ABSTRACT
Background: In patients with cirrhosis, esophageal variceal hemorrhage (EVH) is a devastating consequence of portal hypertension (PH). Upper endoscopy is considered the gold standard for the detection and diagnosis of esophageal varices (EVs), despite being invasive and costly. This study was aimed at identifying and evaluating the diagnostic accuracy of noninvasive tools in predicting EVs in patients with compensated cirrhosis. Methods: This cross-sectional study included 50 patients with compensated cirrhosis at the Tygerberg Hospital Gastroenterology Clinic in Cape Town between November 2022 and May 2023. We collected clinical, anthropometric, and laboratory data from patients' physical and electronic charts. All patients underwent an abdominal ultrasound, vibration-controlled transient elastography (VCTE) to assess liver and splenic stiffness, and upper endoscopy. In this comparative study, we evaluated the diagnostic accuracy of different noninvasive tools in detecting EVs in patients with compensated cirrhosis. Results: Of the 50 patients included in the study, 30 (60%) were female and 20 (40%) were male. The patients' age ranged from 18 to 83, with a mean age of 46.6 years. Cirrhosis was mainly due to alcohol use (n = 11, 22%), hepatitis B virus (HBV) infection (n = 11, 22%), and autoimmune hepatitis (n = 10, 20%). The patients included in the study were divided into two subgroups: with (n = 34, 68%) or without (n = 16, 32%) EVs. Statistically significant differences were detected between groups in platelet count (PC), liver stiffness measurement (LSM), spleen stiffness measurement (SSM), portal vein diameter (PVD), bipolar spleen diameter (SBD), aspartate aminotransferase-to-platelet ratio index (APRI), fibrosis-4 index (FIB-4), platelet/bipolar spleen diameter ratio (PSR), liver stiffness-spleen size-platelet ratio (LSPS), liver stiffness-spleen stiffness-platelet ratio score (LS3PS), and spleen stiffness-spleen size-platelet ratio score (SSPS) (p < 0.001). The highest diagnostic precision was observed with SSM (96%), SSPS (96%), LS3PS (94%), LSPS (94%), PSR (94%), and PC (92%). SBD (88%), LSM (86%), APRI (82%), and FIB-4 (82%) had the lowest diagnostic accuracy. Conclusion: SSM and SSPS have the highest diagnostic accuracy for predicting the presence of EVs in patients with compensated cirrhosis. LSPS, LS3PS, and PSR come second at 94%. We recommend SSM and SSPS in institutions with transient elastography equipped with the software necessary to measure splenic stiffness. We introduce and propose LS3PS as a novel composite score for predicting the presence of EVs in patients with compensated cirrhosis. Large-sample-size studies are needed to validate these prediction scores and to allow direct comparison with Baveno VII. These prediction tools can help clinicians avoid unnecessary endoscopic procedures in patients with compensated cirrhosis, especially in developing countries with limited resources such as South Africa.