RESUMEN
OBJECTIVES: Portal hypertensive polyps (PHPs) are incompletely characterized lesions that can be found in the distal stomach of patients with portal hypertension. We aimed to delineate clinical factors associated with the appearance of these rare polyps. MATERIAL AND METHODS: We conducted a cross-sectional study of a cohort with 513 cirrhotic patients comparing patients with and without PHP using descriptive analyses and multivariable logistic regression. To address the problem of missing values, in particular for HVPG and liver stiffness, we used multiple imputation of missing values. RESULTS: The prevalence of macroscopically diagnosed PHP was 3.3% (95% confidence interval 2.0 - 5.4%). In 53% of cases, the correct classification was missed on index gastroscopy. Patients with PHP were older at gastroscopy (65 years vs. 59), had higher hepatic venous pressure gradients (HVPG, 28 mmHg vs. 19 mmHg), higher transient elastography (TE) measurements (50.7 kPa vs. 21.8 kPa) and more often had previous rubber band ligations (RBL, 64.7% vs. 25.8%). The multivariable logistic regression on the outcome macroscopically diagnosed PHP estimated an odds ratio (OR) for HPVG of 1.13 (CI 0.95-1.34), increased liver stiffness of 1.03 (1.00 - 1.07) and previous RBL of 3.84 (1.24 - 11.88), respectively. CONCLUSION: The prevalence of PHPs in the stomach was higher than assumed in previous studies and misclassification was commonly observed. The appearance of these rare polyps is associated with previous RBL and may correlate with severity of PH. Thus, PHPs may be regarded as marker for relevant PH, but clinical significance of these polyps is still uncertain.