Introduction: Atrial fibrillation (AF) is the most common sustained cardiac arrythmia characterised by abnormal electrical impulse conduction throughout the atria. Although many patients with AF respond well to rhythm control therapy (medication to restore sinus rhythm), others experience early AF recurrences. Our understanding of which patients will experience an AF recurrent event whilst on rhythm control therapy are limited and, as such, the genetic determinants of AF may offer some additional insight. The strongest genomic predisposition to AF in general populations, located on chromosome 4q25, also predisposes to recurrent AF after AF ablation and suggests that this predisposition is related to altered electrical function in the atria. Given that PITX2 is the closest gene located to the AF locus, we hypothesised that left atrial PITX2 expression levels are related to recurrent AF.
Methods: PITX2 and PITX2c expression levels were assessed in left atrial appendages from 146 patients undergoing thoracoscopic AF ablation recruited to the AFACT trial. An initial analysis of whole-tissue PITX2 and PITX2c expression levels in 94 samples was carried out, followed by an assessment of PITX2 expression levels in cardiomyocytes from 52 samples. A multivariate analysis considering 15 clinical parameters and PITX2/PITX2c expression was carried out by fitting a logistic regression model with a backward selection. The odd ratios (ORs) and 95% confidence intervals (CI) for selected parameters were quantified and the performance of the model was assessed with the C-statistic.
Results: PITX2 and PITX2c expression were differentially expressed in left atrial appendage whole tissue samples. Neither PITX2 nor PITX2c expression levels were different between patients with or without recurrent AF. Multivariate regression also showed that neither PITX2 nor PITX2c expression levels differentiated patients’ AF recurrence statuses. Enrichment of cardiomyocytes using and anti-pericentriole membrane antibody showed PITX2 expression was enriched in cardiomyocyte nuclei. In cardiomyocytes, stratification of PITX2 expression by AF recurrence outcome showed no differences (p = 0.08, Mann-Whitney U-test). However, in the multivariate analysis, PITX2 and AF duration emerged as the two parameters associated with AF recurrence where decreased levels of PITX2 were associated with an increased chance of recurrence (OR = 0.8131, 95% CI = 0.662, 0.999; AF duration was not significant, p = 0.080). The C-statistic of this model was 0.729, 95% CI 0.584, 0.897).
Conclusions: PITX2 mRNA expression was largely restricted to cardiomyocytes in patients with AF and reduced cardiomyocyte PITX2 expression strongly predicted AF recurrence. Further work is warranted to test whether differences in left atrial cardiomyocyte PITX2 expression levels can be used to guide stratified therapy of patients with AF.