Modeling and simulation analysis of the relationship between lesion recurrence on brain images and clinical recurrence in patients with ischemic stroke.

The objective of current study is to assess the relationship between characteristics of patients with acute ischemic stroke and clinical recurrences to identify predictors for the prognosis by modeling and simulation. Primary endpoint was clinical recurrence of ischemic stroke, and secondary endpoint was occurrence of any of the following clinical recurrence of ischemic stroke, transient ischemic attack, acute coronary syndrome, or vascular deaths. Time to event models were developed by NONMEM(®) using prospectively collected clinical data from 270 patients over 5 years, where 7.0% and 9.3% of them experienced lesion recurrence on MRI at 1 month (LR1M) and clinical recurrence, respectively. Exponential models best described the data. LR1M and diabetes mellitus history were significant predictors for primary endpoint. Times to recurrence for patients with LRIM (+) and diabetes mellitus (+) were predicted to be 0.095 and 0.317 of those for patients with LRIM (-) and diabetes mellitus (-), respectively. LR1M was only predictor for secondary endpoint with predicted time to recurrence in patients with LR1M (+) compared to 0.141 of LR1M (-). Quantitative prediction of clinical recurrence using MRI could improve personalized therapy by identifying patients at risk of recurrence, and could enable efficient clinical trials by stratifying the patients.