OUP user menu

Takotsubo cardiomyopathy and the long-QT syndrome: an insult to repolarization reserve

Elijah R. Behr, Saagar Mahida
DOI: http://dx.doi.org/10.1093/europace/eup081 697-700 First published online: 7 April 2009

Takotsubo cardiomyopathy (TCM) is a transient syndrome with acute, subacute, and chronic phases, and is commonly induced by emotional or physical stress. It is characterized by chest pain, electrocardiogram (ECG) changes, and minimal myocardial enzymatic release mimicking acute myocardial infarction in patients with unobstructed coronary arteries on angiography. Assessment of left ventricular function in patients with TCM reveals apical ballooning and hypokinesia with the preservation of basal contraction. Electrocardiogram changes include ST-segment elevation, the evolution of marked anterior T-wave inversion, and prolongation of the QT interval. These features all appear to resolve with time.1 Wedekind et al.2 report a case of clinically silent pre-existing long-QT syndrome (LQTS) exacerbated by TCM and resulting in life-threatening arrhythmias. This mirrors our own description last year of a similar case3 and prompts an evaluation of the true significance of QT prolongation in this context.

The association between TCM and QT prolongation has been well documented (Table 2). Seth et al.4 reported 12 cases of TCM with an average QTc interval of 478ms. Similarly, Abe et al.5 described 17 patients, most of whom had a prolonged QTc interval in the acute and subacute phases of the condition. The QT interval normalized in all cases between 97 and 191 days from the onset of symptoms. As part of a report on 13 patients by Desmet et al.,6 each patient's ECG during the acute phase was described with QTc intervals ranging from 310 to 674ms. Koji et al. specifically evaluated the QT interval in 10 consecutive cases of TCM although individual patients' measurements were not detailed. They demonstrated mean maximal QTc intervals in the acute (475 ± 46ms), subacute (555 ± 46ms), and chronic phases (444 ± 41ms) of the condition.7 The subacute phase QTc was significantly prolonged compared with acute and chronic phases. It was, therefore, proposed that this temporary QT prolongation reflects the transient myocardial insult of TCM.7

Despite these consistent findings, ventricular arrhythmias in patients with TCM are relatively uncommon. In a review of seven case series containing a total of 180 cases, Bybee et al.1 reported a 1–1.5% incidence of ventricular arrhythmias. In a more recent series of 14 patients, Bonello et al.8 reported two cases of malignant ventricular arrhythmias (one patient with ventricular tachycardia and the other with ventricular fibrillation). A number of individual case reports have also documented the incidence of ventricular tachycardia and ventricular fibrillation in patients with TCM. None of these described, however, the presence of torsades de pointes (TdP).

Torsades de pointes and QTc prolongation as a predictor of risk

More recent case reports have emerged that document TdP in patients with TCM. Table 1 describes the 11 reported cases currently published in the English language. The mean age of these individuals at presentation was 63 years (range 22–87), and the majority were female [9/11 (82%)]. This is similar to the several series described in Bybee et al.'s review whose mean age ranged from 62 to 75 years and whose female proportion ranged from 82 to 100%.1 In TdP cases, in whom the data were available, the mean QTc at presentation was 595ms, the mean maximal QTc recorded was 706ms, and the mean post-TCM QTc was 481ms (Table 2). Comparing paired data where available (paired Student's t-test), the maximal QTc was significantly greater than the presenting QTc (P = 0.0075) and the post-TCM QTc (P = 0.0003), in keeping with Koji's findings of variation of QT interval between acute, subacute, and chronic phases.7 None of these individuals were on QT prolonging drugs as far as can be ascertained. Comparison to the series where TdP was not identified (Table 2) suggests that the QT prolongation described in the TdP cases is greater. Desmet et al.'s6 series reports individuals' acute phase QTc intervals and allows a direct comparison to be made (Table 2). The presenting and maximal QTc intervals in the TdP cases are significantly longer suggesting this as an association with, if not, a potential predictor for the development of TdP. Indeed if a cut-off of acute phase QTc >500ms is utilized then 9 of 11 TdP cases would have been predicted to be at risk (sensitivity 82%) and 11 of 13 non-TdP cases from Desmet's series would have been predicted not to be at risk (specificity 85%).

View this table:
Table 1

Case reports of Takotsubo cardiomyopathy and torsades de pointes

ReferenceAgeM/FPMHFH SDFH LQTSPrior syncopeQT prolonging medicationsPresenting QTc (ms)Max QTc (ms)Post-QTc (ms) (after X days)OutcomeGenetic testing
Wedekind et al.281FPrior QTc prolongation (520 ms), pacemaker, and hypertensionYesYesNoNo690690NSICD implantedNo
Mahida et al.355FDiabetes and hyperlipidaemiaNoNoYesNo510735490 (56)ICD declinedYes
Kurisu et al.987FNoneNoNoYesNS640a880440 (6)Pacemaker implantedNo
Kurisu et al.978MNSNoNoNoNS640a920600 (7)Pacemaker implantedNo
Ghosh et al.1059FAnxiety, hypertension, and alcohol excessNoNoNoNSNS669Normal (NS)DischargedNo
Furushima et al.1161FNSNoNoNoNS740740470 (168)ICD implantedNo
Sasaki et al.1222FNSNoNoNoNo730730520 (22)MexilitineNo
Denney et al.1332MObesity, migraine, asthma, and rhinitisYesNoNoNob416467406 (4)β-BlockerNo
Mofrad et al.1470FNoneNoNoNoNo630c727440 (NS)ICD implantedNo
Nault et al.1576MNoneNoNoNoNo630786Normal (2)DischargedNo
Finstere et al.1675FMitochondrial disorderNoNoNoNS326d425dNSβ-BlockerNo
Death prior to discharge
  • NS, not specified; PMH, previous medical history; FH, family history; TCM, Takotsubo cardiomyopathy; LQTS, long-QT syndrome; SD, sudden death.

  • aIn the presence of high-degree AV Block.

  • bPatient receiving pseudoephedrine.

  • cIn the presence of hypokalaemia.

  • dCalculated from ECGs presented in the report.

View this table:
Table 2

Comparison of QTc intervals in non-torsades de pointes series and the series of torsades de pointes reports

ReferenceNumber of casesMean reported QTc (ms) (95% Confidence interval)Median reported QTc (ms)Range (ms)
Seth et al.412478 (±96)NSNS
Abe et al.517NS500436–581
Desmet et al.613450*,†436310–674
Koji et al.710555 (±46)NSNS
TdP reports presenting QTc (Table 1)10595*635326–740
TdP reports maximal QTc (Table 1)11706730425–920
  • TdP, torsades de pointes; NS, not specified.

  • *P = 0.01 (two-tailed Student's t-test).

  • P = 0.00014 (two-tailed Student's t-test).

Takotsubo cardiomyopathy: another form of acquired long-QT syndrome

The presence of significant QT prolongation at other times than during the TCM episode in at least 4 of 11 TdP cases (36%) also suggests that there is an underlying predisposition towards repolarization abnormality: a reduced repolarization reserve.17 This term was first coined to describe the impact of QT prolonging drugs on predisposed individuals who then developed TdP. It is, therefore, reasonable to regard TCM as a similar transient acquired insult upon myocardial repolarization that may relate to an acute disturbance of cardiac autonomic function18 and precipitates TdP in more vulnerable individuals. In common with drug-induced LQTS, these individuals may carry clinically silent or unexpressed mutations implicated in the congenital LQTS or may harbour multiple common population variants that impair the repolarization reserve in more subtle ways.17 They may also have other predisposing abnormalities such as structural cardiac disease, metabolic disease, or bradycardia.17 There has been little genetic study in TCM at present, but systematic genetic research of TdP cases compared with non-TdP TCM is required to determine whether this phenomenon is driven by genotype and if so whether rare mutations or common variants are more important. This leaves open the future possibility of genetic testing assisting in determining the risk of TdP during acquired insults such as TCM.


In the acute and subacute phases, it therefore appears reasonable to monitor the QT interval closely in individuals with possible TCM and if the QTc prolongs, particularly if it is >500ms, measures should be taken to monitor cardiac rhythm closely and prevent or treat TdP appropriately. As two of the cases were associated with bradycardia due to heart block, a known precipitant of the acquired LQTS,17 temporary transvenous pacing should be considered as a therapeutic option.9

These cases also provide a useful indication to long-term management of these uncommon presentations of TCM. Given its transient nature and the female predominance, unless there are additional features suggestive of high risk LQTS (for example, a post-TCM QTc > 500ms, previous syncope, or previous cardiac arrest) then it is reasonable to advise treatment with a β-blocker alone. Otherwise an ICD may offer more appropriate long-term management. In the two cases of TdP and TCM that were complicated by high-degree AV block, the patients in question appeared to do well with permanent pacing alone after some improvement in their QT interval.


These reports suggest that severe prolongation of the QT interval in TCM may be a marker for the risk of sudden death and should be monitored for in all suspicious cases. This may reflect an underlying abnormality of repolarization that may be genetic in basis and carry a risk of sudden death. Further research is required to study this hypothesis.

Conflict of interest: none declared.


  • The opinions expressed in this article are not necessarily those of the Editors of Europace or of the European Society of Cardiology.