Clinical outcome of patients with mild pre-stroke morbidity following endovascular treatment: a HERMES substudy

Background Analyses of the effect of pre-stroke functional levels on the outcome of endovascular therapy (EVT) have focused on the course of patients with moderate to substantial pre-stroke disability. The effect of complete freedom from pre-existing disability (modified Rankin Scale (mRS) 0) versus predominantly mild pre-existing disability/symptoms (mRS 1–2) has not been well delineated. Methods The HERMES meta-analysis pooled data from seven randomized trials that tested the efficacy of EVT. We tested for a multiplicative interaction effect of pre-stroke mRS on the relationship between treatment and outcomes. Ordinal regression was used to assess the association between EVT and 90-day mRS (primary outcome) in the subgroup of patients with pre-stroke mRS 1–2. Multivariable regression modeling was then used to test the effect of mild pre-stroke disability/ symptoms on the primary and secondary outcomes (delta-mRS, mRS 0–2/5–6) compared with patients with pre-stroke mRS 0. Results We included 1764 patients, of whom 199 (11.3%) had pre-stroke mRS 1–2. No interaction effect of pre-stroke mRS on the relationship between treatment and outcome was observed. Patients with pre-stroke mRS 1–2 had worse outcomes than those with pre-stroke mRS 0 (adjusted common OR (acOR) 0.53, 95% CI 0.40 to 0.70). Nonetheless, a significant benefit of EVT was observed within the mRS 1–2 subgroup (cOR 2.08, 95% CI 1.22 to 3.55). Conclusions Patients asymptomatic/without disability prior to onset have better outcomes following EVT than patients with mild disability/symptoms. Patients with pre-stroke mRS 1–2, however, more often achieve good outcomes with EVT compared with conservative management. These findings indicate that mild preexisting disability/symptoms influence patient prognosis after EVT but do not diminish the EVT treatment effect.