Risk factors for Takotsubo syndrome following cardiac surgery: A case–control study

Takotsubo syndrome following cardiac surgery is a rare occurrence. However, early diagnosis is essential to prevent treatment which could increase the left ventricular outflow tract pressure gradient in patients with Takotsubo syndrome, and lead to cardiogenic shock. Therefore, our study aimed to identify the incidence of Takotsubo syndrome after cardiopulmonary bypass and the associated risk factors and prognosis.


Background
Takotsubo syndrome was rst described in Japan in 1991 (1). The term 'Takotsubo' in traditional Japanese means 'octopus trap with a wide base' (2). Owing to the shape of the left ventricle, Takotsubo syndrome is also sometimes referred to as ampullary-shaped cardiomyopathy. Although the pathophysiology of Takotsubo syndrome is not fully understood yet, sympathetic over-activation, excessive catecholamine release, and coronary microcirculatory dysfunction, induced by emotional or physical stress, have been suggested as important contributing factors (3)(4)(5).
Certain types of surgeries may act as a physiological stressor, triggering Takotsubo syndrome (6). In 2007, Itoh et al. (7) reported on Takotsubo syndrome following cardiac surgery (mitral valve repair), leading the authors to suggest that myocardial stunning, caused by cardiopulmonary bypass, might be a cause of Takotsubo syndrome (7). Takotsubo syndrome preceded by cardiac surgery, however, is a very rare occurrence with only a few cases having been reported in the literature. Despite the low prevalence of Takotsubo syndrome after cardiac surgery, accurate diagnosis of the syndrome, and its differential diagnosis from low cardiac output syndrome following cardiac surgery, is essential as the use of catecholamines and beta-receptor agonists may increase the left ventricular out ow tract (LVOT) pressure gradient in patients with Takotsubo syndrome and lead to cardiogenic shock (8). In addition, catastrophic complications, such as ventricular tachycardia (3.0%) or ventricular rupture (0.2%), may occur without proper management based on early detection of Takotsubo syndrome (9). However, early detection of Takotsubo syndrome is challenging due to the rarity of the syndrome and lack of knowledge about the associated risk factors. Therefore, our aim in this study was to identify the risk factors and prognosis of Takotsubo syndrome after cardiopulmonary bypass..

Study population
From February 2007 to July 2017, a total of 9,037 consecutive cardiac surgeries were performed at Asan Medical Center, Seoul, Korea. Among these cases, 3,264 patients who were diagnosed with coronary artery disease pre-operatively or who underwent heart transplantation were excluded. Of the remaining 5,773 patients, Takotsubo syndrome was revealed postoperatively in 52 patients (0.9%). To perform a case-control study, we randomly selected 104 patients of the remaining 5,721 patients (without Takotsubo syndrome) as the control group (with a 2:1 ratio to the Takotsubo syndrome) to identify risk factors. The analysis of risk factors was performed retrospectively. The study was approved by the Institutional Review Board of Asan Medical Center and the requirement for informed consent was waived (IRB number: 2017 − 1144) owing to the retrospective design of our study.

Data Collection
Takotsubo syndrome was diagnosed based on the criteria of the Mayo clinic (6, 10). ECG was performed every day postoperatively during hospitalization. Cardiac enzymes were routinely checked every day until those decreased to a normal range. Transthoracic echocardiography (TTE) was routinely performed at least 4-days after surgery and performed immediately if patients showed unstable vital signs.
Once Takotsubo syndrome was diagnosed, follow-up TTE was performed on a daily or weekly basis, depending on the severity of LV dysfunction. All TTE was reviewed by dedicated cardiologists of our institute.

Statistical Analysis
Statistical analyses were performed using R program (version 3.1.2). Data are expressed as the mean ± standard deviation for continuous variables, with categorical variables reported as a count and percentage. Comparison between the Takotsubo syndrome group and the control group was performed using Student's t-test or the Mann-Whitney U-test for continuous variables and the Chi-squared or Fisher's exact test for categorical variables. For all analyses, a P-value < 0.05 was considered statistically signi cant. Univariable logistic regression analyses were performed to identify risk factors for Takotsubo syndrome in the entire study cohort. Clinical relevant variables with a P < 0.15 on univariable analysis were included in a multivariable analysis, using a backward elimination method to retain predictive variables.

Baseline Characteristics
The baseline characteristics for the Takotsubo syndrome and control groups are summarized in Table 1. The mean interval from the cardiac surgery to the diagnosis of Takotsubo syndrome was 5.0 ± 4.5 days. The mean age was similar in both groups, but with a greater proportion of females in the Takotsubo syndrome than control group (P = 0.08). In the Takotsubo group, the majority of patients (69.2%) had undergone mitral valve operation, compared to 32.7% in the control group (p < 0.05). With regard to intraoperative variables, the time of cardiopulmonary bypass or aortic cross-clamping was comparable between the two groups. Of note, vasoactive inotropic agents, with the exception of vasopressin, were used in a greater proportion of patients in the Takotsubo syndrome than control group (P < 0.05). The length of intensive care unit (ICU) stay and the rate of early mortality were similar in both groups. Troponin-I was increased up to mean value of 108.4 ± 118.8 ng/ml and newly developed electrocardiographic abnormalities (ST-segment changes or T-wave inversion) were documented in 94% (n = 49) of the patients.
In the Takotsubo group, the left ventricular systolic function was recovered to normal values on echocardiography at a mean duration of 7.1 ± 1.3 months after diagnosis (Fig. 1).

Discussion
To the best of our knowledge, this is the rst case-control study to report Takotsubo syndrome after cardiac surgery. We also identi ed that of the patients who developed Takotsubo syndrome, 69.2% had undergone mitral valve surgery. We identi ed atrio-ventricular valve surgery, and the immediate postoperative use of epinephrine or dobutamine as speci c risk factors for Takotsubo syndrome following cardiac surgery. Our ndings are consistent with those previously reported. In the literature, 14 cases of Takotsubo syndrome after cardiac surgery have been reported, with the majority of these cases (n = 10, 71.4%) occurred after mitral valve surgery (2,7,(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22). Like all other surgeries, cardiac surgery can cause emotional and/or physiological stress which may trigger Takotsubo syndrome. In addition, cardiopulmonary bypass needed during cardiac surgery induces a systemic in ammatory response which may also cause coronary microvascular dysfunction and excessive catecholamine release (7). Direct manipulation and incision of the heart also increases its vulnerability to catecholamine-induced cardiac toxicity (21). We note, however, that the reason why Takotsubo syndrome occurs more frequently after mitral valve surgery than other cardiac surgery is unknown.
Several differential diagnoses for Takotsubo syndrome have previously been presented. Of these, papilloannular discontinuity after mitral valve replacement, which leads to a spherical LV and, thus, decreases LV systolic function, should be differentiated from 'transient' Takotsubo syndrome (12,23). Other causes of myocardial stunning after cardiac surgery, such as suboptimal myocardial protection or coronary air embolism, should also be considered as differential diagnoses of Takotsubo syndrome. These differential diagnoses should be identi ed intra-operatively, using transesophageal echocardiography, based on the characteristic echocardiographic nding of Takotsubo syndrome, namely apical ballooning that extends beyond a single coronary territory (15).
In 2008, Takotsubo syndrome, induced by pharmacological stress (epinephrine infusion), was rst described by Wong et al. (11). In the following year, Abraham at al. reported that exogenous catecholamines and beta-receptor agonists (epinephrine, dobutamine), like endogenous catecholamines, could induce Takotsubo syndrome (24). Previous ndings that excessive release of catecholamine may cause Takotsubo syndrome (3,4) are convincing and are consistent with our own results that immediate postoperative use of epinephrine and dobutamine as risk factors for Takotsubo syndrome following cardiac surgery. Previous reports on the occurrence of Takotsubo syndrome after cardiac surgery indicated that, in most cases, the syndrome developed several minutes to several days after the release of cross-clamping (2, 7, 11, 12, 14-22). Of note, Ohata et al. did report on a case of Takotsubo syndrome that developed 1-month after aortic valve replacement in a 70-year-old female (13). In our study, intra-Conclusions operative transesophageal echocardiography was performed in all cases, with TTE routinely performed at least 4 days after surgery. We do note that in the case of unstable vital signs, or if the patient complained of chest discomfort, TTE was immediately performed. The median interval between the cardiac surgery and the diagnosis of Takotsubo syndrome in our study cohort was 4.0 (Interquartile range, 3.0-5.0) days.
Therefore, our imaging protocol would have been appropriate to detect Takotsubo syndrome in most cases.
Takotsubo syndrome has been known to occur mostly in elderly women (25). Ueyama et al. reported that, in rats, the use of estradiol might be protective against LV dysfunction induced by emotional stress (26).
In the present study, there were more female patients in Takotsubo group (75.0%) than in control group (59.6%), although this difference was not signi cant (P = 0.08). However, female sex was identi ed as a risk factor for Takotsubo syndrome after cardiac surgery on univariable analysis, but was not retained as an independent risk factor on multivariable analysis.
Characteristically, Takotsubo syndrome is a transient heart failure syndrome (27), with a mortality rate of 3.2% having been reported in a previous case-summary study (25). The mortality rate of Takotsubo syndrome after cardiac surgery in our study cohort was 3.8% and, thus, was comparable to previously published data. Of note, the mortality rate in the Takotsubo group (3.8%) was similar to that in the control group (7.7%; P = 0.58). Madias reported a low prevalence of diabetes mellitus among patients with Takotsubo syndrome (28). Based on this nding, Madias suggested that diabetic autonomic neuropathy and decreased catecholamine release might have protective effect against Takotsubo syndrome. Madias did report that the prevalence of hypertension in patients with Takotsubo syndrome was comparable to that in the general population. In contrast, in our study, the prevalence of hypertension was lower in the Takotsubo syndrome than control group, while the prevalence of diabetes mellitus was similar between the two groups. In our multivariable analysis, hypertension was identi ed as having a protective factor against Takotsubo syndrome after cardiac surgery. It is possible that anti-hypertensive medications, such as beta-receptor antagonists, taken before cardiac surgery may lower the risk of Takotsubo syndrome following cardiac surgery. However, data about anti-hypertensive medications prescribed to the patients in our study cohort by their primary-care physician could not be accurately identi ed and, thus, analysis of the plausible protective role of anti-hypertensive medications was not possible.

Limitations
This study is a retrospective observational study of patients enrolled from a single center. Although we did use a case-controlled design, in the absence of randomization, effects of selection or detection bias cannot be denied. In addition, due to the retrospective design of the study, the causal relationship between the development of Takotsubo syndrome and the use of epinephrine or dobutamine could not be determined. Finally, we could not completely exclude confounding effect of medications taken preoperatively, a factor which should be assessed in subsequent, larger, cohort studies.
Takotsubo syndrome occurred in about 1% of patients who underwent cardiac surgery under cardiopulmonary bypass. Immediate postoperative use of epinephrine and doputamine, as well as atrioventricular valve surgery were factors associated with the development of Takotsubo syndrome after cardiac surgery. The resulting left ventricular dysfunction was reversible in the majority of patients, without an increase in the rate of mortality. The ndings of our study will need to be veri ed in a larger, multi-center, prospective studies.

Consent to publication
Not applicable.

Availability of data and materials
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.