Higher Nutritional Indices Linked to Lower Bleeding Risk After Stroke Thrombectomy

In acute ischemic stroke patients undergoing mechanical thrombectomy, higher baseline nutritional status is associated with significantly lower odds of developing hemorrhagic transformation, according to a retrospective cohort study published in BMC Neurology.

Researchers evaluated clinical registry records from 236 acute ischemic stroke patients who received mechanical thrombectomy between November 2018 and May 2023 at the Affiliated Hospital of Xuzhou Medical University.

Patient nutritional status was assessed through two clinical metrics: the prognostic nutritional index, or PNI, and the geriatric nutritional risk index, or GNRI.

Follow-up computed tomography performed between 24 and 36 hours after mechanical thrombectomy classified patients by the presence or absence of hemorrhagic transformation, which was documented in 45 of the 236 patients.

To control for baseline confounding, investigators conducted propensity-score matching based on admission National Institutes of Health Stroke Scale scores, Alberta Stroke Program Early CT Scores, and successful recanalization.

In separate multivariable logistic regression models further adjusted for eosinophil count and thrombin time to address residual confounding after matching, higher PNI was associated with decreased odds of hemorrhagic transformation (odds ratio per unit increase, 0.831; 95% confidence interval, 0.762–0.907; P < 0.001). Higher GNRI was likewise associated with reduced odds of hemorrhage (odds ratio per unit increase, 0.935; 95% confidence interval, 0.895–0.977; P = 0.003).

Receiver operating characteristic curve analyses evaluated the discrimination of both indices for hemorrhagic transformation, demonstrating an area under the curve of 0.768 (95% confidence interval, 0.678–0.857) for PNI and 0.693 (95% confidence interval, 0.590–0.796) for GNRI.

The study authors noted that because the investigation was a retrospective cohort analysis, these associations support PNI and GNRI as potential risk-stratification markers but do not establish causality, highlighting the need for larger prospective studies with external validation.

The investigation evaluated the relationship between nutritional status and hemorrhagic transformation following mechanical thrombectomy in patients with acute ischemic stroke. Drawing on clinical registry data from the Affiliated Hospital of Xuzhou Medical University collected between November 2018 and May 2023, the researchers studied a cohort of 236 patients undergoing mechanical thrombectomy. Within this cohort, 45 patients developed post-procedural hemorrhagic transformation. Follow-up computed tomography performed between 24 and 36 hours after mechanical thrombectomy provided the radiological basis for classifying each patient by the presence or absence of hemorrhagic transformation.

To evaluate baseline nutritional status, the study utilized the prognostic nutritional index and the geriatric nutritional risk index. To address confounding differences among patients, the team implemented propensity-score matching based on three key parameters: admission National Institutes of Health Stroke Scale scores, Alberta Stroke Program Early CT Scores, and successful recanalization.

After propensity-score matching, investigators constructed separate multivariable logistic regression models for the prognostic nutritional index and the geriatric nutritional risk index. These models incorporated additional adjustments for eosinophil count and thrombin time to manage residual confounding remaining after matching. In the adjusted regression analyses, higher scores on both nutritional indices were associated with reduced odds of hemorrhagic transformation. Each unit increase in the prognostic nutritional index corresponded to an odds ratio of 0.831 (95% confidence interval, 0.762–0.907; P < 0.001). For the geriatric nutritional risk index, each unit increase corresponded to an odds ratio of 0.935 (95% confidence interval, 0.895–0.977; P = 0.003).

Receiver operating characteristic curve analyses evaluated the discrimination of both indices for post-thrombectomy hemorrhagic transformation. The prognostic nutritional index achieved an area under the receiver operating characteristic curve of 0.768, with a 95% confidence interval of 0.678 to 0.857. Meanwhile, the geriatric nutritional risk index demonstrated an area under the receiver operating characteristic curve of 0.693, with a 95% confidence interval of 0.590 to 0.796. These statistical analyses quantified how effectively each metric discriminated between patients who experienced hemorrhagic transformation and those who did not.

Regarding study limitations and clinical interpretation, the authors noted that lower nutritional indices were associated with hemorrhagic transformation in this retrospective cohort. Because the study analyzed retrospective registry data, these documented associations support the prognostic nutritional index and the geriatric nutritional risk index as potential risk-stratification markers but do not establish causality. The investigators stated that larger prospective studies with external validation are required to verify these findings.

Subject of Research: Nutritional status and hemorrhagic transformation risk after mechanical thrombectomy in acute ischemic stroke

Article Title: Nutritional status and risk of hemorrhagic transformation in acute ischemic stroke patients treated with mechanical thrombectomy

Article References: Zhang, W., Zhang, W., Ding, C., Yao, Q., Yang, X., & Wang, Y. (2026). Nutritional status and risk of hemorrhagic transformation in acute ischemic stroke patients treated with mechanical thrombectomy. BMC Neurology. https://doi.org/10.1186/s12883-026-05444-w

Image Credits: AI Generated

DOI: 10.1186/s12883-026-05444-w

Keywords: acute ischemic stroke, mechanical thrombectomy, hemorrhagic transformation, prognostic nutritional index, geriatric nutritional risk index, nutritional status, computed tomography, propensity-score matching, NIHSS, ASPECTS, recanalization, multivariable logistic regression, eosinophil count, thrombin time, receiver operating characteristic, stroke neurology