Risk factors of unexplained early neurological deterioration after treatment for ischemic stroke due to large vessel occlusion: a post hoc analysis of the HERMES study

R. Bourcier*, M. Goyal, K.W. Muir, H. Desal, D.W.J. Dippel, C.B.L.M. Majoie, W.H. van Zwam, T.G. Jovin, P.J. Mitchell, A.M. Demchuk, R.J. van Oostenbrugge, S.B. Brown, B. Campbell, P. White, M.D. Hill, J.L. Saver, C. Weimar, R. Jahan, F. Guillemin, S. BracardO. Naggara, HERMES Trialists Collaboration

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

3 Citations (Web of Science)


Background Early neurological deterioration (END) after endovascular treatment (EVT) in patients with anterior circulation acute ischemic stroke (AIS) is associated with poor outcome. END may remain unexplained by parenchymal hemorrhage (UnEND). We aim to analyze the risk factors of UnEND in the medical management (MM) and EVT arms of the HERMES study. Methods We conducted a post-hoc analysis of anterior AIS patients who underwent EVT for proximal anterior occlusions. Risk factors of UnEND, defined as a worsening of >= 4 points between baseline National Institutes of Health Stroke Scale (NIHSS) and NIHSS at 24 hours without hemorrhage, were compared between both arms using mixed logistic regression models adjusted for baseline characteristics. An interaction analysis between the EVT and MM arms for risk factors of UnEND was conducted. Results Among 1723 patients assessable for UnEND, 160 patients experienced an UnEND (9.3%), including 9.1% (78/854) in the EVT arm and 9.4% (82/869) in the MM arm. There was no significant difference in the incidence of UnEND between the two study arms. In the EVT population, independent risk factors of UnEND were lower baseline NIHSS, higher baseline glucose, and lower collateral grade. In the MM population, the only independent predictor of UnEND was higher baseline glucose. However, we did not demonstrate an interaction between EVT and MM for baseline factors as risk factors of UnEND. UnEND was, similarly in both treatment groups, a significant predictor of unfavorable outcome in both the EVT (p<0.001) and MM (p<0.001) arms. Conclusions UnEND is not an uncommon event, with a similar rate which ever treatment arm is considered. In the clinical scenario of AIS due to large vessel occlusion, no patient-related factor seems to increase the risk for UnEND when treated by EVT compared with MM.
Original languageEnglish
Pages (from-to)221-226
Number of pages6
JournalJournal of Neurointerventional Surgery
Issue number3
Early online date15 Feb 2022
Publication statusPublished - Mar 2023


  • stroke
  • thrombectomy
  • thrombolysis
  • TIME
  • PA

Cite this