Reduced doses NOAC show similar efficacy and bleeding risk but possibly higher mortality risk
Effectiveness and safety of reduced dose non-vitamin K antagonist oral anticoagulants and warfarin in patients with atrial fibrillation: propensity weighted nationwide cohort study.Literature - Nielsen PB, Skjøth F, Søgaard M et al., - BMJ. 2017; 10;356:j510. doi: 10.1136/bmj.j510
- Mean age varied markedly across exposure to the different drugs, for instance the average age for apixaban users was 83.9 years, while the average age of warfarin users was 71.0.
- Renal disease was more often seen in patients treated with apixaban (9.5%) and rivaroxaban (9.1%), than in those receiving dabigatran (3.9%) or warfarin (8.3%).
- Patients treated with apixaban had most comorbidities, and consequently, the CHA2DS2-VASc score was highest in these patients (4.3), and slightly lower in the dabigatran (3.8) and rivaroxaban (3.6) groups and lowest for warfarin-treated patients (3.0).
Inverse probability of treatment weight was applied to account for baseline differences, which resulted in comparable distribution of variables between treatment groups in this cohort, thus allowing for comparisons.
- The highest weighted event rate for Ischaemic stroke or systemic embolism (SE) was seen for apixaban (4.8%) and the lowest for dabigatran (3.3%), similar to rivaroxaban (3.5%) and warfarin (3.7%). Trends for the rate of ischaemic stroke or SE after one year or 2.5 years as compared with warfarin were not significant.
- Weighted event rates for bleeding outcomes were similar for apixaban (5.1%), rivaroxaban (5.6%) and warfarin (5.1%) and lower for dabigatran (4.1%).
- Weighted Cox regression analyses showed lower bleeding event rates for dabigatran (HR: 0.80, 95%CI: 0.70-0.92), apixaban (HR: 0.96, 95%CI: 0.73-1.27) and rivaroxaban (HR: 1.06, 95%CI: 0.87-1.29) compared with warfarin after one year.
- Higher risk of all-cause mortality was seen for patients treated with NOACs. Crude one-year mortality risks were 8.6% for warfarin, 12.2% for dabigatran, 21.2% for rivaroxaban and 25.4% for apixaban. The weighted hazard ratios were 1.48 (95%CI: 1.31-1.67) for apixaban, 1.04 (95%CI: 0.96-1.13) for dabigatran and 1.52 (i5%CI: 1.36-1.70) for rivaroxaban.
The subgroup of patients with an indication for dose reduction (n=21949), showed worse outcomes than the main analysis.
- Apixaban showed higher one-year risk of ischaemic stroke or SE than warfarin (HR: 1.24, 95%CI: 1.00-1.55), while rivaroxaban showed lower rates (HR: 0.63, 95%CI: 0.47-0.85).
- Apixaban (HR: 0.78, 95%CI: 0.61-0.99) and dabigatran (HR: 0.81, 95%CI: 0.69-0.94) showed lower bleeding rates than warfarin, and rivaroxaban was similar to warfarin (HR: 1.00, 95%CI: 0.81-1.24).
- Apixaban and rivaroxaban were associated with a higher risk of all-cause mortality (HR: 1.23, 95%CI: 1.10-1.36 and HR: 1.48, 95%CI: 1.32-1.67) respectively. No difference in mortality rate was seen with dabigatran (HR: 0.93, 0.84-1.02).
These propensity-weighted real world data on the effectiveness and safety of low dose NOAC regimens as compared with warfarin in patients with AF showed a non-significant trend towards higher thromboembolic events and similar rates of bleeding for apixaban (2.5 mg). Low-dose rivaroxaban (15 mg) and dabigatran (110 mg) showed a non-significant trend towards lower one year rates of ischaemic stroke or SE, and similar (rivaroxaban) and lower (dabigatran) rates of bleeding.
While standard dosing regimes have consistently favoured a NOAC over warfarin with regard to mortality, some of the current analyses on reduced doses indicated an increased mortality risk. Although inverse probability treatment weighting allows for causal inference on average treatment effects, it is possible that some unmeasured residual confounding and selective prescribing behaviour remain.