We conducted an ambispective cohort study reviewing patients with a definite diagnosis of TTS included in the Spanish National Takotsubo Registry (RETAKO Registry) between June 2002 and March 2017. Patients were included prospectively from 2012 onwards, and were allowed to be included retrospectively. The first patient included was discharged in June 2002 and the last in March 2017. Inclusion criteria and methodology are described in depth in the original manuscript.3

Briefly, inclusion criteria were as per Modified Mayo Criteria22: (1) akinesia or dyskinesia of the apical and/or midventricular segments of the left ventricle with regional wall motion abnormalities that extended beyond the distribution of a single epicardial vessel; (2) absence of angiographic obstructive coronary artery disease, which was defined as a luminal obstruction of more than 50%; (3) new electrocardiographic abnormalities together with cardiac troponin elevation; (4) absence of pheochromocytoma or myocarditis; and (5) complete recovery of left ventricular function during follow-up.

The study protocol recommended an electrocardiogram at admission, at 24 and 48 hours; an early echocardiogram; and blood tests, including initial and peak troponin. Study protocol also included a three-month clinical and echocardiogram follow-up and at least a 12-month clinical follow-up. Follow-up data were allowed to be continued past the 12-month mark. All other clinical decisions, including treatment and timing of tests and discharge were left to the discretion of the treating physician.

Data were collected in a dedicated online data input application and included: baseline characteristics; past medical history; relevant blood tests; electrocardiographic, echocardiographic and angiographic findings, and medical treatment before, during and after hospitalization.

The protocol was submitted and received approval of the Medical Ethics Committee at the Hospital Clínico San Carlos in Madrid, Spain. The study was conducted in compliance with the Declaration of Helsinki and applicable local requirements. A written informed consent was obtained prior to the inclusion in the database.

The objective of this study was to determine the prognostic impact of APT at discharge at 12-month mandatory follow-up and beyond. The null hypothesis was that being discharged on APT was not related to survival free of the primary endpoint of death; or survival free of the secondary combined endpoint of recurrence or readmission to a cardiology ward; or survival free of the secondary combined endpoint of death, recurrence or readmission.

Antiplatelet therapy status at discharge (APT cohort) was defined as a patient treated with aspirin and/or any P2Y12 inhibitor at discharge for an index hospitalization for TTS. Patient not receiving APT at discharge (non-APT cohort) was defined as a patient reported as not being prescribed aspirin or any P2Y12 inhibitor at discharge. All-cause death during follow-up was defined as the event of death being reported during the follow-up period. Recurrence of TTS was defined as the reappearance of similar symptoms reported during the follow-up period3. Readmission was defined as a readmission in a cardiology ward reported during the follow-up period.3 At least one echocardiogram record after discharge was deemed necessary to achieve survival status, even if further follow-up was not available.

Chronic kidney disease was defined as baseline serum creatinine >1.5 mg/dL.3 Mitral regurgitation was considered greater than mild when graded II-IV.3 Based on prior literature, advanced age was defined as an age >75 years for the multivariate analysis. Other continuous variables were transformed in dichotomous variables by using the third quartile of the distribution as a cut-off point.

Statistical analyses were performed using SPSS Statistics 20.0 software (SPSS Inc., Chicago, IL, USA). Continuous variables were explored for normal distribution with the Kolmogorov-Smirnov test. Variables following normal distribution were expressed as mean (standard deviation) and non-normally distributed variables were expressed as median (inter-quartile range (IQR)). Continuous variables were also explored using box plots to identify extreme values leading to data entry errors. Categorical variables were expressed as count (percentage). Comparisons between continuous variables were performed with t-student or U-Mann-Whitney tests as appropriate. Comparisons between categorical variables were performed with the chi-square or Fisher's exact tests as appropriate. All p values were two-tailed, with statistical significance set at a level of <0.05. Cases with missing values were excluded from analyses.

To evaluate the association between the exposition variable (APT at discharge) and the primary and secondary endpoints, we conducted a multivariate Cox survival regression model. Also, the results were plotted using Kaplan-Meier survival curves and compared with the log-rank test. To avoid biases derived from unequal follow-up in comparison groups, we tested the proportional hazards assumption with Kaplan-Meier survival curves and Cox regression models using time dependent covariables for the exposition variable. If hazards were not proportional over time for the exposition variable, we then identified the landmark in time in which the change in hazards occurred and expressed the hazard up to this point. In the first step and in order to adjust for possible confounders, exploratory univariate analyses, evaluating all variables in the study, were performed using the log-rank test. Variables with p values <0.05 were included in a backward stepwise Cox survival regression model. Results were reported as hazard ratios (HR), together with the 95% confidence intervals (95% CI) and p values.

Although TTS pathophysiology is incompletely understood, mechanisms involved in TTS are multiple. In most cases psychological or physical stress trigger the cascade of events in TTS.3,4 It has been observed that an exaggerated response of the sympathetic system to stress generates an intense catecholamine drive that is related to myocardial stunning and an increment in platelet activation.5–8 Also, the elevation of catecholamine levels is associated with a worsening of endothelial function, showing an increment in endothelial dysfunction in TTS patients as compared to a matched control group.14

Some studies have shown that catecholamine stimulation causes increased platelet activation and up-regulation of fibrinogen receptors which is an essential step in platelet aggregation,9,12 and it has been demonstrated that patients admitted for a TTS acute event presented higher blood levels of catecholamines compared to patients presenting with an ACS.7,15 Hoever, this platelet activation may not be exclusively limited to the acute phase. In one observational study exploring platelet function during the acute phase and at follow-up, Nuñez-Gil et al. demonstrated that TTS patients, who had higher levels of adrenaline in the acute phase, displayed increased platelet reactivity at three months.15 In addition, these biological findings (platelet reactivity and endothelial dysfunction) could have a clinical translation, as highlighted by the fact that several studies have shown that the incidence of thrombotic events during follow-up in TTS were comparable to ACS.3,4 Therefore, the prolongation of APT with aspirin or PY212 inhibitors could be a therapeutic option in TTS long-term management. Also, this is reasonable since the clinical profile of these patients points to an important atherosclerotic risk.

To date, the evidence supporting long-term treatment for TTS is poor due to the lack of randomized clinical trials, and only treatment with ACEI or ARB has been associated with a reduction in TTS recurrences in a meta-analysis of observational studies.23 This means that there is no consensus regarding long-term management in this population.24

Despite pathophysiological data relating platelet activation to events in TTS, no study has demonstrated clinical benefit in extending APT in the long-term. A recent meta-analysis,18 including observational studies,17,19–21 observed that APT in the long-term increases all-cause mortality and a composite outcome of all-cause death, TTS recurrence, stroke/transient ischemic attack or myocardial infarction/coronary artery disease progression. Nevertheless, the conclusions of this meta-analysis should be interpreted with caution. First of all, the study conducted by D’Ascenzo et al.,21 which accounts for more than 90% of the weight for overall mortality and for the combined endpoint, could have conditioned the results. Likewise, it is striking that, although statistical significance was not reached in the meta-analysis, the probability stroke/transient ischemic attack or myocardial infarction/coronary artery disease progression was higher in the group of patients treated with APT, taking into account that these patients present a high proportion of cardiovascular risk factors and an elevated cardiovascular risk.

In our research, the vast majority of patients received simple APT with aspirin and only a minority of patients received P2Y12 inhibitors or dual APT, showing reduction in mortality, recurrence or readmission. In line with previous data reported by Dias,16 showing a reduction in major cardiovascular adverse events during hospitalization, our results would support lengthening antiplatelet therapy up to 25 months, mainly with aspirin. It is noteworthy that, in contrast to the work of D’Ascenzo et al.,21 in which the follow-up reached five years, our data show that the optimal prolongation for APT could be around two years. This leads us to hypothesize that there could be a temporary period after the presentation of a TTS in which APT could offer the greatest benefits, minimizing the risks.

Firstly, this study is an observational study with inherent biases. Our data refer to the population of Spain and therefore should be considered carefully before being extrapolated to other geographical areas. Secondly, the clinical criteria for prescribing APT at discharge are unknown. Patients prescribed APT at discharge also received other heart failure-like medications, which might indicate that some clinicians preferred treating TTS based on the evidence of ACS and heart failure. Thirdly, patients in the APT cohort had a better clinical profile than the non-APT cohort. However, after multivariate analysis, APT remained an independent factor for good prognosis. Fourthly, time to follow-up was uneven between cohorts and there was no data on adherence to APT. However, the highest proportion of patient losses occurred in the non-APT cohort, which could penalize and underestimate the beneficial effect of APT. Fifthly, bleeding and other APT-related complications were not specifically checked during follow-up. This might affect the net clinical benefit favoring the APT group.