Advertisement

A Nomogram to Predict the Risk of Stent Dysfunction After TIPS in Patients with Hepatitis B Cirrhosis

Open AccessPublished:February 22, 2022DOI:https://doi.org/10.1016/j.acra.2022.01.016

      Rationale and Objectives

      To develop and validate a nomogram for the prediction of stent dysfunction after transjugular intrahepatic portosystemic shunt (TIPS) placement in patients with hepatitis B cirrhosis.

      Materials and Methods

      From 2012 to 2020, 355 patients with hepatitis B cirrhosis who underwent TIPS placements were enrolled in this study. A multivariable logistic regression analysis was applied to determine independent risk factors for the nomogram construction. Discrimination, calibration, and clinical usefulness of the prediction model were assessed by using receiver operating characteristic curves, calibration scatter plots, and a decision curve analysis (DCA).

      Results

      Independent factors for TIPS stent dysfunction included diabetes, previous splenectomy, the shunting branch of the portal vein, and stent position, which were used to construct the nomogram. The AUC values in the training and validation cohorts were 0.817 (95% CI: 0.731-0.903) and 0.804 (95% CI: 0.673-0.935), respectively, which suggested a good predictive ability. The calibration curves in both cohorts revealed good agreement between the predictions and actual observations. The DCA curve indicated that when the threshold probability ranged from 2% to 88%, the nomogram could provide clinical usefulness and a net benefit.

      Conclusion

      The nomogram that we developed could be conveniently used to predict TIPS stent dysfunction in patients with hepatitis B cirrhosis.

      Key Words

      Abbreviations:

      TIPS (Transjugular intrahepatic portosystemic shunt), CI (Confidence interval), DCA (Decision curve analysis), CDUS (Color doppler ultrasound), PVT (Portal vein thrombosis), INR (International normalized ratio), CTP (Child–Turcotte–Pugh), MELD (Model for end-stage liver disease), PV (Portal vein), SD (Standard deviation), OR (Odds ratio)

      INTRODUCTION

      Transjugular intrahepatic portosystemic shunt (TIPS) has been widely used in the treatment of portal hypertensive complications in patients with cirrhosis (
      • Hung ML
      • Lee EW.
      Role of transjugular intrahepatic portosystemic shunt in the management of portal hypertension: review and update of the literature.
      ,
      • Bureau C
      • Thabut D
      • Oberti F
      • et al.
      Transjugular intrahepatic portosystemic shunts with covered stents increase transplant-free survival of patients with cirrhosis and recurrent ascites.
      ,
      • Brunner F
      • Berzigotti A
      • Bosch J.
      Prevention and treatment of variceal haemorrhage in 2017.
      ). However, stent dysfunction remains the main complication of TIPS (
      • Hayek G
      • Ronot M
      • Plessier A
      • et al.
      Long-term outcome and analysis of dysfunction of transjugular intrahepatic portosystemic shunt placement in chronic primary budd-chiari syndrome.
      ,
      • Li YH
      • Xu ZY
      • Wu HM
      • et al.
      Long-term shunt patency and overall survival of transjugular intrahepatic portosystemic shunt placement using covered stents with bare stents versus covered stents alone.
      ,
      • Patidar KR
      • Sydnor M
      • Sanyal AJ.
      Transjugular intrahepatic portosystemic shunt.
      ). The previously reported 1-year dysfunction rate after TIPS placement ranged from 12.8% to 53.2% (
      • Bureau C
      • Garcia-Pagan JC
      • Otal P
      • et al.
      Improved clinical outcome using polytetrafluoroethylene-coated stents for TIPS: results of a randomized study.
      ,
      • Wan YM
      • Li YH
      • Xu ZY
      • et al.
      Comparison of TIPS alone and combined with partial splenic embolization (PSE) for the management of variceal bleeding.
      ,
      • Qi X
      • Han G.
      Transjugular intrahepatic portosystemic shunt in the treatment of portal vein thrombosis: a critical review of literature.
      ,
      • Bai M
      • He CY
      • Qi XS
      • et al.
      Shunting branch of portal vein and stent position predict survival after transjugular intrahepatic portosystemic shunt.
      ). Stent dysfunction may induce the recurrence of portal hypertensive complications, such as gastroesophageal variceal bleeding and ascites, which significantly affect the patient's quality of life and reduce survival time (
      • Weber CN
      • Nadolski GJ
      • White SB
      • et al.
      Long-term patency and clinical analysis of expanded polytetrafluoroethylene-covered transjugular intrahepatic portosystemic shunt stent grafts.
      ,
      • Wan YM
      • Li YH
      • Xu Y
      • et al.
      Predictors of shunt dysfunction and overall survival in patients with variceal bleeding treated with transjugular portosystemic shunt creation using the fluency stent graft.
      ). Therefore, it is necessary to explore the relevant factors to comprehensively evaluate the function of the TIPS stent and to assess the prognosis of the patient.
      Stent dysfunction after TIPS placement is related to multiple risk factors (
      • Hayek G
      • Ronot M
      • Plessier A
      • et al.
      Long-term outcome and analysis of dysfunction of transjugular intrahepatic portosystemic shunt placement in chronic primary budd-chiari syndrome.
      ,
      • Cura M
      • Cura A
      • Suri R
      • et al.
      Causes of TIPS dysfunction.
      ,
      • Yang C
      • Liu J
      • Shi Q
      • et al.
      Effect of splenectomy on the outcomes in patients with cirrhosis receiving transjugular intrahepatic portosystemic shunt.
      ,
      • Perarnau JM
      • Le Gouge A
      • Nicolas C
      • et al.
      Covered vs. uncovered stents for transjugular intrahepatic portosystemic shunt: a randomized controlled trial.
      ,
      • Huang Z
      • Yao Q
      • Zhu J
      • et al.
      Efficacy and safety of transjugular intrahepatic portosystemic shunt (TIPS) created using covered stents of different diameters: a systematic review and meta-analysis.
      ). Previous studies have demonstrated that acute thrombosis within the stent, as well as pseudointimal hyperplasia and bile leakage, were the main causes of stent dysfunction (
      • Cura M
      • Cura A
      • Suri R
      • et al.
      Causes of TIPS dysfunction.
      ,
      • Ducoin H
      • El-Khoury J
      • Rousseau H
      • et al.
      Histopathologic analysis of transjugular intrahepatic portosystemic shunts.
      ,
      • Clark W
      • Golkar F
      • Luberice K
      • et al.
      Uncovering the truth about covered stents: is there a difference between covered versus uncovered stents with transjugular intrahepatic portosystemic shunts.
      ). However, the clinical risk factors associated with stent dysfunction have not been fully elucidated. Additionally, there are currently no tools to integrate these factors for predicting individual stent dysfunction. Currently, nomograms are widely used as prognostic devices to provide superior individual disease risk estimation and guidance for optimal treatment decisions (
      • Wei L
      • Champman S
      • Li X
      • et al.
      Beliefs about medicines and non-adherence in patients with stroke, diabetes mellitus and rheumatoid arthritis: a cross-sectional study in China.
      ,
      • Shariat SF
      • Karakiewicz PI
      • Suardi N
      • et al.
      Comparison of nomograms with other methods for predicting outcomes in prostate cancer: a critical analysis of the literature.
      ). However, nomograms that can predict stent dysfunction after TIPS placement in hepatitis B cirrhosis patients have not been reported. We hypothesized that some important clinical indicators could serve as independent predictors for TIPS stent dysfunction. Our study aimed to develop a novel nomogram to predict TIPS stent dysfunction based on independent predictors. To our knowledge, this is the first study to describe a nomogram to predict stent dysfunction after TIPS placement in patients with hepatitis B cirrhosis.

      MATERIALS AND METHODS

      Study Population

      This retrospective study was conducted as a single-center study. Due to the fact that the main etiology of liver cirrhosis is infection with hepatitis B virus in China, which accounts for approximately 77% of the total number of cirrhosis patients, this study included patients with hepatitis B cirrhosis as the study population. From 2012 to 2020, a total of 355 patients with hepatitis B cirrhosis who underwent TIPS procedures in our hospital were retrospectively enrolled (Fig 1). Patients meeting the following criteria were included: (1) hepatitis B liver cirrhosis; (2) portal hypertensive complications; and (3) TIPS placement being performed by using a single Viator stent or a bare stent combined with a fluency stent. The exclusion criteria were as follows: (1) liver cirrhosis due to other causes; (2) malignant tumor; (3) cavernous transformation of the portal vein; (4) other serious systemic diseases; and (5) incomplete clinical data or follow-up time < 1 year. The study was approved by the institutional ethics committee, and all of the participants signed informed consent forms.
      Figure 1
      Figure 1Selection flowchart for consecutive patients who underwent TIPS procedures between 2012 and 2020. TIPS, transjugular intrahepatic portosystemic shunt.

      Stent Dysfunction Assessment

      The following events suggested that stent dysfunction may exist (
      • Hayek G
      • Ronot M
      • Plessier A
      • et al.
      Long-term outcome and analysis of dysfunction of transjugular intrahepatic portosystemic shunt placement in chronic primary budd-chiari syndrome.
      ,
      • Wan YM
      • Li YH
      • Xu ZY
      • et al.
      Comparison of TIPS alone and combined with partial splenic embolization (PSE) for the management of variceal bleeding.
      ,
      • Huang Z
      • Yao Q
      • Zhu J
      • et al.
      Efficacy and safety of transjugular intrahepatic portosystemic shunt (TIPS) created using covered stents of different diameters: a systematic review and meta-analysis.
      ): (1) the recurrence of portal hypertensive complications; and (2) color Doppler ultrasound (CDUS) showing that the blood flow velocity within the stent was more than 250 cm/s or less than 50 cm/s. Stent dysfunction was defined as a shunt stenosis greater than 50%, and it was finally confirmed by using portal venography.

      Data Collection and Follow up

      The following clinical data were collected and analyzed: age, indications for TIPS (acute variceal bleeding, recurrent variceal bleeding, refractory ascites), sex (male/female), splenectomy (yes or no), portal vein thrombosis (PVT, yes/no), diabetes (yes/no), complete blood count, serum sodium, international normalized ratio (INR), serum creatinine, total bilirubin, albumin, the Child-Turcotte-Pugh (CTP) scores, CTP classification, the model for end-stage liver disease (MELD) scores, shunting branch of the PV (right/left), types of stents (single Viatorr stent/Fluency stent/bare stent combination), and initial stent position (optimal/suboptimal). The optimal initial stent position was defined when the following criteria were met (
      • Bai M
      • He CY
      • Qi XS
      • et al.
      Shunting branch of portal vein and stent position predict survival after transjugular intrahepatic portosystemic shunt.
      ,
      • Clark TW
      • Agarwal R
      • Haskal ZJ
      • et al.
      The effect of initial shunt outflow position on patency of transjugular intrahepatic portosystemic shunts.
      ): (1) the proximal end of the stent extended to the hepatocaval junction; and (2) the distal end of the stent was parallel to the vascular wall of the PV (the angle between the tangent line of the distal end of the stent and the vascular wall of the PV was less than 20°). Otherwise, the stent position was defined as suboptimal (Fig 2).
      Figure 2
      Figure 2The angle between the tangent line of the distal end of the stent and the vascular wall of the PV was 16.5°, but the cephalic end of the stent did not extend to the hepato-caval junction (a). The cephalic end of the stent extended to the hepato-caval junction but the angle between the tangent line of the distal end of the stent and the vascular wall of the PV was 49.0° (b). The cephalic end of the stent did not extend to the hepato-caval junction, and the angle between the tangent line of the distal end of the stent and the vascular wall of the PV was 80.5° (c). PV, portal vein.
      The follow-up time was 1 year, and the primary end point was stent dysfunction. All of the patients were followed for clinical evaluation (medical history inquiry and physical examination), blood biochemistry, and CDUS (diameter, flow velocity, and direction of flow in the PV and stent) at 1, 3, 6, and 12 months or at the time of clinical symptom recurrence. When stent dysfunction was suspected, angiography was performed to assess the stent patency.

      Statistical Analysis

      Patients (n = 355) were randomly divided into a training cohort (n = 249) and a validation cohort (n = 106) by using R 4.0.2 software at a ratio of 7:3. Baseline patient demographic and clinical features were compared between the two groups via the t test for continuous variables and via the chi-square test for categorical variables. Subsequently, in the training cohorts, variables with a p value < 0.1 in the univariate analysis were entered into the multivariate logistic regression analysis by using the stepwise selection method to identify the independent predictive factors for stent dysfunction. Based on the identified independent predictive factors, a nomogram was constructed to predict the probability of 1-year stent dysfunction after TIPS procedures. Subsequently, we internally validated the nomogram by using the validation cohorts.
      The receiver operating characteristic curve and the AUC were used to evaluate the predictive discrimination of the nomogram. Calibration curves were plotted to determine the concordance of the nomogram. Moreover, we utilized decision curve analysis (DCA) to assess the clinical usefulness and net benefits of the nomogram.
      Data are presented as the mean and standard deviation or count (percentage). Statistical analysis was performed via SPSS software (version 21, IBM Corporation, Armonk, NY) and the programming language R (version 4.0.3) for Windows. A two-sided p value < 0.05 was considered to indicate statistical significance.

      RESULTS

      Patient Characteristics

      A total of 355 patients with hepatitis B cirrhosis were included in this study. Approximately 70% of the patients were randomly selected as the training cohort to develop a prediction model, and the remaining 30% of the patients were used to test the model as the validation cohort. Figure 1 presents the patient selection flowchart. There were 32 patients with stent dysfunction in the training cohort and 17 patients in the validation cohort. The incidence of stent dysfunction was 12.9% and 16.0% in the training cohort and validation cohort, respectively. There were no statistically significant differences between the two cohorts (Table 1).
      Table 1Distribution of Demographic and Clinical Information
      VariablesTraining Cohort n = 249 (70%)Validation Cohort n = 106 (30%)p
      Age (years)54.0 ± 11.856.3 ± 10.20.079
      Sex
      Male159660.776
      Female9040
      Indications for TIPS0.904
      Acute variceal bleeding3212
      Recurrent variceal bleeding20086
      Refractory ascites178
      Diabetes0.918
      Yes6025
      No18981
      PVT0.141
      Yes3823
      No21183
      Splenectomy0.754
      Yes5321
      No19685
      White blood cells (109/L)5.2 ± 4.35.3 ± 5.60.800
      Hemoglobin (g/L)76.7 ± 13.176.4 ± 12.50.841
      Platelets (109/L)90.9 ± 66.180.5 ± 66.40.179
      Serum sodium (mmol/L)139.6 ± 5.2139.7 ± 6.00.937
      INR1.37 ± 0.311.44 ± 0.310.086
      Serum creatinine (μmol/L)66.6 ± 19.170.3 ± 26.70.141
      Total bilirubin (μmol/L)25.7 ± 21.126.4 ± 20.70.754
      Albumin (g/L)31.8 ± 8.030.4 ± 6.40.093
      CTP scores7.7 ± 1.47.8 ± 1.40.372
      CTP classification0.205
      A5114
      B17277
      C2615
      MELD scores11.3 ± 3.411.9 ± 3.00.087
      Shunting branch of the PV0.928
      Left16871
      Right8135
      Types of stents0.143
      A single viatorr stent8829
      Fluency stent/bare stent combination16177
      Initial stent position0.847
      Optimal17676
      Suboptimal7330
      Stent dysfunction0.426
      Yes3217
      No21789
      CTP, Child–Turcotte–Pugh; INR, International normalized ratio; MELD, Model for end-stage liver disease; PV, Portal vein; PVT, Portal vein thrombosis; TIPS, transjugular intrahepatic portosystemic shunt.

      Risk Factors for Dysfunction

      The univariate logistic regression analysis showed that age, splenectomy, diabetes, complete blood count, INR, shunting branch of the PV, and initial stent position were correlated with the risk of stent dysfunction (Table 2). The multivariate logistic regression analysis demonstrated that diabetes (yes, OR = 3.531 [1.294-9.637], p = 0.014), splenectomy (yes, OR = 4.912 [1.702-14.176], p = 0.003), shunting branch of the PV (right, OR = 3.999 [1.629-9.816], p = 0.002), and initial stent position (suboptimal, OR = 6.635 [2.495-17.642], p < 0.001) were independent risk factors for stent dysfunction (Table 3).
      Table 2Univariate Logistic Regression in the Training Cohort
      VariablesStent Dysfunction n = 32 (12.9%)Stent Patency n = 217 (87.1%)OR95% CIp
      Age58.1 ± 9.753.4 ± 12.01.0361.002-1.0710.037
      Sex1.4430.680-3.0610.339
      Male18 (56.3%)141 (65.0%)
      Female14 (43.7%)76 (35.0%)
      Indications for TIPS1.0710.278-4.1340.920
      Acute variceal bleeding4 (12.5%)28 (12.9%)
      Recurrent variceal bleeding26 (81.3%)174 (80.2%)
      Refractory ascites2 (6.2%)15 (6.9%)
      Diabetes2.8911.338-6.2480.007
      Yes14 (43.7%)47 (21.7%)
      No18 (56.3%)170 (78.3%)
      PVT2.5371.069-6.0200.035
      Yes9 (28.1%)29 (13.4%)
      No23 (71.9%)188 (86.6%)
      Splenectomy3.0281.382-6.6340.006
      Yes12 (37.5%)42 (19.4%)
      No20 (62.5%)175 (80.6%)
      White blood cells (109/L)4.8 ± 2.45.2 ± 4.50.9760.887-1.0740.621
      Hemoglobin (g/L)79.3 ± 12.976.4 ± 13.11.0180.988-1.0480.241
      Platelets (109/L)113.5 ± 70.787.5 ± 64.91.0051.000-1.0100.043
      Serum sodium (mmol/L)138.7 ± 5.1139.7 ± 5.30.9660.903-1.0340.318
      INR1.3 ± 0.21.4 ± 0.30.2150.047-0.9920.049
      Serum creatinine (μmol/L)65.4 ± 20.266.8 ± 18.90.9960.976-1.0160.695
      Total bilirubin (μmol/L)24.3 ± 16.725.9 ± 21.70.9960.977-1.0150.694
      Albumin (g/L)30.7 ± 4.532.0 ± 8.40.9720.915-1.0330.365
      CTP scores7.8 ± 1.37.6 ± 1.41.1160.856-1.4530.418
      CTP classification1.6690.836-3.3290.146
      A3 (9.4%)48 (22.1%)
      B25 (78.1%)147 (67.8%)
      C4 (12.5%)22 (10.1%)
      MELD scores10.6 ± 3.211.4 ± 3.40.9190.811-1.0410.183
      Shunting branch of the PV4.9952.272-10.9830.000
      Left12 (37.5%)157 (72.4%)
      Right20 (62.5%)60 (27.6%)
      Types of stents1.4630.645-3.3170.362
      A single viatorr stent9 (28.1%)79 (36.4%)
      Fluency stent/bare stent combination23 (71.9%)138 (63.6%)
      Initial stent position5.1572.365-11.2480.000
      Optimal13 (40.6%)163 (75.1%)
      Suboptimal19 (59.4%)54 (24.9%)
      CTP, Child–Turcotte–Pugh; INR, International normalized ratio; MELD, Model for end-stage liver disease; PV, Portal vein; PVT, Portal vein thrombosis; TIPS, transjugular intrahepatic portosystemic shunt.
      Table 3Multivariate Logistic Regression in the Training Cohort
      VariablesSEOR95% CIp
      Age0.0201.0210.981-1.0620.307
      Diabetes0.5123.5311.294-9.6370.014
      PVT0.5832.4430.779-7.6550.125
      Splenectomy0.5414.9121.702-14.1760.003
      Platelets0.0031.0020.996-1.0090.534
      INR0.9340.2790.045-1.7410.172
      Shunting branch of the PV0.4583.9991.629-9.8160.002
      Initial stent position0.4496.6352.495-17.6420.000
      INR, International normalized ratio; PV, Portal vein; PVT, Portal vein thrombosis.

      Development and Validation of the Nomogram

      The nomogram was constructed based on the independent risk factors (Fig 3). The AUC values in the training and validation cohorts were 0.817 (95% CI: 0.731-0.903) and 0.804 (95% CI: 0.673-0.935), respectively (Fig 4), which indicated good predictive stability. Moreover, the calibration curves demonstrated good agreement between the observed outcome and prediction in both cohorts (Fig 5). DCA indicated that when the threshold probability ranged between 2% and 88%, the use of the nomogram to identify stent dysfunction in hepatitis B cirrhosis patients provided a net benefit (Fig 6).
      Figure 3
      Figure 3The nomogram for predicting the probability of stent dysfunction.
      Figure 4
      Figure 4The AUC values in the training and validation cohorts were 0.817 (95% CI: 0.731-0.903) and 0.804 (95% CI: 0.673-0.935), respectively. (Color version of figure is available online.)
      Figure 5
      Figure 5Calibration curves for the training (a) and validation cohorts (b).
      Figure 6
      Figure 6The DCA for the prediction nomogram. The y-axis measures the net benefit. The horizontal black line represents the assumption that no patients should take the necessary measures, whereas the gray line represents the assumption that all of the patients should take the necessary measures. The blue line represents the prediction nomogram, which showed that when the threshold probability was between 2% and 88%, the nomogram could provide clinical usefulness and net benefit. DCA, decision curve analysis. (Color version of figure is available online.)

      DISCUSSION

      In this study, the univariate logistic regression analysis, followed by multivariate logistic regression analysis, were performed to identify independent risk factors for stent dysfunction. Eventually, four variables (splenectomy, diabetes, shunting branch of the PV, and initial stent position) were selected as independent risk factors for stent dysfunction in hepatitis B cirrhosis patients.
      In the previous research, splenectomy has been confirmed as an independent risk factor for TIPS stent dysfunction (
      • Yang C
      • Liu J
      • Shi Q
      • et al.
      Effect of splenectomy on the outcomes in patients with cirrhosis receiving transjugular intrahepatic portosystemic shunt.
      ,
      • Qi X
      • He C
      • Guo W
      • et al.
      Transjugular intrahepatic portosystemic shunt for portal vein thrombosis with variceal bleeding in liver cirrhosis: outcomes and predictors in a prospective cohort study.
      ,
      • Luo X
      • Zhao M
      • Wang X
      • et al.
      Long-term patency and clinical outcome of the transjugular intrahepatic portosystemic shunt using the expanded polytetrafluoroethylene stent-graft.
      ). Qi et al. found that patients who underwent splenectomy had a significantly higher rate of shunt dysfunction after TIPS placement (
      • Qi X
      • He C
      • Guo W
      • et al.
      Transjugular intrahepatic portosystemic shunt for portal vein thrombosis with variceal bleeding in liver cirrhosis: outcomes and predictors in a prospective cohort study.
      ,
      • Luo X
      • Zhao M
      • Wang X
      • et al.
      Long-term patency and clinical outcome of the transjugular intrahepatic portosystemic shunt using the expanded polytetrafluoroethylene stent-graft.
      ). Furthermore, Yang et al. and Dong et al. reported that patients with splenectomy had a high incidence of portal vein thrombosis, which can markedly affect TIPS stent patency (
      • Yang C
      • Liu J
      • Shi Q
      • et al.
      Effect of splenectomy on the outcomes in patients with cirrhosis receiving transjugular intrahepatic portosystemic shunt.
      ,
      • Dong F
      • Luo SH
      • Zheng LJ
      • et al.
      Incidence of portal vein thrombosis after splenectomy and its influence on transjugular intrahepatic portosystemic shunt stent patency.
      ). These results are consistent with the findings of our study. Moreover, in our study, a history of diabetes was first identified as an independent risk factor for TIPS stent dysfunction. This is likely explained by the hypercoagulable state of diabetic patients, which may induce acute thrombosis. Hyperglycemia in diabetes damages vascular endothelial cells and increases platelet activity, and abnormal lipid and protein metabolism may increase clotting and viscosity (
      • Chen Z
      • Ding S
      • Yuan Y
      • et al.
      Relationship between perioperative cardiovascular events and glycated hemoglobin in diabetic patients undergoing noncardiac surgery.
      ).
      In addition, we identified the shunting branch of the PV as being an independent risk factor for TIPS stent dysfunction. Several studies have shown that the rate of long-term patency of the left branch of the portal vein is significantly higher than that of the right branch (
      • Bai M
      • He CY
      • Qi XS
      • et al.
      Shunting branch of portal vein and stent position predict survival after transjugular intrahepatic portosystemic shunt.
      ,
      • Clark TW
      • Agarwal R
      • Haskal ZJ
      • et al.
      The effect of initial shunt outflow position on patency of transjugular intrahepatic portosystemic shunts.
      ,
      • Chen SL
      • Hu P
      • Lin ZP
      • et al.
      The effect of puncture sites of portal vein in TIPS with ePTFE-covered stents on postoperative long-term clinical efficacy.
      ). The main reasons for this effect are as follows: (1) the stent shunt has better plasticity and is less likely to produce pseudointimal hyperplasia and stenting stricture due to the straighter puncture trajectory between the middle hepatic vein and the left portal vein (
      • Uflacker R
      • Reichert P
      • D' Albuquerque LC
      • et al.
      Liver anatomy applied to the placement of transjugular intrahepatic portosystemic shunts.
      ); (2) the distance between the left branch of the portal vein and the hepatic vein is shorter, which facilitates shunt creation and decreases the contact area between the stent and liver parenchyma, thus reducing the probability of pseudointimal hyperplasia and growth of the liver parenchyma into the stent (
      • Haskal ZJ
      • Ring EJ
      • LaBerge JM
      • et al.
      Role of parallel transjugular intrahepatic portosystemic shunts in patients with persistent portal hypertension.
      ); and (3) the laminar shear stress of the left branch of the portal vein causes less turbulence during stent shunting and reduces the risk of thrombosis (
      • Li X
      • Wang XK
      • Chen B
      • et al.
      Computational hemodynamics of portal vein hypertension in hepatic cirrhosis patients.
      ). Therefore, the left portal vein should be chosen for shunt creation in TIPS procedures, which could reduce the incidence of stent dysfunction.
      Furthermore, our study found that the initial stent position was significantly related to TIPS stent dysfunction. The initial stent position was defined as optimal and suboptimal, with the latter having a higher risk of stent dysfunction. When the cephalic end of the stent does not extend to the hepato-caval junction, the possibility of pseudointimal hyperplasia and thrombosis will be increased, which will ultimately lead to stent dysfunction (
      • Ducoin H
      • El-Khoury J
      • Rousseau H
      • et al.
      Histopathologic analysis of transjugular intrahepatic portosystemic shunts.
      ,
      • Rossi P
      • Salvatori FM
      • Fanelli F
      • et al.
      Polytetrafluoroethylene-covered nitinol stent-graft for transjugular intrahepatic portosystemic shunt creation: 3-year experience.
      ,
      • Luo X
      • Nie L
      • Tsauo J
      • et al.
      Parallel shunt for the treatment of transjugular intrahepatic portosystemic shunt dysfunction.
      ). Clark et al. reported that the initial stent position within the hepatic venous outflow is predictive of shunt patency, with TIPS extension to the hepatocaval junction having a longer lifespan than shunts terminating in the HV (
      • Clark TW
      • Agarwal R
      • Haskal ZJ
      • et al.
      The effect of initial shunt outflow position on patency of transjugular intrahepatic portosystemic shunts.
      ). In addition, if the tangent angle between the caudal end of the stent and the wall of the portal vein is too large, it causes chronic damage to the intima, thus leading to stent stenosis or occlusion (
      • Luo SH
      • Chu JG
      • Huang H
      • et al.
      Effect of initial stent position on patency of transjugular intrahepatic portosystemic shunt.
      ).
      Nomograms are a simple and powerful tool for individualized disease-related risk estimations. Nomograms that can predict hepatic encephaloapthy-free survival and 3-month mortality after TIPS procedures have been described in the previous literature. However, to the best of our knowledge, no studies have been demonstrated to date on prediction models for TIPS stent dysfunction (
      • Yin X
      • Zhang F
      • Guo H
      • et al.
      A nomogram to predict the risk of hepatic encephalopathy after transjugular intrahepatic portosystemic shunt in Cirrhotic Patients.
      ,
      • Malinchoc M
      • Kamath PS
      • Gordon FD
      • et al.
      A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts.
      ). In our study, based on independent risk factors related to stent dysfunction, we constructed and validated a nomogram that may help clinicians in monitoring stent dysfunction and in optimizing individualized treatment strategies. Specifically, the risk of TIPS stent dysfunction in each patient can be calculated by using our nomogram and clinicians should pay more attention to patients with a high risk of stent dysfunction. For example, consider a hepatitis B cirrhosis patient who underwent TIPS placement with previous splenectomy, diabetes, left shunting branch of the PV, and suboptimal initial stent position. For this patient, we could use the nomogram to calculate the risk by adding the points assigned to each risk factor. Finally, we obtained a total score of 254, which indicate that the stent dysfunction probability was approximately 63.7%, and it was recommended to increase follow-up times for this patient, especially regarding CDUS for stent dysfunction.
      Nevertheless, the present study had several limitations. First, as a retrospective study, there is an inevitable data bias. Second, the patient population was limited to hepatitis B cirrhosis patients, and we only predicted the probability of 1-year TIPS stent dysfunction. Further research should be performed to predict the long-term patency rate of TIPS stents. Not all of the patients underwent venography, which may have caused TIPS dysfunction to be underestimated in the current cohort. In addition, diabetes is the only newly identified risk factor for stent dysfunction after TIPS, and the other 3 factors are well known. Finally, internal and external validations of the nomogram were performed by using data from the same center, and the included variables were limited. Multicenter studies involving more variables should be conducted to establish a nomogram model for predicting TIPS stent dysfunction.

      CONCLUSIONS

      In summary, we established and validated a nomogram for evaluating the risk of TIPS stent dysfunction in hepatitis B cirrhosis patients. The nomogram provides a promising tool to optimize individual management for hepatitis B cirrhosis patients who underwent TIPS.

      Funding

      This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

      Acknowledgments

      Thanks for our colleagues of Interventional Operating Room.

      References

        • Hung ML
        • Lee EW.
        Role of transjugular intrahepatic portosystemic shunt in the management of portal hypertension: review and update of the literature.
        Clin Liver Dis. 2019; 23: 737-754https://doi.org/10.1016/j.cld.2019.07.004
        • Bureau C
        • Thabut D
        • Oberti F
        • et al.
        Transjugular intrahepatic portosystemic shunts with covered stents increase transplant-free survival of patients with cirrhosis and recurrent ascites.
        Gastroenterology. 2017; 152: 157-163https://doi.org/10.1053/j.gastro.2016.09.016
        • Brunner F
        • Berzigotti A
        • Bosch J.
        Prevention and treatment of variceal haemorrhage in 2017.
        Liver Int. 2017; 37 (Suppl 1): 104-115https://doi.org/10.1111/liv.13277
        • Hayek G
        • Ronot M
        • Plessier A
        • et al.
        Long-term outcome and analysis of dysfunction of transjugular intrahepatic portosystemic shunt placement in chronic primary budd-chiari syndrome.
        Radiology. 2017; 283: 280-292https://doi.org/10.1148/radiol.2016152641
        • Li YH
        • Xu ZY
        • Wu HM
        • et al.
        Long-term shunt patency and overall survival of transjugular intrahepatic portosystemic shunt placement using covered stents with bare stents versus covered stents alone.
        Clin Radiol. 2018; 73: 580-587https://doi.org/10.1016/j.crad.2018.01.014
        • Patidar KR
        • Sydnor M
        • Sanyal AJ.
        Transjugular intrahepatic portosystemic shunt.
        Clin Liver Dis. 2014; 18: 853-876https://doi.org/10.1016/j.cld.2014.07.006
        • Bureau C
        • Garcia-Pagan JC
        • Otal P
        • et al.
        Improved clinical outcome using polytetrafluoroethylene-coated stents for TIPS: results of a randomized study.
        Gastroenterology. 2004; 126: 469-475https://doi.org/10.1053/j.gastro.2003.11.016
        • Wan YM
        • Li YH
        • Xu ZY
        • et al.
        Comparison of TIPS alone and combined with partial splenic embolization (PSE) for the management of variceal bleeding.
        Eur Radiol. 2019; 29: 5032-5041https://doi.org/10.1007/s00330-019-06046-6
        • Qi X
        • Han G.
        Transjugular intrahepatic portosystemic shunt in the treatment of portal vein thrombosis: a critical review of literature.
        Hepatol Int. 2012; 6: 576-590https://doi.org/10.1007/s12072-011-9324-5
        • Bai M
        • He CY
        • Qi XS
        • et al.
        Shunting branch of portal vein and stent position predict survival after transjugular intrahepatic portosystemic shunt.
        World J Gastroenterol. 2014; 20: 774-785https://doi.org/10.3748/wjg.v20.i3.774
        • Weber CN
        • Nadolski GJ
        • White SB
        • et al.
        Long-term patency and clinical analysis of expanded polytetrafluoroethylene-covered transjugular intrahepatic portosystemic shunt stent grafts.
        J Vasc Interv Radiol. 2015; 26 (quiz 1265): 1257-1265https://doi.org/10.1016/j.jvir.2015.04.005
        • Wan YM
        • Li YH
        • Xu Y
        • et al.
        Predictors of shunt dysfunction and overall survival in patients with variceal bleeding treated with transjugular portosystemic shunt creation using the fluency stent graft.
        Acad Radiol. 2018; 25: 925-934https://doi.org/10.1016/j.acra.2017.11.020
        • Cura M
        • Cura A
        • Suri R
        • et al.
        Causes of TIPS dysfunction.
        AJR Am J Roentgenol. 2008; 191: 1751-1757https://doi.org/10.2214/AJR.07.3534
        • Yang C
        • Liu J
        • Shi Q
        • et al.
        Effect of splenectomy on the outcomes in patients with cirrhosis receiving transjugular intrahepatic portosystemic shunt.
        J Gastroenterol Hepatol. 2021; https://doi.org/10.1111/jgh.15543
        • Perarnau JM
        • Le Gouge A
        • Nicolas C
        • et al.
        Covered vs. uncovered stents for transjugular intrahepatic portosystemic shunt: a randomized controlled trial.
        J Hepatol. 2014; 60: 962-968https://doi.org/10.1016/j.jhep.2014.01.015
        • Huang Z
        • Yao Q
        • Zhu J
        • et al.
        Efficacy and safety of transjugular intrahepatic portosystemic shunt (TIPS) created using covered stents of different diameters: a systematic review and meta-analysis.
        Diagn Interv Imaging. 2021; 102: 279-285https://doi.org/10.1016/j.diii.2020.11.004
        • Ducoin H
        • El-Khoury J
        • Rousseau H
        • et al.
        Histopathologic analysis of transjugular intrahepatic portosystemic shunts.
        Hepatology. 1997; 25: 1064-1069https://doi.org/10.1002/hep.510250503
        • Clark W
        • Golkar F
        • Luberice K
        • et al.
        Uncovering the truth about covered stents: is there a difference between covered versus uncovered stents with transjugular intrahepatic portosystemic shunts.
        Am J Surg. 2011; 202: 561-564https://doi.org/10.1016/j.amjsurg.2011.06.021
        • Wei L
        • Champman S
        • Li X
        • et al.
        Beliefs about medicines and non-adherence in patients with stroke, diabetes mellitus and rheumatoid arthritis: a cross-sectional study in China.
        BMJ Open. 2017; 7e017293https://doi.org/10.1136/bmjopen-2017-017293
        • Shariat SF
        • Karakiewicz PI
        • Suardi N
        • et al.
        Comparison of nomograms with other methods for predicting outcomes in prostate cancer: a critical analysis of the literature.
        Clin Cancer Res. 2008; 14: 4400-4407https://doi.org/10.1158/1078-0432.CCR-07-4713
        • Clark TW
        • Agarwal R
        • Haskal ZJ
        • et al.
        The effect of initial shunt outflow position on patency of transjugular intrahepatic portosystemic shunts.
        J Vasc Interv Radiol. 2004; 15: 147-152https://doi.org/10.1097/01.rvi.0000109401.52762.56
        • Qi X
        • He C
        • Guo W
        • et al.
        Transjugular intrahepatic portosystemic shunt for portal vein thrombosis with variceal bleeding in liver cirrhosis: outcomes and predictors in a prospective cohort study.
        Liver Int. 2016; 36: 667-676https://doi.org/10.1111/liv.12929
        • Luo X
        • Zhao M
        • Wang X
        • et al.
        Long-term patency and clinical outcome of the transjugular intrahepatic portosystemic shunt using the expanded polytetrafluoroethylene stent-graft.
        PLoS One. 2019; 14e0212658https://doi.org/10.1371/journal.pone.0212658
        • Dong F
        • Luo SH
        • Zheng LJ
        • et al.
        Incidence of portal vein thrombosis after splenectomy and its influence on transjugular intrahepatic portosystemic shunt stent patency.
        World J Clin Cases. 2019; 7: 2450-2462https://doi.org/10.12998/wjcc.v7.i17.2450
        • Chen Z
        • Ding S
        • Yuan Y
        • et al.
        Relationship between perioperative cardiovascular events and glycated hemoglobin in diabetic patients undergoing noncardiac surgery.
        Biomed Res Int. 2020; (2020)3645374https://doi.org/10.1155/2020/3645374
        • Chen SL
        • Hu P
        • Lin ZP
        • et al.
        The effect of puncture sites of portal vein in TIPS with ePTFE-covered stents on postoperative long-term clinical efficacy.
        Gastroenterol Res Pract. 2019; (2019)2935498https://doi.org/10.1155/2019/2935498
        • Uflacker R
        • Reichert P
        • D' Albuquerque LC
        • et al.
        Liver anatomy applied to the placement of transjugular intrahepatic portosystemic shunts.
        Radiology. 1994; 191: 705-712https://doi.org/10.1148/radiology.191.3.8184050
        • Haskal ZJ
        • Ring EJ
        • LaBerge JM
        • et al.
        Role of parallel transjugular intrahepatic portosystemic shunts in patients with persistent portal hypertension.
        Radiology. 1992; 185: 813-817https://doi.org/10.1148/radiology.185.3.1438768
        • Li X
        • Wang XK
        • Chen B
        • et al.
        Computational hemodynamics of portal vein hypertension in hepatic cirrhosis patients.
        Biomed Mater Eng. 2015; 26 (Suppl 1): S233-S243https://doi.org/10.3233/BME-151310
        • Rossi P
        • Salvatori FM
        • Fanelli F
        • et al.
        Polytetrafluoroethylene-covered nitinol stent-graft for transjugular intrahepatic portosystemic shunt creation: 3-year experience.
        Radiology. 2004; 231: 820-830https://doi.org/10.1148/radiol.2313030349
        • Luo X
        • Nie L
        • Tsauo J
        • et al.
        Parallel shunt for the treatment of transjugular intrahepatic portosystemic shunt dysfunction.
        Korean J Radiol. 2013; 14: 423-429https://doi.org/10.3348/kjr.2013.14.3.423
        • Luo SH
        • Chu JG
        • Huang H
        • et al.
        Effect of initial stent position on patency of transjugular intrahepatic portosystemic shunt.
        World J Gastroenterol. 2017; 23: 4779-4787https://doi.org/10.3748/wjg.v23.i26.4779
        • Yin X
        • Zhang F
        • Guo H
        • et al.
        A nomogram to predict the risk of hepatic encephalopathy after transjugular intrahepatic portosystemic shunt in Cirrhotic Patients.
        Sci Rep. 2020; 10: 9381https://doi.org/10.1038/s41598-020-65227-2
        • Malinchoc M
        • Kamath PS
        • Gordon FD
        • et al.
        A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts.
        Hepatology. 2000; 31: 864-871https://doi.org/10.1053/he.2000.5852