Individualized portal pressure gradient threshold based on liver function categories in preventing rebleeding after TIPS

Background The evidence in Portal pressure gradient (PPG) < 12 mmHg after transjugular intrahepatic portosystemic shunt (TIPS) for preventing rebleeding mostly comes from observations in uncovered stents era. Moreover, association between Child–Pugh classes and post-TIPS hepatic encephalopathy (HE) has indicated that tolerance of PPG reduction depends on liver function. This study aimed to investigate the optimal PPG for covered TIPS and explore the optimal threshold tailored to the Child–Pugh classes to find individualized PPG to balance rebleeding and overt HE. Methods This multicenter retrospective study analyzed rebleeding, OHE, and mortality of patients associated with post-TIPS PPGs (8, 10, 12, and 14 mmHg) in the entire cohort and among different Child–Pugh classes. Propensity score matching (PSM) and competing risk analyses were performed for sensitivity analyses. Results We included 2100 consecutively screened patients undergoing TIPS. In all patients, PPG < 12 mmHg reduced rebleeding after TIPS (p = 0.022). In Child–Pugh class A, none of the PPG thresholds were discriminative of clinical outcomes. In Child–Pugh class B, 12 mmHg (p = 0.022) and 14 mmHg (p = 0.037) discriminated rebleeding, but 12 mmHg showed a higher net benefit. In Child–Pugh class C, PPG < 14 mmHg had a lower rebleeding incidence (p = 0.017), and exhibited more net benefit than 12 mmHg. Conclusion Different PPG standards may be required for patients with different liver function categories. A PPG threshold < 12 mmHg might be suitable for patients in Child–Pugh class B, while < 14 mmHg might be optimal for patients in Child–Pugh class C. Graphical Abstract Supplementary Information The online version contains supplementary material available at 10.1007/s12072-023-10489-x.


Introduction
Transjugular intrahepatic portosystemic shunt (TIPS) is commonly used to treat portal hypertension and its complications. The role of TIPS in the management of portal hypertension complications has been studied. It is recommended as the first-line treatment for patients with acute variceal bleeding and as the treatment choice for patients who fail standard β-blocker/endoscopy to prevent variceal rebleeding [1,2]. The reduction in the portal pressure gradient (PPG) because of TIPS causes the following two major concerns: how to achieve efficient prevention of rebleeding and how to avoid excessive hepatic encephalopathy (HE) caused by over-shunt.
According to past practices and the Baveno VII recommendation [1,3], a target PPG of < 12 mmHg or > 50% of the baseline is expected to prevent rebleeding [4][5][6][7][8][9]. However, most studies on hemodynamic targets were performed before the introduction of covered stents and have not been adequately updated [4,5]. HE risk is still high with covered stents for TIPS, even when using the current PPG standard [10][11][12], limiting their widespread clinical application. Therefore, it is important to explore the PPG threshold suitable for covered stents for TIPS to prevent rebleeding and reduce overt HE (OHE) risk [13].
There is accumulating evidence that Child-Pugh classes /scores are one of the best predictors of post-TIPS HE [14][15][16]. High Child-Pugh classes /scores have shown the greatest potential for predicting post-TIPS HE, indicating different tolerances to PPG reduction in patients with different liver function classes. Therefore, we hypothesized that the optimal PPG threshold should be adjusted according to the Child-Pugh classes.
In this multicenter retrospective study, we investigated the effect of PPG reduction in patient clinical outcomes after TIPS to examine the beneficial effect of PPG after covered TIPS and to explore whether an individualized PPG threshold is needed in patients with different Child-Pugh classes.

Patients
This retrospective study enrolled 3633 patients who received treatment using TIPS for variceal bleeding at six academic hospitals in China between January 2010 and June 2020. The disposition of patients is shown as a flow chart in Supplementary Fig. 1. All hospitals had a gastroenterology/liver unit, and experienced clinicians that perform TIPS regularly. All patients received etiological treatment with continued antiviral therapy, alcohol abstinence, or ursodeoxycholic acid depending on the cause of cirrhosis. And all patients were treated with medication or endoscopic therapy before TIPS according to the guidelines of portal hypertension [1]. Lactulose or rifaximin was used to prevent hepatic encephalopathy for 3 months after TIPS.

Study design
The primary endpoint of this study was all-cause rebleeding. The secondary endpoints were all-cause mortality and development of OHE. Physical examination, biochemical and hematological tests, and abdominal ultrasound data were obtained. Follow-up time was defined as the interval from the date of admission to death, liver transplantation, last visit, or study end (September 6, 2021). OHE was defined as HE grade ≥ 2 according to the West Haven-modified criteria [17,18].

TIPS procedure and PPG measurement
Participating centers are fully experienced in TIPS procedure and conformed to the TIPS Clinical Practice Guidelines by Chinese Medical Association [19]. An expanded polytetrafluoroethylene-covered stent with a suitable diameter (6, 8, or 10 mm), Viatorr TIPS endoprosthesis (Gore, Inc.; Flagstaff, AZ, USA), or Fluency stent (Bard, Inc., Tempe, AZ, USA) was deployed using the transjugular approach and was initially dilated to 6, 8, or 10 mm. The selection of stent diameter is mainly based on the operator's experience and patient's characteristics of pre-TIPS PPG, pre-TIPS HE, liver function and the availability of stents. PPG was obtained preoperatively (baseline PPG) and immediately after stent placement (immediate PPG). PPG was measured as the difference between the pressure measured in the portal vein (PV) and inferior vena cava (rather than the right atrium) with a 5.5F Fogarty Thru-Lumen Embolectomy Catheter (Edwards Lifesciences LLC, Irvine, USA). All TIPS procedures were performed under local anesthesia (In China, it is rare for TIPS to be performed under general anesthesia, except for emergency TIPS; it avoids the influence of deep sedation on PPG). Permanent tracings from all pressure measurements were obtained using a recalibrated Mac-Lab electromechanical transducer and a polygraph (GE Healthcare, Freiburg, Germany).
In addition, all patients enrolled in this study underwent TIPS for secondary prophylaxis of variceal bleeding, which avoided the influence of emergency vasoconstrictive drugs on PPG measurement. Therefore, the immediate PPG in this study could represent the actual PV pressure in these patients. PPG was measured according to uniform procedures at each medical center, and all PPG measurements that did not follow the established procedures were excluded.

Statistical analysis
Quantitative variables were reported as medians (quartiles) and compared using the Mann-Whitney U test because they were non-normally distributed. Qualitative variables were reported as frequencies (percentages) and compared using the χ 2 test. Multivariate Cox proportional hazards analysis was used to identify independent risk factors, and hazard ratios (HR) with 95% confidence intervals (95% CI) were reported. The cumulative incidences of OHE and rebleeding were estimated in a competing risk setting, where death or liver transplantation competed with the event of interest. When mortality was estimated, liver transplantation and non-cirrhotic death were considered as competing risks. Decision curve analysis (DCA) was performed to determine which of the two potential PPGs was better.
Propensity score matching (PSM) was performed between the different post-PPG groups to provide a balanced baseline. We took a 1: n PSM match, where n was as large as possible to reduce data loss. The nearest neighbor matching method with a 0.4 caliper was used to construct the control group. The standard mean difference (SMD) was used to assess whether the groups were comparable after PSM. SMD < 0.2 indicated a small difference between groups [20]. The PSM details for each group are presented in Supplementary Tables 2-17. All results with a two-sided p < 0.05 were considered statistically significant. All analyses were performed using SPSS v.25 (IBM; Armonk, NY, USA), GraphPad Prism v.8.0.2 (San Diego, CA, USA), and R v.4.1.0 (http:// www.R-proje ct. org/) with the Matchit, survival, and cmprsk packages.

Patient characteristics
We screened 3663 consecutive patients with cirrhosis who underwent TIPS for secondary prophylaxis of variceal bleeding and were admitted to the six participating hospitals between January 2010 and June 2020; finally, 2100 patients were included in the study. All patients with non-standard PPG measurements, incomplete basic information, advanced liver cancer, and cardiopulmonary diseases affecting survival were excluded.  The median PPG decreased from 24 to 8.8 mmHg in the entire cohort. Difference of the portal pressure according to stent diameters was shown in Supplementary Table 18. Demographic and baseline characteristics are shown in Table 1.

Clinical outcomes
Rebleeding occurred in 409 patients during the follow-up period, resulting in a cumulative incidence of rebleeding of 19

Other PPG thresholds of 8, 10, and 14 mmHg in all patients
We also explored other PPG values (8, 10, and 14 mmHg) as potential thresholds. A post-TIPS PPG < 8 or 10 mmHg could not achieve less rebleeding without increasing the OHE (Supplementary 4-7). When stratified by 14 mmHg, a significant difference was observed between the two groups for rebleeding (39.12 vs. 51.22%, p = 0.006, HR = 1.489, 95% CI 1.120-1.979) before PSM, but not in OHE and mortality ( Fig. 2 and Supplementary Fig. 8-9). Decision curve analysis (DCA) showed that using 12 mmHg as a cut-off in terms of rebleeding could achieve more net benefit than using 14 mmHg (Fig. 5). The effect of PPG reduction on clinical outcomes in patients with different liver function categories

PPG < 12 mmHg in different liver function categories
As shown in Fig. 1, the outcomes of different Child-Pugh class patients were significantly different. Therefore, we further analyzed the PPG thresholds in different Child-Pugh classes.
In patients graded Child-Pugh A, setting 12 mmHg as the cutoff value for post-TIPS PPG was not discriminative of rebleeding, OHE, or mortality before or after PSM ( Supplementary Fig. 10). Competing risk analysis also showed consistent results ( Supplementary Fig. 11).

Other PPG thresholds of 8, 10, and 14 mmHg in different liver function categories
None of the tested PPG thresholds (8, 10, and 14 mmHg) showed significant differences in rebleeding, OHE, or mortality in patients graded as Child-Pugh class A (Supplementary Fig. 16-21).
In Child-Pugh class C, no significant difference was observed in rebleeding, OHE, or mortality between groups when stratified by 8 mmHg or 10 mmHg (Supplementary Fig. 28 . 4 and Supplementary  Fig. 32). Competing risk analysis showed similar results for these outcomes (Supplementary Fig. 33).
Therefore, as both 12 and 14 mmHg could reduce rebleeding in Child-Pugh class B and C patients, we compared their net benefits in rebleeding by DCA to determine the optimal threshold. DCA showed that 12 mmHg was marginally better than 14 mmHg in patients graded Child-Pugh class B, while 14 mmHg was superior to 12 mmHg in patients graded Child-Pugh class C (Fig. 5).

Discussion
This study confirms the prognostic value of PPG on patient outcomes, including rebleeding and OHE, the study revealed the necessity of tailoring PPG thresholds according to liver function, and identifies the optimal PPG thresholds in patients with different Child-Pugh grades. The advantages of the current study are as follows: (1) being conducted on the largest real-world TIPS cohort to date; (2) updating TIPS threshold in the era of covered stents with lower shunt dysfunction rate and possibly more stable post-procedure PPG variation; (3) applying multiple statistical methods including PSM and competing risk to validate the robustness of our findings; and (4) all the patients enrolled in this study underwent TIPS for secondary prophylaxis of rebleeding under local anesthesia, which avoids the influence of emergency vasoconstrictive drugs and sedation on PPG measurement. Therefore, immediate PPG can represent the real situation of PV pressure in this study.
A target threshold of 12 mmHg was proposed and implemented over 30 years ago [4,5,8,9,13,[21][22][23][24], and has since been recommended by Baveno VII [1], derived from uncoated TIPS. However, this has not yet been confirmed for covered TIPS. Therefore, it is important to explore the PPG threshold suitable for covered-stent TIPS. Moreover, the optimal PPG threshold should be tailored according to Child-Pugh classes because patients with different liver function classes have different PPG reduction tolerances.
Bosch et al. hypothesized that higher PPGs, such as 14 mmHg, might be more appropriate for covered stents than lower PPGs since their well-maintained patency would reduce the need to immediately lower post-TIPS PPG, which is required to counter gradually reduced shunt diameter and increased PPG with bare stents [13]. A recent cohort study of 216 patients confirmed this hypothesis, the study found that post-procedure PPGs higher than 12 mmHg did not necessarily indicate higher variceal rebleeding rates after 12 months [25]. Therefore, we explored this issue using a larger TIPS cohort.
In this study, patients with PPGs > 12 mmHg after TIPS had a higher incidence of rebleeding than those with PPGs < 12 mmHg. However, this result was not robust after PSM. PPGs > 14 mmHg also had a higher incidence of rebleeding than PPGs < 14 mmHg, but provided a net benefit higher than that of > 12 mmHg. However, this result was not robust after PSM. Nevertheless, when the entire cohort was stratified according to Child-Pugh class, an indicator of liver function, we found that 12 mmHg did not affect rebleeding or OHE rates in Child-Pugh class A patients. However, 12 mmHg affected rebleeding in Child-Pugh class B and C patients. These results indicate that a one-size-fits-all strategy may not be suitable for setting a PPG threshold, with liver function being important in tailoring PPG targets for different patients [26,27]. Therefore, we explored several PPG thresholds to propose an appropriate threshold for each Child-Pugh class.
Neither 10 nor 14 mmHg PPG affected rebleeding or OHE after TIPS in Child-Pugh class A patients, perhaps reflecting their relatively low cumulative incidence, increasing the difficulty in identifying significant differences. However, the results showed that patients had numerically lower rebleeding rates with PPGs < 10 mmHg than those with PPGs > 10 mmHg, which was not observed at PPG thresholds of 12 and 14 mmHg, indicating that 10 mmHg might be a more appropriate threshold than 12 and 14 mmHg. Attia et al. reported that patients with lower Child-Pugh scores could tolerate lower PPGs than those with higher Child-Pugh scores [27]. Therefore, we hypothesized that a PPG threshold of < 10 mmHg might be suitable for patients with Child-Pugh class A. Unfortunately, the small number of patients with PPG values in our cohort did not allow powered statistical tests.
A PPG threshold of 12 mmHg, but not 14 mmHg, affected rebleeding and OHE rates in Child-Pugh class B patients, indicating that a PPG threshold of 12 mmHg remains appropriate for them. In contrast, a PPG threshold of 14 mmHg appeared to be effective in reducing the rebleeding risk in Child-Pugh class C patients. These patients had a high incidence of OHE (> 50%), which we believe causes the lack of a significant OHE effect. Furthermore, a lower PPG threshold generally indicates more shunting, which might further reduce liver perfusion and exacerbate liver damage based on severely impaired function [6,[28][29][30]. Therefore, we hypothesized that a PPG of < 14 mmHg is better than a PPG of < 12 mmHg for Child-Pugh class C patients.
Although the percentage of pressure reduction is an important rebleeding indicator, several studies have explored its effective range [5][6][7][8][9]31]. In this study, post-TIPS PPGs remained high in patients with high pre-TIPS PPGs (1065 had pre-TIPS PPGs ≥ 24 mmHg), even when the reduction exceeded a certain percentage (e.g., a 50% reduction of 40 mmHg was 20 mmHg). Consequently, these patients were still exposed to adverse outcomes, and meaningful results could not be obtained.
Our study has several limitations. First, it was a retrospective study with incomplete information on post-TIPS liver function and the causes of rebleeding or death. However, we included consecutive patients to avoid excessive bias. Second, to date, the PPG threshold of < 12 mmHg has been the goal of TIPS. Therefore, only approximately 16% of patients in this study had post-TIPS PPG ≥ 12 mmHg, increasing the difficulty of identifying significant differences in PSM analyses when the sample size was further reduced.
In conclusion, a PPG threshold < 12 mmHg after TIPS may not be optimal for every patient. Different PPG thresholds may be required in patients with different liver function categories. Moreover, although a PPG threshold of < 12 mmHg may be suitable for Child-Pugh class B patients, a PPG threshold of < 14 mmHg is more suitable for Child-Pugh class C patients. Given the retrospective nature of this study, these results should be confirmed in future prospective studies with larger cohorts.
Author contributions All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by CZ, CT, YX, JT, GW, YZ, HW, HX, JX, FZ, LZ, GH, MZ, BW, PL, WW, CC. The first draft of the manuscript was written by CZ and YX, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding This work was supported by grants from the National Natural Science Foundation of China (81970533) and the National Natural Science Foundation of China (82000566).

Data availability All data collection was approved by Research Ethics
Committee of participating center. The data are not publicly available due to privacy and ethical restrictions. Ethical approval All procedures were performed in accordance with the Declaration of Helsinki.

Conflict of interests
Informed consent Written informed consent was obtained from all patients.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.