- Email: [email protected]
Clinical Characteristics and Four-Year Outcomes of Patients in the Rhode Island Takotsubo Cardiomyopathy Registry
Clinical Characteristics and Four-Year Outcomes of Patients in the Rhode Island Takotsubo Cardiomyopathy Registry Richard A. Regnante, MDa,*, Ryan W. Zuzek, MDa, Steven B. Weinsier, MDa, Syed R. Latif, MDb, Russell A. Linsky, MDa, Hanna N. Ahmed, MD, MPHb, and Immad Sadiq, MDa The aim was to establish a registry of patients with a diagnosis of Takotsubo cardiomyopathy (TC) to help learn more about the characteristics, treatment strategies, and natural history of this disease. Data for patients with TC diagnosed from July 2004 to April 2008 at 2 major hospitals in Rhode Island were obtained. A data set was created that included baseline demographics and characteristics, hospital, course, and clinical outcomes. TC was diagnosed in 70 patients during the study period. Postmenopausal women comprised 95% of the cohort. Six patients presented with cardiogenic shock, 9 required intubation, 3 experienced sustained ventricular arrhythmias, and 1 patient died of cardiac causes. Average ejection fraction was 37% at cardiac catheterization. Troponin-I was increased in all except 1 patient. Follow-up echocardiography showed full recovery of wall motion abnormalities, with an average ejection fraction of 59%. Most patients were treated using standard cardiovascular medications for acute coronary syndrome, and 43% were discharged on warfarin therapy because of severe apical wall motion abnormalities. Univariate analysis suggested that long-term use of angiotensin-converting enzyme inhibitors before the onset of TC was protective against cardiogenic shock, sustained ventricular arrhythmia, and death. Consecutive cases grouped into different seasons showed a statistically significant spike in the occurrence of TC during the summer months. In conclusion, the acute phase of this condition may lead to critical illness and death, and use of an angiotensin-converting enzyme inhibitor may have a protective effect. Overall long-term prognosis and recovery of left ventricular function were excellent. © 2009 Elsevier Inc. All rights reserved. (Am J Cardiol 2009;103:1015–1019) The medical community has pondered for some time about the possible association between psychological stress and cardiac events. Recently, a distinct syndrome composed of symptoms and electrocardiographic changes indistinguishable from an acute coronary syndrome has been recognized. Most of these patients underwent emergent cardiac catheterization and no obstructive coronary lesions were found. Characteristic left ventriculograms showing mid and apical segment akinesis with basal segment hyperkinesis suggested the diagnosis of Takotsubo cardiomyopathy (TC). TC was first described in Japanese reports in 1990 and received its name from the octopus-trapping pot, which has a narrow neck and round bottom that resembles the left ventriculogram during systole.1– 6 Our aim was to establish a registry of patients with TC diagnosed at the 2 major hospitals in Rhode Island to help us learn more about the clinical characteristics, different treatment strategies, longterm outcomes, and natural history of this disease.
a
Department of Internal Medicine, Division of Cardiology, The Miriam Hospital, Warren Alpert Medical School of Brown University; and bDepartment of Internal Medicine, Warren Alpert Medical School of Brown University, Providence, Rhode Island. Manuscript received July 23, 2008; revised manuscript received and accepted December 2, 2008. *Corresponding author: Tel: 401-793-4102; fax: 401-793-4049. E-mail address: [email protected] (R.A. Regnante). 0002-9149/09/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2008.12.020
Methods Patients with acute coronary syndromes who underwent emergent cardiac catheterization with findings consistent with the proposed Mayo criteria for TC were enrolled in the study.3 Informed consent was obtained as outlined in the institutional review board–approved protocol. We analyzed medical records and angiographic data for patients enrolled from July 2004 to April 2008 at the 2 major hospitals in Rhode Island. A data set was created that included demographics; baseline clinical characteristics; imaging data, including cardiac catheterization; laboratory data; medical treatment regimen; hospital course; follow-up; and clinical outcomes. Electrocardiographically, we defined ST-segment elevation as deviation ⬎1 mm higher than the baseline in ⱖ2 contiguous leads. We defined T-wave inversions when they were present in ⱖ2 contiguous leads and represented a change from a previous electrocardiogram (ECG) tracing, if available. Patients were considered to have coronary spasm if angiographically they had a segment of coronary narrowing not induced by a catheter that resolved after intracoronary injection of nitroglycerin. Coronary artery dominance was defined by which vessel supplied both the posterior descending artery and posterolateral artery, with codominance defined as the posterior descending artery arising from the right coronary and the posterolateral branch arising from the circumflex artery. Ejection fraction after contrast left ventriculography was calculated using left ventricular www.AJConline.org
1016
The American Journal of Cardiology (www.AJConline.org) Table 2 Precipitating factors
Table 1 Patient clinical characteristics (n ⫽ 70) Age (yrs) Caucasian Women Hypertension Hyperlipidemia Diabetes mellitus Tobacco abuse Never Former Current Depression or general anxiety disorder Chronic obstructive pulmonary disease or asthma Presenting symptoms Chest pain alone Chest pain and shortness of breath Shortness of breath alone Syncope with or without chest pain or shortness of breath Other Electrocardiography ST-Segment elevation Dynamic T-wave changes only No changes Cardiac catheterization Ejection fraction (%) Left ventricular end-diastolic pressure (mm Hg) Coronary spasm Coronary dominance Right Left Codominant
67 ⫾ 11 61 (87%) 67 (95%) 46 (66%) 34 (49%) 10 (14%) 37 (53%) 18 (26%) 15 (21%) 26 (37%) 20 (29%) 36 (51%) 18 (26%) 10 (14%) 4 (6%) 2 (3%) 29 (41%) 29 (41%) 12 (17%) 37 ⫾ 12% 21 ⫾ 9 0 59 (84%) 10 (14%) 1 (1%)
Emotional Bad news about family member Domestic argument Domestic/financial stressor Wallet stolen Bad driving directions New diagnosis of personal disease Heated town meeting Upcoming operation Unemployed adult son living at home Stressful business meeting Sending parent to nursing home Directing a symphony New job anxiety Lost job Physical Severe physical illness Chronic obstructive pulmonary disease exacerbation Fall Tooth extraction Minor motor vehicle accident Compression fracture Postsurgical Drug overdose Postsurgical pain Multiple bee stings Allergic reaction Colonoscopy preparation No identifiable stressor
26 (37%) 6 6 3 1 1 1 1 1 1 1 1 1 1 1 21 (30%) 4 4 3 2 1 1 1 1 1 1 1 1 23 (33%)
4100 4000 Number of cases
analytical software with the Heartlab system (Agfa Heartlab Inc., Greenville, South Carolina). Shock was defined as systolic blood pressure ⬍90 mm Hg with signs of end-organ hypoperfusion requiring the use of vasopressor agents. Sustained ventricular arrhythmia was defined as ventricular fibrillation or ventricular tachycardia lasting ⱖ30 seconds and requiring chemical cardioversion or defibrillation. Hypertension was considered present if blood pressure was consistently ⬎140/90 mm Hg or patients were on treatment with a preexisting antihypertensive agent. Hyperlipidemia was considered present if total cholesterol was ⬎200 mg/dl or patients were on a preexisting antihyperlipidemic regimen. Stressors were defined as physical illness or a significant emotional event up to 1 week preceding the index event. For patients prospectively enrolled in the registry, follow-up telephone calls were made to each respective patient at 30 days, 6 months, and annually thereafter. Follow-up records from outside cardiologists were also obtained. Recurrence of TC was defined using the same criteria as the index event, confirmed using coronary angiography and left ventriculography. Deaths that occurred were investigated using review of medical records and/or death certificate and telephone calls to family members when death occurred outside the hospital. Continuous data were summarized as mean ⫾ SD unless otherwise noted. Frequencies and percentages were used to describe categorical variables. Univariate analysis was performed using SAS software (SAS Institute Inc., Cary, North
3900 3800 3700 3600 3500 3400 June, July, Sept, Oct, Dec, Jan, March, Aug Nov Feb April, May Figure 1. Cardiac catheterization volume by season.
Carolina) with Fisher’s exact test. Statistical significance was defined as p ⬍0.05. To analyze for potential seasonal variations in the incidence of TC, consecutive patients from September 1, 2004, to August 31, 2007, were grouped into 3-month intervals signifying the different seasons during a 3-year period according to date of diagnosis. Seasons were classified as December through February (winter), March through May (spring), June through August (summer), and September through November (fall). Chi-square goodness-of-fit test was used to study differences in seasonal variation between groups.
Cardiomyopathy/Rhode Island Takotsubo Cardiomyopathy Registry
1017
Figure 2. Number of cases of TC by season. Table 4 Medications
Table 3 Clinical outcomes Intubated Shock/vasopressors/intra-aortic balloon pump Sustained ventricular arrhythmias Mean peak troponin-I Median length of stay (d) Follow-up ejection fraction using echocardiography (%) TC recurrence Death Thromboembolic events
9 (13%) 6 (9%) 3 (4%) 6.9 ng/ml ⫾ 6 4 days 59 ⫾ 7% 2 3 (1 cardiac, 2 noncardiac) 0
Results Clinical characteristics of patients are listed in Table 1. During 4 years, 19,180 patients underwent diagnostic cardiac catheterization and 70 patients with presumed acute coronary syndrome met all the proposed Mayo criteria for a diagnosis of TC. This represented a frequency of 0.4%, and 95% of this group were women. Of interest, preexisting hypertension was present in 66% of patients, and a preadmission diagnosis of depression or generalized anxiety disorder was present in 37% of patients. A clearly identified preceding stressor was present in 67% of patients, with 55% of these representing a physical stressor and 45% representing an emotional stressor. Table 2 lists presenting stressors for the cohort. Chest pain and/or shortness of breath were the presenting symptom in 91% of patients. ST-Segment elevations were present on the index ECG in 41% of patients. Dynamic T-wave inversions were present on the index ECG in 41% of patients. The remaining 17% had no significant ECG changes. Troponin-I was increased in 99% of patients, with an average peak of 6.9 ng/ml, which was out of proportion to the extensive wall motion abnormalities seen at cardiac catheterization. Dates of presentation were grouped into 3-month intervals signifying the different seasons, with an equal number of seasons each year. There were 15,324 diagnostic cardiac catheterizations performed during this period. A total of 59 consecutive cases of TC were identified during this period.
Before admission* Aspirin Plavix  Blocker Calcium channel blocker ACE-I or angiotensin receptor blocker Diuretic Statin Warfarin Discharge† Aspirin Plavix  Blocker Calcium channel blocker ACE-I or angiotensin receptor blocker Diuretic Statin Warfarin
23 (33%) 4 (6%) 18 (26%) 6 (9%) 27 (39%) 15 (22%) 17 (25%) 0 54 (86%) 3 (5%) 54 (86%) 5 (8%) 47 (67%) 13 (21%) 33 (52%) 27 (43%)
* One patient was transferred from another hospital and preadmission medications were never able to be documented. † Discharge medications were not available for 4 patients and an additional 3 patients were pending discharge at the time of writing this report.
Despite a decrease in the overall volume of catheterizations (Figure 1) during the summer months, there was a distinct spike in the occurrence of TC in each of the 3 years (Figure 2). During a 3-year period, this clustering of cases in the summer months was noted to be statistically significant (p ⬍0.001). At the time of cardiac catheterization, most patients had apical dyskinesis, with the exception of 2 patients with a left ventriculogram that showed midanterior and midinferior akinesis with apical sparing. Right coronary dominance was present in 84% of patients, and 14% were left dominant. One patient was codominant. Average ejection fraction calculated using left ventriculography was 37%. Average left ventricular end-diastolic pressure was 21 mm Hg. No patient had coronary spasm. Clinical outcomes of the group are listed in Table 3. Three patients experienced sustained ventricular tachycardia or ventricular fibrillation. Each of these patients was successfully resuscitated and had normal neurologic func-
1018
The American Journal of Cardiology (www.AJConline.org)
Table 5 Univariate analysis for predictors of death, cardiogenic shock, and arrhythmias (ventricular fibrillation or sustained ventricular tachycardia) Variable Men Hypertension Hyperlipidemia Diabetes mellitus Chronic obstructive pulmonary disease Current smoker Emotional trigger Non-Caucasian race Preexisting aspirin use Preexisting ACE-I use Preexisting angiotensin receptor blocker use Preexisting -blocker use Preexisting calcium channel blocker use Preexisting statin use Preexisting diuretic use Chest pain on presentation Shortness of breath on presentation Intubation at time of presentation Right coronary artery dominance T-Wave abnormalities on index ECG ST-Segment elevation on index ECG
Meeting End Point (%)
p Value
1.4% 8.6% 5.8% 2.9% 2.9% 8.6% 8.6% 2.9% 1.5% 0% 4.6% 2.9% 1.5% 1.5% 1.5% 11.6% 8.7% 7.1% 11.4% 5.7% 7.1%
0.406 0.493 0.513 0.652 0.494 0.745 0.308 0.628 0.082 0.026 0.088 0.715 1.00 0.271 0.434 1.00 0.302 0.004 0.361 1.00 1.00
tion at the time of discharge. There were 3 deaths. One patient died within the first 24 hours of acute heart failure after a failed extubation, and the family chose not to have the patient reintubated. Two patients died of noncardiac causes before adequate follow-up was obtained. Follow-up echocardiographic data were available for 54 patients, and all showed full recovery of wall motion abnormalities and left ventricular ejection fraction to an average of 59%. Serial outpatient echocardiograms were not obtained and thus it was impossible for us to determine the exact time frame for full recovery of ventricular function. However, most echocardiograms obtained within 1 month after the hospitalization showed complete resolution of wall motion abnormalities. Two patients presented with recurrences of TC. No thromboembolic events were observed in any patient regardless of treatment regimen. Table 4 lists preadmission and discharge medications for patients. Most of our patients were treated using standard cardiovascular medications for patients with acute coronary syndromes, with 43% discharged on warfarin therapy because of the presence of severe apical wall motion abnormalities. Table 5 lists univariate analysis findings of patients. We used a common end point of death, cardiogenic shock, or ventricular arrhythmia to signify critical illness, and 11 patients met this criteria. We found that intubation at time of the index presentation was a strong predictor of critical illness. None of the patients on preexisting treatment with angiotensin-converting enzyme inhibitors (ACE-Is) developed critical illness. Discussion Our data emphasized many of the same findings of case series in other parts of the world.7–9 Our cohort of patients
was predominantly older women with a paucity of cardiovascular risk factors. Presentation was frequently preceded by a physical or emotional stressor, but this was not present in all cases. Interestingly, there was a wide spectrum of illness severity. A subset of patients presented with or developed cardiogenic shock or sustained ventricular arrhythmias. Some of these patients required care in the coronary care unit while on treatment with ventilators, vasopressors, or intra-aortic balloon counterpulsation. If patients were adequately supported during the first 48 hours of their illness, they all had an excellent prognosis. A fair portion of the cohort had a diagnosis of hypertension and was on preexisting treatment with ACE-Is or angiotensin receptor blockers. Interestingly, use of a preexisting ACE-I seemed to carry a protective effect against the development of critical illness. There was no clear consensus about the appropriate treatment of patients with TC in Rhode Island, but there was a marked increase in prescribing aspirin,  blockers, ACE-Is or angiotensin receptor blockers, and statins during the hospitalization. These findings were consistent with what 1 would expect to see in treatment after an acute coronary syndrome. A large number of patients were also discharged on warfarin therapy for severe apical wall motion abnormalities. To our knowledge, there were no data to support the use of particular medication regimens in patients with TC. Interestingly, ⬎¼ of patients with a TC diagnosis were on preexisting treatment with a  blocker. Use of  blockers has been proposed as a possible treatment to potentially reduce TC recurrence. Our results did not suggest that -blocker use was protective for TC recurrence, but only 26% of our cohort were using them at baseline and our recurrence rate was too low to assess statistical significance. We observed a thought-provoking trend in the time of year when TC was most often diagnosed. Among the many interesting epidemiologic characteristics of myocardial infarctions was the seasonal variation. The National Registry of Myocardial Infarction (NRMI) and other registries have shown that the occurrence of myocardial infarction had a marked peak in the winter months, with the nadir in the summer months.10,11 In contrast to the seasonal variation for acute myocardial infarction (AMI) reported in the NRMI registry, our cohort of patients showed a distinct spike in the occurrence of TC during the summer months in each of the 3 years. This could not be explained by a decrease in the volume of cardiac catheterizations performed during the different seasons. We found that our catheterization laboratories performed fewer catheterizations during the summer months. TC did not appear to follow the typical seasonal pattern of AMI as seen in the NRMI. This observation suggested the possibility that TC may have an alternative pathophysiologic mechanism from a patient with a AMI, and this phenomenon deserves further study. The present study represented a relatively small cohort of patients with a relatively uncommon medical condition. This cohort of patients was enrolled as registry patients, and their treatment and outcomes were observed. Treatments were empirical and may reflect biases that we cannot account for.
Cardiomyopathy/Rhode Island Takotsubo Cardiomyopathy Registry 1. Pilgrim TM, Wyss TR. Takotsubo cardiomyopathy or transient left ventricular apical ballooning syndrome: a systematic review. Int J Cardiol 2008;124:283–292. 2. Gianni M, Dentali F, Grandi AM, Summer G, Hiralal R, Lonn E. Apical ballooning syndrome or Takotsubo cardiomyopathy: a systematic review. Eur Heart J 2006;27:1523–1529. 3. Ahmed H, Linsky R, Weinsier S, Regnante R, Sadiq I. Takotsubo cardiomyopathy: a review and future perspective. Future Cardiol 2008;4:23–32. 4. Bybee KA, Kara T, Prasad A, Lerman A, Barsness G, Wright S, Rihal C. Systematic review: transient left ventricular apical ballooning: a syndrome that mimics ST-segment elevation myocardial infarction. Ann Intern Med 2004;141:858 – 864. 5. Sharkey SW, Lesser JR, Zenovich AG, Maron M, Lindberg J, Longe T, Maron B. Acute and reversible cardiomyopathy provoked by stress in women from the United States. Circulation 2005;111: 472– 476.
1019
6. Kawai S, Kitabatake A, Tomoike H. Guidelines for diagnosis of Takotsubo (ampulla) cardiomyopathy. Circ J 2007;71:990 –992. 7. Abe Y, Kondo M, Matsuoka R, Araki M, Dohyama K, Tanio H. Assessment of clinical features in transient left ventricular apical ballooning. J Am Coll Cardiol 2003;41:737–742. 8. Regnante R, Ahmed H, Linsky RA, Weinsier S, Sadiq I. Characteristics and outcomes of patients in the Rhode Island Takotsubo Cardiomyopathy Registry. Catheter Cardiovasc Intervent 2007;69(suppl):S14 –S15. 9. Elesber AA, Prasad A, Lennon RJ, Wright S, Lerman A, Rihal C. Four-year recurrence rate and prognosis of the apical ballooning syndrome. J Am Coll Cardiol 2007;50:448 – 452. 10. Ornato J, Peberdy M, Chandra N, Bush D. Seasonal pattern of acute myocardial infarction in the National Registry of Myocardial Infarction. J Am Coll Cardiol 1996;28:1684 –1688. 11. Spencer F, Goldberg R, Becker R, Gore J. Seasonal distribution of acute myocardial infarction in the Second Registry of Myocardial Infarction. J Am Coll Cardiol 1996;31:1226 –1233.
a
Department of Internal Medicine, Division of Cardiology, The Miriam Hospital, Warren Alpert Medical School of Brown University; and bDepartment of Internal Medicine, Warren Alpert Medical School of Brown University, Providence, Rhode Island. Manuscript received July 23, 2008; revised manuscript received and accepted December 2, 2008. *Corresponding author: Tel: 401-793-4102; fax: 401-793-4049. E-mail address: [email protected] (R.A. Regnante). 0002-9149/09/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2008.12.020
Methods Patients with acute coronary syndromes who underwent emergent cardiac catheterization with findings consistent with the proposed Mayo criteria for TC were enrolled in the study.3 Informed consent was obtained as outlined in the institutional review board–approved protocol. We analyzed medical records and angiographic data for patients enrolled from July 2004 to April 2008 at the 2 major hospitals in Rhode Island. A data set was created that included demographics; baseline clinical characteristics; imaging data, including cardiac catheterization; laboratory data; medical treatment regimen; hospital course; follow-up; and clinical outcomes. Electrocardiographically, we defined ST-segment elevation as deviation ⬎1 mm higher than the baseline in ⱖ2 contiguous leads. We defined T-wave inversions when they were present in ⱖ2 contiguous leads and represented a change from a previous electrocardiogram (ECG) tracing, if available. Patients were considered to have coronary spasm if angiographically they had a segment of coronary narrowing not induced by a catheter that resolved after intracoronary injection of nitroglycerin. Coronary artery dominance was defined by which vessel supplied both the posterior descending artery and posterolateral artery, with codominance defined as the posterior descending artery arising from the right coronary and the posterolateral branch arising from the circumflex artery. Ejection fraction after contrast left ventriculography was calculated using left ventricular www.AJConline.org
1016
The American Journal of Cardiology (www.AJConline.org) Table 2 Precipitating factors
Table 1 Patient clinical characteristics (n ⫽ 70) Age (yrs) Caucasian Women Hypertension Hyperlipidemia Diabetes mellitus Tobacco abuse Never Former Current Depression or general anxiety disorder Chronic obstructive pulmonary disease or asthma Presenting symptoms Chest pain alone Chest pain and shortness of breath Shortness of breath alone Syncope with or without chest pain or shortness of breath Other Electrocardiography ST-Segment elevation Dynamic T-wave changes only No changes Cardiac catheterization Ejection fraction (%) Left ventricular end-diastolic pressure (mm Hg) Coronary spasm Coronary dominance Right Left Codominant
67 ⫾ 11 61 (87%) 67 (95%) 46 (66%) 34 (49%) 10 (14%) 37 (53%) 18 (26%) 15 (21%) 26 (37%) 20 (29%) 36 (51%) 18 (26%) 10 (14%) 4 (6%) 2 (3%) 29 (41%) 29 (41%) 12 (17%) 37 ⫾ 12% 21 ⫾ 9 0 59 (84%) 10 (14%) 1 (1%)
Emotional Bad news about family member Domestic argument Domestic/financial stressor Wallet stolen Bad driving directions New diagnosis of personal disease Heated town meeting Upcoming operation Unemployed adult son living at home Stressful business meeting Sending parent to nursing home Directing a symphony New job anxiety Lost job Physical Severe physical illness Chronic obstructive pulmonary disease exacerbation Fall Tooth extraction Minor motor vehicle accident Compression fracture Postsurgical Drug overdose Postsurgical pain Multiple bee stings Allergic reaction Colonoscopy preparation No identifiable stressor
26 (37%) 6 6 3 1 1 1 1 1 1 1 1 1 1 1 21 (30%) 4 4 3 2 1 1 1 1 1 1 1 1 23 (33%)
4100 4000 Number of cases
analytical software with the Heartlab system (Agfa Heartlab Inc., Greenville, South Carolina). Shock was defined as systolic blood pressure ⬍90 mm Hg with signs of end-organ hypoperfusion requiring the use of vasopressor agents. Sustained ventricular arrhythmia was defined as ventricular fibrillation or ventricular tachycardia lasting ⱖ30 seconds and requiring chemical cardioversion or defibrillation. Hypertension was considered present if blood pressure was consistently ⬎140/90 mm Hg or patients were on treatment with a preexisting antihypertensive agent. Hyperlipidemia was considered present if total cholesterol was ⬎200 mg/dl or patients were on a preexisting antihyperlipidemic regimen. Stressors were defined as physical illness or a significant emotional event up to 1 week preceding the index event. For patients prospectively enrolled in the registry, follow-up telephone calls were made to each respective patient at 30 days, 6 months, and annually thereafter. Follow-up records from outside cardiologists were also obtained. Recurrence of TC was defined using the same criteria as the index event, confirmed using coronary angiography and left ventriculography. Deaths that occurred were investigated using review of medical records and/or death certificate and telephone calls to family members when death occurred outside the hospital. Continuous data were summarized as mean ⫾ SD unless otherwise noted. Frequencies and percentages were used to describe categorical variables. Univariate analysis was performed using SAS software (SAS Institute Inc., Cary, North
3900 3800 3700 3600 3500 3400 June, July, Sept, Oct, Dec, Jan, March, Aug Nov Feb April, May Figure 1. Cardiac catheterization volume by season.
Carolina) with Fisher’s exact test. Statistical significance was defined as p ⬍0.05. To analyze for potential seasonal variations in the incidence of TC, consecutive patients from September 1, 2004, to August 31, 2007, were grouped into 3-month intervals signifying the different seasons during a 3-year period according to date of diagnosis. Seasons were classified as December through February (winter), March through May (spring), June through August (summer), and September through November (fall). Chi-square goodness-of-fit test was used to study differences in seasonal variation between groups.
Cardiomyopathy/Rhode Island Takotsubo Cardiomyopathy Registry
1017
Figure 2. Number of cases of TC by season. Table 4 Medications
Table 3 Clinical outcomes Intubated Shock/vasopressors/intra-aortic balloon pump Sustained ventricular arrhythmias Mean peak troponin-I Median length of stay (d) Follow-up ejection fraction using echocardiography (%) TC recurrence Death Thromboembolic events
9 (13%) 6 (9%) 3 (4%) 6.9 ng/ml ⫾ 6 4 days 59 ⫾ 7% 2 3 (1 cardiac, 2 noncardiac) 0
Results Clinical characteristics of patients are listed in Table 1. During 4 years, 19,180 patients underwent diagnostic cardiac catheterization and 70 patients with presumed acute coronary syndrome met all the proposed Mayo criteria for a diagnosis of TC. This represented a frequency of 0.4%, and 95% of this group were women. Of interest, preexisting hypertension was present in 66% of patients, and a preadmission diagnosis of depression or generalized anxiety disorder was present in 37% of patients. A clearly identified preceding stressor was present in 67% of patients, with 55% of these representing a physical stressor and 45% representing an emotional stressor. Table 2 lists presenting stressors for the cohort. Chest pain and/or shortness of breath were the presenting symptom in 91% of patients. ST-Segment elevations were present on the index ECG in 41% of patients. Dynamic T-wave inversions were present on the index ECG in 41% of patients. The remaining 17% had no significant ECG changes. Troponin-I was increased in 99% of patients, with an average peak of 6.9 ng/ml, which was out of proportion to the extensive wall motion abnormalities seen at cardiac catheterization. Dates of presentation were grouped into 3-month intervals signifying the different seasons, with an equal number of seasons each year. There were 15,324 diagnostic cardiac catheterizations performed during this period. A total of 59 consecutive cases of TC were identified during this period.
Before admission* Aspirin Plavix  Blocker Calcium channel blocker ACE-I or angiotensin receptor blocker Diuretic Statin Warfarin Discharge† Aspirin Plavix  Blocker Calcium channel blocker ACE-I or angiotensin receptor blocker Diuretic Statin Warfarin
23 (33%) 4 (6%) 18 (26%) 6 (9%) 27 (39%) 15 (22%) 17 (25%) 0 54 (86%) 3 (5%) 54 (86%) 5 (8%) 47 (67%) 13 (21%) 33 (52%) 27 (43%)
* One patient was transferred from another hospital and preadmission medications were never able to be documented. † Discharge medications were not available for 4 patients and an additional 3 patients were pending discharge at the time of writing this report.
Despite a decrease in the overall volume of catheterizations (Figure 1) during the summer months, there was a distinct spike in the occurrence of TC in each of the 3 years (Figure 2). During a 3-year period, this clustering of cases in the summer months was noted to be statistically significant (p ⬍0.001). At the time of cardiac catheterization, most patients had apical dyskinesis, with the exception of 2 patients with a left ventriculogram that showed midanterior and midinferior akinesis with apical sparing. Right coronary dominance was present in 84% of patients, and 14% were left dominant. One patient was codominant. Average ejection fraction calculated using left ventriculography was 37%. Average left ventricular end-diastolic pressure was 21 mm Hg. No patient had coronary spasm. Clinical outcomes of the group are listed in Table 3. Three patients experienced sustained ventricular tachycardia or ventricular fibrillation. Each of these patients was successfully resuscitated and had normal neurologic func-
1018
The American Journal of Cardiology (www.AJConline.org)
Table 5 Univariate analysis for predictors of death, cardiogenic shock, and arrhythmias (ventricular fibrillation or sustained ventricular tachycardia) Variable Men Hypertension Hyperlipidemia Diabetes mellitus Chronic obstructive pulmonary disease Current smoker Emotional trigger Non-Caucasian race Preexisting aspirin use Preexisting ACE-I use Preexisting angiotensin receptor blocker use Preexisting -blocker use Preexisting calcium channel blocker use Preexisting statin use Preexisting diuretic use Chest pain on presentation Shortness of breath on presentation Intubation at time of presentation Right coronary artery dominance T-Wave abnormalities on index ECG ST-Segment elevation on index ECG
Meeting End Point (%)
p Value
1.4% 8.6% 5.8% 2.9% 2.9% 8.6% 8.6% 2.9% 1.5% 0% 4.6% 2.9% 1.5% 1.5% 1.5% 11.6% 8.7% 7.1% 11.4% 5.7% 7.1%
0.406 0.493 0.513 0.652 0.494 0.745 0.308 0.628 0.082 0.026 0.088 0.715 1.00 0.271 0.434 1.00 0.302 0.004 0.361 1.00 1.00
tion at the time of discharge. There were 3 deaths. One patient died within the first 24 hours of acute heart failure after a failed extubation, and the family chose not to have the patient reintubated. Two patients died of noncardiac causes before adequate follow-up was obtained. Follow-up echocardiographic data were available for 54 patients, and all showed full recovery of wall motion abnormalities and left ventricular ejection fraction to an average of 59%. Serial outpatient echocardiograms were not obtained and thus it was impossible for us to determine the exact time frame for full recovery of ventricular function. However, most echocardiograms obtained within 1 month after the hospitalization showed complete resolution of wall motion abnormalities. Two patients presented with recurrences of TC. No thromboembolic events were observed in any patient regardless of treatment regimen. Table 4 lists preadmission and discharge medications for patients. Most of our patients were treated using standard cardiovascular medications for patients with acute coronary syndromes, with 43% discharged on warfarin therapy because of the presence of severe apical wall motion abnormalities. Table 5 lists univariate analysis findings of patients. We used a common end point of death, cardiogenic shock, or ventricular arrhythmia to signify critical illness, and 11 patients met this criteria. We found that intubation at time of the index presentation was a strong predictor of critical illness. None of the patients on preexisting treatment with angiotensin-converting enzyme inhibitors (ACE-Is) developed critical illness. Discussion Our data emphasized many of the same findings of case series in other parts of the world.7–9 Our cohort of patients
was predominantly older women with a paucity of cardiovascular risk factors. Presentation was frequently preceded by a physical or emotional stressor, but this was not present in all cases. Interestingly, there was a wide spectrum of illness severity. A subset of patients presented with or developed cardiogenic shock or sustained ventricular arrhythmias. Some of these patients required care in the coronary care unit while on treatment with ventilators, vasopressors, or intra-aortic balloon counterpulsation. If patients were adequately supported during the first 48 hours of their illness, they all had an excellent prognosis. A fair portion of the cohort had a diagnosis of hypertension and was on preexisting treatment with ACE-Is or angiotensin receptor blockers. Interestingly, use of a preexisting ACE-I seemed to carry a protective effect against the development of critical illness. There was no clear consensus about the appropriate treatment of patients with TC in Rhode Island, but there was a marked increase in prescribing aspirin,  blockers, ACE-Is or angiotensin receptor blockers, and statins during the hospitalization. These findings were consistent with what 1 would expect to see in treatment after an acute coronary syndrome. A large number of patients were also discharged on warfarin therapy for severe apical wall motion abnormalities. To our knowledge, there were no data to support the use of particular medication regimens in patients with TC. Interestingly, ⬎¼ of patients with a TC diagnosis were on preexisting treatment with a  blocker. Use of  blockers has been proposed as a possible treatment to potentially reduce TC recurrence. Our results did not suggest that -blocker use was protective for TC recurrence, but only 26% of our cohort were using them at baseline and our recurrence rate was too low to assess statistical significance. We observed a thought-provoking trend in the time of year when TC was most often diagnosed. Among the many interesting epidemiologic characteristics of myocardial infarctions was the seasonal variation. The National Registry of Myocardial Infarction (NRMI) and other registries have shown that the occurrence of myocardial infarction had a marked peak in the winter months, with the nadir in the summer months.10,11 In contrast to the seasonal variation for acute myocardial infarction (AMI) reported in the NRMI registry, our cohort of patients showed a distinct spike in the occurrence of TC during the summer months in each of the 3 years. This could not be explained by a decrease in the volume of cardiac catheterizations performed during the different seasons. We found that our catheterization laboratories performed fewer catheterizations during the summer months. TC did not appear to follow the typical seasonal pattern of AMI as seen in the NRMI. This observation suggested the possibility that TC may have an alternative pathophysiologic mechanism from a patient with a AMI, and this phenomenon deserves further study. The present study represented a relatively small cohort of patients with a relatively uncommon medical condition. This cohort of patients was enrolled as registry patients, and their treatment and outcomes were observed. Treatments were empirical and may reflect biases that we cannot account for.
Cardiomyopathy/Rhode Island Takotsubo Cardiomyopathy Registry 1. Pilgrim TM, Wyss TR. Takotsubo cardiomyopathy or transient left ventricular apical ballooning syndrome: a systematic review. Int J Cardiol 2008;124:283–292. 2. Gianni M, Dentali F, Grandi AM, Summer G, Hiralal R, Lonn E. Apical ballooning syndrome or Takotsubo cardiomyopathy: a systematic review. Eur Heart J 2006;27:1523–1529. 3. Ahmed H, Linsky R, Weinsier S, Regnante R, Sadiq I. Takotsubo cardiomyopathy: a review and future perspective. Future Cardiol 2008;4:23–32. 4. Bybee KA, Kara T, Prasad A, Lerman A, Barsness G, Wright S, Rihal C. Systematic review: transient left ventricular apical ballooning: a syndrome that mimics ST-segment elevation myocardial infarction. Ann Intern Med 2004;141:858 – 864. 5. Sharkey SW, Lesser JR, Zenovich AG, Maron M, Lindberg J, Longe T, Maron B. Acute and reversible cardiomyopathy provoked by stress in women from the United States. Circulation 2005;111: 472– 476.
1019
6. Kawai S, Kitabatake A, Tomoike H. Guidelines for diagnosis of Takotsubo (ampulla) cardiomyopathy. Circ J 2007;71:990 –992. 7. Abe Y, Kondo M, Matsuoka R, Araki M, Dohyama K, Tanio H. Assessment of clinical features in transient left ventricular apical ballooning. J Am Coll Cardiol 2003;41:737–742. 8. Regnante R, Ahmed H, Linsky RA, Weinsier S, Sadiq I. Characteristics and outcomes of patients in the Rhode Island Takotsubo Cardiomyopathy Registry. Catheter Cardiovasc Intervent 2007;69(suppl):S14 –S15. 9. Elesber AA, Prasad A, Lennon RJ, Wright S, Lerman A, Rihal C. Four-year recurrence rate and prognosis of the apical ballooning syndrome. J Am Coll Cardiol 2007;50:448 – 452. 10. Ornato J, Peberdy M, Chandra N, Bush D. Seasonal pattern of acute myocardial infarction in the National Registry of Myocardial Infarction. J Am Coll Cardiol 1996;28:1684 –1688. 11. Spencer F, Goldberg R, Becker R, Gore J. Seasonal distribution of acute myocardial infarction in the Second Registry of Myocardial Infarction. J Am Coll Cardiol 1996;31:1226 –1233.