Jianjun Luo

Background and Aims: Hepatic venous pressure gradient (HVPG) is the gold standard for diagnosing clinically significant portal hypertension in patients with viral‐ and alcohol‐related cirrhosis, but its accuracy is limited in other aetiologies. This study aimed to evaluate the correlation of portal venous pressure measurements in patients with non‐alcoholic, non‐viral‐related decompensated cirrhosis, using fine‐needle puncture (FN‐PVP), direct portal vein catheterisation (D‐PVP) and wedged hepatic vein balloon occlusion (WHVP). Methods: This study retrospectively included patients from the author's centre who simultaneously underwent FN‐PVP, HVPG and transjugular intrahepatic portosystemic shunt (TIPS) placement for variceal bleeding or refractory ascites. The Spearman correlation coefficient (r), linear regression coefficient (R ² ), intraclass correlation coefficient (ICC) and the Bland–Altman method were used to assess the correlation and consistency between the pressure measurements obtained by the different methods. Results: Seventy‐one patients were included, all undergoing FN‐PVP measurements without procedure‐related complications. The correlation and consistency between WHVP and D‐PVP were moderate ( r = 0.487, R ² = 0.25), while FN‐PVP demonstrated excellent correlation and consistency with D‐PVP ( r = 0.945, R ² = 0.914). Similar patterns were observed when comparing HVPG and direct portal pressure gradient (D‐PPG), fine‐needle portal pressure gradient (FN‐PPG) and D‐PPG (r: 0.558 vs. 0.918; R ² : 0.302 vs. 0.87). In patients without intrahepatic venovenous shunts (IHVS), the correlation and consistency between WHVP and D‐PVP showed slight improvement ( r = 0.609, ICCa = 0.523, ICCc = 0.620). Using D‐PPG as the benchmark, FN‐PPG demonstrated excellent classification accuracy, significantly outperforming HVPG. Conclusion: In patients with non‐alcoholic, non‐viral‐related PHT, HVPG may underestimate the true PPG, regardless of the presence of IHVS. In contrast, FN‐PVP can safely and accurately reflect portal pressure and serves as an effective alternative for direct portal pressure measurement.

This study aimed to construct a machine learning (ML) model to facilitate non-invasive identification of moderate-to-severe intrahepatic venovenous shunts (MS-IHVS) before hepatic venous pressure gradient (HVPG) measurement. A total of 605 patients receiving HVPG measurement were retrospectively included for model development. Two validation cohorts were used: an FN-validation cohort (n = 194; retrospectively enrolled) with both wedged hepatic venous pressure (WHVP) and fine-needle (21-22G) portal venous pressure (FN-PVP) measurements, and a TIPS-validation cohort (n = 92; prospectively enrolled with WHVP and direct PVP measurements. Nine ML and eleven stacking ensemble models were developed. Performance was evaluated using area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, specificity, PPV, NPV, F1-score, calibration plots, confusion matrices, and decision-curve analysis. The Shapley Additive Explanation was used for model interpretation. Four predictors were selected for model development, including portal vein thrombosis, spontaneous portosystemic shunts, prothrombin time, and number of endoscopic treatments. Across all single and stacking models, the stacking ensemble model combining Random Forest (RF) and Gradient Boosting (GB) performed best (AUC 0.907, 95%CI 0.858-0.945). A Comprehensive Rank also placed RF + GB first. AUCs were 0.781 in FN-validation and 0.889 in TIPS-validation. Model-based risk stratification showed markedly reduced WHVP-PVP/FN-PVP agreement and correlation (intraclass correlation coefficient for agreement 0.314 and 0.143; r = 0.579 and 0.347) in high-risk group compared with low-risk group. This RF + GB stacking model effectively identified the risk of MS-IHVS in patients with sinusoidal PHT, providing a powerful tool for early detection and avoiding unnecessary invasive examination.

Background and aim Cirrhotic patients receiving trans-jugular intrahepatic portosystemic shunt (TIPS) for preventing re-bleeding may benefit from a shunt diameter of ≤ 8 mm. This study aimed to emulate a target trial comparing decompensation-free survival in patients receiving coaxial stent versus fully- or under-dilated 8 mm TIPS for preventing variceal re-bleeding. Approach Participants were recruited from a single-center, prospective, non-randomized cohort after excluding patients impossible to be assigned to the coaxial stent group. The coaxial stent consisted of a pre-implanted 5-/6 mm balloon-expandable stent with an internal 8 mm Viatorr stent. Causal inference framework was designed based on a directed acyclic graph, incorporating inverse probability weighting (IPW), doubly robust estimation by augmented IPW, mediation analysis, and E value analysis. Results 200 patients entered into target trial emulation, with 47 (23.5%), 101 (50.5%), and 52 (26.0%) in coaxial stent, under-dilated, and fully dilated groups. Coaxial stent feasibility was demonstrated by portacaval pressure gradient adjustability (33 pairs before/after dilation: 22.8 ± 3.0 to 17.9 ± 3.1 mmHg, p < 0.001) and stability (76 pairs without intervention: 17.7 ± 4.2 to 18.5 ± 4.5 mmHg, p = 0.035, r = 0.712, ICC = 0.710). Crude (IPW-weighted) 1-year decompensation-free survival incidences were 57.7% (52.6%), 60.1% (60.0%), and 76.6% (78.9%) for fully dilated, under-dilated, and coaxial stent groups, with doubly robust analysis revealing improved prognosis in coaxial stent versus fully dilated (HR 0.54 [0.25–0.93], p = 0.026, E value = 3.108, direct effect 66.3%) and under-dilated (HR 0.63 [0.34–1.05], p = 0.070, E value = 2.553, direct effect 60.7%) groups. Coaxial stent group demonstrated significantly reduced hepatic encephalopathy risk compared with other groups (p = 0.045/0.026), while other outcomes were comparable. Conclusions The coaxial stent TIPS strategy could effectively achieve controlled expansion of 5–8 mm and may improve overall prognosis in cirrhotic patients with previous variceal bleeding.

Background and Aims: While remeasuring portacaval pressure gradient ( PPG ) after transjugular intrahepatic portosystemic shunt ( TIPS ), it provides superior prognostic information, and its clinical utility is limited by invasiveness. We aimed to develop and validate an explainable machine learning ( ML ) model for predicting short‐term PPG changes and improving prognostic value of PPG in cirrhotic patients undergoing TIPS . Methods: We enrolled 328 and 128 patients (2018–2023) in retrospective training and prospective validation cohorts, respectively. PPG was measured pre‐TIPS, immediately (imePPG) and 2–4 days post‐TIPS (delPPG). All ML models were developed using hyperparameter tuning with 5‐fold cross‐validation. Results: Support vector regression model (SVR_RBF) demonstrated superior performance across feature subsets, with its 6‐feature model achieving optimal balance between robustness and simplicity. In the validation cohort, the final model significantly improved PPG agreement (R‐squared: 0.265–0.617, mean square error: 16.64–4.90, mean absolute error: 3.45–1.65, intraclass correlation coefficient: 0.533–0.841). SHapley Additive exPlanation analysis identified imePPG reduction from baseline as the primary determinant of subsequent PPG changes. Restricted cubic spline analyses revealed significant nonlinear relationships between relative PPG reduction (but not absolute values) and 2‐year further decompensation in the training cohort. Optimal reduction thresholds were defined as U‐shaped curves with hazard ratio (HR) nodes at 1 (imePPG: 30%–70%, delPPG: 20%–55%), which were validated in the prospective cohort showing significant associations for both true ( p = 0.003, HR = 0.45) and predicted ( p = 0.044, HR = 0.57) delPPG, but not for imePPG ( p = 0.196). Conclusions: The six‐feature SVR_RBF model enhances the accuracy and prognostic utility of PPG measurements before TIPS completion, supporting clinical decision‐making without additional invasive assessments.

Background and aim The effect of shunting the left or right branch of the portal vein on the post-transjugular intrahepatic portosystemic shunt (TIPS) hepatic encephalopathy (HE) remains controversial. Therefore, we designed this prospective, randomized, single-blinded trial to compare the incidence of HE after shunting the left or right branch of the portal vein during TIPS in cirrhotic patients to prevent variceal rebleeding. Methods Consecutive patients diagnosed with hepatitis B virus (HBV)-related cirrhosis were randomly assigned in a 1:1 ratio to undergo the TIPS procedure, with the shunt directed to either the left or right branch of the portal vein. The primary outcome was the incidence of overt HE within 2 years. The secondary outcomes included shunt dysfunction, variceal rebleeding, survival, adverse events, and changes in liver function. Results Between June 2019 and December 2020, 130 patients were randomly divided into two groups (left branch group vs right branch group). The 2-year cumulative incidence rates of free overt HE in the left branch group did not demonstrate statistical superiority over the right branch group (78.5% vs 68.6%, respectively; p = 0.236, HR = 1.49, 95% CI 0.76–2.93). Similarly, no statistical differences were observed in stent patency, all-cause bleeding, survival, or adverse events. Conclusion Compared to shunting the right branch of the portal vein, shunting the left branch during TIPS placement did not impact the incidence of post-TIPS HE and other clinical outcomes in cirrhotic patients to prevent variceal rebleeding. The choice of puncture site should be primarily guided by the patient’s liver anatomy and the operator’s experience.

To evaluate the correlation and consistency between hepatic venous pressure gradient(F-HVPG) calculated as the wedged hepatic venous pressure (WHVP) minus free hepatic venous pressure (FHVP), I-HVPG calculated as WHVP minus inferior vena cava pressure (IVCP) in the hepatic segment, and portal pressure gradient (PPG). Data were collected from 112 patients with portal hypertension undergoing transjugular intrahepatic portosystemic shunt (TIPS) along with HVPG measurement. FHVP, IVCP, WHVP, and portal venous pressure (PVP) were collected intraoperatively. Pearson’s correlation and Bland–Altman method were used to assess correlation and consistency. A total of 112 patients were retrospectively collected. The correlation coefficient (r) values (p < 0.001) between FHVP and IVCP, WHVP and PVP, F-HVPG and I-HVPG, F-HVPG and PPG, I-HVPG and PPG were 0.835, 0.717, 0.946, 0.667 and 0.698, respectively; the determination coefficient (R2) values were 0.697, 0.514, 0.895, 0.445 and 0.487, respectively. Bland–Altman plots showed that F-HVPG and I-HVPG had the narrowest 95% limits of agreement. Among patients with FHVP-IVCP > 2 mmHg, the (r) values (p < 0.05) between F-HVPG and I-HVPG, F-HVPG and PPG, I-HVPG and PPG were 0.907, 0.648 and 0.807, respectively; the (R2) values were 0.822, 0.420 and 0.651, respectively. Bland–Altman plots showed that I-HVPG had the narrower 95% limits of agreement with PPG. F-HVPG and I-HVPG demonstrated high correlation and consistency. I-HVPG consistently correlates more closely with PPG than F-HVPG, both in the overall cohort and in patients with FHVP-IVCP > 2 mmHg. These results suggest that I-HVPG may serve as a more reliable alternative. Due to the significant underestimation of the PPG, HVPG measurement should not be used to exclude patients from a TIPS intervention.