SECRET
Year of Publication: 2023
Journal: Stroke
Link: https://doi.org/10.1161/STROKEAHA.123.044113
PDF: https://www.ahajournals.org/doi/10.1161/STROKEAHA.123.044113
PDF: Clinical Question
Is a randomized trial of rivaroxaban vs standard-of-care anticoagulation feasible in cerebral venous thrombosis, and what are the safety and patient-reported outcomes?
Study Overview
Objective
Rivaroxaban 20 mg daily vs standard-of-care anticoagulation (warfarin INR 2-3 or LMWH) — to assess feasibility of recruitment, safety, and functional outcomes in a Canadian phase II CVT trial.
Study Summary
- Phase 2 feasibility trial: 55 patients randomized at 12 Canadian centers, 57% consent rate among eligible patients.
- Primary composite safety outcome (mortality/sICH/major extracranial bleeding) occurred in 1 rivaroxaban vs 0 control patients.
- All participants achieved partial or complete recanalization by day 180.
- At enrollment, both arms showed impaired quality of life, mood, headache impact, fatigue, and cognition; all improved by day 180.
- Mean mRS improved for 81% rivaroxaban vs 67% control; by day 365, 83-84% were mRS 0-1.
- Recruitment feasibility demonstrated, but low event rates suggest a large international phase 3 would be required.
Intervention
Rivaroxaban 20 mg daily (or 15 mg if creatinine clearance <50 mL/min) vs standard-of-care anticoagulation (warfarin INR 2-3 with bridging parenteral anticoagulation until target, or low-molecular-weight heparin), for 180 days with optional extension to 365 days.
Patients per Arm
Rivaroxaban 26; Standard-of-care 27
Bottom Line
In 55 patients with symptomatic CVT at 12 Canadian centers randomized 1:1 to rivaroxaban 20 mg daily or standard-of-care (warfarin or LMWH), recruitment target was met (21.3 patients/year; 57% consent rate) with no major safety concerns. Numerically more bleeding with rivaroxaban (1 sICH, 2 CRNMB) but rates within prior CVT literature. Among the 24 patients per arm with follow-up imaging, all achieved at least partial recanalization by day 180; patient-reported outcomes (quality of life, mood, headache, fatigue, cognition) all improved over time. Feasibility confirmed but larger trials needed.
Major Points
- Multicenter Canadian phase 2 feasibility trial at 12 comprehensive stroke centers (Field 2023)
- N=55 randomized (27 to rivaroxaban, 28 to standard-of-care); 53 included in the mITT cohort (26 rivaroxaban, 27 standard-of-care) after 2 post-randomization exclusions (1 rivaroxaban withdrew prior to first dose; 1 control determined not to have CVT)
- Recruitment: 21.3 participants/year (feasibility met); 57% of eligible candidates consented
- Inclusion within 14 days of CVT diagnosis; no lead-in parenteral anticoagulation required (distinct from RE-SPECT CVT and EINSTEIN-Jr)
- 62.2% randomized within 1-4 days of diagnosis; 41.5% within 2 days
- Rivaroxaban started at 20 mg daily (no 15 mg BID loading typical for acute DVT/PE) — safety signal still favorable
- Primary composite safety outcome (mortality/sICH/major extracranial hemorrhage) at 180 d: 1 rivaroxaban (sICH) vs 0 standard
- Any intracranial hemorrhage and CRNMB numerically more frequent with rivaroxaban (consistent with ACTION-CVT and RE-SPECT signals)
- Among the 24 patients per arm with follow-up venography by day 180, all achieved at least partial recanalization (De Sousa grade 1B or better) — 24/24 (100%) in each arm
- Patient-reported outcomes: reduced QoL (EQ-5D), low mood (PHQ-9), high headache impact (HIT-6), fatigue (FAS), impaired cognition (MoCA) at baseline; all markedly improved by day 180
- By day 365, 82.6% (rivaroxaban) and 84.0% (control) achieved mRS 0-1
- Conclusion: feasibility established; larger DOAC-vs-warfarin trial impractical due to growing DOAC preference in practice; hierarchical composite win-ratio design suggested for future CVT trials
Design
Study Type: Phase 2 multicenter prospective open-label blinded-endpoint 1:1 parallel-group randomized feasibility trial (NCT03178864)
Randomization: 1
Blinding: Open-label; blinded central adjudication of primary endpoints
Follow-up Duration: 180 days primary; 365 days optional extension
Sample Size: 55
Analyzed: 53
Analysis: Descriptive statistics; feasibility = recruitment rate; composite safety endpoint; patient-reported outcome trajectories
Inclusion Criteria
- Age ≥18 years
- Within 14 days of imaging (CTV or MRV)-confirmed diagnosis of CVT
- Suitable for oral anticoagulation in judgment of treating physician
- Parenchymal brain scan (non-contrast CT or MRI) within 72 hours prior to randomization
- Informed consent from participant or legally authorized representative
Exclusion Criteria
- Known antiphospholipid antibody syndrome
- Anticipated invasive procedure (e.g., thrombectomy, hemicraniectomy)
- Unable to swallow due to depressed level of consciousness
- Estimated creatinine clearance <30 mL/minute
- Pregnancy or breastfeeding
- Another concurrent medical condition requiring mandatory antiplatelet or anticoagulant use
- Any contraindication to anticoagulation
- Concomitant use of strong CYP3A4 inhibitors
Baseline Characteristics
| Characteristic | Control | Active |
|---|---|---|
| N | 27 | 26 |
| Age | Median 48.0 years (IQR 39.0-58.0) | Median 48.5 years (IQR 37.0-64.25) |
| Female | 17/27 (63.0%) | 18/26 (69.2%) |
| Baseline imaging | CT/CTV 85.2%, MR/MRV 14.8% | CT/CTV 76.9%, MR/MRV 23.1% |
| NIHSS median (IQR) | 0 (0-1) | 0 (0-1) |
| NIHSS 0-4 | 25/27 (92.6%) [NIHSS 0: 18/27 (66.7%); NIHSS 1-4: 7/27 (25.9%)] | 22/26 (84.6%) [NIHSS 0: 15/26 (57.7%); NIHSS 1-4: 7/26 (26.9%)] |
| NIHSS 5-9 | 2/27 (7.4%) | 4/26 (15.4%) |
| Any intracranial hemorrhage at baseline (Heidelberg) | 8/26 (30.8%) | 13/25 (52.0%) |
| Symptom onset to enrollment, d (median, IQR) | 9.0 (3-14.0) | 8.0 (6.75-14.0) |
| Diagnosis to enrollment, d (median, IQR) | 3.0 (1-7) | 4.0 (2-6) |
| Received bridging anticoagulation before randomization | 26/27 (96.3%) | 25/26 (96.2%) |
| Presenting headache | 24/27 (88.9%) | 23/26 (88.5%) |
| Seizure at presentation | 10/27 (37.0%) | 13/26 (50.0%) |
| Oral contraceptive use | 8/27 (29.6%) | 7/26 (26.9%) |
| mRS ≥1 at baseline | 21/27 | 21/26 |
| Venous edema/infarct | 8/25 (32.0%) |
Arms
| Field | Control | Rivaroxaban |
|---|---|---|
| N | 27 | 26 |
| Intervention | Warfarin dosed to INR 2.0-3.0 with bridging parenteral anticoagulation (LMWH or UFH) for 2 consecutive days after INR reached target, OR LMWH alone per investigator; continued for 180 days with optional extension | Rivaroxaban 20 mg daily orally (15 mg if CrCl <50 mL/min), taken with food; no lead-in parenteral anticoagulation required; continued for 180 days with optional extension |
| Duration | 180 days + optional 185 additional days | 180 days + optional 185 additional days |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Primary feasibility: annual recruitment rate; primary safety: composite of all-cause mortality, symptomatic intracranial hemorrhage, or major extracranial hemorrhage at 180 days | Primary | 0/27 (0%, 95% CI 0-12.8) | 1/26 (3.8%, 95% CI 0.1-19.6) - one spontaneous subdural hemorrhage at 3.5 months | 3.8% difference in proportions (95% CI -3.5 to 11.2) | Descriptive (not powered for significance) |
| Any intracranial hemorrhage 180 d | Secondary | 0/27 (0%) | 2/26 (7.7%, 95% CI 0.9-25.1) | Diff 7.7% (95% CI -2.65 to 17.9) | Descriptive |
| Clinically relevant nonmajor bleeding 180 d | Secondary | 0/27 (0%) | 2/26 (7.7%) - menorrhagia, post-hysterectomy infection-related | Diff 7.7% | Descriptive |
| Combined major + CRNMB 180 d | Secondary | 0/27 (0%) | 3/26 (11.5%, 95% CI 2.45-30.1) | Diff 11.5% (95% CI -1.05 to 23.8) | Descriptive |
| VTE recurrence 180 d | Secondary | 0/27 (0%) | 1/26 (3.8%) - asymptomatic new CVT | Diff 3.8% | Descriptive |
| Any partial/complete recanalization by 180 d (De Sousa grade 1B+) | Secondary | 24/24 (100%) with 180-d imaging | 24/24 (100%) with 180-d imaging | No difference | Same rate |
| Complete recanalization by 180 d (De Sousa grade 3) | Secondary | 12/24 (50%) | 9/24 (37.5%) | Diff -12.5% (95% CI -40.35 to 15.3) | Descriptive |
| mRS 0-1 at 180 days | Secondary | 21/26 (80.8%) | 17/24 (70.8%) | Diff -10% (95% CI -37.0 to 9.1) | Descriptive |
| mRS shift improvement (≥1 category) among baseline mRS >0 at day 180 | Secondary | 14/21 (66.7%) | 17/21 (81.0%) | Diff 14.3% (95% CI -12.0 to 41.0) | Descriptive |
| mRS 0-1 at 365 days | Secondary | 21/25 (84.0%) | 19/23 (82.6%) | Diff -1.4% (95% CI -22.5 to 19.7) | Descriptive |
| EQ-5D-5L mean change baseline to day 180 | Secondary | 0.73 → 0.89 | 0.76 → 0.89 | Mean diff 0.01 (95% CI -0.08 to 0.09) | Descriptive |
| PHQ-9 mean change baseline to day 180 | Secondary | 9.15 → 3.80 | 6.52 → 3.81 | Mean diff -2.55 (95% CI -5.46 to 0.36) | Descriptive |
| HIT-6 headache impact mean change baseline to day 180 | Secondary | 56.78 → 44.08 | 59.70 → 47.00 | Mean diff 0.9 (95% CI -5.83 to 7.63) | Descriptive |
| FAS fatigue mean change baseline to day 180 | Secondary | 25.83 → 19.72 | 23.21 → 18.57 | Mean diff -1.72 (95% CI -8.02 to 4.57) | Descriptive |
| MoCA <26 at baseline (impaired cognitive performance) | Secondary | 16/24 (66.7%) | 10/17 (58.8%) | Baseline impairment; not an adverse event | Descriptive |
| All-cause mortality 180 d | Adverse | 0/27 | 0/26 | No deaths in either arm | |
| All-cause mortality 365 d | Adverse | 1/27 (3.7%) | 0/26 | 1 unrelated death (during separate admission after scalp bleed) | |
| Symptomatic intracranial hemorrhage 180 d | Adverse | 0 | 1 (spontaneous subdural, surgical evacuation at 3.5 mo) | Isolated event | |
| Major extracranial hemorrhage 180 d | Adverse | 0 | 0 | None | |
| Any intracranial hemorrhage 365 d | Adverse | 1/27 (3.7%) | 2/26 (7.7%) | Diff 4.0% (95% CI -8.55 to 16.5) | Descriptive |
| Asymptomatic hemorrhagic progression (any imaging) | Adverse | Not separately reported | 1 worsening of baseline hemorrhagic lesion at 24 h (asymptomatic) | Early safety signal | |
| Symptomatic traumatic subdural hemorrhage 365 d (additional, control) | Adverse | 1 (with scalp bleeding, later unrelated death) | 0 | Trauma-related | |
| CRNMB events (cumulative to 365 d) | Adverse | 1 | 2 | Diff 4.0% (95% CI -8.55 to 16.5) | Numerically more with rivaroxaban |
| VTE recurrence 365 d | Adverse | 1/27 (3.7%) - ischemic stroke in patient with JAK2 V617F mutation switched to rivaroxaban | 1/26 (3.8%) - asymptomatic new CVT | Diff 0.1% (95% CI -10.15 to 10.4) | Descriptive |
Subgroup Analysis
Patient-reported outcomes improved similarly in both arms: mean EQ-5D-5L rose 0.76→0.89 (rivaroxaban) and 0.73→0.89 (control) by day 180; PHQ-9 mood scores fell from mild-to-moderate depression ranges to normal; HIT-6 headache impact fell from severe to mild-moderate; FAS fatigue scores improved similarly. MoCA cognition scores were impaired at baseline (approximately half scored <26) but high rates of follow-up non-completion limit conclusions. The primary pattern: CVT causes substantial early symptom burden that largely resolves over 6 months regardless of anticoagulant choice. The paper reports no notable differences between rivaroxaban and control in outcome trajectories; no formal effect-modification analysis by baseline intracranial hemorrhage was reported. The rivaroxaban arm did have numerically more baseline hemorrhage (52% vs 31%), reflecting random variation in small samples.
Criticisms
- Small sample size (n=55) precludes superiority/noninferiority conclusions; powered only for feasibility
- Rivaroxaban 20 mg once daily (vs 15 mg BID standard for DVT/PE) — pharmacokinetic modeling supports this but direct comparison lacking
- High rate of missing cognitive data on follow-up limits MoCA trajectories
- Numerically higher bleeding with rivaroxaban (1 sICH, 2 CRNMB) warrants attention but confidence intervals are very wide
- Baseline characteristics unbalanced (52% vs 31% intracranial hemorrhage in rivaroxaban vs control) reflects small sample and may influence outcomes
- Open-label design with blinded endpoint adjudication partially mitigates bias but does not fully eliminate it
- Canadian single-country enrollment; generalizability limited. Bayer Canada provided rivaroxaban in-kind but had no role in study conduct or analysis; primary funding was from independent Canadian public/foundation sources
Funding
Canadian Institutes of Health Research; Canadian Partnership for Stroke Recovery; Canadian Stroke Consortium; Heart and Stroke Foundation of Canada; Michael Smith Health Research BC; Vancouver Coastal Health Research Institute. Design supported by the Clinical Trials Methodology Course (NIH-NINDS R25NS088248). Bayer Canada provided rivaroxaban in-kind but had no role in the conduct of the study or the analysis.
Based on: SECRET (Stroke, 2023)
Content summarized and formatted by NeuroTrials.ai.