Rupture précoce de la barrière hémato-encéphalique après un infarctus cérébral : méthodologie d’évaluation, prédicteurs et évolution clinique
Year: 2026
Type: dissertation
Abstract: This thesis investigates the clinical significance of post-contrast FLAIR (pcFLAIR) enhancement of cerebrospinal fluid (HARM) and ocular structures (GLOS), which reflect early disruption of brain barriers following stroke, by combining a systematic review with an analysis of a thrombectomy cohort. The review (38 studies) shows that the timing of assessment is the main determinant of what pcFLAIR reveals: early on, HARM—especially sulcal—is consistent with acute opening of the blood–brain barrier; later, the imaging primarily reflects gadolinium redistribution. It highlights the heterogeneity of definitions and assessment protocols, and for the first time proposes a model integrating the glymphatic system into the observed enhancement patterns. In the retrospective study conducted on the ETIS registry (n = 229), HARM and GLOS were common (˜49% and 42%). Their detection increased with age, higher serum creatinine levels, and a shorter interval between baseline FLAIR and pcFLAIR; conversely, single-pass recanalization reduced the frequency of HARM, suggesting less endothelial injury. After adjustment, HARM was independently associated with early neurological deterioration and ischemic growth, with no robust link to hemorrhagic transformation or 90-day outcome; GLOS had no independent prognostic value. No association was found between HARM and poor collateral status, nor between the Fazekas score and HARM/GLOS. Overall, HARM appears to be a relevant early marker in the thrombectomy setting, whereas GLOS likely reflects clearance dynamics. Prospective multi-timepoint protocols (early and delayed pcFLAIR), integrating perfusion imaging (PWI) and DTI-ALPS, could help distinguish leakage from delayed clearance and refine the prognostic value of these markers beyond a binary interpretation.
Source: HAL
Author: Bousfiha, Camile
Institution: Université Paris Cité
Language: French
Subfield: Epidemiology
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7141778666