Where the Clot Sits Shapes How Patients Die After Stroke Thrombectomy

Patients whose large-vessel strokes strike the back of the brain face a markedly different pattern of fatal complications after endovascular treatment than those whose clots sit in the front, according to a multicenter cohort study drawing on the ANGEL-ACT registry, which tracked consecutive patients treated at 111 hospitals across China. Among 1,671 eligible patients with acute ischemic stroke caused by large vessel occlusion—1,300 with anterior circulation occlusions and 371 with posterior circulation occlusions—228 deaths, or 13.6 percent, occurred within 90 days of treatment. All-cause mortality was substantially higher in the posterior circulation group, at 19.4 percent versus 12.0 percent in the anterior group.

The study, published in Frontiers in Neurology, grouped adjudicated causes of death into prespecified categories and applied the Fine-Gray subdistribution hazard model, censoring survivors at 90 days and treating deaths from other causes as competing events. Cumulative incidence functions were estimated with the Aalen-Johansen estimator. The starkest divergence appeared in deaths attributed to cerebral edema.

The 90-day cumulative incidence of cerebral edema-related death was 9.16 percent in posterior circulation occlusions versus 4.15 percent in anterior circulation occlusions. After adjustment, the subdistribution hazard ratio for anterior versus posterior circulation was 0.47, with a 95 percent confidence interval of 0.30 to 0.73 and a P value of 0.001; Gray's test yielded a P value below 0.001. Patients with posterior circulation occlusions thus faced roughly double the risk of dying from brain swelling, an association that held up statistically in the complete cohort.

Fatal hemorrhagic transformation, the conversion of a blocked-vessel stroke into a bleeding one, told the opposite story numerically but not statistically. It was more frequent in anterior circulation occlusions, at 3.77 percent versus 1.89 percent in posterior circulation occlusions, yet the association did not reach significance in the complete cohort: the adjusted subdistribution hazard ratio was 1.76 with a 95 percent confidence interval of 0.78 to 3.98 and a P value of 0.17, while Gray's test returned P equal to 0.076. This pattern was consistent across all sensitivity analyses, meaning hemorrhagic transformation accounted for a larger share of anterior circulation deaths without corresponding to a statistically significant increase in risk in the complete cohort.

Two other causes also split along territorial lines. Pneumonia-related death was significantly more frequent in posterior circulation occlusions, with an adjusted subdistribution hazard ratio of 0.34 for anterior versus posterior circulation, a 95 percent confidence interval of 0.17 to 0.68, and a P value of 0.003; this association remained significant after Bonferroni correction for multiple comparisons. Recurrent stroke-related death showed a similar direction, with an adjusted subdistribution hazard ratio of 0.29, a 95 percent confidence interval of 0.10 to 0.86, and a P value of 0.026, but this association did not survive Bonferroni correction, so the authors treat it with appropriate caution.

The authors conclude that fatal complications after endovascular treatment differ by occlusion territory, and that posterior circulation occlusion is independently associated with a higher risk of death from cerebral edema, pneumonia, and—although the association did not survive Bonferroni correction—recurrent stroke. Hemorrhagic transformation accounted for a larger share of anterior circulation deaths, but this did not correspond to a statistically significant increase in risk in the complete cohort. The study was designed to quantify these differences rather than to explain the mechanisms behind them, and the reported results are framed in terms of association rather than mechanism.

Several limitations shape how the findings should be read. The study was a retrospective multicenter cohort based on registry data rather than randomized assignment, so unmeasured differences between anterior and posterior circulation patients could contribute to the observed gaps. The posterior circulation group was also much smaller, 371 patients versus 1,300, and the Bonferroni correction, a conservative adjustment for testing multiple causes of death, further narrows what can be claimed with confidence: only the cerebral edema and pneumonia associations survived it.

What the study establishes is that the territory of a large vessel occlusion is associated not merely with whether patients die after endovascular treatment but with how they die within 90 days.

Subject of Research: Cause-specific 90-day mortality after endovascular treatment for anterior versus posterior circulation large vessel occlusion

Article Title: Cause-specific mortality after endovascular treatment for anterior vs. posterior circulation large vessel occlusion: a multicenter cohort study

Article References: Li, M., Wan, Y., Mo, Y., Li, J., Zhang, X., He, X., Chen, H., Huang, M., Guo, J., Gao, F., Mo, D., Ma, N., Miao, Z., Wang, A., & Deng, Y. (2026). Cause-specific mortality after endovascular treatment for anterior vs. posterior circulation large vessel occlusion: a multicenter cohort study. Frontiers in Neurology, 17, Article 1948536. https://doi.org/10.3389/fneur.2026.1948536

Image Credits: AI Generated

DOI: 10.3389/fneur.2026.1948536

Keywords: acute ischemic stroke, large vessel occlusion, endovascular treatment, posterior circulation, cerebral edema, hemorrhagic transformation, pneumonia, competing risks, Fine-Gray model, ANGEL-ACT registry, cause-specific mortality, Cause-specific