Introduction: Angiotensin-receptor blockers may exert favourable anti-arrhythmic effects in atrial fibrillation (AF), but their mechanisms are not fully understood. In this study, we tested the hypotheses that (i) candesartan reduces atrial fibrillatory rate and (ii) fibrillatory rate and its response to candesartan are related with the outcome of cardioversion. For this purpose, a post hoc subanalysis of the randomized, placebo-controlled CAPRAF (Candesartan in the Prevention of Relapsing Atrial Fibrillation) trial was performed.
Methods and results: Patients with AF undergoing electrical cardioversion were randomized to receive candesartan 8 mg once daily (n = 58) or matching placebo (n = 66) and no additional class I or III anti-arrhythmic drugs. Fibrillatory rate was determined from ECG lead V1 at baseline and at the day of cardioversion using spatiotemporal QRST cancellation and time-frequency analysis. The median time on treatment was 29 days. Candesartan reduced fibrillatory rate [399 +/- 48 vs. 388 +/- 49 fibrillations/min (fpm), P = 0.04], but not placebo (402 +/- 58 vs. 402 +/- 61 fpm, P = 0.986). Candesartan effects were only observed if the baseline fibrillatory rate was high [>420 fpm: 445 +/- 21 vs. 415 +/- 49 fpm, P = 0.006 vs. intermediate (360-420 fpm): 397 +/- 19 vs. 391 +/- 37 fpm, P = 0.351 vs. low (<360 fpm): 326 +/- 26 vs. 338 +/- 29 fpm, P = 0.179]. Cardioversion success was 100% in patients with an on-treatment rate <360 fpm vs. 83% in patients with higher rates (P = 0.02). Risk for AF recurrence was similar in patients with low (64%), intermediate (75%), or high on-treatment rates (63%, P = 0.446) and was also independent of candesartan effects on the fibrillatory rate.
Conclusion: In patients with persistent AF, candesartan decreases the fibrillatory rate, but this effect is restricted to patients with high baseline fibrillatory rates and is not associated with improved cardioversion outcome. Fibrillatory rates <360 fpm are associated with successful cardioversion, but not with AF recurrence.