Cardiogenic Parameters Effectively Predict Hemorrhagic Transformation in Patients with Non-Large Artery Atherosclerosis Infarction

Abstract Background: Non-large artery atherosclerosis (LAA) infarcts are relatively rare compared to LAA type infarcts. Hemorrhagic transformation (HT) is one of the complications of acute ischemic stroke (AIS), and we aim to investigate the risk factors for the development of HT in patients with non-LAA type AIS.Patients and methods：From January 2015 to January 2020, we included a total of 52 patients with non-LAA type AIS who met the criteria for the occurrence of HT, for which 136 patients without HT were matched. Patients were followed up every 3 months by phone or in the clinic, with a minimum of 1 year of follow-up per patient. The risk factors associated with prognosis were derived after univariate analysis and multifactorial logistic regression analysis, and a nomogram model was developed based on these risk factors. The accuracy of the model was evaluated by creating a calibration plot and the receiver operating characteristic curve (ROC).Results: Through univariate analysis and logistics regression analysis, we found that the patient's creatine kinase-MB (CK-MB), ejection fraction (EF) and platelets (PLT) were independently related to the HT in patients with non-LAA type AIS. We accordingly developed smoothing curves to predict the probability of HT and found that the HT rate increased with increasing CK-MB and decreased with increasing EF and PLT. The nomogram based on these three factors had an area under curve (AUC) of 0.875 for the development group and 0.889 for the external validation group. The calibration plot showed a good prediction accuracy.Conclusions: Patients with non-LAA type AIS are relatively uncommon and presenting with HT is even rarer. Our nomogram built on cardiogenic parameters and PLT can accurately predict the occurrence of HT in these non-LAA type AIS patients, which has good clinical significance.