This study provides unique insights into the association between SDB and spheric cardiac remodelling after acute myocardial infarction: firstly, in contrast to the no SDB and CSA groups, patients with OSA exhibited a significant increase in systolic and diastolic sphericity index within 12 weeks after acute myocardial infarction. Secondly, the number of obstructive apnoeas and hypopnoeas was positively correlated with systolic sphericity index, whereas the number of central apnoeas and hypopnoeas was not. Data were robust after multiple regression analysis accounting for demographics and risk factors for spheric remodelling after acute myocardial infarction. Thirdly, no significant association between wall thickness, formation of cardiac aneurysm and SDB was observed.
To our best knowledge, an association between SDB in general and post-infarction spheric cardiac remodelling, measured by sphericity index, has not previously been described. After STEMI, up to 66% of patients are affected by OSA [
23] and 12% by CSA [
24]. Previous analyses in patients with acute myocardial infarction have revealed that the increase in systolic and diastolic left ventricular volumes is more pronounced in patients with SDB compared to those without. In patients with SDB no differences between OSA and CSA were observed [
15]. Left ventricular function is diminished in both CSA and OSA patients after STEMI compared to patients without SDB [
25‐
27] and an improvement in AHI over time is associated with an improvement in left ventricular function [
24]. In OSA patients without acute myocardial infarction, higher AHI values are associated with increased cardiac wall thickness [
28]. Data from patients with CSA are lacking. Thus, previous research has identified negative post-STEMI remodelling of the left ventricle in SDB. However, spheric remodelling, which has additional prognostic value [
8], has not been specifically addressed.
Effect size and potential clinical impact
The sphericity index can be used to detect cardiac remodelling with 100% sensitivity and 90% specificity and is more accurate than left ventricular volumes and ejection fraction [
22]. A threshold of 0.29 in diastolic sphericity index is associated with a higher probability of developing heart failure [
7,
29]. In the current analysis, this pathologic threshold was considerably exceeded at baseline, indicating clinically relevant spheric cardiac remodelling after myocardial infarction. Mannaerts et al. used 3D echocardiography to detect a sphericity index of 0.32 in patients with cardiac remodelling after STEMI compared to 0.22 in patients without cardiac remodelling, which is in line with our findings using CMR [
22].
A more spherical left ventricle, assessed by ventriculography, is associated with reduced capacity and diminished ejection fraction. In addition, sphericity index is an independent predictor of survival in patients with coronary artery disease [
30]. Survival after myocardial infarction is lower in patients with a higher sphericity index, which is used as a surrogate marker for congestive heart failure [
8]. In this context, it is notable that the sphericity index increased considerably in our OSA cohort, which reflects potential detrimental effects on cardiac remodelling in this specific group of patients.
Pathophysiology
The main pathophysiological differences between the CSA and OSA patients are negative intrathoracic pressure swings in the OSA group due to breathing efforts against the occluded pharynx, elevated blood pressure in the OSA cohort and arousals [
31].
Data indicate that OSA-associated negative intrathoracic pressure swings contribute to increased left ventricular transmural pressure. In OSA patients with congestive heart failure, continuous positive airway pressure increases intrathoracic pressure [
14], reduces left ventricular volume and left ventricular transmural pressure, impedes cardiovascular complications and improves cardiac function [
11,
32‐
34]. Similar data in patients with CSA are lacking.
Moreover, the analysis evaluating the association between obstructive apnoeas and sphericity index showed higher numeric beta coefficients (effect size) compared to the analysis with obstructive apneas and hypopneas. This finding may underline, that in particular obstructive apnoeas and to a lower extend hypopnoeas may contribute to the intrathoracic negative pressure changes and might affect cardiac remodeling.
In addition, arousals from sleep may contribute to increased afterload and thus spheric cardiac remodeling. The association between the frequency of arousals may differ between obstructive and central respiratory events, because arousals occur in OSA to terminate apnoeas and activate pharyngeal muscles in order to re-open the occluded upper airways, while in CSA the association between arousals and respiratory events is modest [
35]. In the present study we found a similar association between central and obstructive respiratory events with arousals from sleep (Figure S1a, b).
Activation of the sympathetic nervous system seems to be lower in CSA patients with congestive heart failure compared to those with OSA [
36]. Several studies have revealed elevated blood pressure in OSA patients [
37,
38]. This results in higher cardiac afterload and may lead to increased cardiac remodelling and a consequent deterioration in left ventricular mechanics [
39]. In contrast, CSA has a relatively weak influence on blood pressure [
12]. Our study also observed that higher blood pressures are more strongly associated with the severity of OSA, rather than CSA.
Thus, the increased left ventricular transmural pressure (negative intrathoracic pressure swings plus increased arterial blood pressure) in OSA patients, but not in CSA patients, may promote spheric cardiac remodelling and thinning of the left ventricular wall in the region of the myocardial infarction. Accordingly, in the present analysis, sphericity index is correlated with the quantity of obstructive apnoeas and hypopnoeas, but not central apnoeas and hypopnoeas. These results are robust, even after accounting for potentially confounding clinical factors such as age, sex and BMI.
The prevalence of sleep apnoea after myocardial infarction is high (approximately 54%) [
40]. Within 12 weeks after myocardial infarction, the recovery of cardiac function is associated with a reduction in sleep apnoea, whereas severity of sleep apnoea is not changed in patients with persistent limited cardiac function [
40]. Therefore, sleep apnoea should be re-evaluated when cardiac function changes.
Limitations
The results from this sub-analysis must be interpreted in the light of several limitations. The sample size is relatively small, since only patients with infarcted left anterior descending artery were included. Furthermore, cardiac aneurysm is a rare event. However, the selected sample is appropriate considering that sphericity index was first evaluated in patients with anterior wall myocardial infarction. A direct causal relationship cannot be inferred due to the observational nature of the study design. Neither intrathoracic pressures nor oesophageal pressures were recorded to quantify the postulated negative intrathorathic pressure swings during obstructive apneas [
31]. Larger, prospective, randomised trials are now required to verify these findings. Further data will be generated by the interventional TEAM-ASV I study (NCT02093377, adaptive servoventilation versus control in patients with acute myocardial infarction and sleep-disordered breathing).