International Journal of Biomedical Imaging (Jan 2023)

Prediction of Esophageal Varices in Viral Hepatitis C Cirrhosis: Performance of Combined Ultrasonography and Clinical Predictors

  • Puwitch Charoenchue,
  • Wittanee Na Chiangmai,
  • Amonlaya Amantakul,
  • Wasuwit Wanchaitanawong,
  • Taned Chitapanarux,
  • Suwalee Pojchamarnwiputh

DOI
https://doi.org/10.1155/2023/7938732
Journal volume & issue
Vol. 2023

Abstract

Read online

Objectives. This study is aimed at evaluating the diagnostic performance of clinical predictors and the Doppler ultrasonography in predicting esophageal varices (EV) in patients with hepatitis C-related cirrhosis and exploring the practical predictors of EV. Methods. We conducted a prospective study from July 2020 to January 2021, enrolling 65 patients with mild hepatitis C-related cirrhosis. We obtained clinical data and performed grayscale and the Doppler ultrasound to explore the predictors of EV. Esophagogastroduodenoscopy (EGD) was performed as the reference test by the gastroenterologist within a week. Results. The prevalence of EV in the study was 41.5%. Multivariable regression analysis revealed that gender (female, OR=4.04, p=0.02), platelet count (11 cm, OR=3.64, p=0.02), and absent right hepatic vein (RHV) triphasicity (OR=3.15, p=0.03) were significant predictors of EV. However, the diagnostic accuracy indices for isolated predictors were not good (AUROC=0.63–0.66). A combination of these four predictors increases the diagnostic accuracy in predicting the presence of EV (AUROC=0.80, 95% CI 0.69-0.91). Furthermore, the Doppler assessment of the right hepatic vein waveform showed good reproducibility (κ=0.76). Conclusion. Combining clinical and Doppler ultrasound features can be used as a screening test for predicting the presence of EV in patients with hepatitis C-related cirrhosis. The practical predictors identified in this study could serve as an alternative to invasive EGD in EV diagnosis. Further studies are needed to explore the diagnostic accuracy of additional noninvasive predictors, such as elastography, to improve EV screening.