Cangrelor Without Aspirin; DOACs for Cerebral Thrombosis: Biplane Weekly No. 5

An analysis of the French ETIS registry asked whether intravenous aspirin adds anything when cangrelor is given during thrombectomy. Among 644 patients in 34 centres who received intraoperative cangrelor between 2018 and 2025, 226 also had aspirin and 418 did not.

Favourable 90-day outcome, mRS 0–2 or equal to the pre-stroke score, was 57.1% with aspirin and 58.8% without (adjusted OR 0.93, 95% CI 0.67–1.30). Symptomatic intracranial haemorrhage was 6.2% versus 9.2% (adjusted OR 0.66, 95% CI 0.38–1.13) and mortality 22.3% versus 23.9% (adjusted OR 0.91), neither clearly different. In patients without pre-stroke antiplatelet therapy, symptomatic haemorrhage was 8.9% with cangrelor alone versus 4.6% with aspirin (adjusted OR 0.49, 95% CI 0.26–0.93), a finding the authors call unexpected and uncertain. The analysis was observational with propensity weighting, the interval for the primary outcome still allows moderate benefit or harm, and the authors do not claim equivalence.

The accompanying editorial frames peri-procedural antiplatelet therapy in thrombectomy as a field navigating limited evidence.

Two new analyses address anticoagulant choice in cerebral venous thrombosis, both observational.

A target trial emulation from a Chinese multicentre registry compared initial DOAC with warfarin in 2,476 patients (942 DOAC, 1,534 warfarin). Symptomatic recurrent venous thromboembolism at one year was 3.74% with DOACs and 5.47% with warfarin after overlap weighting, a risk difference of −1.73 percentage points (95% CI −3.76 to 0.30). That met the prespecified non-inferiority margin of 1.5 percentage points, but superiority was not shown. Major bleeding occurred in none of the DOAC patients and one warfarin patient; the authors treat the bleeding findings as exploratory. Most DOAC patients took rivaroxaban (701 of 942), so the result says little about apixaban or edoxaban.

A subanalysis of the CLOT-VENUS registry compared unfractionated heparin (220 patients) with enoxaparin (139) in the acute phase. Unfractionated heparin was associated with worse six-month outcome (adjusted OR 2.63 for mRS 3–6, higher meaning worse), though the interval was not reported in the abstract. Only 64% of aPTT values were in range, and the difference diminished when analyses were limited to patients with consistently therapeutic aPTT. That points to dosing quality as much as to the drug itself.

A Bayesian network meta-analysis pooled 15 randomised trials of reperfusion treatment 4.5–24 hours after onset, with eight trials in the primary network restricted to confirmed intracranial large-vessel or basilar occlusion. The authors compared thrombectomy, tenecteplase plus thrombectomy, alteplase and tenecteplase alone.

Thrombectomy against control gave an OR of 2.93 (95% CrI 1.40–6.53) for functional independence, mRS 0–2, a mortality OR of 0.71 and a higher symptomatic haemorrhage risk (OR 2.65, 95% CrI 1.28–5.81). Adding tenecteplase to thrombectomy was inconclusive: OR 1.10 (95% CrI 0.34–3.52) for tenecteplase plus thrombectomy versus thrombectomy alone, and tenecteplase alone was not clearly different from control.

The credible intervals are wide, and the abstract gives no patient totals. The result fits ATTENTION LATE, covered here on 12 October, in which tenecteplase before late basilar thrombectomy showed no benefit.

Two post hoc analyses of the ATTENTION trial of thrombectomy in basilar artery occlusion appeared this week. Both are exploratory.

The first compared baseline imaging in 340 patients, 277 with CT and 63 with MRI. Modality was not associated with 90-day mRS 0–3 (OR 0.99, 95% CI 0.50–1.95), and it did not significantly delay randomisation (mean difference 5.23 minutes) or puncture (13.16 minutes). With only 63 MRI patients and modality not randomised, the interval is wide.

The second split the 340 patients by onset-to-puncture time (under or from 6 hours) and by cause. Favourable outcome, mRS 0–3 at 90 days, was numerically more frequent with thrombectomy than with medical management in all four strata: under 6 hours, 46.7% versus 20.4% in non-atherosclerotic occlusion (RR 2.29, 95% CI 1.25–4.18) and 40.4% versus 26.1% in atherosclerotic occlusion; from 6 hours, 56.9% versus 26.3% in atherosclerotic and 39.5% versus 21.7% in non-atherosclerotic occlusion. Only the first of these was individually significant, and the three-way interaction was not (P=0.662), so the analysis gives no basis for different thrombectomy time windows by cause.

Two reports this week test how far MMA embolisation for chronic subdural haematoma extends beyond the randomised trials.

A 16-centre retrospective cohort of 600 patients with unilateral collections, analysed with propensity matching, compared standalone bilateral embolisation (54 patients) with unilateral (546). Rescue surgery during the index admission was 0% with bilateral and 9.5% with unilateral treatment (P<0.001), but a post hoc composite of any retreatment did not differ (P=0.37 unmatched). The bilateral group was small, had lower baseline severity and more general anaesthesia, and the authors do not support routine bilateral embolisation.

A single-centre cohort of 214 consecutive patients asked who the trials would have enrolled. Eligibility ranged from 6.1% for EMPROTECT and 13.5% for EMBOLISE to 78.9% for MEMBRANE; 28.0% met none of the five trials' criteria. Recurrence (16.8%), rescue surgery (16.8%) and complications (5.6%) did not differ by eligibility after correction for multiple testing.

A systematic review of five observational cohorts compared microsurgical clipping (402 patients) with Woven EndoBridge embolisation (396). In the good-quality concurrent comparisons, clipping was associated with higher angiographic occlusion (RR 1.30, 95% CI 1.14–1.49). Morbidity (RR 0.86, 95% CI 0.33–2.25) and mortality (RR 0.69, 95% CI 0.37–1.28) intervals were too wide to show a difference, and adverse-event reporting was too inconsistent for formal comparison.

Larger exploratory estimates, such as RR 1.71 for complete occlusion, come from fair-quality studies with external WEB comparators and ceiling-level clipping results, and likely inflate the gap. Follow-up duration is not stated in the abstract. The estimate that held in the better-quality comparisons, RR 1.30, is the one to weigh.

The Save ChildS Pro prospective registry reports 103 children aged 0–17 treated with thrombectomy (median age 11). Overall, multiple passes were associated with worse 90-day mRS (common OR 3.47, 95% CI 1.63–7.38), but multiple passes also reached eTICI 2c–3 less often (30.0% versus 82.1%).

Among the 50 children with eTICI 2c–3, good outcome (mRS 0–2) was 93.8% after a single pass and 94.4% after multiple passes, a risk difference of −0.6% (95% CI −15% to +20%). The authors conclude that the quality of the final recanalisation, rather than first-pass effect itself, drives outcome. The subgroup is too small to exclude a clinically relevant effect of pass count, but in these children extra passes that reached eTICI 2c–3 were not followed by a lower rate of good outcome than single-pass results.