New serum markers for predicting esophageal varices: Is it a reality?


  • Giovanni Silva

    Corresponding author
    • Department of Internal Medicine, Botucatu School of Medicine, Sao Paulo State University – UNESP, Botucatu-SP, Brazil
    Search for more papers by this author


Dr Giovanni Silva, Department of Internal Medicine, Botucatu School of Medicine, Sao Paulo State University, UNESP, Botucatu, SP, Brazil. Email:

Liver cirrhosis is invariably associated with hemodynamic disturbances manifested as portal hypertension (PH) and concomitant splanchnic vasodilation. PH is the main cause of complications in patients with chronic liver disease. Its consequences are bleeding from gastroesophageal varices, ascites, hepatopulmonary syndrome, and hepatic encephalopathy.[1]

Understanding of the pathophysiology of PH may be important both for the introduction of effective pharmacological therapy and possibly also for the prediction of the development of esophageal varices. Ohm's law (ΔPA = Q × R) explains why PH occurs. The meanings are ΔPA = intrahepatic pressure, Q = blood flow from systemic circulation, and R = intrahepatic vascular resistance. Obviously, increasing either or both results in an elevation of portal pressure. Current knowledge about the mechanisms of increased resistance to portal blood flow and of the formation of portal-systemic collaterals indicates that hepatic vascular resistance is modulated by adjustment to the increased hepatic vascular tone; the latter is attributable to hepatic endothelial dysfunction, and the abnormal angiogenesis resulting from liver inflammation and fibrogenesis, while flow increases as a result of the hyperkinetic splanchnic circulation, contributing to the formation of varices.[2]

Gastroesophageal varices are present in more than 50% of patients with PH and are more likely as liver disease progresses.[1, 3] Bleeding from esophageal varices occurs at a rate of 5–15% per year in untreated patients. The risk factors for bleeding are variceal size, decompensated cirrhosis, and the presence of stigmata at endoscopy (red wale marks).[1] Currently, the American Association for the Study of the Liver recommends that all patients undergo endoscopy to assess the presence, the size, and the aspect of varices at the time of the diagnosis of cirrhosis. If no varices are present at index endoscopy, this procedure should be repeated at 2–3 years in compensated cirrhosis and annually in decompensated cirrhosis.[4] Therefore, there is considerable interest in developing models to predict the presence of large varices by nonendoscopic methods. Several studies have evaluated the noninvasive markers of esophageal varices in patients with cirrhosis, such as the platelet count, FibroTest, spleen size, portal vein diameter, transient elastography of the liver, and more recently, transient elastography of the spleen.[5]

Traditionally, increased splanchnic blood flow in cirrhosis has been attributed to an overproduction of endogenous vasodilators and a decreased vascular reactivity to vasoconstrictors, while the formation of collaterals has been considered a mechanical consequence of the increased portal pressure. However, more recent studies have demonstrated that angiogenesis is important for the formation of the new blood vessels.[6] Evidence supporting a role for angiogenesis in the pathogenesis of PH is overexpression of the potent angiogenic factor, vascular endothelial growth factor (VEGF).

Although the mechanisms of PH are complex, it can be explained as follows:[2, 6, 7]

  • 1The hepatic vasodilators: nitric oxide (NO) is a powerful endogenous vasodilator that modulates the intrahepatic vascular tone. In the cirrhotic liver, the synthesis of NO is insufficient to compensate for the activation of vasoconstrictors. Carbon monoxide (CO) is the initial product of heme oxidation by the enzyme, heme oxygenase (HO-1), and is an important modulator of intrahepatic vascular resistance.
  • 2Splanchnic vasodilators: increased blood flow is one of the factors that contribute to PH. Portal venous inflow is increased in cirrhosis, mainly because of the splanchnic vasodilatation. The main substances involved are NO and glucagon. At the same time, CO also contributes to splanchnic vasodilatation. Therefore, HO-1 is involved and also has a role in stimulating VEGF production, resulting in the development of a hyperdynamic splanchnic circulation. Other substances like prostaglandins and endocannabinoids are also vasodilators and are present in this context.
  • 3The formation of collateral vessels is a complex process involving the opening, dilatation, and hypertrophy of pre-existing vascular channels. The consequence of collaterals is an incremental increase of vascular resistance, and this may be important in determining the portal pressure. Indeed, this is influenced by vascular endothelial VEGF-dependent angiogenesis.
  • 4Vasoconstrictors: the main vasoconstrictor agents in the hepatic vascular bed are endothelin, angiotensin II, norepinephrine, and vasopressin. All are involved in modulating vascular tone, thereby increasing hepatic resistance.
  • 5Endothelial dysfunction: the intrahepatic vascular bed also exhibits endothelial dysfunction, which diminishes the availability of NO as well as increasing the production of endothelial-derived vasoconstrictors, such as prostaglandin H2.

Therefore, beside the structural alterations (fibrosis, nodule formation), there is a complex dynamic component that contributes to increasing hepatic vascular resistance and splanchnic vasodilatation. Understanding this pathophysiology has revealed markers that are associated with the presence of PH and varices. On the other hand, variation of the genes that encode proteins involved in systemic and splanchnic vasodilatation have been found to be associated with the presence of esophageal varices.[8] Therefore, variation of these genes could play a role in addition to predicting the presence of esophageal varices, as well as their likelihood of bleeding.

In this issue of the journal, Yang and colleagues have analyzed 951 patients with cirrhosis of various etiologies. The main aim was to evaluate additional blood markers and genetic risk for the prediction of the presence of esophageal varices in cirrhosis. Also, they performed a 2-year follow-up to evaluate predictors for esophageal varices (EV) bleeding. The authors also studied another 650 independent patients to confirm the association between genetic variants and presence of EVs, namely for validation cohort.[9]

The factors analyzed in this study included plasma levels of soluble CD163 (sCD163), VEGF and HO-1, genetic polymorphisms of HO-1, VEGF, and vascular endothelial growth factor receptor 2 (VEGFR2). Soluble CD163 is a specific marker of activated macrophages, another potential biomarker for PH in cirrhosis. The activation of Kupffer cells may be involved in PH by the release of vasoconstrictor substances. Recently, Grønbaek et al. have shown that sCD163 plasma concentration in cirrhosis is almost three times higher than in controls, and sCD163 was an independent predictor of the hepatic venous pressure gradient.[10] Yang et al.[9] found that serum sCD163 level was elevated in patients with cirrhosis complicated by esophageal varices, and this marker could potentially be used to predict the presence of EVs in clinical practice.

The genetic elements of the Yang study identified that patients with esophageal varices showed significantly higher frequencies of risk genotypes of HO-1, SS in (GT)n repeat, and AA in T(-413)A than corresponding wild-type genotypes. Further, cirrhotic patients carrying C(+405)G GG and C(+936)T TT risk genotype of VEGF, and Val(-297)Ile Ile/Ile risk genotype of VEGFR2 had a higher likelihood of developing EVs than those carrying wild-type genotype. Therefore, in addition to traditional markers (platelet count and splenomegaly), the authors showed that high serum sCD163 level and these polymorphisms in the HO-1 and VEGF predict the presence of esophageal varices in patients with cirrhosis. Further, the combination of platelet count, serum sCD163 level, and those risk haplotypes of HO-1 and VEGF conferred higher predictive values for varices than platelet count alone. Finally, patients with these same risk haplotypes (HO-1 and VEGF) have a higher chance of esophageal variceal bleeding than those not carrying these haplotypes.

In conclusion, the presence of esophageal varices and the likelihood of their bleeding could potentially be predicted by selected serum and genetic markers, associated with clinicopathological markers like platelet count. However, it is necessary to study these genetic polymorphisms in other populations, as these can vary in different geographic regions. In addition, it may be valuable to examine the utility of these novel serum and genetic markers in relation to spleen transient elastography for the prediction of esophageal varices and the likelihood of variceal bleeding.