Non‐invasive tools for the prediction of esophageal varices in cirrhosis: Can advanced ultrasound techniques spare endoscopy?

Le résumé fourni par la source

Variceal bleeding is one of the major causes of death in patients affected by hepatic cirrhosis, together with hepatic failure and hepatocellular carcinoma (HCC).1 An early diagnosis and staging of esophageal varices are therefore essential to provide effective treatment and prevent acute hemorrhage. Upper gastrointestinal endoscopy is undoubtedly the gold standard for diagnosis of varices, as it not only can reliably demonstrate the presence and grade of varices, but it also allows endoscopic treatment when necessary.2 Nevertheless, endoscopy remains an invasive and uncomfortable procedure for patients. The procedure is also costly, and serious complications such us perforation and hemorrhage can occur, however rarely.3 In addition, the risk of variceal development in early and compensated cirrhosis is low, and many endoscopies might be unnecessary.4 For these reasons, in last decades a lot of effort has been made to stratify cirrhotic patients using non-invasive indexes of portal hypertension (PH).5, 6 The aim is to carefully select patients at high risk of variceal development and to decrease the number of needless invasive procedures. In this scenario, technological innovation of different ultrasonographic techniques represented an authentic revolution compared to more traditional indexes of PH such us platelet count and spleen diameter. In fact, multimodal ultrasound imaging allows an accurate and reproducible non-invasive evaluation of PH (Figure 1), which represents the leading cause of variceal development. However, the current American Association for the Study of Liver Diseases (AASLD) guidelines7 consider endoscopy avoidable only for patients with platelet count >150 000/mm3 and liver stiffness <20 kPa. In this study, the authors tested the accuracy of Doppler hemodynamic liver index (HDLI) and point shear wave liver elastography (pSWE) in predicting the presence of esophageal varices using endoscopy as the reference standard.8 They also evaluated the performance of these techniques in identifying the different stages of varices (F1, F2, or F3) and the varices needing treatment (VNT), defined as F2-3 grade and/or presence of high-risk stigmata of bleeding such us red color sign or platelet-fibrin plug. HDLI showed 66% sensitivity and 92.3% specificity in predicting varices with the cutoff set up to 0.66 (AUROC 0.876) and 64.8% sensitivity and 96.1% specificity in predicting VNT with the cutoff set up to 0.71 (AUROC 0.858). pSWE showed 83% sensitivity and 84.6% specificity in predicting varices with the cutoff set up to 14.5 kPa (AUROC 0.884) and 79.5% sensitivity and 92.8% specificity in predicting VNT with the cutoff set up to 22.5 kPa (AUROC 0.93). The combination of the two techniques reached 94.3% sensitivity and 78.1% specificity in the identification of varices and 92.8% sensitivity and 89.2% specificity in the identification of VNT. The authors also reported a strong positive correlation between both HDLI and pSWE and varices grade (0.839 and 0.888 respectively). Based on these findings, the authors suggest avoiding routine endoscopy in patients with HDLI <0.66 and liver stiffness at pSWE <14.5 kPa, as the risk of varices is very low and the risk of VNT extremely low. They also suggest avoiding endoscopy if HDLI <0.71 and liver stiffness <22.5 kPa in specific situations such us limited resource settings or apprehensive patients. This very interesting cross-sectional study confirms the effectiveness of ultrasound in predicting esophageal varices and it adds three important innovations. Firstly, it supports the use pSWE, while most previous studies evaluated liver stiffness by using transient elastography (TE).9-11 Secondly, the authors demonstrated for the first time a correlation between HDLI and varices grade. And finally, by combining pSWE and HDLI the authors overcame the limit of single techniques, with a real benefit in term of diagnostic accuracy. We believe this study confirms further the crucial role of ultrasound in the follow-up of liver cirrhosis. The diagnostic power of one ultrasound examination allows physicians to detect ascites, identify and characterize new-onset nodules with contrast-enhanced ultrasound (CEUS), evaluate the amount of hepatic fibrosis with shear wave elastography (SWE), and estimate the level of PH with Doppler indexes. Furthermore, data relative to liver stiffness and Doppler study could be combined, as the present study suggests, to predict the presence and the severity of esophageal varices with high accuracy. This rich equipment of ultrasonographic tools avoids overuse of expensive and invasive procedures and improves patient compliance to the follow-up. There is no agreement on the appropriate frequency of non-invasive evaluation of PH in cirrhotic patients, and longitudinal studies are lacking. In clinical practice, we think it is reasonable that patients coming in for their routine six-monthly HCC screening ultrasound are also evaluated for PH. If this evaluation is integrated with clinical and laboratory data, an even more complete clinical picture of the cirrhotic patient can be obtained. This strategy could accommodate both patient convenience and health economics needs. The main limits of this approach are the reliance on single operator skills and experience, and the difficulty in obtaining good images and measures in patients with abdominal obesity or breathing difficulties. However, even with these limitations, ultrasound remains one of the most powerful and effective assets for the hepatologist. Open Access Funding provided by Universita degli Studi di Sassari within the CRUI-CARE Agreement. [Correction added on 19 May 2022, after first online publication: CRUI-CARE funding statement has been added.] The authors declare no conflict of interest. No additional data has been provided.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.