ABC | Volume 115, Nº2, August 2020

Original Article Almeida Junior et al. Takotsubo Multicenter Registry (REMUTA) Arq Bras Cardiol. 2020; 115(2):207-216 In relation to triggering factors, emotional stress was not present in the majority of patients. In the Intertak registry, the largest registry of TTS published to date, 5 the rate was 27.7%, showing that the absence of an emotional factor preceding the clinical manifestation absolutely does not exclude this diagnosis. Furthermore, TTS preceded by physical stress generally has a secondary cause and a worse prognosis. BNP was shown to be elevated in our population. In the Intertak registry, 5 the average value was 6 times the cutoff limit for the test. These values are greater than those observed in patients with acute coronary syndrome but lower than those of the general population with decompensated heart failure, such as in the BREATHE registry where it was 1075 (518; 1890). We observed a lower prevalence of STE than the 43.7% seen in the Intertak registry. 5 On the other hand, out study had a higher rate of STD (7.7% in the Intertak registry). A multicenter Japanese registry of TTS showed an elevated rate of STE of approximately 74% and negative T waves in 70% of cases. These data show that typical ischemic changes may be absent in 25% to 70% of cases. The degree of ventricular dysfunction, as reflected by LVEF, was equal to that observed in the Intertak registry 5 (41% ± 11.8%). The mid-apical pattern was by far the one most found, which is in consonance with the literature. It is noteworthy that, in our registry, the biventricular pattern held third place. This is not well described in other studies on TTS, and it shows that we should pay more attention to the assessment of the RV in this pathology. Non-negligible rates of complications such as moderate to severe mitral regurgitation, pericardial effusion, intra-ventricular thrombus, and left ventricular outflow tract obstruction were observed, showing that ventricular dysfunction is not the only problem and that cardiac involvement may by more complex in some cases. Another point worth underscoring is that practically one third of our patients were discharged from the hospital without an improvement in ventricular function on pre-discharge control echocardiogram. It is worth underlining that, although, by definition, dysfunction ventricular is reversible in TTS, there is no specific time for this improvement. In our sample, median length of hospital stay was 7.5 days. This population should receive closer outpatient follow-up in order to verify if longer ventricular function recovery time has a prognostic impact. The Swedeheart study, which evaluated 302 patients with TTS, found cardiogenic shock in 5% of cases and cardiac arrest in 3%, while these rates were much lower in our registry. The use of inotropic and diuretic medications was also 7% and 20%, respectively, in the Swedeheart study, 9 which were much lower than the rates observed in our registry, showing, once again, the much greater severity in our cohort. Mortality in the Swedeheart study 9 was 4% in 30 days; in the Japanese registry, it was 6.3% during the in-hospital period, and in the Intertak registry 5 it was 5.6% over 1 year. In addition to the significantly higher mortality in our study, in comparison with international studies, we also observed a high rate of complications such as shock, acute pulmonary edema, need for invasive and non- Regarding in-hospital clinical evolution, we observed that 40.5% of patients required non-invasive mechanical ventilation, and 25.7% required invasive mechanical ventilation. Acute pulmonary edema was observed in 24.1% of patients, and circulatory shock was observed in 17.4%. Ventricular arrhythmia was present in 8.5% of patients; cardiorespiratory arrest was present in 12.7%, and mechanical circulatory support was used in 7.7% of cases (Figure 3). In-hospital mortality was observed to be 10.6%, and mortality after 1 year was 16.5% (Figure 4). Only 1 patient with primary TTS progressed to in-hospital death (0.91%), in contrast with 17 patients in the secondary TTS group (28.3%). Table 4 shows univariate analysis of clinical predictors and complementary exams with their statistical significance. For the variables with non-normal distribution (troponin, BNP, and ejection fraction), we used the Mann-Whitney test, and only ejection fraction calculated by the Teichholz method showed significant different between the death and survival groups (p = 0.001). In multivariate analysis of predictors of death (forward stepwise logistic regression), we observed that secondary TTS (p = 0.035 and OR: 4.5) and cardiogenic shock (p < 0.001 and OR: 13.2) were independent predictors of mortality, while the presence of chest pain was a protective factor (p < 0.011 and OR: 0.14). Table 5 shows this analysis. The survival curve of these predictors is shown in Figure 5. Discussion This study is the first multicenter registry of TTS in Brazil. The most important observations after analysis of data were the following: 1) The majority of clinical and epidemiological characteristics are similar to those in international registries, namely, predominantly elderly women, with chest pain and dyspnea as the most prevalent symptoms; 2) Emotional stress was found in only 38% of cases; 3) We observed an elevated rate of secondary TTS; 4) In-hospital mortality was elevated, as was mortality after 1 year of follow-up, and 5) Secondary TTS and shock were independent predictors of mortality, while chest pain was a protective factor. Table 3 – Echocardiographic variables Variable (n) Result LVEF Teichholz (143) (median/IQR) 48 (40; 62) LVEF Simpson (87) (median/IQR) 40 (35; 48) Moderate to severe MR (167) (%) 6.6 LV or RV thrombus (167) (%) 3.0 Pericardial effusion (167) (%) 4.8 LVOT obstruction (166) (%) 4.2 Reversal of LV dysfunction (132) (%) 68.2 LV: left ventricle; LVEF: left ventricular ejection fraction; LVOT: left ventricular outflow tract; MR: mitral regurgitation; RV: right ventricle 210