There is an urgent need for non-invasive methods for predicting portal hypertensive gastropathy (PHG). This study aims to develop and validate a non-invasive method based on clinical parameters for predicting PHG in patients with liver cirrhosis (LC).
The overall survival (OS) and hepatocellular carcinoma (HCC)-free survival were evaluated in LC patients, both with and without PHG. A prediction model for PHG was then constructed based on a training dataset that contained data on 492 LC patients. The discrimination, calibration, and clinical utility of the predicting nomogram were assessed using the C-index, calibration plot, and decision curve analysis. Internal validation was conducted using a bootstrapping method, and further external validation using data on the 208 other patients.
LC patients with PHG had a worse prognosis compared with those without PHG. A nomogram was constructed using clinical parameters, such as age, hemoglobin content, platelet count and Child-Pugh class. The C-index was 0.773 (95% CI: 0.730–0.816) in the training cohort, 0.761 after bootstrapping and 0.745 (95% CI: 0.673–0.817) in the validation cohort. The AUC values were 0.767, 0.724, and 0.756 in the training, validation and total cohorts, respectively. Well-fitted calibration curves were observed in the training and validation cohorts. Decision curve analysis demonstrated that the nomogram was clinically useful at a threshold of 15%.
The nomogram constructed to predict the risk of developing PHG was found to be clinically viable. Furthermore, PHG is an independent risk factor for OS of LC, but not for the occurrence of HCC.