Esophageal and gastric variceal bleeding (EGVB) is one of the most common complications and one of the main causes of death in patients with liver cirrhosis. It is the key to improve the survival rate of patients with liver cirrhosis to early predict the risk of esophageal and gastric variceal bleeding and actively take preventive measures. As the "gold standard" for the diagnosis of EGVB, gastroscopy leads to poor patient compliance due to its invasive operation. Therefore, how to early apply non-invasive indexes to predict the risk of variceal rupture is the focus of current research.
Early studies found the predictive value of the ratio of PLT to spleen diameter, because varicose veins and hypersplenism were the results of portal hypertension, and splenomegaly was the main clinical manifestation of hypersplenism. In addition, the PLT of patients with liver cirrhosis was affected by many factors, except hypersplenism, the number of thrombopoietin would be greatly reduced when hepatocytes were damaged. Through a prospective study of 150 patients, Baig et al. showed that the ratio of PLT to spleen diameter is more accurate than the two alone.
In the earlier study, Giannini et al. believed that the ratio of PLT and spleen longest diameter could predict whether patients with liver cirrhosis have esophageal varices, and that the ratio 909 is an effective index to predict the presence or absence of esophageal varices. When the ratio is lower than 909, the positive predictive value and negative predictive value reach 96% and 100% respectively. Subsequently, a large number of studies have verified it. In addition, Albreedy et al. found that the area ratio of platelet and spleen can reflect EGVB, showed good sensitivity (100%) and specificity (88%).
In the early days, ultrasonic Doppler was the main examination equipment, which could indirectly evaluate the risk of EGVB by measuring the diameter of liver and spleen. With the development of TE, it has been widely accepted because of its advantages of non-invasive, simple, fast, repeatability and safety, and has become one of the important non-invasive methods for evaluating liver fibrosis, However, the application of TE in patients with ascites needed to be carefully explained. Therefore, more studies tended to predict the risk of esophagogastric varices in cirrhotic patients without ascites in the compensatory period through the indicators measured by TE. The study confirmed that TE technology was a valuable and meaningful non-invasive examination for the diagnosis of esophageal varices, and the LSM value was closely related to the degree of esophageal varices. In a prospective cohort, Kazemi et al. reported that 13.9 kPa and 19.0 kPa are the critical values of varices and VNT. Li et al. study results showed that through evaluation by TE and gastroscopy, 22.8 kPa, 30.6 kPa and 34.6 kPa were the best critical values of LSM for mild, moderate and severe esophageal varices. Since then, several studies have been published on the LSM prediction critical values, but the reported results vary widely.
The 2015 Baveno VI Consensus introduced the important innovations in the management of patients with compensated liver cirrhosis. The guidelines proposed that LSM<20 kPa combined with PLT> 150,000/mm3 can be used as a new standard to identify varicose veins with a lower risk of bleeding. In this part of patients, endoscopy screening could be avoided. Since then, a number of studies have confirmed the effectiveness of this risk classification rule based on this standard. However, there were still studies that have not reached the same conclusion. A meta-analysis was conducted to integrate the evidence of LSM and PLT for the identification of esophageal varices, and 15 studies were included. The results suggested that compared with patients with high LSM value or low PLT, patients with low LSM value and normal PLT had low risk of varicose veins, and the heterogeneity between the studies was low. It was confirmed that PLT combined with LSM value could be used to identify the risk of venous bleeding.
However, according to this standard, only about 20% of the patients did not need endoscopy. Therefore, in the recent study, the American Liver Research Association proposed a new standard that endoscopy is not required when the LSM< 25kpa and the PLT>110×109/L. Subsequently, different studies verified the diagnostic value of the two standards. In a study of 1035 patients with compensatory advanced chronic liver disease, the extended standard (51.7%) could avoid more endoscopy than the original standard (27.6%), but the extended standard also missed more high-risk varices (6.8%, 3.8%). According to the etiological stratification of liver diseases (hepatitis B, hepatitis C, alcoholic fatty liver and nonalcoholic fatty liver), the negative predictive values of the original Baveno VI standard were 0.92, 1.00, 1.00 and 1.00, and the negative predictive values of the extended standard were 0.92, 0.96, 0.92 and 0.93. Similarly, in a large meta-analysis of 1000 patients with chronic liver disease, the Baveno VI standard would allow 262 patients to avoid screening endoscopy, but 6 patients would be missed. On the contrary, using the expanded Baveno VI standard would result in 428 patients avoided screening endoscopy, but 20 HRV patients will be missed. Subsequently, Vilar et al. added the spleen diameter to the index. By enrolling 518 patients with compensated chronic liver disease from 5 clinical research centers in Europe and Canada, all patients measured the LSM value by TE. The study showed that when liver stiffness × spleen diameter/PLT (LSPS)<1.33, the risk of VET in patients is <5%. In addition, there are several validation studies performed in the Asian cohort especially in the recent years[19-23]. We summarized and presented these studies in table 4. These studies are mainly concentrated in China, Singapore, and Korea. The etiology of the involved patients in different studies is different, including viral hepatitis and/or fatty liver. The endoscope spare rate in expanded Baveno VI standard (40%-60%) was significantly higher than that in Baveno VI standard (20%-40%). But missing rate in expanded Baveno VI standard is correspondingly higher.
However, the current standards were mostly derived from Western countries, and the subjects included in the study were mostly alcohol and hepatitis C patients with liver cirrhosis. The etiology spectrum of liver cirrhosis was not the same in our country. Therefore, it was necessary to find a more suitable decision threshold for our country. In this large multi-center cohort study, based on the BAVENO VI guidelines, we combined the LSM value detected by TE and PLT to explore the best standard and best cut-off value (LSM<25kpa, PLT>120×109) to identify patients with compensatory cirrhosis with low risk of clinically significant varicose veins, so as to safely avoid screening endoscopy in 23.4% of patients.
However, our study also has certain limitations. First, although our data came from multiple clinical research centers, the number of study patients with compensated liver cirrhosis were much smaller than that of domestic patients, and larger data were needed to explore and verify the standard. In addition, our study did not provide instantaneous elastography. Fibrotouch detection was closely related to the technique of the inspector. In addition, factors such as obesity and alcohol consumption may have a certain impact on the LSM value, which would increase the LSM value. In such cases it may cause differences in our study standards. Therefore, whether the standards we were exploring can be applied to clinical practice still needs to be further improved and verified.
In short, the combination of TE and PLT can identify patients with compensatory cirrhosis with low risk of varicose veins. When the LSM <25 kPa and PLT >120×109/L, endoscopy screening was avoided in 23.4% of patients.