Residual alterations of cardiac and endothelial function in patients who recovered from Takotsubo cardiomyopathy

Abstract Introduction Takotsubo cardiomyopathy (TCM) is characterized by transient left ventricle dysfunction. Hypothesis A residual cardiac and endothelial dysfunction is present in patients who recovered from TCM. Methods In this single‐center prospective study, patients with prior TCM were included and followed for 6.4 ± 1.6 years. All underwent comprehensive cardiac function assessment, including tissue Doppler imaging (TDI) and 2‐dimensional strain (2DS) echocardiography at their first visit. The number of circulating endothelial progenitor cells and levels of proangiogenic vascular endothelial growth factor (VEGF) and its receptor (VEGF‐R) were measured. All measurements were compared with healthy controls. Results Forty‐two women (age 58. ±8.6 years, LVEF 58.1 ± 6.1%) comprised the TCM group. Patients post‐TCM had significantly lower early velocities E′ (6 (5.0–8.0) vs. 9 (7.0–11.0) cm/s, p = .001) by TDI and higher E/E′ ratio (p = .002), lower LV global average longitudinal strain (LGS) (−18.9 ± 3.5% vs. −21.7 ± 2.3%, p = .002) and RV LGS (−20.1 ± 3.9% vs. −23.4 ± 2.8%, p = .003) were evident. There was a trend toward a higher VEGF‐R (p = .09) along with decreased VEGF/VEGF‐R ratio representing inadequate VEGF production. In‐hospital mortality was not reported and only two non‐cardiac deaths occurred at long‐term follow‐up. Conclusions Altered TDI and 2DS indices suggest residual biventricular myocardial injury in post‐TCM patients with the apparent LV function recovery. Inappropriate production of VEGF and VEGF‐R were observed, suggesting a possible underlying endothelial dysfunction in these patients.


| INTRODUCTION
Takotusbo cardiomyopathy (TCM) is an acute syndrome occurring mainly in postmenopausal women. 1 It is induced in the vast majority of cases by emotional or physical stress. It is characterized by transient left ventricle (LV) wall motion abnormalities with no evidence of coronary artery disease. 1,2 The pathogenesis of TCM is not completely understood and several mechanisms have been proposed including catecholamine-induced myocardial stunning, coronary spasm and microvascular dysfunction. 3,4 Recent studies have suggested that TCM is associated with endothelial dysfunction (ED) representing an important link between stress and transient myocardial dysfunction. [4][5][6] Endothelial progenitor cells (EPCs) play an important role in vasculogenesis and endothelial homeostasis. 7 Different mobilization of EPCs has been described in patients with dilated, hypertrophic and peripartum cardiomyopathy compared with healthy individuals. [8][9][10][11] No published data exists on endothelial function biomarkers in post TCM patients.
The incidence of TCM is underestimated due to under-diagnosis of self-limiting transient cardiac dysfunctions, account for up to 2% of patients with acute coronary syndrome. 12 Data on TCM recurrence is limited. 13 Five-year recurrence varies from 5% to 22% 14 and up to 20% at 10 years. 15 InterTAK registry showed a comparable long-term mortality risk with acute coronary syndrome patients. 16 Pellicca et al.
showed 3.5% annual rate of total mortality, and 1.0% annual rate of recurrence during a median follow-up of 28 months. 3 Few data have been published on residual changes in cardiac function of patients with prior recovered TCM. Only a few studies describe residual abnormalities using speckle tracking echo imaging and magnetic resonance imaging (MRI) showing systolic and diastolic deformation abnormalities that may persist despite normalization of global LV function at follow-up of 4-20 months. [17][18][19] In our study, we evaluated endothelial function measures and possible residual myocardial injury by comprehensive echocardiographic techniques in patients with apparent cardiac function recovery post-acute TCM.

| Patient population
Forty-two women with prior TCM were followed at our heart failure clinic and were enrolled at their first visit. All met the accepted diagnostic criteria for TCM according to the Mayo Clinic and the European Society of Cardiology-Heart Failure Association criteria. 14,16,20 All underwent a comprehensive evaluation of the LV and RV systolic and diastolic function assessment by echocardiography, tissue Doppler imaging (TDI), and two-dimensional speckle tracking imaging (2DS). At six-month follow-up, patients underwent stress echocardiography to assess LV contractile reserve. Blood samples were collected from 36 patients for circulating EPCs (CD34+, CD34 +/KDR), VEGF, and VEGF-R at time of enrollment.
All measures were compared with data from the control group that included healthy females, carefully chosen to be as close as possible in terms of age, comorbidities and BMI ( Table 1).
The study was approved by our institution's medical ethics review board and informed consent was obtained from all patients for being included in the study.

| Echocardiography
Echocardiographic measurements were obtained according to the principles described in the Recommendations for Chamber Quantification. 21 Ejection fraction measurements were based on biplane Simpson's method. Peak velocities of early (E) and late (A) diastolic filling and deceleration time were assessed by pulse-wave Doppler.
Early (E 0 ), late (A 0 ) and systolic (S 0 ) diastolic velocities were assessed by TDI. Contractile reserve by stress echocardiography (Bruce protocol) was defined as an increase in LVEF ≥5% at target heart rate. The heart rate recovery (HRR) was calculated as the difference between peak heart rate and heart rate at the first minute of recovery.

| 2DS analysis
For 2DS analysis, the LV and RV myocardium images were obtained with a frame rate > 50 Hz. 2DS and strain rate (SR) were performed by offline semiautomatic analysis. Raw data were stored digitally as  short-axis view was measured at the mitral valve and apical levels. RV longitudinal strain was assessed from the free wall; and the base, mid and apical segments were averaged.

| Isolation of EPC's by florescence-activated cell sorting, VEGF and VEGF-R ELISA
In all patients, the number of CD34+, CD34+/KDR+ cells were quantified and VEGF and VEGF-R (sFlt1) levels were obtained in the post-TCM women and controls.

| Isolation of peripheral blood mononuclear cells
Human peripheral blood mononuclear cells were isolated from whole blood by a ficoll gradient. Peripheral blood mononuclear cells were taken for fluorescence-activated cell sorting (FACS) staining, and plasma was removed and stored at À80 C for an enzyme-linked immunosorbent assay (ELISA).

| Statistical analysis
After evaluation of normality by Kolmogorov-Smirnov tests, continuous variables that were normally distributed are presented as mean and SD. Continuous variables that were not normally distributed are presented as median (25th-75th percentile). A two-sided t-test was performed to determine the statistical significance of differences in parameters between women with post-TCM and controls, and a Fisher exact test was used for categorical variables. A p value <.05 was considered significant. Original data were also compared between the two groups using a nonparametric Mann-Whitney U test.
Pearson's correlation coefficient was used for correlation analysis.
Statistical analysis was performed using IBM SPSS version 21.0 (Armonk, NY).

| RESULTS
Forty-two women (mean age 58.8 ± 8.6 years) with TCM were enrolled at their first visit after their index event in the outpatient heart failure clinic at 3.5 ± 2.8 months from the acute presentation.
The demographic characteristics of these patients at time of enrolment are presented in Table 1. Thirty-six women with similar age (56.3 ± 8.4 years), comorbidities and medications comprised the control group.

| Clinical presentation at the index event
Clinical characteristics and complications at TCM presentation are presented in Table

| Outcomes at long-term follow-up
All post-TCM patients were followed at our heart failure clinic (clinical and echocardiography) every 6 months after their acute presentation.
Mean follow-up was 6.9 ± 1.6 years (median 5.7, max. 10 years min. 2 years). Fifty-one percent had a follow-up of more than 5 years.
No stroke, heart failure, death or TCM recurrence, occurred during the first 30 days after admission. No cardiovascular mortality occurred and only two patients died, one from respiratory failure due to severe COPD and the other due to sepsis. Recurrence of TCM occurred in three women, 1, 4 and 6 years after the index event.

| DISCUSSION
In our study, we demonstrated that patients with prior TCM had residual biventricular cardiac dysfunction, including lower LV early

| Residual cardiac impairment
Available data suggest that LVEF assessment by standard echocardiography may not be sensitive enough to mirror complete recovery of LV function. TDI and 2D tracking imaging more accurately assesses both LV and RV function. The 2DS technique for quantification of myocardial strain, provides data on longitudinal and circumferential myocardial function and rotation. 23 These deformation indices were reported to be sensitive indicators of subtle changes in LV function. 24,25 In our study we found that patients with prior TCM had signifi- Contrary to these reports, supporting our findings, Neil et al.
showed that in 36 TCM patients, recovery of the LVEF was obtained within 3 months and GLS, SR, and apical twist improved significantly.
However, they found that GLS remained reduced compared to controls even at the three-month follow-up. 28 Subsequently, this group compared 52 patients with TCM to 44 healthy control subjects using MRI. 18 At 4-months after the TCM acute episode, there were still significant GLS impairment and apical circumferential strain (p<0.01), but with no differences in TDI parameters compared with healthy controls despite LVEF normalization.
Nowak et al. investigated LV function recovery at 6 months after discharge among TCM patients. 17 These investigators described only residual abnormalities in LV apical rotation, significantly lower mean systolic apical rotation, mean peak early diastolic rotation rate compared to the healthy controls, and no difference in diastolic parameters on TDI and GLS. Recently, Scally et al. compared 37 patients with prior (>12-month) TCM after 20 (range 13-39) months to 37 age-sex and comorbidity-matched controls. 19 This study showed that despite normal LVEF and serum biomarkers (BNP), patients with prior TCM at a median time of 20 months from the index event had impaired cardiac deformation indices by means of both reduced apical circumferential strain.
RV involvement has been described in 24% of patients at acute presentation of TCM, usually in patients in whom the LVEF was significantly lower. 29 However, serious complications such as cardiogenic shock and pulmonary edema were not more frequent in these patients.. 30

| Endothelial dysfunction in prior TCM patients
The precise mechanisms leading to TCM remain unclear. An endothelial imbalance that may interplay between emotional or physical stress and myocardial dysfunction in TCM may be involved in the pathophysiology of this syndrome. 5,6,33 This hypothesis was evaluated by Naegele M et al. 5 Endothelial dysfunction was measured on the bra- EPCs are peripherally increased in response to any vascular injury by mobilization of cytokines such as VEGF and others, and are potential biomarkers for cardiomyopathies. 34 EPCs elevation was described in congestive heart failure and the severity was associated with elevation in the early phase and depression in the advanced phases of the disease. 10,11 In our recently published study on women who recovered from peripartum cardiomyopathy, we found significantly higher sFlt1 (VEGF-R) levels with a clear trend toward lower circulating CD34+/KDR+ levels compared to healthy controls, suggesting an endothelial imbalance. 9 In the current study, we found a trend for

| Clinical assessment and outcomes
The prognosis and recurrence rate has been recently shown to be worse than initially thought. Recently, a high in-hospital mortality has been reported up to 5%. 15,36,3737 A number of reports describe a relatively high long-term mortality and recurrence rates. Pelicca et al.

| CONCLUSION
Residual alteration of cardiac function by tissue Doppler and speckletracking imaging were found in patients recovered from TCM despite normalized LVEF. Inappropriate production of VEGF and VEGF-R were observed even after LV recovery in patients with Takotsubo cardiomyopathy, suggesting a possible underlying endothelial dysfunction in these patients. Future larger studies are needed to evaluate residual endothelial dysfunction in this patient population.

| LIMITATIONS
The main limitation of our study is a relatively small number of patients including females only. Therefore, the differences in endothelial function biomarkers did not reach statistical significance and showed only a trend compared with controls. Despite it, all underwent an extensive workup and were followed more than 5 years and we were able to report complete recurrence and mortality data.

ACKNOWLEDGMENTS
We thank Estela Derazne (Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv) for additional biostatistical analysis and her helpful comments.