Recanalisation quality, not first-pass effect, drives outcomes in paediatric mechanical thrombectomy: a multicentre prospective study.

INTRODUCTION

In adult endovascular thrombectomy (EVT), first-pass success is associated with better outcomes, but whether this "first-pass effect" also holds true in childhood stroke remains unknown. We investigated the association of thrombectomy pass counts with 90-day outcomes after paediatric EVT.

PATIENTS AND

METHODS

We analysed prospectively collected data from the international Save ChildS Pro registry (ages 0-17 years). All EVT-treated patients were included irrespective of the reperfusion grade achieved. Successful reperfusion was defined as expanded thrombolysis in cerebral infarction (eTICI) ≥ 2b and complete/near-complete reperfusion as eTICI 2c-3. The primary outcome was ordinal 90-day mRS in a proportional odds model. Single-pass and multiple-pass patients were then compared with each other within the subgroup achieving eTICI 2c-3. Logistic regression for mRS > 2 was fitted with pass count and eTICI grade as covariates.

RESULTS

Of 103 patients (median age 11 years [IQR 6-15]; 55/102 [53.9%] male), median passes were 2 (IQR 1-3); 39 (37.9%) had 1 pass, 61 (59.2%) more than 1 and 3 (2.9%) recanalised without a thrombectomy pass. Multiple passes were associated with lower eTICI 2c-3 rates (30.0% vs 82.1%; P < .001), and worse 90-day mRS (common odds ratio 3.47; 95% CI, 1.63-7.38; P = .001). However, among patients achieving eTICI 2c-3 (n = 50), we detected no difference by pass count (mRS 0-2: 93.8% vs 94.4%; risk difference - 0.6%, 95% CI, -15% to +20%; P = 1.000). After accounting for eTICI, pass count was not independently associated with mRS > 2, while eTICI remained significant.

CONCLUSIONS

In this analysis of the first-pass effect in paediatric thrombectomy, multiple passes were associated with worse functional outcomes, but this effect appeared to be largely mediated through recanalisation quality. Subgroup analyses were underpowered to exclude a clinically relevant residual effect of pass count, and these findings should be considered hypothesis-generating.