*3.2. Cardiovascular Hospitalizations*

From the 51 patients hospitalized for cardiovascular reasons during the follow-up, 15 were hospitalized more than once.

More than 40% of all cardiovascular-related hospitalizations occurred within the first year after the first diagnosis (Figure 4).

**Figure 4.** Time free from hospitalization for cardiovascular reasons according to the Kaplan-Meier curves.

EF at discharge and the heart rate at discharge resulted as independent predictors of cardiovascular-related hospitalization according to the multivariate Cox regression model (Table 4 and Figure S1c,d).

**Table 4.** Multivariable Cox-proportional hazard model for cardiovascular hospitalization.


The univariate model included: The male gender, age, body mass index, hypertension, diabetes, dyslipidaemia, chronic kidney disease, chronic obstructive pulmonary disease, OSAS, hyperthyroidism, hypothyroidism, type of arrhythmia, NYHA class on admission and at discharge, heart rate on admission and at discharge, BNP on admission and at discharge, Troponin I on admission and at discharge, LVEF on admission and at discharge, iLAV, rate or rhythm control (dummy variable) and pharmacological treatment at discharge (with each drug class from Table 2 considered as a separate variable). The complete model is detailed in Supplementary Table S3.

BNP: brain natriuretic peptide; CI: confidence interval; iLAV: indexed left atrial volume; LVEF: left ventricular ejection fraction; NYHA: New York Heart Association; OSAS: obstructive sleep apnoea.

Comparing TIS for each drug class considered in Table 2 showed no differences between the patients with and without cardiovascular hospitalizations (all *p* > 0.05).

#### **4. Discussion**

The main message of the present study is that, while TCM is associated with an overall good prognosis, TCM patients do recur over a long-time follow-up.

This represents an outstanding difference between HF patients and pure TCM patients as the former is known to be a progressive, worsening condition commonly culminating in the patient's exitus. Although under-recognized, this study shows that almost 10% of all hospitalizations for acute HF meet the diagnostic criteria for TCM. Therefore, early recognition of the possible triggering arrhythmia is of paramount importance as it can lead to treatment strategies which can favour patient recovery. A clinical suspicion of TCM should arise in all patients presenting with new and quickly worsening symptoms of HF, a low overall cardiovascular risk profile and the recent evidence of high-rate arrhythmia. In these cases, a prompt reduction of the heart rate (either through rate control drugs or restoration of sinus rhythm) should be performed as soon as possible, and possibly even while the diagnostic workup for the exclusion of structural heart disease is still in progress. A cardioversion attempt should be made (when feasible) in order to prevent further deterioration of the systolic function and catheter ablation should be taken into serious consideration [10]. Moreover, in these patients, a sleep study and polysomnography should be performed as soon as possible, even during the hospital stay, as potentially able to unmask OSAS.

The rate of TCM recurrence is higher between the fifth and the sixth year after diagnosis. It can only be speculated that this could be due to the natural progressive reduction of the patients' adherence to treatment over time.

Our multivariate analysis found three major independent predictors of TCM recurrence. The most important was a concomitant diagnosis of OSAS, which increased the risk of recurrence 5-fold. It was hypothesized that this could be related to the fact that OSAS can alter the physiological parasympathetic modulation of the heart during sleep leading to sympathetic excitation and favouring ventricular and atrial ectopic beats [16,17]. Moreover, OSAS has been described to be an independent risk factor for AF and has been shown to decrease the success rate of antiarrhythmic drugs, electrical cardioversion and catheter ablation [18], potentially leading to TCM recurrence. Despite the lack of information regarding the actual adherence to non-invasive ventilation, it is noted that half of the patients with OSAS and TCM recurrence were not treated with continuous positive airway pressure at all. Therefore, it appears important to educate patients affected by OSAS on the importance of non-invasive ventilation while offering the best treatment strategy in order to improve long-term compliance.

Another striking result that warrants discussion is that the heart rate at discharge is associated with an increased risk of TCM recurrence. More precisely, for each increased beat per minute, the risk of recurrence increases by 5%. Moreover, this association proved to be linear, at least within the ranges of the heart rate seen in our population, and holds true independently of the rhythm at discharge, the treatment strategy and the class of medications used. To make an example, a patient with a lenient rate control strategy (110 bpm) has a 2.5-fold risk of TCM recurrence when compared to the same patient undergoing a strict rate control strategy (80 bpm). This is in contrast to the known evidence that both the heart rate targets are considered similarly effective in preventing adverse events in patients with AF [19]. The reasons for such a striking difference can be found in the different pathophysiological mechanisms. In the RACE II trial, AF patients with severe HF or with recent decompensation were excluded, thus leaving only patients without HF or with stable mild symptoms for at least three months [20]. In this setting, it has already been demonstrated that the actual benefit from the heart rate reduction and sinus rhythm restoration could be counterbalanced by the increased likelihood of adverse effects due to anti-arrhythmic drugs [21] and, therefore, pushing too hard on heart rate reduction could produce no further clinical benefits. On the other hand, it is well known that the heart rate is a risk factor in patients with HF, even when the sinus rhythm is present. Dysfunctional myocardium is energetically depleted and myocardial exerted force is negatively associated with the rate of contraction [22]. In an HF setting, such as the one of TCM occurrence, reducing the heart rate improves contractility, extends coronary diastolic filling time, reduces energy expenditure and improves cardiac output [23]. Moreover, the benefits of a reduced heart rate are consistent over the years, due to the positive modifications of the extracellular matrix and myocytes properties [24], resulting in a reduced risk of cardiovascular events and HF recurrences over a long follow-up. Regarding BNP, a small study already demonstrated that a NT-proBNP drop after four weeks was able to identify TCM with a sensitivity of 84% and a specificity of 95% [25]. In our population, this study found that BNP during the acute phase is an independent predictor of recurrences. This adds evidence to the notion that the patients with pure TCM may benefit from a continuation of HF treatment even after

normalization of LVEF in order to prevent recurrences, even if the usefulness, duration and safety of HF treatment in TCM still represent an unexplored grey area.

Although recent reviews and small case series [1,26,27] have hypothesized the TCM recurrences may be characterized by a more severe onset of the condition, our prospective study on a larger population, actually showed that the recurrences are characterized by a higher LVEF and a reduced heart rate. This could be related to the rate-control strategy and to the continuation of HF treatment after discharge. In our population, 15 out of 17 patients had AF as the trigger of TCM recurrence. Therefore, it is feasible to hypothesize that the progressive nature of AF could contribute to the risk of TCM recurrence. Furthermore, when compared to other supraventricular arrhythmias, such as atrial flutter or atrioventricular node re-entry tachycardia, currently available pharmacological and non-pharmacological rhythm control strategies for AF are surely less effective in obtaining an optimal and long-lasting restoration of the sinus rhythm [10].

Regarding major clinical events, there were a few and potentially unrelated to the combination between HF and tachyarrhythmia. Finally, in terms of cardiovascular related hospitalizations, almost half of this study's population was re-hospitalized, even though by definition, none had structural heart disease. Most hospitalizations occurred during the first year after the event and were related to rhythm control procedures, such as elective cardioversions and catheter ablations. Moreover, 16 hospitalizations were due to the recurrence of TCM. The heart rate at discharge confirmed its predicting value along with the LVEF at discharge. Similar to the fact that heart rate reduction has been demonstrated to be beneficial in HF [23], our data confirm the role of the rate control in the pathophysiology of this peculiar, reversible form of systolic dysfunction. This strengthens the message that, in pure TCM, the lower the heart rate at discharge, the better the long-term prognosis.
