Optimizing early neurological deterioration prediction in acute ischemic stroke patients following intravenous thrombolysis: a LASSO regression model approach

Abstract

Background: Acute ischemic stroke (AIS) remains a leading cause of disability and mortality globally among adults. Despite Intravenous Thrombolysis (IVT) with recombinant tissue plasminogen activator (rt-PA) emerging as the standard treatment for AIS, approximately 6-40% of patients undergoing IVT experience Early Neurological Deterioration (END), significantly impacting treatment efficacy and patient prognosis.

Objective: This study aimed to develop and validate a predictive model for END in AIS patients post rt-PA administration using the Least Absolute Shrinkage and Selection Operator (LASSO) regression approach.

Methods: In this retrospective cohort study, data from 531 AIS patients treated with intravenous alteplase across two hospitals were analyzed. LASSO regression was employed to identify significant predictors of END, leading to the construction of a multivariate predictive model.

Results: Six key predictors significantly associated with END were identified through LASSO regression analysis: previous stroke history, Body Mass Index (BMI), age, Onset to Treatment Time (OTT), lymphocyte count, and glucose levels. A predictive nomogram incorporating these factors was developed, effectively estimating the probability of END post-IVT. The model demonstrated robust predictive performance, with an Area Under the Curve (AUC) of 0.867 in the training set and 0.880 in the validation set.

Conclusion: The LASSO regression-based predictive model accurately identifies critical risk factors leading to END in AIS patients following IVT. This model facilitates timely identification of high-risk patients by clinicians, enabling more personalized treatment strategies and optimizing patient management and outcomes.

Keywords: LASSO regression; acute ischemic stroke (AIS); early neurological deterioration (END); intravenous thrombolysis (IVT); predictive modeling.

FIGURE 1

Flow chart of the study design. This figure illustrates the process of participant selection, data collection, and analysis phases for the retrospective cohort study. It details the inclusion and exclusion criteria and the division of participants into training and validation sets.

FIGURE 2

(A) LASSO regression coefficient profiles: displaying the progression of coefficients of various predictors as the regularization parameter (lambda) is increased. Each line represents a different predictor variable in the LASSO regression model. (B) Selection of Lambda in LASSO Regression: This graph shows the cross-validation curve for model tuning. The lambda value with the minimal cross-validation error is highlighted, indicating the optimal level of penalization for the LASSO model.

FIGURE 3

Predictive nomogram for early neurological deterioration (END) in acute ischemic stroke patients following intravenous thrombolysis (IVT): this nomogram integrates key predictive factors identified through LASSO regression, including stroke history, BMI, Age, OTT, Glucose, and SII, to estimate the probability of END post-IVT. BMI, body mass index; OTT, onset to treatment time; SII, Systemic Immune Inflammation Index.

FIGURE 4

(A) Receiver operating characteristic (ROC) Curve of the predictive model in the training dataset: This figure depicts the ROC curve evaluating the performance of the predictive model for early neurological deterioration post-intravenous thrombolysis in the training dataset. It highlights the area under the curve (AUC) score, demonstrating the model’s discriminative capability. (B) Receiver operating characteristic (ROC) curve of the predictive model in the validation dataset: displayed here is the ROC curve assessing the efficacy of the predictive model for early neurological deterioration post-intravenous thrombolysis in the validation set. The area under the curve (AUC) value is provided as a measure of the model’s predictive accuracy.

FIGURE 5

(A) Calibration plot for the training dataset: This figure demonstrates the calibration of the predictive model in the training set, comparing predicted probabilities of early neurological deterioration post-intravenous thrombolysis with observed outcomes. (B) Calibration plot for the validation dataset: comparison of predicted probabilities of early neurological deterioration post-intravenous thrombolysis with actual outcomes in the validation dataset, indicating the model’s calibration accuracy.

FIGURE 6

(A) Decision curve analysis (DCA) for the training dataset: this analysis illustrates the net benefits of the predictive model in the training set across different threshold probabilities. (B) Decision curve analysis (DCA) for the validation dataset: the plot shows the clinical usefulness of the model in the validation dataset by assessing net benefits at various probability thresholds.