LAUSR

The long awaited preliminary results of the CABANA study were presented by Dr Douglas Packer as a late breaking trial at the Heart Rhythm Society Scientific Sessions in Boston in May 2018. The study, which commenced in 2009, recruited 2204 patients with atrial fibrillation (AF) (new or under-treated) (42% paroxysmal, 47% persistent, and 10% long-standing persistent) from 110 centres in 10 countries, and randomized them to ablation vs. anti-arrhythmic drugs (AAD). Approximately 30% of patients were asymptomatic or minimally symptomatic, 30% moderately symptomatic and 40% severely symptomatic. Patients under the age 65 had to have at least one cardiovascular (CV) risk factor for stroke. The primary endpoint was a composite of all-cause mortality, disabling stroke, serious bleeding or cardiac arrest, and secondary endpoints were all cause mortality and death or CV hospitalization (primary and secondary endpoints had been reversed 5 years into the study due to low event rates and limited recruitment). All patients received anticoagulation based on contemporary guidelines. Of note, 102 patients randomized to ablation did not receive it (in part due to financial constraints in some geographic distributions), while 301 patients randomized to AAD crossed over to ablation. Rhythm control with 1–4 drugs was attempted in 87% of patients in the AAD group, in addition to rate control. The headline news was that, by intention to treat (ITT), the primary outcome did not differ between groups [8% for ablation vs. 9.2% for AAD, hazard ratio (HR) 0.86 (0.65–1.15)]. For the secondary endpoints, all-cause mortality was not statistically different [5.2% vs. 6.1%, HR 0.85 (0.6–1.21)], but death or CV hospitalization favoured ablation [51.7% vs. 58.1%, HR 0.83 (0.74–0.93)]. Subgroup analyses for the primary endpoint suggested benefits from ablation in younger patients (<75, statistically significant under 65), as well as benefits in patients with heart failure or cardiomyopathy. Ablation complications were very low and, despite likely use of older technology during part of the study, ablation was associated with a 47% reduction in AF by ITT, with 60% of patients in this arm free from AF (19% repeat ablations). There has already been much debate regarding whether one should simply consider the outcomes by ITT or look deeper into study results. In a study where event rates proved substantially lower than power calculations had predicted, but more importantly one with so many crossovers, we need to also seriously examine the results of per treatment (pre-specified), and per protocol analyses. On treatment analysis indicates a 33% reduction in the primary endpoint [7% vs. 10.9%, HR 0.67 (0.50–0.89), P1⁄4 0.006], while for secondary endpoints, all-cause mortality was reduced by 40% [4.4% vs. 7.5%, HR 0.60 (0.42– 0.86), P1⁄4 0.005], with a similar (to ITT) 17%, significant relative risk reduction in death or CV hospitalization. Per protocol analysis also suggests better outcomes with ablation, with a significant 27% reduction in primary composite endpoint. While controversy may continue for some time regarding interpretation of this hugely important study, we have much to learn from careful assessment and interpretation of the totality of these important findings.