Background
Idiopathic portal hypertension (IPH) is a relatively rare disease characterized by portal hypertension (PH) in the absence of causative disease, such as cirrhosis, chronic liver disease and occlusion of the extrahepatic portal vein or hepatic vein [
1]. The main clinical signs of this disease are PH and portal hypertension-related complications, including variceal bleeding, splenomegaly, hypersplenism, ascites and hepatic encephalopathy. These are also the main factors that affect the prognosis of patients [
1]. Up to now, the pathophysiological mechanisms of IPH are poorly understood and the progression cannot be presented. Therefore, the important treatment for managing PH and its related complications as recommended in IPH [
2]. Gastroesophageal varices (GEVs) are a progressive condition in IPH, and the accurate evaluation of the severity of GEVs is important for the prognosis, surveillance and management of IPH.
Upper gastrointestinal endoscopy (UGE) is considered the gold standard for predicting the severity of GEVs. However, it has several limitations: invasive, carry complications risk, costly and require a specific expertise [
3]. Several noninvasive parameters based on noninvasive ultrasonic elastography technologies and/or laboratory markers, such as liver stiffness (LS), spleen stiffness (SS), spleen stiffness-to-liver stiffness ratio (SS/LS), LS spleen-diameter-to-platelet-ratio score (LSPS), portal hypertension risk score (PH risk score) and varices risk score, are used to predict the severity of GEVs in patients with chronic liver disease [
4‐
6], but their diagnostic performance remains unknown in IPH. Very few studies, which also have small sample sizes, have been performed to evaluate the diagnostic performance of LS by using transient elastography (TE) in IPH patients [
7]. In addition, the performance of SS measurement based on elastography technologies has been demonstrated in patients with chronic liver disease by many studies [
8,
9].
Two-dimensional shear wave elastography (2D-SWE) was demonstrated to be a more effective noninvasive tool by several studies, as compared to TE, acoustic radiation force impulse imaging (ARFI), and point shear wave elastography (pSWE), and it could obtain a higher success rate in patients with obesity, ascites and narrow intercostal windows [
9]. Meanwhile, it could acquire more accurate tissue stiffness values, because of combining B-mode imaging with a color-coded tissue stiffness map in real time, so that organ capsule, vessels and bile ducts can be effectively avoided. In addition, the cutoff values of different techniques have obvious specificity [
10,
11].
Therefore, in our study, we aimed to clarify the diagnostic performance of SS by using 2D-SWE for predicting the presence of high-risk varices (HRV) in IPH patients compared with hepatitis B virus infected (HBV-infected) patients. Herein, the current retrospective single-center comparative study was designed.
Methods
Study design
This was a retrospective single-center comparative study that aimed to assess the performance of SS for predicting the presence of HRV in IPH patients and compared to the performance in HBV-infected patients. UGE was used as the gold standard, and SS was compared with LS, SS/LS, LSPS, PH risk score and varices risk score. Between December 2015 and December 2021, a total of 188 patients were enrolled, including patients with IPH (45.7%, 86 of 188) and HBV (54.3%, 102 of 188). The retrospective study was carried out in accordance with releveant guidelines and regulations or declaration of Helsinki and was approved by the ethics committee of the First Affiliated Hospital of Fourth Military Medical University. Informed consent was waived for the retrospective single-center comparative study. All authors accessed the study data and reviewed and approved the final manuscript.
Patient population
The inclusion criteria were as follows: (1) age 18–75 years; (2) IPH diagnosed by liver biopsy; (3) HBV-infected patients with HBsAg positive more than 6 months and no other chronic liver disease; and (4) UGE, noninvasive examinations (2D-SWE examinations and abdominal Doppler US) and laboratory tests within 7 days. The exclusion criteria were as follows: (1) previous treatment with non-selective beta-blockers, shunt placement, surgical treatment, band ligation, liver transplantation, splenectomy and overt hepatic encephalopathy; (2) intrahepatic or extrahepatic malignancies; (3) portal vein thrombosis or cavernous transformation diagnosed by Doppler US or computed tomographic (CT); (4) companied with other chronic liver disease, including autoimmune hepatitis, any other viral hepatitis and alcoholic hepatitis; (5) 2D-SWE examination failed; (6) missing important laboratory data; and (7) female patients who were pregnant or lactating. Finally, the demographic and clinical information of the patients were recorded, including gender and age.
Abdominal US and two-dimensional shear wave elastography examinations
Abdominal US and 2D-SWE examinations were conducted by using the Aixplorer system (SuperSonic Imagine; Aix-en-Provence, France) with a convex broadband transducer (SC6-1, frequency of 1–6 MHz). All ultrasound-related examinations were performed by two experienced sonographers who conducted at least 1000 abdominal US and 1000 2D-SWE examinations and were blinded to the clinical information and serological results.
All patients fasted for at least 4 h before the examinations. During the examinations, each patient was placed in the dorsal decubitus position with their arms maximally lifted, which allowed for full view of the epigastrium. Firstly, spleen size, the diameters and velocity of portal venous system, and heart rate were obtained by using conventional US, and then the flow volumes in portal venous system were calculated. Secondly, the LS (a 4 × 3 cm box) and SS (a 3 × 3 cm box) were measured by using 2D-SWE through the right and left intercostal windows, respectively. Patients needed to hold their breath (neither at full inspiration nor at full expiration at the end of expiration) for approximately 5 s, and effective 2D-SWE images were acquired, in which the region of interest (ROI) filled at least 85% of the color map and was stabilized for approximately 5 s. Then, the activated Q-box system (diameter range 5–20 mm) was placed in a parenchyma location, avoiding large vessels, biliary tracts and focal lesions, and its depths were 2 cm below the organ’s capsule. According to early reports [
12], at least five 2D-SWE successful measurements were performed, and then median values were calculated in liver and spleen for each patient, respectively. Finally, the mean value of Young’s modulus was used for statistical analysis.
Endoscopic assessment
All UGE examinations were performed by two endoscopists with more than 8 years of experience. The GEVs results were recorded as the LDRF classification described by Li et al. [
13], which was used in the National Clinical Research Center for Digestive Diseases and First Affiliated Hospital of Fourth Military Medical University of Digestive Diseases (the high-level teaching hospital in China). The definition of HRV that was used in our hospital was previously described by Karagiannakis et al. and the definitions were as follows: esophageal varices sizes at least 5 mm, varices with red wales irrespective of size, and varices with any gastric varices [
8].
Serological data
All serological data (including liver function, blood counts and coagulation tests) were extracted from the institutional electronic medical records. On the basis of biochemical markers, the noninvasive scores were calculated as reported previously as follows: SS/LS = spleen stiffness value/liver stiffness value [
14], LSPS = [LS (by using either TE or SWE and given in kilopascals) × spleen diameter (in centimeters)]/platelet count ratio (× 10
9/L) [
4], PH risk score = − 5.953 + 0.188 × LS (by using either TE or SWE and given in kilopascals) + 1.583 × sex (1: male; 0: female) + 26.705 × spleen diameter (in millimeters)/platelet count (× 10
9/L) ratio [
5,
6], varices risk score = − 4.364 + 0.538 × spleen diameter (in millimeters) − 0.049 × platelet count (× 10
9/L) − 0.044 × LS + 0.001 × [LS × platelet count (× 10
9/L)] [
5].
Statistical analysis
The continuous variables were expressed as the means ± standard deviations (SD) or medians [interquartile ranges (IQR)], depending on whether the variables followed a normal or non-normal distribution, whereas the categorical variables were expressed as numbers and percentages, when appropriate. For the analysis of the participants’ baseline characteristics, the continuous variables between groups were analyzed by Student’s t test or the Mann–Whitney U test, when appropriate. Categorical variables were compared by the Chi-square test or Fisher’s exact test, when appropriate. The diagnostic performance of noninvasive parameters for predicting the presence of HRV was estimated by receiver operating characteristic (ROC) curves. Differences between the areas under the ROC curves (AUCs) were compared by using the DeLong test. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive diagnostic likelihood ratio (LR+) and negative diagnostic likelihood ratio (LR−) were calculated. All statistical analyses were two sided, and p values less than 0.05 indicated statistical significance.
Statistical analyses were performed using SPSS software (version 26; IBM, Armonk, NY, USA) and MedCalc software (V.11.2; 2011 MedCalc Software bvba, Mariakerke, Belgium).
Discussion
This retrospective single-center study focused on patients with IPH by using 2D-SWE to predict the presence of HRV, as compared to HBV-infected patients. Accurate evaluation of the presence of HRV is of great importance. Herein, for the first time, we analyzed the diagnostic performance of SS by using 2D-SWE in patients with IPH compared to those with HBV.
For evaluating the presence of HRV, SS showed the highest performance compared with other noninvasive parameters (AUC: 0.98) in IPH. Virginia Hernándea-Gea et al. regarded the natural history of patients with non-cirrhotic portal hypertension and found that the SS was markedly increased in the early stages of the disease [
15], this finding could support our conclusion. Additionally, these pathophysiological studies have been demonstrated by the findings of several studies. At the early stage of IPH, the gross histological features of the liver are associated with the intrahepatic vascular alterations, which belong to Glisson’s sheath. The intrahepatic vein branches present sclerotic, vein wall thicken, obliteration, and early slight lymphoid cell infiltration of the portal tracts and branches [
16]; furthermore, the liver function tends to be typically preserved or slightly deranged [
1]. Our results regarding liver function are consistent with those previously reported (Table
2). The above natural history leads to the main clinical features that are associated with PH at the early stage of IPH, including splenomegaly, hypersplenism, and variceal vein [
1]. Therefore, the SS is noticeably changed in patients with IPH.
In HBV-infected patients, the SS was the best potential noninvasive parameter for evaluating the presence of HRV (AUC 0.96). On the one hand, SS provided the highest AUC compared with LS and SS/LS; on the other hand, although no significant difference was observed between the AUCs of SS, LSPS, PH risk score and varices risk score, the SS by using 2D-SWE could be more easily performed in clinical compared with other parameters. When patients are in the early stage of HBV-infections, the hallmarks of liver are mostly present in the hepatocytes, as opposed to the portal tracts in IPH [
17,
18]. Additionally, the volume of hepatocytes accounts for more than 90% of the total volume of the liver, with liver function being more severely affected in HBV-infected patients than in those with IPH. In the liver, there is mostly inflammation, thick fibrous septa, and small nodules, which are the most important factors for the LS increase [
19], as found in our study (Table
2). However, at the later stages, with the progression of hepatocytes death, extracellular matrix deposition, and vascular reorganization, the pathological hallmark of the liver is pseudolobule formation, which includes regenerative nodules, fibrous septa, and microvascular clotting [
16]. Finally, the irreversible histological aberrations mentioned above drive the increased intrahepatic resistance to the onset of complications of portal hypertension. As discussed above, increased portosystemic collaterals flow and the complications will appear, including upper gastrointestinal variceal bleeding, splenomegaly, hypersplenism, portosystemic collaterals and ascites (Table
2). With the progression of portal pressure, the severity of portal pressure partially depends on extrahepatic elements that are closely related to blood flow, including splanchnic vasodilatation, hyperdynamic circulation, and portosystemic collaterals [
20]. Finally, the correlation between LS and portal hypertension may be lost. Kumar and Reiberger et al. showed that the relationship between LS and portal pressure will be lost with increased portal pressure in cirrhosis [
19‐
22]. In a study of HCV-infected patients, the researcher found that there was a correlation between LS and the presence of GEV, but no relationship between LS and the GEV’s size was observed [
23]. The above conclusions agree with the findings from this study.
The study has several limitations. Firstly, the sample size was small because IPH is a rare disease. Secondly, this study is a retrospective single-center study. As described above, these limitations may limit the representativeness of the conclusions. However, our threshold has a strict quality control: (1) the UGE examinations were performed by two experienced specialists; (2) the US and 2D-SWE examinations of all patients were conducted by two experienced sonographers. The above characteristics contributed to avoid inter observer variation. Lastly, the study was a retrospective analysis. Hence, further well-designed prospective multicenter study will be needed to verify the conclusions in this study.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.