Abstract
In patients with paroxysmal atrial fibrillation, cryoballoon ablation (CBA) and radiofrequency ablation (RFA) represent two therapeutic approaches supported by increasing literature. While both these ablation techniques play a role during different stages of the patient's therapeutic pathway, their use as first-line is being increasingly recognized. This scoping review comparatively examined the evidence of effectiveness for these two ablation techniques. Our analysis was limited to the evaluation of the end-point of time to recurrence of atrial fibrillation (or other forms of atrial arrhythmias), which was the primary end-point in most clinical trials. The method used for pooling the information from clinical trials (Shiny method) is original and based on an artificial intelligence (AI) method that reconstructs individual patient data from published Kaplan-Meier time-to-event curves. Because a network meta-analysis has been published on this same clinical material, one objective of the present work was to compare the meta-analytic results with those generated by the Shiny method. A standard literature search was conducted on PubMed/Medline. Only randomized studies comparing CBA versus medical therapy, RFA versus medical therapy, or CBA versus RFA in previously untreated patients were eligible. Trials presenting a Kaplan-Meier curve to present the above-mentioned end-point were included. Patient-level data were reconstructed by application of the Shiny method. These individual patient data were then analyzed by standard statistical testing based on hazard ratio (HR) for risk of recurrence and medians of time to recurrence. Our analysis compared the two ablation treatments and medical therapy. A total of five trials were identified through our literature search. Information from these trials was pooled according to the three treatments (CBA: three trials, n = 365; RFA: two trials, n = 99; medical therapy: five trials, n = 457). CBA showed higher effectiveness than medical therapy (HR, 0.51; 95% confidence interval (CI): 0.38 to 0.67). In comparison with medical therapy, RFA showed a numerical trend that remained far from statistical significance (HR, 0.89; 95% CI: 0.62 to 1.27). Medians for time to recurrence were 14.1 months (95% CI: 10.0 to not reached) for RFA and 11.5 months (95% CI: 9.3 to 25.3) for medical therapy. This parameter was not reached for CBA. The current evidence from five randomized trials suggests that CBA ranks first in effectiveness, followed by RFA and medical therapy. In our comparison between the results generated by the Shiny method with those published in the previous meta-analysis, the Shiny method confirmed its ability to account for the length of follow-up in individual trials, whereas the meta-analytic approach confirmed its ability to account for the effects of randomizations performed in the trials.