Andexanet alpha versus four-factor prothrombin complex concentrate in DOACs anticoagulation reversal: an updated systematic review and meta-analysis

Abstract

Background: There is currently a lack of evidence for the comparative effectiveness of Andexanet alpha and four-factor prothrombin complex concentrate (4F-PCC) in anticoagulation reversal of direct oral anticoagulants (DOACs). The primary aim of our systematic review was to verify which drug is more effective in reducing short-term all-cause mortality. The secondary aim was to determine which of the two reverting strategies is less affected by thromboembolic events.

Methods: A systematic review and meta-analysis was performed.

Results: Twenty-two studies were analysed in the systematic review and quantitative synthesis. In all-cause short-term mortality, Andexanet alpha showed a risk ratio (RR) of 0.71(95% CI 0.37-1.34) in RCTs and PSMs, compared to 4F-PCC (I² = 81%). Considering the retrospective studies, the pooled RR resulted in 0.84 (95% CI 0.69-1.01) for the common effects model and 0.82 (95% CI 0.63-1.07) for the random effects model (I² = 34.2%). Regarding the incidence of thromboembolic events, for RCTs and PSMs, the common and the random effects model exhibited a RR of 1.74 (95% CI 1.09-2.77), and 1.71 (95% CI 1.01-2.89), respectively, for Andexanet alpha compared to 4F-PCC (I² = 0%). Considering the retrospective studies, the pooled RR resulted in 1.21 (95% CI 0.87-1.69) for the common effects model and 1.18 (95% CI 0.86-1.62) for the random effects model (I² = 0%).

Conclusion: Considering a large group of both retrospective and controlled studies, Andexanet alpha did not show a statistically significant advantage over 4F-PCC in terms of mortality. In the analysis of the controlled studies alone, Andexanet alpha is associated with an increased risk of thromboembolic events.

Clinical trial registration: PROSPERO: International prospective register of systematic reviews, 2024, CRD42024548768.

Keywords: 4F-PCC; Andexanet alpha; Anticoagulation reversal; DOACs; Haemorrhage; Meta-analysis.

Fig. 1

PRISMA flowchart

Fig. 2

Risk of bias of the included studies. For the controlled trials, the RoB2 was used to assess the risk of bias. In all remaining included studies, Robins-I was utilized. Patient selection was the primary cause of bias in the studies that were included. For further details see the main text

Fig. 3

Forest plots for all-cause short-term mortality. A Forest plot for the controlled (RCT and PSM) studies. The random effects model exhibited a risk ratio (RR) of 0.71 (95% CI 0.37–1.34). Since the confidence interval crosses the unit, the difference was not statistically significant. B Forest plot for the retrospective studies. The pooled RR resulted in 0.84 (95% CI 0.69–1.01) for the common effects model and 0.82 (95% CI 0.63–1.07) for the random effects model

Fig. 4

Funnel plot for all-cause short-term mortality. A Funnel plot for the controlled (RCT and PSM) studies. Egger’s test was not statistically significant (p = 0.47). B Funnel plot for the retrospective studies. Egger’s test was statistically significant (p = 0.03). Some studies show an excess reduction of mortality in favour of the Andexanet alpha group

Fig. 5

Forest plots for thromboembolic events. A For RCTs and PSMs, the common and the random effects model exhibited a RR of 1.74 (95% CI 1.09–2.77], and 1.71 (95% CI 1.01–2.89), respectively. B Forest plot for the retrospective studies. The pooled RR resulted in 0.84 (95% CI 0.69–1.01) for the common effects model and 0.82 (95% CI 0.63–1.07) for the random effects model. B Considering the retrospective studies, the pooled RR resulted in 1.21 (95% CI 0.87–1.69) for the common effects model and 1.18 (95% CI 0.86–1.62) for the random effects model

Fig. 6

Funnel plot for thromboembolic events. A Funnel plot for the controlled (RCT and PSM) studies. Egger’s test was not statistically significant (p = 0.59). B Funnel plot for the retrospective studies. Egger’s test was not statistically significant (p = 0.22)