number from the apical site to the basal site.

Figure 4

Simultaneous recording of the surface electrocardiogram in leads III, V3, and V5, and unipolar electrograms (A) and 12-lead electrocardiogram (B) during atrial extrastimulation of 580 ms at a basic cycle length of 1000 ms. Endo 1 to Endo 4 and Epi 1 to Epi 6 are the same as in Figure 3.

Activation recovery interval restitution curves were resolved using a single exponential decay function. Diastolic interval (DI) was defined as the interval between the recovery time from the last basic beat and that of extrastimulation. Data were fitted to the following equation: ARI (t) =ARImax − ΔARI × e−(t/τ), where ARImax is the ARI during the plateau of restitution, ARI (t) the ARI of the DI preceding extrastimulation, and ΔARI and τ the amplitude and time constant, respectively. Restitution kinetics of the ARIs revealed that ARImax and ΔARI were greater in Epi than in Endo at the same level of LV. ARImax and ΔARI were greater at the apical site than at the basal site in either Epi or Endo.

Figure 1C shows an ECG at 6 months after the onset of cardiomyopathy. The QT interval remarkably improved, and the T-wave became biphasic in leads II, III, aVF, and V2–6.

The patient was treated with a β-blocker, and an implantable cardioverter defibrillator (ICD) was implanted, but she did not receive an ICD shock.

Discussion

Transient LV apical ballooning syndrome, also known as takotsubo cardiomyopathy, is a recently described novel acute cardiac syndrome.1 In takotsubo cardiomyopathy, QT prolongation is often associated with deep negative T-waves due to unknown mechanism. In the present case, we demonstrated that atrial premature stimulation simultaneously altered the T-wave morphology (from negative to biphasic) in several surface ECG leads and changed the ARI gradients both from the LV basal site to the apical site and from the epicardium to the endocardium. Several previous studies showed that action potential duration is longer in the endocardium than in the epicardium in normal animal hearts.3,4 Azarov et al.5 recently showed that the ARI is longer in the LV base than in the apex of rabbit heart. Laurita et al.6 previously demonstrated that a reversal in the epicardial repolarization gradient of guinea pig hearts was reflected on the surface ECG by a change in the polarity of the T-wave. Although LV repolarization gradients in humans have not been well studied, the results of this case were different from several animal studies using a normal heart. This might suggest that T-wave abnormalities in takotsubo cardiomyopathy are related to abnormal ventricular repolarization gradients. The change in the T-wave morphology (from negative to biphasic) in leads II, III, aVF, and V3–6 on the surface ECG during atrial premature stimulation might be affected by a change in the distribution of the whole ventricular repolarization. To elucidate the mechanism of the change of T-wave polarity, higher resolution mapping is required, but such mapping cannot usually be attempted in human EPS.

In restitution curves of ARIs, the results seemed to be compatible with the change of ARI at atrial extrastimulation, demonstrated in Figures 3 and 4.

Conflict of interest: none declared.

References

View Abstract