Right Ventricular Abnormalities in TakotsuboCardiomyopathy

Ana Clara Rodrigues, M.D., Laise Guimaraes, M.D., Edgar Lira, M.D., Wercules Oliveira, M.D.,Claudia Monaco, M.D., Adriana Cordovil, M.D., Claudio H. Fischer, M.D., Marcelo Vieira, M.D., andSamira Morhy, M.D.

Echocardiography Laboratory MDP, Hospital Israelita Albert Einstein, Sao Paulo, Brazil

Background: Takotsubo cardiomyopathy, described as transient regional contractile abnormalities lim-ited to the apical and mid-segments of the left ventricle (LV), has also been reported to involve basaland/or mid LV segments (inverted Takotsubo); fewer reports, however, have addressed right ventricular(RV) dysfunction. Aim: To assess the distribution of regional abnormalities and RV involvement in Takot-subo cardiomyopathy and compare it to the literature. Methods and Results: We evaluated 23 patientswith both classical and inverted presentations (19 female, aged 64 19 years), including 2 recur-rences, totaling 25 episodes. Classical Takotsubo was observed in 15 patients, while 10 had the invertedform. LV ejection fraction (EF) was lower for classical compared to inverted presentation (30 7 vs.45 4%, P < 0.001) with higher troponin values (1.3 1.4 vs. 0.5 0.6, P = 0.034). RV abnormali-ties were found in 7 patients (28%), mainly with classical presentation (6 patients), presenting with midand apical RV impairment. One patient with inverted Takotsubo had mid-RV involvement. Patients withRV involvement had lower left ventricular ejection fraction (LVEF) (28 10% vs. 40 10%, P = 0.02),but not when adjusted for presentation type. Overall rate of complications was higher for classical com-pared to inverted presentation, and not inuenced by RV involvement. Conclusion: RV contractileabnormalities may follow the same LV regional distribution in Takotsubo cardiomyopathy; the type ofpresentation rather than the presence of RV dysfunction seems to be responsible for an increased risk ofcomplications and severity of functional impairment. (Echocardiography 2013;30:1015-1021)

Key words: Takotsubo, echocardiography, right ventricle

Takotsubo cardiomyopathy (stress-inducedcardiomyopathy) was rst described as transientregional contractile abnormalities involving theapical and mid-segments of the left ventricle (LV)with hypercontractile basal segments, usuallyrelated to triggers such as emotional or physicalstress.1 Classically, the regional abnormalities donot correspond to any coronary artery territory,and no signicant coronary artery stenosis isfound. In spite of the amount of myocardiumaffected, the prognosis is usually favorable, withrapid spontaneous recovery. An inverted patternof stress-induced transient regional dysfunctionhas also been reported, with contractile abnor-malities limited to the basal and/or mid segmentsof the LV, and a normally contracting apex.24

The association of right ventricular (RV) dysfunc-tion in patients with Takotsubo59 has beenreported, with emphasis on poorer outcome inthis situation.7 However, studies of Takotsubo

cardiomyopathy focusing specically on RV con-tractile abnormalities are less frequent.6,7 To givefurther insights into demographics and clinicaland laboratory features of RV involvement inTakotsubo cardiomyopathy we assessed a caseseries of patients and compared it with ndingsfrom the available literature.

Methods:From July 2006 up to August 2011, all patientswho fullled the diagnosis of Takotsubo cardio-myopathy were enrolled in the study. Criteria forstress-induced cardiomyopathy were the follow-ing: (1) acute chest pain/discomfort or dyspneawith elevation of cardiac enzymes and ECGchanges (ST and T changes) and; (2) transient LVcontractile abnormalities not related to a singlecoronary artery; (3) absence of signicant(>50%) luminal narrowing by coronary arteriog-raphy or normal coronary tomography within2 days of the onset of the symptoms. Patientswere classied into 2 presentations: Classical Takotsubo: akinesia/hipokinesia of

mid and apical LV segments

Address for correspondence and reprint requests: Ana ClaraRodrigues, M.D., Echocardiography Laboratory MDP, Hospi-tal Israelita Albert Einstein, Av Albert Einstein, 627 CEP05652900, Sao Paulo- SP Brazil. Fax: 55 11 21515974;E-mail: claratude@yahoo.com

1015

2013, Wiley Periodicals, Inc.DOI: 10.1111/echo.12215 Echocardiography

Inverted Takotsubo: contractile abnormali-ties limited to the basal and/or mid seg-ments of the LV, and no apicalabnormalities

All patients had echocardiograms undertakenup to 12 hours after the onset of symptoms. Theexaminations were repeated at least once untilLV functional recovery. Left ventricular ejectionfraction was obtained by Simpsons modied ruleor visual estimation. Images were reviewed by 2experienced echocardiographers to assess RVinvolvement; at least 2 echocardiographic planeswith clear visualization of the RV were necessaryfor analysis (apical four-chamber and subcostalplane). The RV was divided into basal, midand apical segments. Patients with incompleterecordings or suboptimal images were excludedfrom the study. Clinical characteristics (age, gen-der, onset of symptoms) and the condition thatpreceded onset as a possible triggering factorwere analyzed. Cardiac enzyme release (creatinekinase MB [CKMB] and troponin T) was checkedfor determination of peak values, and a 12-leadECG (Page Writer Trim III, Philips MedicalSystems, Andover, MA, USA) was examined forall patients. All echocardiographic examinationswere reviewed up to normalization of ventricularfunction. Clinical (arrhythmias, pulmonaryedema, cardiogenic shock) and echocardio-graphic (LV thrombus, pericardial effusion) com-plications, the use of intravenous catecholamineor assist devices and in-hospital death wereexamined by review of medical records.

In addition, we searched PubMed database toidentify original English language articles in peer-reviewed journals with the terms Takotsubo,apical ballooning, stress cardiomyopathy, andampulla cardiomyopathy. Search was limited tohumans and adults and to studies including morethan 5 consecutive patients in a case series.We selected studies that assessed RV contractileabnormalities in Takotsubo patients, includingdata on demographic characteristics, symptoms,clinical and laboratory presentation, RV charac-teristics, complications, and outcome. A formalmeta-analysis was not used because of the heter-ogeneity of the methodology and outcomeassessment among the studies, therefore anarrative synthesis of the collected data wereundertaken.

Statistical Analysis:Data are expressed as mean SD. The differencesin continuous variables comparing cardiacfunction in the acute and subacute periods weretested using paired Student t-test. MannWhitneytest was used to test the differences between the

groups. Categorical measures were comparedwith chi-square test or Fishers exact test. AP-value 5 ng/mL) wasobserved in 15 of the 25 episodes (60%) and ele-vated troponin T (>0.03 ng/mL) was seen in allbut 1 patient (mean value = 0.94 1.2 ng/mL).Coronary angiography (Allura Xper FD20,Philips Healthcare, Bothell, WA, USA) was nor-mal except for 2 patients (40% lesion). Threepatients undertook coronary computed tomog-raphy (2 female under 32 years and a 57-year-old male after liver transplantation), withnormal ndings. Regression of wall-motion abnor-malities took 12 6 days (ranging from 3 to43 days), with minor segmental abnormalitiespersisting in 3 patients. The LV ejection fractionimproved from 34 13% at baseline to 63 6%(P < 0.00001). Pulmonary artery systolic pressurewas increased (>35 mmHg) in 14 patients (56%),and measured 40 9 mmHg. Transient left

1016

Rodrigues, et al.

intraventricular pressure gradient was docu-mented in only 1 patient.

Classical Takotsubo presentation, with akine-sia of mid and apical segments of the LV, wasobserved in 15 patients, while 10 patients hadthe inverted form of Takotsubo, including the 2patients with recurrence: 4 patients had akinesiaof both basal and mid- LV segments; 3 had akine-sia of all mid segments; and 3 had akinesia of allbasal segments. At the onset of the disease, LVejection fraction was signicantly lower (30 7vs. 45 4%, P < 0.001) for classical comparedto the inverted Takotsubo presentation, andthese patients also expressed higher troponin val-ues. Pleural effusion was common (56% of thepatients, occasionally existing prior to the onsetof Takotsubo (3 patients), and more frequent inpatients with classical presentation (P = 0.04).Demographics and cardiac markers of classical

and inverted Takotsubo patients are displayed onTable I.

Right Ventricular Involvement:Results from the present series regarding RVinvolvement are displayed in Table II. In thisstudy, most patients with RV dysfunction werefemale. RV abnormalities were present in 7patients, and predominantly associated with clas-sical Takotsubo presentation (6 patients). Apicaland mid-RV akinesia was seen in 5 patients andakinesia restricted to the RV apex (Fig. 1)observed in 2 patients. Only 1 patient with theinverted presentation of Takotsubo had RVinvolvement, and likewise, the affected segment(mid ventricular) was analogous to that observedfor the LV (Fig. 2). A lower left ventricular ejec-tion fraction (LVEF) was found for patients withRV dysfunction compared to those without it

TABLE I

Characteristics of Classical Compared to Inverted TakotsuboPresentation

VariableClassical(n = 15)

Inverted(n = 10) P

Age (years) 63 11 61 5 NSFemale (n) 11 (73%) 9 (90%) NSBody surfacearea (m2)

1.66 0.1 1.65 0.1 NS

Troponin T(ng/mL)

1.3 1.4 0.5 0.6 0.03

CKMB (ng/mL) 5.9 4.2 4.5 2.9 NSLeft atrium (mm) 38 4 34 5 0.056PASP (mmHg) 43 6 36 7 NSInitial LVEF (%) 30 7 45 4

(28 10% vs. 40 10%, P = 0.02). However,when controlled for the type of presentation(patients with classical Takotsubo), there was nosignicant difference in LVEF, but a trend forlower LVEF (P = 0.08) in patients with RV involve-ment. Cardiac injury markers were similar forpatients with and without RV impairment. Arapid recovery of RV contractile abnormalitieswas the rule (5.7 4.0 days, ranging from 1 to13 days) and observed in all 7 patients. Regres-sion of wall-motion abnormalities was earlier forthe right than for the LV (P < 0.05).

The literature search identied a total of 11potentially relevant studies assessing Takotsubocardiomyopathy, with 5 meeting eligibility crite-ria (2 from the United States, 1 from SouthKorea, 1 from Germany, and 1 from multiplecountries).59 Evaluation was undertaken eitherby echocardiography (2 studies) or magnetic res-onance (MR) (3 studies); data from the pooledstudies are shown in Table III.

Clinical Outcome and Complications:Overall rate of complications was higher(P < 0.05) for classical (Table IV) as compared toinverted Takotsubo presentation, and increasedin patients with lower LVEF. One patient had car-diogenic shock requiring intravenous dobuta-mine infusion and 1 had a resuscitated cardiacarrest. There was 1 episode of atrial brillationrequiring cardioversion, and 1 pulmonary embo-lism. An apical LV thrombus was observed in 2patients and a small pericardial effusion withspontaneous resolution in 5. Pleural effusion wascomparable for patients with and without RVcontractile abnormalities.

In our series, RV involvement was not associ-ated to a higher prevalence of cardiovascular com-plications. During follow-up (13 14 months)there were 3 deaths (3, 6, and 9 months after theonset of Takotsubo) due to malignancy (breastcancer, lymphoma, and lung cancer, respec-tively). Mortality was not higher for patients withRV dysfunction.

Discussion:Takotsubo cardiomyopathy has been describedas transient regional contractile abnormalitiesinvolving the apical and mid segments of the LV,usually related to triggers such as emotional orphysical stress.1 Subsequently, involvement ofmyocardial regions other than the apex wasreported, with contractile abnormalities limitedto the basal and/or mid segments of the LV and anormally contracting apex, suggesting a newclinical entity.3,4 Reports on RV involvementassociated to Takotsubo are less frequent59 andmainly described in patients with classicalTakotsubo variant. In our series, RV regionaldysfunction was found in 7 patients (28%,considering 25 episodes), a nding comparableto the reported prevalence of RV abnormalities(ranging from 21 to 34%).59 Accordingly, mostpatients with RV abnormalities were female; onthe other hand, compared to studies thatreported on patients age,59 patients with RVinvolvement in our series were younger; of note,mean age in our series was also lower(64 19 years old).

Interestingly, stress-induced regional dysfunc-tion involving the RV may follow the same pat-tern of contractile dysfunction found in the LV: in6 patients with classical Takotsubo there was RVapical involvement (isolated or in associationwith mid-RV segment), whereas in 1 patient withinverted Takotsubo (and contractile abnormali-ties involving the mid segments of the LV) thecorresponding mid-RV segment was affected. RVapical involvement in patients with classicalTakotsubo has been characterized in previousreports.7,10 Accordingly, in the 3 studies report-ing the distribution of LV wall-motion abnormali-ties,57 RV dysfunction was associated withclassical Takotsubo (apical ballooning) in 2reports, with only one study describing the asso-ciation with inverted (mid-ventricular variant)presentation.5 Elesber,7 in patients with classicalTakotsubo, also found a similar pattern of wall-motion abnormalities for both the left and RV.Detailed distribution of RV wall-motion abnor-malities was reported in only 1 of the 5 selectedstudies,6 but correspondence with LV abnormali-ties was not described.

Of interest, regression of RV dysfunction wasrapid, and complete normalization of contractile

Figure 2. A modied apical echocardiographic view to focuson the right chambers, depicting involvement of mid seg-ment of RV free wall (arrows), sparing the apex and basal seg-ment. The LV also showed akinesia of mid ventricularsegments in this patient with inverted Takotsubo (bedsideechocardiography). RV = Right ventricle; LV = left ventricle;RA = right atrium.

1018

Rodrigues, et al.

abnormalities observed in all 7 patients; in theparticular patient with inverted pattern ofTakotsubo and RV mid-wall akinesia, a repeatechocardiogram undertaken after 24 hours nolonger showed RV contractile abnormalities.Accordingly, it is likely that early RV involvementmay be missed in some patients either because aRV angiography is not regularly undertaken, orechocardiography is performed later, when RVabnormalities have resolved. In previous studies,there was either incomplete7 or no adequatefollow-up for RV performance.5,6,8

In our series, LVEF was lower in patients withRV involvement, but contrary to other reports,when controlled for presentation type, there wasno signicant difference regarding this variable.Data on initial LVEF were reported in 4 studies,59

with considerable variability among them (mea-surements ranging from 26 to 43%) and highervalues mainly observed for MR measured EF.6,9 Inthese reports, RV dysfunction was associated with

a lower LVEF; however, the effect of presentationtype on LVEF was not taken into account, and amultivariate analysis undertaken in only onestudy.7 Of note, cardiac injury markers were simi-lar for patients with and without RV impairmentin our series. These data were reported in onlyone study6 and likewise, no difference was seenfor patients with RV abnormalities. Indeed, it ispossible that patients with extensive LV myocar-dial involvement, as in classical Takotsubo, willmost likely have associated RV involvement.

In spite of the amount of myocardiumaffected, the long-term prognosis is generallyfavorable in Takotsubo cardiomyopathy, withspontaneous functional recovery. In our seriesthere were 2 in-hospital noncardiac deaths, bothassociated with malignancy, and one later death;complications, however, were fairly common,with 2 patients experiencing serious complica-tions (cardiac arrest and a cardiogenic shock). Inaddition, 2 patients with apical akinesia had LV

TABLE III

Characteristics of Studies Reporting on RV Involvement in Takotsubo Cardiomyopathy

Studies Sharkey et al.5 Hagui et al.6 Elesber et al.7 Lee et al.8 Eitel et al.10 Present Series

Examination MR/TTE/CT MR TTE TTE MR TTEn 136 (95 with MR) 34 25 56 239 25RV dysfunction (%) 23(24%) 9 (26%) 8 (32%) 12 (21%) 81 (34%) 7 (28%)Age (years) 70 8 70 13 73 12 58 25Female Gender (n) 8 (89%) 8 (100%) 6 (86%)Classical/inverted form 18/5 9/0 8/0 6/1Initial LVEF (%) 26 8 40 6 29 9 43.1 8.5 28 10Final LVEF (%) 58 9 65 7Troponin T (ng/ml) 9.7 11.2 1.5 1.2Pleural effusion (n/%) 6 (67%) 62% 4 (57%)CHF (n/%) 6 (75%) 4 (57%)Thrombus (n) 0 0LVOT (n) 0 1Death (n) 0 2 2

RV = right ventricular; MR = magnetic resonance; TTE = transthoracic echocardiography; CT = computed tomography;LVEF = left ventricular ejection fraction; CHF = congestive heart failure; LVOT = left ventricular outow tract obstruction.

TABLE IV

Complications of Takotsubo Cardiomyopathy

Classical(n = 15)

Inverted(n = 10)

RV Involvement(n = 7)

Pericardial effusion 3 2 0Thrombus 2 0 0Shock/sudden death 2 0 2Atrial brillation 1 0 1PE 1 0 0CHF 8 2 4Death 2 1 2Total 19 5 9

RV = Right ventricular; CHF = congestive heart failure; PE = pulmonary embolism.

1019

RV Abnormalities in Takotsubo Cardiomyopathy

thrombus, and 1 patient had a pulmonary embo-lism. Dyskinetic/akinetic left or RV segments canfacilitate intracavitary thrombus formation, andare a potential risk for thromboembolic complica-tions.

Predictably, complications were more com-mon for classical Takotsubo presentation, proba-bly resulting from more severe initial LVdysfunction. A higher prevalence of complica-tions for patients with RV involvement asreported in one study7 was not observed in thepresent series, suggesting again that complica-tions may be associated to a lower LVEF andindeed be related to the larger amount of LVmyocardium involved. Contrary to otherreports,6,9 we did not observe a higher preva-lence of pleural effusion in patients with RVimpairment6; in fact, 3 patients showed pleuraleffusion on chest tomography before the onsetof Takotsubo. In agreement with a previousreport,7 mortality was not higher for patientswith RV dysfunction in this study.

Pathophysiology of Takotsubo has not beenclearly established, but an exaggerated sympa-thetic stimulation with catecholamine cardio-toxicity1 is a key theory among several othersproposed. This theory holds true for most casesof Takotsubo, including those with RV involve-ment, as regional differences in adrenergic sensi-tivity or innervation11 could explain distinctclinical presentations and regional dysfunction,both for the left and the right ventricle. Interest-ingly, the 2 patients with recurrent episodes ofTakotsubo had different forms of the disease: onehad classical Takotsubo followed by invertedform with associated RV involvement and theother started with an inverted form followed byclassical presentation of Takotsubo. This does notagree with the theory of contractile abnormalitiesresulting exclusively from a different distributionof the sympathetic nerves, unless individual sensi-tivity or density of receptors changes from 1episode to the other. It actually implies a func-tional, not structural, response from adrenergicreceptors, with the distribution (or density) ofadrenergic receptors in the myocardium a poten-tially dynamic process. Varying LV wall-motionabnormalities in the same patient with Takotsubohave also been described.12 Other proposedmechanisms (thrombosis of a left anteriordescending artery that wraps around the apexwith rapid resolution by endogenous brinoly-sis13 and transient obstruction of LV outow tractin response to stress14) have been described aspotentially implicated in the pathogenesis ofTakotsubo. Only 1 patient in our series had anintra-ventricular gradient; additionally, neithertheory explains fully the presence of RVapical involvement or apical-sparing contractile

abnormalities. Takotsubo is most likely a hetero-geneous disorder, with multiple causes playing arole in its pathogenesis; it has gained attentionduring the past few years, however, alternativepresentations are possibly under recognized.Careful evaluation of the RV should bring moreattention to regional abnormalities in this cham-ber; however, as recovery may rapidly occur, ear-lier examinations are crucial to observe RVcontractile abnormalities.

Limitations:A major limitation is the small number ofpatients, resulting from the fact that Takotsubo isan uncommon disease. Moreover, due to com-plex geometry, assessment of the RV using echo-cardiography is limited, and mainly qualitative orsemiquantitative7; however, for serial cardiacevaluation it is more suitable than other examina-tions; tissue Doppler velocities and tricuspidannular plane systolic excursion (TAPSE) havebeen used for RV evaluation; however, RVinvolvement was restricted to RV apical or midsegments, with hypercontractile basal segments,limiting the use of these techniques in thispopulation.

Conclusion:RV involvement is relatively common in Takotsubocardiomyopathy and may follow the same distri-bution of LV regional abnormality. At present, itremains unclear if RV involvement can be de-nitely associated to a higher rate of complications.

References1. Bybee KA, Kara T, Prasad A, et al: Systematic review:

Transient left ventricular apical ballooning: A syndromethat mimics ST-segment elevation myocardial infarction.Ann Intern Med 2004;141:858865.

2. Reuss CS, Lester SJ, Hurst RT, et al: Isolated left ventricu-lar basal ballooning phenotype of transient cardiomyopa-thy in young women. Am J Cardiol 2007;99:14511453.

3. Van de Walle SO, Gevaert SA, Gheeraert PJ, et al:Transient stress-induced cardiomyopathy with aninverted takotsubo contractile pattern. Mayo Clin Proc2006;81:14991502.

4. Ennezat PV, Pesenti-Rossi D, Aubert JM, et al: Transientleft ventricular basal dysfunction without coronary steno-sis in acute cerebral disorders: A novel heart syndrome(inverted Takotsubo). Echocardiography 2005;22:599602.

5. Sharkey SW, Windenburg DC, Lesser JR, et al: Natural his-tory and expansive clinical prole of stress (tako-tsubo)cardiomyopathy. J Am Coll Cardiol 2010;55:333341.

6. Haghi D, Athanasiadis A, Papavassiliu T, et al: Rightventricular involvement in Takotsubo cardiomyopathy.Eur Heart J 2006;27:24332439.

7. Elesber AA, Prasad A, Bybee KA, et al: Transient cardiacapical ballooning syndrome: Prevalence and clinicalimplications of right ventricular involvement. J Am CollCardiol 2006;47:10821083.

8. Lee PH, Song JK, Sun BJ, et al: Outcomes of patients withstress-induced cardiomyopathy diagnosed by echocardi-

1020

Rodrigues, et al.

ography in a tertiary referral hospital. J Am Soc Echocardi-ogr 2010;23:766771.

9. Eitel I, von Knobelsdorff-Brenkenhoff F, Bernhardt P, et al:Clinical characteristics and cardiovascular magneticresonance ndings in stress (takotsubo) cardiomyopathy.JAMA 2011;306:277286.

10. Donohue D, Ahsan C, Sanaei-Ardekani M, et al: Earlydiagnosis of stress-induced apical ballooning syndromebased on classic echocardiographic ndings and correla-tion with cardiac catheterization. J Am Soc Echocardiogr2005;18:1423.

11. Pierpont GL, DeMaster EG, Cohn JN: Regional differencesin adrenergic function within the left ventricle. Am J Phys-iol 1984;246:H824H829.

12. Blessing E, Steen H, Rosenberg M, et al: Recurrence of tak-otsubo cardiomyopathy with variant forms of leftventricular dysfunction. J Am Soc Echocardiogr 2007;20:439e411439 e432.

13. Ibanez B, Navarro F, Farre J, et al: Tako-tsubo syndromeassociated with a long course of the left anteriordescending coronary artery along the apical diaphrag-matic surface of the left ventricle. Rev Esp Cardiol2004;57:209216.

14. Mahmoud RE, Mansencal N, Pilliere R, et al: Prevalenceand characteristics of left ventricular outow tractobstruction in Tako-Tsubo syndrome. Am Heart J2008;156:543548.

1021

RV Abnormalities in Takotsubo Cardiomyopathy

Copyright of Echocardiography is the property of Wiley-Blackwell and its content may not becopied or emailed to multiple sites or posted to a listserv without the copyright holder'sexpress written permission. However, users may print, download, or email articles forindividual use.