Calcium and postoperative atrial fibrillation: round up the usual suspects!

Abstract

Postoperative atrial fibrillation (AF) is a frequent and burdensome complication in patients undergoing coronary revascularization and heart valve surgery. Whilst it is recognized that haemodynamic stress and the inflammatory response associated with cardiac surgery may play an important role in triggering the arrhythmia, patients with a diagnosis of postoperative AF who are discharged from hospital in sinus rhythm are more likely to develop AF and ischaemic stroke on long-term followup. Such findings imply that paroxysmal AF episodes brought about by acute illness or stress may still reflect the presence of a pre-existing substrate subtending future arrhythmia, myocardial dysfunction and cardioembolic complications. To date, atrial electrical remodelling and fibrosis have been a main focus of research in AF, even though both factors are more robustly identifiable as a consequence rather than a primary cause of AF. In contrast, large prospective cohorts have identified impaired left atrial function (i.e. a lower emptying fraction) in individuals in sinus rhythm as an independent predictor of incident AF and ischaemic stroke. As for all observational findings, whether or not left atrial function is causally associated with AF (vs. being a biomarker of other known or unknown AF or stroke risk factors)] remains to be conclusively established, as do the mechanisms by which reduced atrial contraction may lead to AF. In this issue of Cardiovascular Research, Fakuade et al. report a reduction in calcium transient amplitude and intracellular stores, and slower rate of calcium reuptake in atrial myocytes isolated from the right atrial appendage of patients in sinus rhythm who develop AF after cardiac surgery compared with those who did not. These findings would be expected to result in impaired contraction and indeed in a subgroup of patients, the authors show that preoperative left atrial systolic performance in patients who develop AF after surgery is lower than that recorded in those who maintain sinus rhythm postoperatively. Based on our current knowledge of the mechanisms regulating intracellular calcium handling in cardiomyocytes, one could easily identify a number of ‘usual suspects’ that may account for this phenotype, some of which were swiftly acquitted. For instance, differences in postoperative heart rhythm were not associated with changes in peak and overall calcium influx through the L-type calcium channel (under basal conditions or in the presence of isoprenaline) or with differences in intracellular calcium buffering properties, t-tubule density, sodium-calcium exchanger (NCX) function or sarcolemmal calcium ATP-ase activity. Importantly, differences in ryanodine receptor (RyR) phosphorylation or calcium leak did not account for these findings. A slower rate of intracellular calcium decay in paced cardiomyocytes in the absence of differences in NCX function would point to a reduced SR calcium-ATPase (SERCA2A) activity. SERCA2A is the master regulator of intracellular calcium cycling and its activity is chiefly responsible for the rate of calcium reuptake in the SR, which in turn influences the calcium buffering reserve of contractile proteins and the intracellular calcium stores, and plays an important role in both cardiomyocyte contraction and relaxation, particularly under unloaded conditions. The relationship between SERCA2A activity and arrhythmia is more complex and may vary depending on concomitant changes in diastolic calcium and the threshold for SR calcium leak/release from the RyR. In previous publications, SERCA2A activity has been reported to be higher in atrial myocytes from patients with postoperative or paroxysmal AF and the combination of raised RyR phosphorylation and increased SERCA2A activity has been recently shown to increase the probability of AF induction in Langendorff-perfused, aged mouse hearts. Even in the presence of severe heart failure, SERCA2A expression has not been shown to be reduced in human atrial myocytes and, taking all evidence into consideration, one would conclude that a difference in SERCA2A activity, when present, is more commonly ascribed to its modulation by regulatory peptides or by post-translational modifications than explained by differences in its total protein content. Here, however, the atrial phenotypic differences in patients who develop postoperative AF were attributed to a reduction in SERCA2A protein. Whilst the average data show a significantly lower SERCA2A protein content in right atrial appendage homogenates of patients who develop postoperative AF, the reduction is modest and the overlap between groups considerable.