The quandary of study design in evaluating effectiveness of contact force sensing catheter in atrial fibrillation ablation: Authors’ reply

We read with interest the letter by Natale et al.1 pertaining to our recent meta-analysis. Our study demonstrated that contact force (CF) guidance is associated with improved safety and efficacy of atrial fibrillation (AF) in observational studies but that these benefits are not reproduced in meta-analysis of randomized controlled trials (RCTs).2 Natale et al. highlighted several limitations of current RCTs that may have underestimated the potential benefits of CF guidance in our meta-analysis. In particular, they identified significant heterogeneity between trials with regards to CF targets as well as potential operator ‘training effects’. Although these are significant limitations that we also acknowledge in our article, it is equally important to note that the perceived benefits of CF guidance in observational studies may be largely driven by confounders and bias. Specifically, limitations of current observational studies include lack of blinding during outcome assessment, inconsistencies in follow-up, significant imbalances in baseline clinical characteristics, and temporal bias due to non-contemporaneous control groups.2 Nonetheless, the findings of our meta-analysis are not intended to imply that CF is an unimportant parameter during AF ablation. To the contrary, an abundance of preclinical data has shown CF is a critical determinant of lesion quality. Excessively high CF values have been associated with perforation whereas insufficient CF has been linked with an increased risk of electrical reconnection and arrhythmia recurrence.3 What our meta-analysis does show, however, is that the availability of real-time CF data alone does not necessarily lead to an improvement in clinical outcomes. We agree with Natale et al. that our findings are at least partially reflective of deficiencies in the way data obtained from CF-sensing catheters has been interpreted and applied to date. Of note, in several RCTs, operators who were not blinded to CF data spent a greater proportion of time within the CF target range, without a corresponding improvement in freedom from AF or reduction in major complications. This discordance between improved ability to titrate CF and lack of improved outcomes may be due to the selection of sub-optimal CF target ranges that are not in keeping with contemporary guidelines.4 It may also reflect the need to integrate CF with other variables impacting lesion quality, such as power and radiofrequency duration. Novel markers of lesion quality incorporating CF with other ablation parameters have been shown to be stronger predictors of AF recurrence than CF alone.5 Thus, overall we agree with Natale et al. that CF-sensing technology should not be abandoned based on our meta-analysis of current RCTs. However, positive findings from observational studies should not be the basis for widespread adoption of CF-sensing catheters given the inherent biases of these non-randomized study designs. Rather, carefully designed RCTs using contemporary CF targets and sophisticated ablation indices are required to rigorously evaluate the impact of CF-sensing technology on clinical outcomes of AF ablation. Conflict of interest: S.K. is supported by the NSW Health Early to Mid-Career Fellowship and research grants from Biotronik and Abbott Medical, Viertel Foundation Project Support, Perpetual Ramaciotti Health Investment Grant, Westmead Hospital Charitable Trust Project Support Grant, Westmead Hospital Research and Education Network Research Grant, Cardiac Society of Australia and New Zealand Bayer Young Investigator Grant, and the National Heart Foundation Vanguard Grant. None of the research grants are relevant to the present manuscript. The remaining authors have nothing to disclose.