Portal hypertensive gastropathy (PHG) is used to describe the endoscopic appearance of gastric mucosa in patients with cirrhotic portal hypertension, with a characteristic mosaic-like pattern with or without red spots. The severity of PHG can vary from mild to severe, and patients with PHG are at an increased risk of acute and chronic gastrointestinal bleeding. According to the study by Kim et al, severe PHG showed a significantly high-risk of mortality and reduced expected survival time than none or mild PHG. Therefore, we need to detect PHG as soon as possible.However, as an invasive examination, endoscopy examination may massive gastrointestinal bleeding or gastric perforation. Most patients are afraid and disable to tolerate it, which significantly reduces the real morbidity of PHG and delays the time for diagnosis and treatment. Therefore, there is a need to find effective non-invasive methods that can predict patients with PHG in the early stage, especially which require treatment.

Cirrhosis is the end-stage of every chronic liver disease. Histologically, cirrhosis is characterized by regenerative nodules surrounded by fibrous bands, which can be divided into compensated phase and decompensated one. Compensated phase is referred to as asymptomatic stages, also known as compensated advanced chronic liver disease (cACLD). Decompensated cirrhosis is a rapidly progressive phase marked by the development of complications of portal hypertension (PH) and/or liver dysfunction. As the disease progresses, the presence of PH and liver dysfunction can lead to the appearance of ascites, esophagogastric variceal bleeding (EVB), hepatic encephalopathy (HE) and so on, marking the transition from a compensated to a decompensated phase.

Portal hypertensive gastropathy (PHG) is one of the common complications of cirrhosis as well, which is considered to be a potential cause of upper gastrointestinal bleeding (UGIB). PHG is used to describe the endoscopic appearance of gastric mucosa in cirrhosis patients, with a characteristic mosaic-like pattern with or without red spots, which was significantly underestimated due to its absence of typical clinical manifestations. The pathogenesis and risk factors of PHG in patients with liver cirrhosis are still controversial, but the pathogenesis are generally believed to be related to PH and advanced liver dysfunction. The presence of PHG is usually combined with esophageal varices (EVs). The severity degree of PHG can vary from no/mild to severe, and patients with PHG are at an increased risk of acute and chronic gastrointestinal bleeding . The overall prevalence of PHG varies from 4% to 98% . PHG accounts for 0.8-40% of UGIB in patients with cirrhosis and 8-50% of non-variceal UGIB, among which 2-12% of PHG can cause acute upper gastrointestinal bleeding (AUGIB), up to 95% of which is associated with severe PHG and may be life-threatening. The prevalence of chronic UGIB in patients with PHG ranges from 3 to 26%, which can lead to iron-deficiency anemia. The proportion of UGIB that also can be attributed to mild PHG (range, 3.5-31%). The mortality rate related to PHG bleeding is nearly 12.5%.

According to the study by Kim et al., severe PHG showed a significantly high risk of mortality and reduced expected survival time than none or mild PHG. At present, most of the researches and guidelines usually do not recommend primary prevention for patients with mild PHG, however NSBB as primary prevention should be used for mild PHG combined with small EV (guidelines recommend don't need primary prevention for small EV). The risk of UGIB with severe PHG is higher than that with mild PHG, which means that NSBB should also be considered for prevention even without EVs. For PHG patients with chronic iron deficiency anemia, after excluding anemia caused by other causes, not only should use NSBB to reduce portal venous pressure and the risk of bleeding, but also should pay attention to iron supplementation. From the above, obviously, it is of great important to detect patients with PHG as soon as possible. Nowadays, upper gastrointestinal endoscopy (UGIE) is the gold standard for measuring PHG. However, as an invasive examination, endoscopy examination may massive gastrointestinal bleeding or gastric perforation. Most patients are afraid and disable to tolerate it, which significantly reduces the real morbidity of PHG and delays the time for diagnosis and treatment. Therefore, there is a need to find effective non-invasive methods that can predict patients with PHG in the early stage, especially which require treatment.

Recently, there had been significant efforts in the last decade to assess the utility of non-invasive techniques and methods for the evaluation of liver fibrosis (LS) and portal pressure as well as the presence of EVs in patients with cirrhosis. Recently, the most promising technique is the measurement of LS by transient elastography (TE), either alone or combined with other parameters, had confirmed to be evaluated the seviety of liver fibrosis in patients with chronic liver disease, predicted the presence of clinically significant portal hypertension (CSPH) and ruled out high-risk esophageal varices (HEVs) in patients with cirrhosis. The use of TE alone can effectively assess liver fibrosis and portal hypertension by liver stiffness measurement (LSM), but the diagnostic accuracy of EVs is limited.The Baveno VI report guidelines acknowledged this application and recommend that LS combined with platelet count (PLT) can rule out HEVs in patients with cACLD, it means that LS<20 kPa plus PLT>150×109/L had a very low risk of having HEVs and therefore do not require routine screening endoscopy. The criteria were validated and expanded extent to allow more people to avoid endoscopic screening. According to the studies by Kim et al, LS-spleen diameter to platelet ratio score (liver stiffness×spleen diameter/platelet count, LSPS) , LS combined with other parameters，was clearly superior to the performance of LS alone for predicting EVs in patients with hepatitis B-related cirrhosis, and subsequently validated as a reliable predictor for CSPH in patients with cirrhosis by several studies. The PH risk score (-5.953+0.188×LS + 1.583 ×sex (1: male; 0: female) +26.705×spleen diameter/platelet count) has been firstly developed by Berzigotti et al.for diagnosis of CSPH in patients with cirrhosis of any cause. In addition, platelet count to spleen diameter ratio (PSR) and aspartate transaminase to platelet ratio (APRI) have also been confirmed by some studies that can partly predict the seviety of liver fibrosis and EVs.

Some recent studies have reported that non-invasive methods also can predict PHG in patients with cirrhosis to some extent. Zhang et al.established a model for predicting PHG combined with ultrasound examination related indicators, but the model was complex and diagnostic accuracy was poor. Yang et al.proposed a VAP scoring system combined spleen volume, platelet count and albumin, which can predict the presence of EVs and PHG in patients with chronic liver disease, but the results of this study have not been validated. Mandhwani et al.explored the efficacy of PSR and right liver lobe diameter to albumin ratio (PLAR) in predicting PHG in patients with cirrhosis, PSR was the better predictor better than PLAR. More recently, a study suggested that the accuracy of LS in predicting the presence of PHG in patients with hepatitis B-related cirrhosis was good. However, few high-quality studies used non-invasive methods combined LS and other parameters to predict the PHG in patients with liver cirrhosis, considering portal hypertension was the common pathogenesis of EVs and PHG, we speculated that the above non-invasive methods combined LS and other routine parameters are also suitable for predicting PHG in patients with cirrhosis.

The aim of this study was to establish a new predictive model (PHG risk score) for diagnosing PHG in patients with cirrhosis, then to compare the diagnostic accuracy of PHG risk score, the non-invasive models combined with LS and other non-invasive models combined with routine parameters in predicting PHG in patients with cirrhosis.

Inclusion Criteria:

1. aged 18-80 years old.
2. confirmed patients with cirrhosis : Cirrhosis is diagnosed using standard laboratory, radiological, and physical examination findings, or by liver histology in equivocal cases
3. paired noninvasive tests (blood tests, TE, and CT/enhanced CT) and gastroscopy within 3 months;
4. The nature of the present study was explained to all patients and informed consent was obtained from each case.

Exclusion Criteria:

1. the presence of severe ascites that might significantly hamper the accurate assessment of LSM , obesity (body mass index>35kg/m2) or intercostal space was too narrow;
2. severe cardiopulmonary diseases, renal failure, acute illness, infectious diseases, acute liver failure or cholestasis or hepatic congestion;
3. malignant tumors or other serious medical diseases (such as hepatocellular carcinoma, gastric cancer, pancreatic cancer and colon cancer) that may reduce expected survival time;
4. non-cirrhotic portal hypertension and those with portal vein thrombosis;
5. previous or ongoing treatment for portal hypertension, spleen diameter<4cm or splenectomy;
6. pregnant women and patients with implanted pacemakers;
7. patients can not cooperate with the endoscopy, CT/enhanced CT, FibroTouch /FibroScan and other related examination.

In this cross-sectional study , we will collect all patients with cirrhosis who meet the inclusion and exclusion criteria criteria coming to The Second Affiliated Hospital, Xi'an Jiaotong University since March 2018 to December 2018 .