Risk-stratified effectiveness of direct oral anticoagulants versus vitamin K antagonists in thrombotic antiphospholipid syndrome: a prospective cohort study

Q: What is the clinical evidence from this study?

Study design: Cohort. Population: Thrombotic antiphospholipid syndrome (APS) (n=413). Intervention: Rivaroxaban vs. Warfarin (vitamin K antagonist). Primary outcome: First recurrent thrombotic event (arterial or venous thrombosis) (HR 0.69, 95% CI 0.27 to 1.75, p=0.44).

RuiQi Li; Yangzhong Zhou; Yuan Zhao; Hui Jiang; Liying Peng; Qian Wang; Xinping Tian; Mengtao Li; Xiaofeng Zeng; Can Huang; Jiuliang Zhao

doi:10.1136/lupus-2026-002288

November 1, 2026Lupus Science & MedicineOpen Access

Risk-stratified effectiveness of direct oral anticoagulants versus vitamin K antagonists in thrombotic antiphospholipid syndrome: a prospective cohort study

View Full Paper

Ask AI

Bookmark

Share

Key result

Rivaroxaban was not associated with a different overall risk of recurrent thrombosis compared to warfarin in thrombotic APS, though risk increased in triple-positive patients.

Figures

Why the study?

The comparative effectiveness and safety of rivaroxaban versus warfarin in thrombotic antiphospholipid syndrome, including baseline risk-stratified treatment effects, needed evaluation.

Does rivaroxaban compared with warfarin alter the risk of recurrent thrombosis or major bleeding in patients with thrombotic antiphospholipid syndrome?

Key Points

To compare the effectiveness and safety of rivaroxaban versus warfarin in thrombotic antiphospholipid syndrome and assess baseline risk-stratified treatment effects.
Prospective, single-centre cohort study of 413 patients with thrombotic antiphospholipid syndrome receiving either rivaroxaban (n=89) or warfarin (n=324) at Peking Union Medical College Hospital.
Median follow-up was 25.1 months (IQR, 8.5–44.4), comprising 198.5 patient-years for rivaroxaban and 833.1 patient-years for warfarin.
Inverse probability of treatment weighting using covariate balancing propensity scores was applied, and hazard ratios were estimated with weighted Cox proportional hazards regression models.
Recurrent thrombosis occurred in 6 rivaroxaban-treated patients (6.7%; 3.02/100 patient-years) and 26 warfarin-treated patients (8.0%; 3.12/100 patient-years), showing no significant weighted difference (HR 0.69, 95% CI 0.27 to 1.75; p=0.44), nor in major bleeding (HR 0.68, 95% CI 0.15 to 3.03; p=0.61).
A significant treatment-by-subgroup interaction occurred according to triple-positive antiphospholipid antibody status (P for interaction=0.007).

Study Design

Type

Cohort (n=413)

Multicenter

No

Structured PICO

Does rivaroxaban compared with warfarin alter the risk of recurrent thrombosis or major bleeding in patients with thrombotic antiphospholipid syndrome?

P

Population

413 patients with thrombotic antiphospholipid syndrome (54.5% female, median age 38 in rivaroxaban arm and 34 in warfarin arm) treated with rivaroxaban or warfarin at a single center in China and followed for a median of 25.1 months.

E

Exposure

Rivaroxaban (n=89, 21.5%) prescribed for secondary thromboprophylaxis over a median follow-up of 21.5 months (IQR 8.3–41.9; 198.5 patient-years), with concomitant therapies including glucocorticoids (36.0%), single antiplatelet therapy (27.0%), and statins (14.6%).

C

Comparator

Warfarin (vitamin K antagonist, n=324, 78.5%) prescribed for secondary thromboprophylaxis with INR monitoring (overall median TTR 45.2%) over a median follow-up of 25.8 months (IQR 8.5–46.1; 833.1 patient-years), with concomitant therapies including glucocorticoids (50.9%), single antiplatelet therapy (28.4%), and statins (20.7%).

O

Outcome

Primary effectiveness outcome: first recurrent thrombotic event (composite of arterial thrombosis [myocardial infarction, stroke/TIA, limb/carotid/subclavian/vertebral/splanchnic/retinal artery thrombosis, splanchnic infarction] or venous thromboembolism [limb DVT, PE, CTEPH, splanchnic/jugular/subclavian/cerebral venous sinus/retinal venous/right atrium thrombosis]). Primary safety outcome: major bleeding (ISTH criteria, including gastrointestinal bleeding, intracranial haemorrhage, and other major bleeding events), clinically evaluated and confirmed by physicians at median 25.1 months follow-up.composite

Main Result

Hazard Ratio: 0.69 (95% CI 0.27–1.75)

Absolute Event Rate: 6.7% vs 8%

p-value: p=0.44

Limitations

Observational design subject to inherent selection biases and residual unmeasured confounding
Potential residual anticoagulant interference during repeat lupus anticoagulant testing
Exclusion of 9 patients (2.1%) due to incomplete baseline data
Small sample size within specific subgroups limiting statistical power
Non-randomised concomitant use of single antiplatelet therapy and statins reflecting physician preference and disease severity
Observational study design subject to inherent selection biases and residual unmeasured confounding
Residual anticoagulant interference in lupus anticoagulant (LA) testing cannot be entirely excluded despite temporary replacement with LMWH
Exclusion of 9 of 422 eligible patients (2.1%) due to incomplete baseline data, which may introduce residual selection bias
Small sample size within specific subgroups limiting statistical power to detect modest differences
Concomitant single antiplatelet therapy (SAPT) and statin use was not randomised and may reflect physician preference and disease severity

Ask about this paper

Answers are grounded in this paper and cite their sources.

Implication

In patients with thrombotic antiphospholipid syndrome, rivaroxaban is associated with an increased risk of recurrent thrombosis in triple-positive aPL patients compared with warfarin, but demonstrates comparable effectiveness and safety in non-triple-positive patients, supporting a risk-stratified approach to anticoagulant selection.

Abstract

Objectives To compare the effectiveness and safety of rivaroxaban versus warfarin in antiphospholipid syndrome (APS) and assess baseline risk-stratified treatment effects.Methods Patients with thrombotic APS were treated with either rivaroxaban or warfarin and followed for recurrent thrombosis and major bleeding. Inverse probability of treatment weighting balanced baseline characteristics. HRs and 95% CIs were estimated using Cox models; prespecified subgroup analyses assessed effect modification.Results Among 413 patients with thrombotic APS, 89 (21.5%) received rivaroxaban, and 324 (78.5%) received warfarin. Median follow-up was 25.1 months (IQR, 8.5–44.4). Total follow-up was 198.5 patient-years for the rivaroxaban group and 833.1 patient-years for the VKA group. Recurrent thrombosis occurred in six rivaroxaban-treated patients (6.7%; 3.02/100 patient-years) and 26 warfarin-treated patients (8.0%; 3.12/100 patient-years). In weighted analyses, rivaroxaban was not associated with a significantly different risk of recurrent thrombosis (HR 0.69, 95% CI 0.27 to 1.75; p=0.44) or major bleeding (HR 0.68, 95% CI 0.15 to 3.03; p=0.61). The association between anticoagulant treatment and recurrent thrombosis differed significantly by triple-positive aPL status (P for interaction=0.007). Among triple-positive patients (n=199), rivaroxaban was associated with a higher risk of recurrent thrombosis than VKA (HR 3.65, 95% CI 1.10 to 12.13). In non-triple-positive patients (n=214), rivaroxaban showed a trend toward a lower risk of recurrent thrombosis (HR 0.16, 95% CI 0.02 to 1.17; p=0.071). Among 89 rivaroxaban-treated patients, no recurrent thrombotic events were observed in those with prior arterial thrombosis (0/22), and these patients more frequently received concomitant single antiplatelet therapy (72.7% vs 11.9%; p<0.001) and statins (36.4% vs 7.5%; p=0.002).Conclusion In this prospective cohort, no statistically significant differences in recurrent thrombosis or major bleeding were observed between rivaroxaban and warfarin. The treatment-by-triple-positivity interaction suggests that the comparative effectiveness of anticoagulants may differ according to aPL profile.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Li et al. (2026) conducted a cohort in Thrombotic antiphospholipid syndrome (APS) (n=413). Rivaroxaban vs. Warfarin (vitamin K antagonist) was evaluated on First recurrent thrombotic event (arterial or venous thrombosis) (HR 0.69, 95% CI 0.27 to 1.75, p=0.44). Rivaroxaban was not associated with a different overall risk of recurrent thrombosis compared to warfarin in thrombotic APS, though risk increased in triple-positive patients.

synapsesocial.com/papers/6af8f7b3016f830a1d01746f https://doi.org/10.1136/lupus-2026-002288

View Full Paper

Ask AI

Bookmark

Share

Risk-stratified effectiveness of direct oral anticoagulants versus vitamin K antagonists in thrombotic antiphospholipid syndrome: a prospective cohort study

Figures

Key Points

Study Design

Structured PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Risk-stratified effectiveness of direct oral anticoagulants versus vitamin K antagonists in thrombotic antiphospholipid syndrome: a prospective cohort study

Figures

Key Points

Study Design

Structured PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading