ABC | Volume 114, Nº3, March 2020

Original Article Campos et al. Takotsubo syndrome recurrence Arq Bras Cardiol. 2020; 114(3):477-483 The prognosis is usually good and characterized as benign by many authors, even though there is a 1-2% risk of ventricular arrhythmias and approximately 2% of in-hospital mortality associated with TTS. 5 Patients with a history of TTS have an annual recurrence rate of 1.5%, though it may be as high as 11% in 4 years. 3,5,6 In recent years, there has been an increase in the number of published TTS-related studies, especially in the USA, Europe, and Japan. Much of the data on this pathology comes from the International TakotsuboRegistry (InterTAKRegistry), an international collaborative network with data from35 cardiovascular centers in 15 different countries. 2,7 However, predictors of TTS recurrence are still not well established. Objective The present study aimed to analyze the main factors associated with TTS recurrence. Methods A systematic review of the literature was proposed, using the PRISMA model. The main databases of international literature – PubMed, Scielo, Lilacs, and Cochrane – were searched. As a search strategy, the following descriptors were used: Takotsubo Syndrome; Left Ventricular Apical Ballooning Syndrome; Takotsubo Cardiomyopathy; Stress Cardiomyopathy; Broken Heart Syndrome. The PubMed MeSH tool was used adding Recurrence as a complementary descriptor. Strategy for articles selection: the selection was carried out in October 2017. All articles published between January 2008 and October 2017 were initially included for further appreciation. First, the titles were evaluated, followed by the abstract, and finally, a careful analysis of the complete article was conducted in order to identify its quality and relevance to the proposed objective. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) criteria 8 for assessing the methodological quality of observational studies were used, with a minimum of 80% compliance of the 22 items in the STROBE checklist to be included in this study. This whole process was carried out by two researchers. Articles were also searched based on the references of the selected articles. Only articles characterized as prospective cohort study, retrospective cohort study, case control or case series study were selected. Articles in which the text was not in English were excluded. Only studies reporting at least 3% recurrence rate and/or five or more patients with recurrence were included, so that there could be a significant analysis of the recurrence predictors. Results Initially, a total of 164 articles were identified. Four other studies were identified and selected from the references of the initially identified articles. At the end of the analysis of the studies, six were selected to compose this systematic review (Figure 1 and Table 1). Globally, the recurrence rate, before excluding the four articles that did not meet the criteria of at least 3% recurrence rate and/or five or more patients with recurrence, ranged from 0.2 to 5% per year. The global recurrence rate, considering the selected studies, was 3.8% in a follow up that ranged from 5 to 17 years. Table 2 shows the main information for each selected article. Looi et al. 9 studied a prospective cohort study of 100 patients diagnosed with TTS by the Mayo criteria. From these, seven patients (7%) had a recurrence and one presented with four recurrent episodes. Recurrences occurred between 99 and 679 days after the first episode. All recurrences occurred within two years after the first episode, being more frequent in the first year. In four of the seven patients who presented with recurrences (57%), the initial and subsequent events were triggered by emotional stress. Four of the seven patients who had recurrences were already using a BB in the second episode. Templin et al. 3 presented a case-control study with 1,750 TTS patients, according to theMayo criteria. From these, 455 patients were matched, by age and gender, with patients diagnosed with acute coronary syndrome (ACS) andwho had their data obtained through the Zurich Acute Coronary Syndrome Registry. During a 17-year follow-up period, 57 patients with TTS had recurrences, representing a rate of 1.8% recurrence per patient‑year. The second episode occurred from 25 days to 9.2 years after the first one. A total of 29 of 57 patients with recurrences (50.8%) were on BB therapy at the time of their recurrent episode. In the retrospective cohort study by Patel et al., 10 224 patients diagnosed with TTS had their data obtained through the Mayo Clinic database over a 10-year period. Only 7 recurrent episodes were documented. None of the men had TTS recurrence. During a mean follow-up of 3.5 years, 2 women under 50 years of age (16%) and 5 women aged 50 years old or older (3%) developed TTS recurrence (p = 0.017). Elesber et al. 11 studied a retrospective cohort and analyzed data from 100 patients diagnosed with TTS over a period of 16 years and 11 months. Recurrence rate was 11.4% at a mean follow-up time of 4.4 ± 4.6 years, being higher in the first 4 years (2.9% per year), and decreasing to about 1.3% per year in subsequent segment time. There was no difference between patients with or without recurrences in relation to the use of: aspirin (60 versus 67%; p = 0.67); angiotensin converting enzyme inhibitor (ACEI)/angiotensin II receptor blocker (ARB) (60 versus 51%; p = 0.59); BB (80 versus 52%; p = 0.10); or statins (40 versus 33%; p = 0.67). In another prospective cohort study, 12 23 patients who underwent coronary angiography were diagnosed with TTS according to the Mayo criteria, over a period of 7 years and 11 months. Five patients (21.7%) developed recurrent TTS and one patient presented with 2 recurrent episodes. The mean time to a recurrent episode was 105.4 ± 83 days, and the recurrence rate was higher in the first 3 months. Compared with patients with no recurrences, those with a recurrent episode were older (71.4 versus 65.7 years), had lower ejection fraction (36.5 versus 44.2%), higher systolic blood pressure (139 versus 128.4 mmHg) and higher peak troponin levels (8.1 versus 2.5 μg/ml). Three of the five patients who presented with a recurrence were on BB. 478