Post-Endovascular Thrombectomy Perfusion Imaging (CT and MR) as a Prognostic Biomarker and Patient-Selection Tool for Adjuvant Therapy: A Systematic Review, Meta-Analysis, and Clinical Threshold Framework.

BACKGROUND AND PURPOSE: Despite high rates of macrovascular recanalization, approximately half of patients with large vessel occlusion stroke fail to achieve functional independence after endovascular thrombectomy (EVT). Residual tissue-level perfusion abnormalities - detectable on post-procedural CT perfusion (CTP) - have been proposed as a biomarker of futile recanalization. We aimed to establish post-EVT perfusion imaging (by CT or MR perfusion) as a surrogate marker predictive of functional independence on the modified Rankin Scale - by synthesizing evidence-based perfusion thresholds and defining the optimal acquisition timing. Analysis was restricted to studies performing perfusion imaging exclusively after EVT, so that thresholds reflect post-procedural reperfusion status rather than pre-treatment ischemic burden. METHODS: We systematically searched MEDLINE, EMBASE, and the Cochrane Library from January 2018 through April 2026 for cohort studies performing perfusion imaging after EVT, reporting at least one quantitative CT- or MR-perfusion parameter alongside functional outcome (modified Rankin Scale [mRS] at any post-procedural time point) or neurological outcome (National Institutes of Health Stroke Scale [NIHSS]), and enrolling ≥10 patients. Both anterior and posterior circulation occlusions were eligible. The review was conducted in accordance with PRISMA 2020 (checklist provided as supplementary material). Risk of bias was assessed using the ROBINS-I tool. Pooled odds ratios (ORs) were estimated using the DerSimonian-Laird random-effects estimator. A pre-specified sensitivity analysis excluded our proof-of-concept cohort (which received adjunctive cerebrolysin) to isolate the pooled estimate from any treatment-related influence. RESULTS: Eight independent post-EVT perfusion imaging studies (670 patients; five CT-perfusion, two MR-perfusion, and one mixed CT/MR cohort) met inclusion criteria. Residual hypoperfusion was present in 21-53% of angiographically successful reperfusions and was consistently associated with reduced odds of functional independence (pooled OR 0.28, 95% CI 0.15-0.51; I²=6%). The pre-specified sensitivity analysis excluding the cerebrolysin-treated cohort yielded a pooled OR of 0.30 (95% CI 0.16-0.54; I²=0%), confirming that the pooled effect is not driven by the adjuvant-treated subgroup. A Tmax >6 s volume <3.5 mL at 30-90 minutes post-EVT was the most consistently validated CT-perfusion threshold (adjusted OR 3.5, 95% CI 1.6-7.8); for MR perfusion, an rCBV-defined impaired-microvascular-reperfusion volume >5 mL was the corresponding threshold. The optimal perfusion acquisition window is 30-120 minutes post-EVT. CONCLUSIONS: Post-EVT perfusion imaging (CT or MR) provides tissue-level reperfusion information that complements modified Thrombolysis in Cerebral Infarction (mTICI) grading and identifies biologically distinct patient subgroups. Standardized post-EVT perfusion imaging at 30-120 minutes, applied with the proposed threshold framework, should be considered as an eligibility criterion and stratification variable in future trials of post-EVT adjuvant therapy across the pharmacological spectrum.