Good collaterals and better outcomes after EVT for basilar artery occlusion: A systematic review and meta-analysis

Background: Stroke caused by acute basilar artery occlusion (BAO) is devastating with high dependency and mortality. Recent trials have demonstrated the efficacy of endovascular treatment (EVT) for acute BAO, while pretreatment collaterals may be a valuable prognostic indicator for post-EVT outcomes. Aims: To systematically review and synthesize evidence on the associations between pretreatment collateral status and outcomes after EVT in acute BAO. Methods: We retrieved relevant full-text articles published in English since 1 January 2010, reporting associations between pretreatment collateral status and outcomes after EVT for BAO, by searching MEDLINE and Embase. The primary outcome was favorable or good 90-day functional outcome (modified Rankin Scale [mRS] 0–2 or 0–3); secondary outcomes included successful recanalization, symptomatic intracranial hemorrhage, final infarct volume, and 90-day mortality. Risk ratios (RRs) with 95% confidence intervals (CIs) for good versus poor collaterals on the outcomes were synthesized using random-effects models. Subgroup and sensitivity analyses were conducted for the primary outcome. Results: Overall, 29 primary studies (2995 participants) were included in qualitative review, among which 16 studies (1447 participants) were meta-analyzed. With different imaging modalities and methods to grade the collateral status, good collaterals were found in 33–85% of patients in the individual primary studies (I2 = 95.2%, p < 0.001), with a pooled proportion of 51% (95% CI: 40–62%) across all studies. Good pretreatment collaterals were associated with a doubled rate of favorable/good 90-day functional outcome (RR = 2.03, 95% CI: 1.63–2.51, p < 0.001), a higher rate of successful recanalization (RR = 1.23, 95% CI: 1.04–1.45, p = 0.015), and reduced 90-day mortality (RR = 0.59, 95% CI: 0.43–0.81, p = 0.001) after EVT for BAO. None of the primary studies reported the associations of good collaterals with the other secondary outcomes. Subgroup analyses revealed possibly more prominent protective effect of good pretreatment collaterals over the primary outcome, in studies with longer time windows in patient eligibility criteria for EVT (p = 0.028 for between-subgroup heterogeneity). Conclusions: In patients with BAO receiving EVT, good pretreatment collateral status was associated with a higher chance of favorable 90-day functional outcome, despite the various methods in grading the collateral circulation. Efforts are needed for more standardized collateral assessment in BAO, for more reliable and generalizable investigations of its clinical implications.