Even in normal atria in sinus rhythm inter-atrial and site-specific conduction heterogeneities are present, which may lead to spatial non-uniformity of conduction anisotropy.22
In the presence of triggers with sufficient incidence, AF can be induced both in normal and in abnormal atria. The wave pattern and type of AF may vary in complexity,9
depending on the severity of the underlying substrate.23
Although inducibility of AF by pacing in normal atrial structure is dependent on the protocol of pacing,24
it has a predictive value for the occurrence of POAF. 25,26
This suggests higher susceptibility for AF at least in some of the patients who subsequently develop POAF.27
Furthermore, AF occurring as a consequence of cardiac surgery in patients without a history of AF postulates a significant pathophysiological role for the surgical intervention itself.
Early pro-arrhythmic environment
In a recent review, we discussed the acute and chronic factors contributing to initiation of POAF, and emphasized that different mechanisms are responsible for AF in a post operative setting.28
Inflammation in the acute post-operative phase has been subject of extensive research as it has been linked to several local and systemic pro-arrhythmic effects.
First of all, direct atrial trauma, e.g. the venous cannulation through the right atrium, has been shown to induce an inflammatory reaction leading to inhomogeneity in conduction in a canine model.29
The canine sterile pericarditis model, in which application of sterile talc and subsequent pericarditis enhanced AF susceptibility, supports this hypothesis.30
Administration of prednisone in this model reduced inflammation and as expected prevented AF.31
Similarly colchicine was able to reduce the incidence of POAF, as a result of a reduction of the post-cardiotomy syndrome in humans.32
Secondly, systemic inflammatory effects expressed as increased C-reactive protein levels, and therefore complement activation during cardio pulmonary bypass and during the acute post operative phase, were demonstrated to coincide with the peak POAF incidence, advocating a direct immune reaction mediated by the complement system. 33
Indeed, the use of cardio pulmonary bypass is correlated with POAF incidence in most studies.5,34
Thus as expected, oral corticosteroids reduce not only the post-operative inflammatory markers, but also the incidence of POAF as demonstrated by an extensive meta-analysis.35
Furthermore, cardiopulmonary bypass might lead to insufficient cooling of the atria during the cardioplegic arrest.36
This may induce a temporary substrate to initiate AF as a consequence of ischemia reperfusion injury.36,37
Indeed, oxidative stress has been shown to cause a transient pro-arrhythmic effect in the post-operative setting.38
This can explain the anti-arrhythmic effects of statins in POAF prevention. Despite the anti-inflammatory effects of the HMG-CoA enzyme inhibitors, statins had no effect in predetermined substrates for AF, such as enlarged left atria.39
In line with the effect of oxidative stress in the direct post-operative occurrence of AF, administration of ascorbate acid preserved the length of the effective refractory period (ERP) during rapid pacing in a canine model, and reduced POAF especially in the very early post-operative phase in humans.40
Similarly, the antioxidant N-acetylcysteine reduced the incidence of POAF in the early post-operative phase.41
These findings suggest that independent of a pre-existing vulnerability to AF, surgery itself can have several pro-arrhythmic consequences on the atrial tissue based on ischemic and inflammatory effects.
Post-operatively, as a counterbalance to anesthesia induced hypotension and cardiac stunning after aortic crossclamping, fluid administration is required to keep up the cardiac output. Especially in hypertrophic ventricles, e.g. in aortic valve stenosis, significant volume therapy is applied for hemodynamic stabilization in the acute post-operative phase. In this setting, atrial enlargement seems to be an important factor in predicting POAF.32
Acute atrial stretch decreases conduction velocity, increases conduction block and thereby the vulnerability of the atrial tissue to reentry.42,44
In addition, premature atrial beats initiate post-operative AF,45
partly aggravated by high sympathetic activation.6
Possibly, they are also due to increased L-type calcium currents (Ica2+) in patients developing POAF.46
Indeed, milrinone, a phosphodiesterase inhibitor inotropic agent, increased the incidence of POAF significantly, presumably through a protein kinase-A activation leading to triggered activity.47
It can be concluded that transient local and systemic changes in the acute post-operative phase after cardiac surgery are linked to POAF. They cause conduction disturbances and may enhance triggers for AF which both together will enhance the susceptibility to AF.
In addition to post-operative acute phase triggers, several long-term mechanisms have been identified to also predict POAF. These mechanisms are known to produce a more sustained substrate for AF, which favors AF maintenance rather than initiating it. The complexity of AF in these patients is therefore higher and recurrences are more likely after discharge, compared to AF in the less pronounced substrates, which are confine to solely acute phase POAF.
First, POAF is a disease of advanced age.1,48
The atrial wall becomes fibrotic with age49 and this has been identified as one of the most important structural substrates for AF perpetuation in the non-operative setting.10,23
Prolonged signal averaged p-wave duration may reflect atrial fibrosis resulting in conduction delays and has been reported to be a predictor of POAF.50-52
However, pre-operative AF was not excluded in all these studies.50,51
Nevertheless this finding suggests an increased AF vulnerability in patients with intra-atrial conduction delays particularly in a pro-arrhythmic environment.
Several co-morbidities have also been shown to increase the chances of POAF development, such as chronic obstructive pulmonary disease,48,53
obesity and systolic dysfunction.1,53
The role of these co-morbidities in the perpetuation of AF has long been established.8,54
Obesity has been associated with ERP shortening in left atria, diastolic dysfunction and left atrial dilation, all of which predispose to AF.5
In congestive heart failure, Sanders et al. demonstrated increased vulnerability to AF due to prolonged p-wave duration, prolonged ERP duration at the right atrial wall and significantly longer sinus node recovery time.27
Accordingly, prolongation of conduction time with a higher percentage of double potentials and longer iso-electric intervals is linked to severely disturbed substrates. 10,23
These pro-arrhythmic changes seem to occur due to chronic dilated atria, as a consequence of congestive heart failure or longstanding valve pathology for example. In line with these findings, left ventricular diastolic dysfunction was determined as an additional predictor of POAF.55
On the biochemical level, stretch activated channels (SAC’s) increase intracellular Na+ concentrations. As a consequence, the Na+/Ca2+ exchanger extrudes less Ca2+ to the extracellular space, thereby enhancing Ca2+ binding to the actin-myosin response. In this setting intracellular calcium overload causes a down regulation of Ica2+-gene expression, subsequently shortening the ERP.56
Also chronic stretch has been associated with altered matrix metalloproteinase expression and angiotensin II mediated fibrosis.10
Thus as expected, preoperative atrial enlargement has been shown to increase risk for POAF.57
Furthermore, preoperative dispersion of refractoriness and a prolonged PR interval predicted POAF after CABG.58
Atrial structural changes have been attributed to the uncontrolled ventricular rate and congestive heart failure leading to atrial stretch, apoptosis and myolysis.59
Although apoptosis has not consistently been found in the atrial tissue of patients developing POAF, myolysis has been advocated as a preexisting structural substrate in these patients.60,61
Another important structural abnormality, which is closely linked to the propagating abilities of the atrial wall, is the gap-junctional connexin (Cx) distribution. Higher and more heterogeneously distributed levels of Cx 40 have been demonstrated in POAF patients. 62
Also local inflammation caused a reduction in Cx 40 and 43 expressions in the epicardial level, compared to normal distribution in non-inflammatory circumstances.63
The exact role of connexin expression changes in POAF still needs to be defined.
From these findings it can be concluded that at least in some of the patients developing POAF after cardiac surgery, severely altered atrial architecture already is present before the operation. These structural alterations make the atria prone to not only POAF but also to AF in general. This is important because POAF might be an expression of the atrial vulnerability that is detected due to continuous monitoring in a period with excessive triggers, a “poor man’s exercise” effect. This strongly implies close monitoring of these patients later on. Figure 1
gives a possible course of POAF development and perpetuation after cardiac surgery and after discharge.
A schematic sketch of incidence of POAF in the early post-operative phase (i.e. the first post-operative week) and during the late post-operative phase (i.e. weeks to years following cardiac surgery) after discharge. In severely aggravated atrial substrates, POAF might continue to exist and reoccur. Red area under the curve represents patients with a preexistent structural substrate, e.g. atrial enlargement or heart failure, which is not easily reversible. White area under the curve represents patients with less severe structural substrate but who are exposed to the pro-arrhythmic environment of the acute-postoperative phase
Treatment strategies also provide additional information for the possible mechanism that is responsible for POAF. The positive effects of class III anti arrhythmic drugs for example, suggest a role for a reentry mechanism, in which shortening of the excitable gap might be the key element in their effectiveness, while beta-adrenergic blockers advocate a more prominent role for the sympathetic activation and triggers for AF.5,64