Latent class analysis to identify phenotype-specific treatment effects of catheter ablation vs drug therapy in patients with atrial fibrillation: A secondary analysis of the CABANA trial.
Wenhui Li, Maria Sudell, Yang Chen et al.
✦ AI-curated · Sources linked
In 30 seconds
This secondary analysis of the randomized CABANA trial used latent class analysis to identify atrial fibrillation phenotypes and assess whether outcomes differed between catheter ablation and drug therapy. Among 2204 patients, three classes were identified, including an older preserved-function, lower-comorbidity group. In intention-to-treat analysis, ablation had the most favorable estimated effect in this group (adjusted hazard ratio 0.56, 95% confidence interval 0.34-0.92), but the treatment-by-class interaction was not statistically significant (P = .070). Per-protocol analysis showed a similar pattern, with interaction P = .047.
Key findings
- Latent class analysis identified three phenotypes among 2204 patients: older multi-morbid symptomatic, older preserved-function lower-comorbidity, and younger male metabolic-risk.
- In intention-to-treat analysis, the ablation estimate was most favorable in class 2 (adjusted hazard ratio 0.56, 95% confidence interval 0.34-0.92); interaction P = .070.
- Per-protocol analysis showed a similar class 2 pattern, with a treatment-by-class interaction P = .047.
Why it matters
Atrial fibrillation encompasses clinically varied patient profiles, and this analysis suggests that treatment effects may not be uniform across them. Identifying phenotypes associated with differing outcome estimates could help frame future research on individualized rhythm-control strategies. The findings are hypothesis-generating rather than definitive evidence that phenotype should determine treatment.
What to keep in mind
This was a secondary analysis using latent classes derived from baseline variables; the intention-to-treat treatment-by-class interaction was not statistically significant.