Integrating novel neuroimaging measurements and circulating biomarkers for the prediction of secondary injury following stroke. the NIMBLE study.

ABSTRACT INTRODUCTION Cerebral edema (CE) and haemorrhagic transformation (HT) induce brain swelling and represents the most feared complications of reperfusion injury in acute ischemic stroke. Haemorrhagic conversion is particularly concerning due to its potential for rapid clinical deterioration or mortality. Additionally, cerebral oedema, while less immediately alarming, can significantly impact patient outcomes. Optimizing the measurement of these complications and identifying early predictive blood biomarkers are essential steps in preventing and effectively treating these harmful phenomena, which can hinder the successful recanalization of occluded vessels in stroke patients. This thesis aims to enhance understanding and prediction of these events through the application of an innovative methods to measure brain swelling on CT and MRI scans, and the evaluation of circulating biomarkers as early indicators of blood-brain barrier disruption and outcome predictors. By optimizing measurement techniques and exploring predictive biomarkers, the thesis seeks to improve the identification and quantification of complications following revascularization treatments, ultimately guiding personalized strategies to prevent and treat adverse outcomes in acute ischemic stroke patients. METHODS This is a monocentric prospective observational study conducted in Careggi University Hospital recruiting between 04/24/2021 and 06/30/2023 with acute (within 12 hours form symptoms onset) IS of anterior circulation irrespective of reperfusion therapies. Blood samples obtained in all patients at hospital presentation included: Serum-Amyloid-A, alfa-2 macroglobuline (A2M), ultrasensitive CRP, S100, Neuron-Specific Enolase (NSE). The functional outcome was poor prognosis of the patient at 3 months, defined as mRS >2. In a subgroup of patients studied with both CT scan, at baseline and 24 hours follow-up, and brain MRI performed 5 days after symptoms onset, a novel approach to evaluate brain swelling (including CE and HT) called quantification of Anatomical Distortion (AD) has been applied. In details, acute ischemic lesions have been manually defined on follow CT scan and on brain MRI (DWI sequences) using the masking tool in FSLeyes, a software included in the FMRIB Software Library (FSL). A three dimensions lesion mask for each patient, covering the infarcted tissue on all CT scan slices has been thus generated. The image registration approach described by Harston et al. (2018) to quantify AD at 24 hours and after 5days in CT and MRI scans respectively has been applied. In both neuroimaging follow-up timepoints, subtracting AD from total lesion volume, we therefore obtained also the corrected infarct volume in both neuroimaging follow-up timepoints. To adjust for ischemic lesion dimension, anatomical distortion (AD) was divided by the corrected infarct volume to obtain a value of relative AD (relAD). Univariate linear regression analyses were conducted to evaluate the association between each radiological outcome (AD, RelAD and corrected infarct volume, obtained in CT and MRI follow-up scans) and selected baseline variables (including circulating biomarkers). Independent variables associated with the outcome at univariate analysis were included in multivariate models. In another analysis, univariate logistic regression models were created to assess the association between baseline biomarkers and clinical variables and the dichotomized mRS score recorded at three months. The risk factors found to be significant on a univariate analysis will enter into a multiple logistic regression model with a backward selection criterion. RESULTS Our study contributes valuable insights into the prognostic value of various factors in ischemic stroke. Notably, we found a strong association between lower ASPECTS and larger corrected infarct volumes, highlighting ASPECTS' clinical relevance as a predictor of stroke severity. Furthermore, baseline NSE levels showed significant associations with corrected lesion volumes, supporting previous literature on NSE's role as a marker of neuronal damage. Additionally, we observed that total anterior circulation infarction clinical presentation (TACI) based on OCSP classification was associated with corrected lesion volume, suggesting TACI's potential as a predictor of lesion size. Our study revealed significant associations between lower ASPECTS, higher NSE levels, and greater anatomical distortion volumes. Moreover, our multivariate analysis with relAD as outcome uncovered a significant association between alpha 2 macroglobulin levels and the extent of anatomical distortion, shedding light on the potential role of these biomarkers in modulating brain structural changes post-stroke and reperfusion injury. In a logistic regression analysis, we assessed the predictive role of pre-stroke mRS, NIHSS at baseline, and baseline blood biomarkers levels, revealing that higher pre-stroke mRS and NIHSS scores, along with increased levels of A2M and S100B, were significantly associated with a higher risk of poor functional outcome 3 months after stroke, corroborating previous studies' findings and emphasizing the importance of incorporating these factors into clinical practice for risk stratification and treatment decision-making in acute stroke management. CONCLUSION: In conclusion, our study demonstrates the stroke relations existing between NSE (and ASPECTS) and AD and corrected lesion volume, and highlighted the predictive value of pre-stroke mRS, NIHSS at baseline, and baseline blood biomarkers (A2M and S100B) for poor outcomes following ischemic stroke. These findings align with previous literature, highlighting the importance of incorporating these factors into clinical practice for risk stratification and treatment decision-making. Understanding the prognostic significance of these variables can enhance stroke outcome prediction models and aid in personalized patient management strategies. Further prospective studies with larger cohorts are warranted to validate these findings and elucidate the underlying mechanisms driving these associations.