Abstract: Objective To analyze the risk factors for hypertension induced by apatinib, establish a prediction model and verify its accuracy. Methods The data of 168 inpatients treated with apatinib in Anqing Municipal Hospital between 2020 and 2023 was retrospectively analyzed. In addition, 121 of these cases treated between 2020 and 2022 were used as the training set while another 47 cases treated in 2023 were selected as the validation set. The independent risk factors for apatinib-induced hypertension were screened using the clinical information of patients in the training set. A prediction model was established and evaluated by the ROC curve. The accuracy of the model was verified using the information of patients in the validation set. Results A total of 51 patients (42.15%) in the training set developed hypertension. Compared with the non-hypertensive group, there was statistically significant difference in gender, age, rates of complications with hypertension, dose and the adoption of immunization and chemotherapy in the hypertension group (P <0.05). Binary logistic regression analysis showed that age, gender, complications with hypertension, and the combination of immunization and chemotherapy were independent risk factors for apatinib-induced hypertension. The area under the ROC curve of the prediction model was 0.850, the sensitivity was 72.50%, and the specificity was 84.30%. The 47 patients in the validation set were selected for cross-validation, and the accuracy of the prediction model was 87.23%. Conclusion Age, gender, complications with hypertension, and the combination of immunization and chemotherapy are independent risk factors for apatinib-induced hypertension. The fitting model can help predict the incidence of apatinib-induced hypertension.

Key words: Apatinib, Hypertension, Vascular Endothelial Growth Factor (Vegf), Risk Factors, Real-World Data, Logistic Regression, Prediction Model