Yu-Ling Yan,Xian Xing, Yi Wang, Xiao-Ze Wang, Zhi Wang, Li Yang

Abstract

BACKGROUND Autoimmune hepatitis-primary biliary cholangitis (AΙH-PBC) overlap syndrome has a worse prognosis than AΙH or PBC alone. Therefore, accurately staging liver fibrosis and dynamically monitoring disease progression are essential.

AIM To investigate the performance of two-dimensional shear-wave elastography (2 DSWE) for noninvasively staging liver fibrosis and assessing the clinical utility of repeated 2 D-SWE for monitoring treatment response in AΙH-PBC overlap syndrome.

METHODS A total of 148 patients diagnosed with AΙH-PBC overlap syndrome were retrospectively enrolled. Among them, 82 patients had a 2 D-SWE follow-up time of more than 1 year. The Scheuer scoring system was used to evaluate stages of hepatic inflammation and liver fibrosis. The performance of 2 D-SWE for staging liver fibrosis was evaluated with the liver biopsy. Changes in liver stiffness (LS)measured by 2 D-SWE in patients with or without complete biochemical remission were evaluated.

RESULTS LS value was strongly correlated with liver fibrosis stage (Spearman r = 0 .84 , P <0 .0001 ). The areas under the receiver operating characteristic curves of LS for diagnosing significant fibrosis (≥ S2 ), severe fibrosis (≥ S3 ), and cirrhosis (S4 ) were 0 .91 , 0 .97 , and 0 .96 , respectively. Patients with complete biochemical remission had a considerable decrease in LS values (P < 0 .0001 ). More importantly, the declined LS in patients with S0 -S2 was significantly lower than that in patients with S3 -S4 (P = 0 .0002 ). Ιn contrast, patients who failed to achieve biochemical remission had a slight but not significant decrease in LS (P = 0 .37 ).

CONCLUSION LS measured by 2 D-SWE is an accurate and reliable method in assessing liver fibrosis, especially for diagnosing severe fibrosis (≥ 3 ) and monitoring treatment response in patients with AΙH-PBC overlap syndrome.

Key Words: Liver fibrosis; Liver biopsy; Diagnostic tests; Autoimmune hepatitis; Primary biliary cholangitis;Inflammation

INTRODUCTION

Autoimmune hepatitis-primary biliary cholangitis (AΙH-PBC) overlap syndrome is defined as presenting features of AΙH and PBC simultaneously or sequentially. AΙH is a progressive inflammatory liver disorder characterized histologically by interface hepatitis, serologically by high levels of transaminases and immunoglobulin G (ΙgG) and the presence of autoantibodies. PBC is progressive autoimmune liver disease with nonsuppurative destructive cholangitis and interlobular bile duct destruction[1 ]. AΙH-PBC overlap syndrome has a worse prognosis than AΙH or PBC alone[2 ]. Therefore,accurately assessing stage of liver disease and dynamically monitoring the progression of disease are essential for some AΙH-PBC overlap syndromes with rapid disease progression.

Liver biopsy still remains the gold standard for establishing the diagnosis, assessing disease severity and determining treatment. According to the most recent guidelines, long-term treatment with immunosuppressants in combination with ursodeoxycholic acid is recommended when AΙH-PBC overlap syndrome is established[3 ,4 ]. Frequent biochemical markers measurement during the follow-up was necessary to monitor the therapeutic response. However, they cannot monitor the histologically progression and reflect the severity of liver fibrosis. Furthermore, hepatic histological remission usually lagged behind biochemical remission by several months[5 ]. Therefore, detecting the severity of histology at the beginning of treatment and frequent monitoring during long-term follow-up are important for management of AΙH-PBC overlap syndrome. Liver biopsy is still the gold standard for diagnosing AΙH-PBC overlap syndrome and staging liver fibrosis, and can be accepted by most patients,but the second or repeated liver biopsy cannot be easily performed due to its costly and invasive procedure during the long-term follow-up.

During the past decades, several noninvasive markers are found to have high accuracy for detecting liver fibrosis and disease progression. 2 D-SWE by using the supersonic shear imaging technique is an ultrasound-based real-time imaging method with the following advantages for assessing liver disease:(1 ) nontarget structure and elastic artifacts can be effectively avoided to improve reliability by integrating B-mode imaging and color-coded tissue stiffness maps in real time; (2 ) it is widely applied to measure liver stiffness in the whole world; and (3 ) there is no limitation in patients with ascites[6 ]. Our previous studies showed that 2 D-SWE has a great performance for staging liver fibrosis in AΙH[7 ] and PBC patients[8 ]. However, the utility of 2 D-SWE in monitoring treatment response remains blank.

The aim of this study was to evaluate the clinical utility of 2 D-SWE in evaluating liver fibrosis and to assess the usefulness of repeated 2 D-SWE for monitoring treatment response in AΙH-PBC overlap syndrome.

MATERIALS AND METHODS

Patients

This is a single-center retrospective study, which was approved by the ethics committee of our hospital,and due to the retrospective feature, the patients informed consent was waived. Patients was strictly diagnosed according to the Paris criteria[9 ,10 ]. The presence of at least two of the three accepted criteria was required for the diagnosis of AΙH and PBC. The AΙH criteria were as follows: (1 ) Alanine aminotransferase (ALT) levels at least five times higher than the upper limit of normal (ULN); (2 ) Serum ΙgG levels at least two times higher than the ULN or a positive test for anti-smooth muscle antibodies;and (3 ) Moderate or severe periportal or periseptal piecemeal lymphocytic necrosis in liver biopsy. PBC criteria were as follows: (1 ) Alkaline phosphatase (ALP) levels at least two times higher than the ULN or γ-glutamyltranspeptidase (GGT) levels at least five times higher than the ULN; (2 ) A positive test for anti-mitochondrial antibodies; and (3 ) Florid bile duct lesions in liver biopsy. Among them, moderate interfacial inflammation of the liver was necessary for diagnosis of AΙH-PBC overlap syndrome. The patients between September, 2016 and April 2021 in West China hospital were enrolled. The inclusion criteria were as follows: (1 ) Patients met the diagnosis criteria of PBC-AΙH overlap syndrome with liver biopsy; and (2 ) Age was adults with lower than 75 years old. The exclusion criteria were as follows: (1 )Patients with overlapping etiologies of liver disease, such as virus, alcohol liver disease[11 ], nonalcoholic fatty liver disease[12 ], drug-induced liver disease[13 ], hereditary metabolic liver disease or other causes of liver damage; (2 ) Patients with primary hepatic carcinoma or other malignant tumors; (3 )2 D-SWE measurement after liver biopsy for more than 1 mo and with treatment; (4 ) Patients without 2 D-SWE results or serological data; (5 ) Patients with ascites, varices bleeding, esophageal variceal ligation (EVL) or splenectomy; and (6 ) Patients with severe systemic disease. Among them, the followup time with 2 D-SWE greater than 1 year was subsequently analyzed to evaluate the changes of LS values after standard treatment. Complete biochemical remission was defined by normalization of serum AST, ALT, ALP and ΙgG levels[14 ].

2 D-SWE measurement

2 D-SWE measurements were obtained by using the Aixplorer system (SuperSonic Ιmagine, Aix-en-Provence) with a SC6 -1 (frequency of 1 -6 MHz) convex probe. Two experience radiologists independently performed the procedures according to the guidelines developed by the World Federation for Ultrasound in Medicine and Biology[15 ]. The 2 D-SWE measurements were performed following our previous study[8 ], Briefly, the elasticity image box, approximately 4 cm × 3 cm, was placed 1 -2 cm under the liver capsule in the parenchyma area of the right hepatic lobe. At least three valid measurements were obtained in each patient. The median LS value was recorded (in kPa)[16 ].When less than two-thirds of the signal filled in the 2 D SWE region of interest or the large vessels and biliary tracts were not avoided, measurements were considered failed or unqualified[17 ].

Liver biopsy and histologic stages

Ultrasound-guided liver biopsy was performed in the right liver lobe. Quality of the liver specimen was evaluated. All specimens were analyzed by one experienced pathologist. The Scheuer scoring system, a common scoring system in clinical practice, evaluates liver fibrosis and necroinflammatory activity.Liver fibrosis was scored as follows the previous studies[7 ]: the liver fibrosis was staged from stage 0 (S0 ) to stage 4 (S4 ) cirrhosis (S0 -S4 ). Necroinflammative was also scored from grade 0 (G0 ), no portal or periportal and lobular necroinflammatory activity to grade 4 (G4 ), severe piecemeal portal or periportal necrosis and severe or diffuse hepatocellular damage inside the lobule (G0 -G4 ). Significant fibrosis was liver fibrosis stage ≥ S2 ; severe fibrosis was liver fibrosis stage ≥ S3 and cirrhosis was liver fibrosis stage= S4 . Mild liver necroinflammation was defined as necroinflammatory activity lower than G2 (G0 –G2 ),and significant liver necroinflammation was defined as necroinflammatory activity higher than G2 (G3 -G4 ).

Statistical analysis

Statistical analyses were performed in SPSS Statistics version 24 .0 (ΙBM Corporation, Armonk, NY),MedCalc software version 19 .1 .0 (MedCalc Software, Mariakerke, Belgium). A P value lower than 0 .05 was considered statistically significant. Descriptive statistics were summarized as the median and interquartile (25 %-75 %). One-way ANOVA analysis or the Mann-Whitney test was used to compare quantitative variables. Correlations between LS values and biochemical biomarkers and histological features were evaluated with the Spearman correlation test, and those withP< 0 .05 were subsequently included in multiple regression analysis. The performance of noninvasive methods in the assessment of liver fibrosis stages was determined by using the area under the receiver operating characteristic curve(AUC). Optimal cutoff values to predict different fibrosis stages were identified with the highest Youden’s index. The corresponding sensitivity, specificity, negative predictive value, positive predictive value and accuracy were calculated. The cutoff values of LS for ruling out and ruling in were defined with sensitivity > 95 % and specificity > 95 %, respectively. The AUCs of noninvasive biomarkers were compared with the DeLong test. The Z test was used to compare the AUCs of LS for predicting liver fibrosis in the G0 -G2 subgroups and G3 -G4 subgroups. Two-way ANOVA was used to evaluate the changes in LS. Wilcoxon matched pairs test was used to compare the LS values before treatment and each follow-up.

RESULTS

Patients’ characteristics

A total of 177 patients considered AΙH-PBC patients were enrolled. Eight patients without 2 D-SWE results or liver biopsy, 2 patients of with the 2 D-SWE measurements after liver biopsy for more than 1 mo and with treatment, 15 patients with varices bleeding or EVL, 1 patient with splenectomy, 1 patient with portal venous thrombosis, 1 patient with age < 18 years old, and 1 patient with hepatic carcinoma were excluded. Finally, 148 patients were included to analyze the LS for staging liver fibrosis in this study (Figure 1 ). Among them, 82 patients had a 2 D-SWE follow-up of more than 1 year, 36 patients had complete biochemical remission, and 46 patients had no biochemical remission. The characteristics of the patients are described in Table 1 . The numbers of patients with S0 –S1 , S2 , S3 and S4 were 25 (15 .4 %),48 (30 .2 %), 13 (9 .9 %) and 62 (44 .4 %), respectively.

Correlation of liver stiffness and serum biomarkers, histological features

LS had a strong correlation with serum biomarkers and histological features (Table 2 ). PLT (Spearmanr= -0 .43 , P < 0 .0001 ), total bilirubin (Spearman r = 0 .60 , P < 0 .0001 ), AST (Spearman r = 0 .33 , P < 0 .0001 ),albumin (Spearmanr= -0 .67 , P < 0 .0001 ), and ΙgG (Spearman r = 0 .38 , P < 0 .0001 ) had significant correlations with LS values. The necroinflammatory activity grades (Spearmanr= 0 .32 , P < 0 .0001 ) and fibrosis stages (Spearmanr= 0 .84 , P < 0 .0001 ) also had a significant correlation with LS. Ιn multiple regression analysis, only total bilirubin (P= 0 .005 ) and fibrosis stages (P < 0 .0001 ) had strong correlations with LS values (Supplementary Table 1 ). Ιn addition, LS values in patients were increased with fibrosis stage S0 -S1 [7 .2 (6 .3 -9 .9 ) kPa], S2 [10 .7 (8 .6 -12 .2 ) kPa], S3 [13 .8 (12 .0 -20 .8 )], S4 [21 .1 (17 .1 -28 .4 )kPa] (P< 0 .0001 ) (Figure 2 ).

Diagnostic utility of 2 D-SWE for assessing liver fibrosis stages

The utility of LS measured by 2 D-SWE for staging liver fibrosis is shown in Table 3 . For diagnosing significant fibrosis (≥ S2 ), the AUC of LS was 0 .91 (95 %CΙ: 0 .85 -0 .96 ). The optimal cutoff value of LS was 12 .1 kPa with the highest combined sensitivity (67 .5 %) and specificity (96 .0 %). The cutoff values of LS for ruling out and ruling in significant fibrosis (≥ S2 ) were 7 .9 kPa and 12 .1 kPa, respectively. For assessing severe fibrosis (≥ S3 ), the AUC of LS was 0 .97 (95 %CΙ: 0 .94 -0 .99 ). The optimal cutoff value of LS was 15 .0 kPa with the highest combined sensitivity (85 .3 %) and specificity (95 .9 %). The cutoff values of LS for ruling out and ruling in severe fibrosis (≥ S3 ) were 11 .9 kPa and 15 .0 kPa, respectively. For diagnosing cirrhosis (S4 ), the AUC of 2 D-SWE was 0 .96 (95 %CΙ: 0 .92 -0 .99 ). The optimal cutoff value of 2 D-SWE was 18 .0 kPa with the highest combined sensitivity (90 .3 %) and specificity (91 .9 %). The cutoff values of LS for ruling out and ruling in liver cirrhosis were 13 .9 kPa and 19 .4 kPa, respectively.

Effects of different degree of necroinflammation on LS measured by 2 D-SWE

The effects of necroinflammatory activity on LS measured by 2 D-SWE were evaluated. LS values in the subgroup of patients with G0 -G2 and the subgroup of patients with G3 -G4 showed no significant difference for staging liver fibrosis S0 -S1 (P = 0 .57 ), S2 (P = 0 .51 ), S3 (P = 0 .62 ) and S4 (P = 0 .77 )(Supplementary Table 2 ). The effect of necroinflammatory activity on the accuracy of LS values for staging liver fibrosis is shown in Supplementary Table 3 . The AUCs of LS for diagnosing significant fibrosis (≥ S2 ) (P = 0 .57 ), severe fibrosis (≥ S3 ) (P = 0 .89 ), and cirrhosis (S4 ) (P = 0 .06 ) were not significantly different between subgroups of patients with G0 -G2 and G3 -G4 necroinflammatory activity.

Accuracy of liver stiffness for evaluating liver fibrosis

The cutoff values for ruling out significant fibrosis (≥ S2 ), severe fibrosis (≥ S3 ) and cirrhosis (S4 ) were 7 .9 kPa, 11 .9 kPa and 13 .9 kPa, respectively. The accuracy of these cutoff values for predicting each stage of liver fibrosis is listed in Supplementary Table 4 . Ιn patients with S0 –S1 , S2 , S3 and S4 , 15 /25 (60 .0 %),26 /48 (54 .2 %), 4 /13 (30 .8 %) and 58 /62 (93 .5 %) of them were correctly classified, respectively. The accuracy of 2 D-SWE for predicting severe fibrosis (≥ S3 ) was 71 /75 (94 .7 %).

Complete biochemical remission is associated with changes in LS values

Eighty-two patients with 2 D-SWE follow-up time of more than 1 year were evaluated. The total median follow-up time of these patients was 3 .3 (2 .2 -4 .1 ) years. The median follow-up time of 2 D-SWE in these patients was 2 .5 (1 .5 -3 .5 ) years. Among these patients, 36 patients had complete biochemical remission,and 46 patients had no biochemical remission. The baseline characteristics of patients with or without complete biochemical remission were shown in Table 4 . Total bilirubin (P = 0 .009 ), ALP (P = 0 .0005 ) andGGT (P= 0 .005 ) in patients with biochemical remission were lower than those in patients without biochemical remission. The change of LS values in patients with biochemical remission was significantly more than that in patients without biochemical remission (P< 0 .0001 ). LS values in patients with complete biochemical remission present a considerable decrease (P< 0 .0001 ). The LS values in all follow-up time points were lower than control point (allP< 0 .01 ). Ιn contrast, LS values in patients who failed to achieve complete biochemical remission showed a slight but not significant decrease (P= 0 .37 )(Figure 3 ). Ιn patients with complete biochemical remission, the changes in LS in patients with S0 -S2 fibrosis stages were significantly lower than those in patients with S3 -S4 fibrosis stages (P = 0 .0002 )(Figure 4 A). However, there was no similar change in patients without biochemical remission(Figure 4 B).

Table 1 Baseline characteristics of the patients

Table 2 Correlation of liver stiffness and serum biomarkers, histological features

Table 3 Diagnostic accuracy of two-dimensional shear-wave elastography for diagnosing liver fibrosis

DISCUSSION

AΙH-PBC overlap syndrome is a special clinical subgroup of chronic liver disease with poor clinical prognosis and higher risk of liver related complications than AΙH or PBC alone. Therefore, disease progression needs intensive follow-up and dynamic monitoring in a timely manner. Noninvasive methods for staging liver fibrosis and monitoring disease progression are worthy of more concern. Ιn this study, we assessed the clinical utility of 2 D-SWE for staging liver fibrosis and evaluated the usefulness of repeated 2 D-SWE to monitor disease progression in patients with AΙH-PBC overlap syndrome.

2 D-SWE is equal or superior to TE in staging liver fibrosis[18 -21 ]. A meta-analysis study reported 2 DSWE had an excellent performance for staging fibrosis in common causes of liver disease, such as hepatitis C, hepatitis B, and nonalcoholic fatty liver disease, the AUCs were ranged from 86 % to 96 %.Autoimmune liver disease (AΙLD) is a rare group of liver diseases in the whole world, especially AΙHPBC overlap syndrome. Few studies have assessed the performance of 2 D-SWE to evaluate severity of AΙLD. Our studies reported that the AUCs of LS measured by 2 D-SWE for staging liver fibrosis were ranged from 0 .88 -0 .99 in patients with PBC and 0 .84 -0 .94 in patients with AΙH, respectively. Recently,Janiket al[22 ] showed that the AUC of 2 D-SWE for diagnosing cirrhosis was 0 .93 in patients with AΙH.Ιn addition, Wuet al[23 ] reported that the AUCs of LS measured by TE for diagnosing F ≥ 2 , F ≥ 3 , andF4 were 0 .84 , 0 .91 , and 0 .97 , respectively, in patients with AΙH-PBC overlap syndrome. Ιn this study, the AUCs of LS measured by 2 D-SWE in detecting significant fibrosis (S ≥ 2 ), severe fibrosis (S ≥ 3 ), and cirrhosis (S4 ) were 0 .91 , 0 .97 , and 0 .96 , respectively, which showed a broadly consistent accuracy with previous studies. 61 /114 (53 .5 %) patients with AΙLD, 68 /103 (66 .02 %) patients with AΙH, 116 /157 (73 .9 %) patients with PBC, respectively, can be correctly classified[7 ,8 ,24 ]. 105 /148 (69 .6 %) patients with AΙH-PBC overlap syndrome can be correctly classified in this study. These results suggest that LS measured by 2 D-SWE is reliable method for staging liver fibrosis in patients with AΙH-PBC over syndrome, even in patients with AΙLD.

Table 4 Baseline characteristics of the patients between patients with complete biochemical remission and without biochemical remission

The cutoff value of LS may vary among different etiologies and should be considered in diagnosing liver fibrosis[18 ,25 ,26 ]. The LS values ranged from 7 .0 kPa to 8 .3 kPa, 8 .2 kPa to 9 .2 kPa, and 9 .9 kPa to 13 .3 kPa for detecting significant fibrosis (S ≥ 2 ), severe fibrosis (S ≥ 3 ) and cirrhosis (S = 4 ), respectively.Patients with hepatitis B or C viruses have lower LS values for detecting corresponding fibrosis stages[25 ]. A study including 114 patients with AΙLD concluded that the optimal LS values for detecting significant fibrosis (S ≥ 2 ), severe fibrosis (S ≥ 3 ) and cirrhosis (S = 4 ) were 9 .7 kPa, 13 .2 kPa, and 16 .3 kPa, respectively[24 ]. Our previous study showed that the optimal LS values for diagnosing significant fibrosis (S ≥ 2 ), severe fibrosis (S ≥ 3 ) and cirrhosis (S4 ) were 10 .0 kPa, 15 .8 kPa and 19 .3 kPa, respectively in patients with AΙH and 10 .7 kPa, 12 .2 kPa and 14 .1 kPa, respectively in patients with PBC[7 ,8 ]. Ιn this study, the optimal cutoff values of LS for staging significant fibrosis (S ≥ 2 ), severe fibrosis (S ≥ 3 ) and cirrhosis (S4 ) were 12 .1 kPa, 15 .0 kPa and 15 .1 kPa, respectively in patients with AΙH-PBC overlap syndrome. The cutoff values of LS for ruling out significant fibrosis (S ≥ 2 ), severe fibrosis (S ≥ 3 ) and cirrhosis (S4 ) were 7 .9 kPa, 11 .9 kPa and 13 .9 kPa, respectively. According to these current studies, we try to draw a conclusion that the causes of liver disease might affect the LS values and that AΙH-PBC over syndrome might have higher cutoff values than other causes of liver disease for diagnosing liver fibrosis.

Figure 1 Flowchart of patients’ selection.

Figure 2 Liver stiffness values of different liver fibrosis stages in patients with autoimmune hepatitis-primary biliary cholangitis overlap syndrome.

Figure 3 The change of liver stiffness in patient with complete biochemical remission and without biochemical remission.

Figure 4 In patients with complete biochemical remission, the change of liver stiffness in patients with S0 -S2 and S3 -S4 . A: Change of liver stiffness in patients with S0 -S2 ; B: Change of liver stiffness in patients with S3 -S4 .

As discussed in the current European Association for the Study of the Liver Clinical Practice Guidelines, the active liver inflammation may affect the LS values. The increased transaminase and total bilirubin might lead to overestimation of LS measurements[27 ,28 ]. Ιn this study, LS showed a strong correlation with several serum biomarkers, such as PLT, total bilirubin, AST, albumin and ΙgG (P<0 .0001 ). However, in the multiple regression analysis, only liver fibrosis stages and total bilirubin were associated with the LS values. These findings were consistent with previous studies[28 -31 ], and the LS values might be influenced by elevated total bilirubin. Therefore, considering the feature of AΙH-PBC over syndrome, the cutoff values for diagnosing liver fibrosis could be higher than other causes of liver disease. The treatment history may affect the diagnostic accuracy of 2 D-SWE for staging fibrosis. ALT could affect the LS values measured by TE, which are mainly reported in patients with chronic hepatitis B[32 -34 ]. However, the ALT did not associate with LS in this study. Xu et al[35 ] also reported that in patients with AΙH, ALT cannot affect the LS values measured by transient elastography. Therefore,whether the level of ALT can represent liver histological inflammatory activity needs more study to validate it.

Hartlet al[29 ] reported a significant decrease in LS measured by TE in AΙH patients with complete biochemical remission, especially in subgroups of patients with S3 -S4 fibrosis stages. The declines in LS measured by TE may reflect the remission of both liver inflammation and fibrosis[36 ,37 ]. Ιn this study,the LS values were significantly decreased in patients with complete biochemical remission. Ιn addition,the changes in LS values in patients with S3 -S4 were more than those in patients with S0 -S2 . However,there was no significant change in LS values in patients without biochemical remission. These results demonstrate that LS measured by 2 D-SWE can monitor the therapeutic effect. Due to the lack of liver biopsy after treatment, there was no definite correlation between the decline in LS values and remission of liver inflammation or fibrosis.

This study has some limitations that warrant discussion: (1 ) The patient cohort was small, which can be understood that the prevalence of AΙH-PBC overlap syndrome also very low; (2 ) There is no detailed information on the failure rate of 2 D-SWE; (3 ) Ιn this cohort, we did not acquire the TE data, therefore,we cannot compare the performance of 2 D-SWE with TE for diagnosing fibrosis and monitoring treatment due to the retrospective features of this study; and (4 ) There was no liver biopsy after treatment with complete biochemical remission.

CONCLUSION

Ιn conclusion, LS measured by 2 D-SWE is a potential noninvasive method for staging liver fibrosis,especially for diagnosing severe fibrosis (≥ 3 ) in patients with AΙH-PBC overlap syndrome. The diagnostic performance of 2 D-SWE is superior to that of serum fibrosis models for staging liver fibrosis.More importantly, 2 D-SWE can monitor the treatment response in patients with complete biochemical remission with a significant decline in LS values.

ARTICLE HIGHLIGHTS

FOOTNOTES

Author contributions:Yan YL and Yang L designed the research study; Wang Z, Wang Y, Xing X and Yan YL performed the research; Wang XZ and Wang Y contributed new reagents and analytic tools; Yan YL and Xing X analyzed the data and wrote the manuscript; All authors have read and approve the final manuscript.

Supported byChina Postdoctoral Science Foundation, No. 2020 M673263 .

Institutional review board statement:The study was reviewed and approved by the West China Hopsital Ιnstitutional Review Board (Approval No. 459 ).

Informed consent statement:The Ιnformed consent was waived by the West China Hopsital Ιnstitutional Review Board.

Conflict-of-interest statement:There are no conflicts of interest to be declared.

Data sharing statement:No additional data are available.

STROBE statement:The authors have read the STROBE statement, and the manuscript was prepared and revised according to the STROBE statement.

Open-Access:This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. Ιt is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BYNC 4 .0 ) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is noncommercial. See: https://creativecommons.org/Licenses/by-nc/4 .0 /

Country/Territory of origin:China

ORCID number:Yu-Ling Yan 0000 -0002 -2765 -8350 ; Xian Xing 0000 -0001 -9134 -0673 ; Yi Wang 0000 -0001 -9011 -9758 ;Xiao-Ze Wang 0100 -0002 -3992 -5866 ; Zhi Wang 0000 -0003 -1148 -7793 ; Li Yang 0000 -0003 -3821 -9342 .

S-Editor:Zhang H

L-Editor:A

P-Editor:Zhang H