Prospective Visual Inspection of the Ventrum of Tongue (VIVOT) Vasculature Predicts the Presence of Esophageal Varices

Abstract

Venous collateral shunting of blood from the splanchnic to systemic venous systems due to portal hypertension increases the pressure in the ventral lingual venous bed. We hypothesized that the appearance of sublingual varices evaluated by pre-endoscopy/bedside visual inspection of ventrum of tongue (VIVOT) might predict the presence of esophageal varices (EVs). Methods: To test this hypothesis, we prospectively enrolled patients with cirrhosis (CP) referred for EV screening for assessment of lingual vasculature after informed consent. Non-cirrhosis control patients were also enrolled. Methods: VIVOT was scored based on the presence of vessels > 2 mm and/or serpiginous veins. VIVOT scores were then correlated with endoscopic findings. Results: A total of 59 patients with cirrhosis (Group 1) were enrolled, as were 62 patients without cirrhosis (Group 2). Group 1 consisted of 100% male patients with mean age 59.5 ± 5.4 years; 39.0% were African American (AA). Group 2 consisted of 86% male patients, 59.0 ± 13 years and 53% AA. Among Group 1 patients, varices were present in 29% (16 esophageal and 3 gastric). There were no demographic differences among Group 1 patients with or without varices. Positive VIVOT scores were associated with EVs on endoscopy in 11 of 16 patients (sensitivity 68.75%). Positive VIVOT findings were present in 8 of 40 patients without EVs (specificity 80%). False-positive VIVOT scores were present in 6 of 62 non-cirrhotic controls. Overall, the positive predictive value among patients with cirrhosis was 59% with a negative predictive value of 84%. Conclusions: VIVOT has modest values in predicting EVs and should not be used alone to stratify patients for endoscopic evaluation when elastography and laboratory tests are available; however, its use in resource-limited settings to identify high-risk patients may be considered.