Should transjugular intrahepatic portosystemic shunt be first-line therapy for preventing esophageal variceal rebleeding? A meta-analysis of clinical outcomes

Abstract

Introduction

According to Baveno VII guidelines, endoscopic band ligation (EBL) plus nonselective beta-blockers (NSBB) is first-line therapy for preventing esophageal variceal rebleeding, with transjugular intrahepatic portosystemic shunt (TIPS) being used in cases of treatment failure. Our study aimed to compare TIPS with EBL plus NSBB in terms of overall mortality and rate of variceal rebleeding.

Methods

We conducted a comprehensive search on Medline, Google Scholar, Embase, and the Cochrane Library up to January 25, 2025. Randomized control trials (RCTs) comparing TIPS with EBL plus NSBB for variceal bleeding were included in the study.

Results

Eight RCTs with 629 patients, 336 receiving TIPS and 293 receiving EBL + NSBB, were included. TIPS had significantly lower rates of variceal rebleeding (odds ratio 0.30, 95% confidence interval 0.19–0.49, P < 0.00001), all-cause rebleeding (P < 0.0001), total episodes of rebleeding (P < 0.0001), new/worsening ascites (P < 0.0001), and peritonitis (P = 0.01). No significant differences were found in mortality, sepsis/pneumonia, hepatic encephalopathy, or hepatocellular carcinoma.

Conclusion

TIPS offers significant advantages over EBL plus NSBB in reducing variceal rebleeding and other complications but does not improve mortality. Further high-powered trials are needed to refine management strategies for variceal rebleeding in cirrhotic patients.

Esophageal varices are a common complication of decompensated liver disease and cirrhosis. Approximately 7% to 8% of liver cirrhosis patients with no history of varices develop varices every year, and the annual risk of variceal bleeding was found to be 5% to 15% according to the Global Burden of Liver Disease Study 2023.¹

Cirrhosis develops after inflammatory stress on the liver parenchyma, which causes fibrosis and loss of hepatic function. Disease evolution toward decompensation involves portal hypertension, which leads to development of dilated venous channels in the gastrointestinal tract, i.e., gastroesophageal varices, which is a sign of decompensated cirrhosis.² If allowed to persist, esophageal varices have the potential to enlarge and advance to variceal rupture and bleeding (at least 15% annually), a serious event associated with significant risk of mortality worldwide.¹

According to the European Society of Gastrointestinal Endoscopy and the American Association for the Study of Liver Disease, first-line management of esophageal varices is endoscopic band ligation (EBL) with or without pharmacological treatment, either a beta-blocker—particularly a nonselective beta blocker (NSBB)—or a vasoactive agent, whereas transjugular intrahepatic portosystemic shunt (TIPS) remains the modality of choice only in cases of refractory bleeds.^3,4 The American Association for the Study of Liver Disease stated that EBL plus NSBB remains first line for secondary prophylaxis of variceal rebleeding and TIPS is second line despite lower rebleeding rates for TIPS. This preference was due to TIPS previously being shown to relate to improvement in mortality and higher rates of hepatic encephalopathy (HE); thus, only patients who had other indications, like those with refractory ascites, were recommended for TIPS placement.⁵ Recent studies aimed to explore whether TIPS could replace the standard treatment, i.e., EBL plus NSBB.⁶ Despite critical advancements in the diagnosis and treatment modalities for esophageal varices, recurrent bleeding or rebleeding after surgical correction is common, and the associated morbidity and mortality remain unacceptably high.

Several randomized controlled trials (RCTs) and meta-analyses have reported the differences in efficacy for treatment of variceal bleeding in patients with cirrhotic portal hypertension separately. Awad et al compared TIPS with EBL and found TIPS to have lower rebleed rates but higher HE risk.⁷ Network meta-analyses have also been conducted comparing TIPS to other modalities.^8,9 However, most of these studies did not directly compare TIPS with EBL plus NSBB. Some compared TIPS with just endoscopic techniques, which include both EBL and sclerotherapy, while others did not include EBL with NSBB explicitly and instead compared various other combinations. To address this gap in the literature, we performed a meta-analysis on outcomes of TIPS vs EBL plus NSBB using data from RCTs, with the primary focus on overall mortality and the rate of variceal rebleeding as well as various side effects, to assess whether TIPS should be considered first-line therapy for rebleeding prophylaxis in all patients.

METHODS

This meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines¹⁰ and did not require ethical approval, as it is a review analysis study. A thorough search of PubMed/Medline, Google Scholar, Embase, and the Cochrane Library was conducted from the beginning of these databases until January 25, 2025, without language restrictions. Our primary outcome was the risk of rebleed after the primary intervention. Our secondary outcomes were mortality, rebleeding episodes, HE, sepsis/pneumonia, new/worsening ascites, peritonitis, and hepatocellular carcinoma.

For the topic of TIPS, the search terms used were TIPS, transjugular portosystemic shunt, TIPSS, and transjugular intrahepatic portosystemic shunt. For the topic of EBL, the search terms included EBL, ligation, variceal band ligation, esophageal band ligation, and band ligation. For beta-blockers, the search terms used were β-blocker, beta blocker, propranolol, nadolol, and carvedilol.⁴ The Boolean operator ‘AND’ was used to combine the main search items (Supplementary Table 1). This and all further tasks, such as data extraction and quality assessment, were done by two reviewers who independently searched the databases to identify all potentially eligible studies, and any disagreements were resolved by consensus or by consulting a third author. We also went through the studies included in similar other meta-analyses to ensure that any eligible articles missed in our primary search were identified.

All RCTs comparing TIPS with EBLT plus NSBB for variceal bleeding were included in the study. Studies that reported a primary outcome, i.e., variceal rebleeding, were included. Observational studies such as cross-sectional, case control, retrospective, and prospective cohort studies were excluded. Studies that explicitly did not compare TIPS with EBL plus NSBB—i.e., any study that compared TIPS with just EBL or NSBB alone or these two interventions combined with any other modality—were excluded. Studies that did not report the primary outcome of variceal rebleeding were excluded.

After searching all the selected databases, we imported the search results into an online research platform, Rayyan 2022.¹¹ Articles were first examined based on title and abstract, after which the selected articles were further filtered by going through full text. We evaluated quality of studies and risk of bias using the Cochrane Collaboration Risk of Bias Tool.¹²

The analysis was conducted using Review Manager Software Version 5.4.¹³ We calculated odds ratios (OR), mean differences, and their corresponding 95% confidence intervals (95% CI) to evaluate the effects of the interventions. A P value < 0.05 was considered statistically significant. The results were presented in forest plots, with dichotomous data expressed as ORs and continuous data as mean differences. The I² statistic was used to determine the extent of heterogeneity between studies. We interpreted the I² statistic in accordance with the criteria specified in the Cochrane Handbook for Systematic Reviews of Interventions, section 10.10.2 (version 6.5, 2024): I² of 0% to 40% might not be important; I² of 30% to 60% may represent moderate heterogeneity; I² of 50% to 90% may represent substantial heterogeneity; and I² of 75% to 100% represents considerable heterogeneity.¹⁴ Since the number of included studies was <10, we used Doi plots and computed LFK indices for each outcome to assess publication bias.¹⁵ We assessed the quality of evidence for each outcome by the GRADE (Grades of Recommendation, Assessment, Development, and Evaluation) approach¹⁶ and constructed a GRADE table for each outcome using the GRADEpro GDT software.¹⁷

Additionally, we performed analyses after creating subgroups based on 1) the severity of liver disease: studies with all ranges of disease severity, i.e., including patients with all Child Pugh scores A, B, and C, and studies with only moderate to severe liver disease, i.e., only patients with Child Pugh score B or C; 2) the type of NSBB: propranolol, propranolol and/or nadolol, and carvedilol; and 3) geographical regions: China, Europe, and multicenter. We also performed sensitivity analyses by excluding the studies with a high risk of bias, i.e., Dunne 2020, Lv 2018,^18,19 and the two oldest included studies, one by one, i.e., Sauer 2002, García-Pagán 2010.^20,21

RESULTS

Our preliminary search of PubMed/Medline, Google Scholar, Embase, and the Cochrane Library yielded 1093 studies. Eight articles^18–25 were deemed eligible according to our criteria. The PRISMA flow diagram for the screening process is shown in Figure 1.

Figure 1.

PRISMA flow diagram.

The final eight articles were all RCTs with a total of 629 participants. These articles were published from 2002 to 2023. Most were centered in China.^19,22,23,25 All eight RCTs compared TIPS with a control group, i.e., EBL plus NSBB. All utilized early TIPS (<72 hours). The most used NSBB was propranolol (n = 6).^19,21–25 The sample size varied from 49¹⁹ to 129²³ participants. The lowest mean follow-up time was 12 months for the experimental group¹⁸ and 10.6 months for the EBL plus NSBB group.²⁰ The longest follow-up times were a mean of 49 months for the experimental group²¹ and a median of 47 months for the EBL plus NSBB group.²² All studies except one²² documented the severity of liver disease and the causes of liver disease in their participants.

The study characteristics are listed in Table 1. The risk of bias assessment can be found in Supplementary Figure 1, and the publication bias results, i.e., Doi plots, can be found in Supplementary Figure 12. The GRADE assessment can be found in Supplementary Table 2.

Table 1.

Characteristics of the included studies

Author (year of Publication)	Country	NSBB used	Patients (TIPS/EBL plus NSBB)	Follow-up duration (months; mean ± SD or median [IQR])
Sauer et al (2002)	Germany	Propranolol	43/42	TIPS = 49.2 ± 3.12 EBL + NSBB = 43.8 ± 3
García-Pagán et al (2010)	Multicenter	Propranolol, nadolol	32/31	16 TIPS = 14.6 ± 8.4 EBL + NSBB = 10.6 ± 9.9
Luo et al (2015)	China	Propranolol	37/36	TIPS = 22.8 ± 7.7 EBL + NSBB = 20.9 ± 8.9
Holster et al (2016)	Netherlands	Propranolol	37/35	23.4 [5.9–30.7]
Lv et al (2017)	China	Propranolol	24/25	TIPS = 30.9 [21.6–42.5] EBL + NSBB = 30.4 [24.6–39.0]
Lv et al (2019)	China	Propranolol	84/45	TIPS = 24.0 [18.1–24.0] EBL + NSBB = 24.0 [9.0–24.0]
Dunne et al (2020)	UK	Carvedilol	29/29	12
Wang et al (2023)	China	Propranolol	50/50	TIPS = 41.6 [34.7–55.1] EBL + NSBB = 46.9 [34.2–59.1]

EBL indicates endoscopic band ligation; IQR, interquartile range; NSBB, nonselective beta-blocker; SD, standard deviation; TIPS, transjugular intrahepatic portosystemic shunt.

Variceal rebleeding

All eight studies reported variceal rebleeding. The total number of participants was 336 in the TIPS group and 293 in the EBL plus NSBB group. Our pooled effect estimate showed that significantly fewer patients experienced variceal rebleeding in the TIPS group as compared to EBL plus NSBB group (OR 0.30, 95% CI 0.19–0.49, P < 0.00001). We observed mild heterogeneity among the studies (I² = 11%, Figure 2). We found major asymmetry in the Doi plot (LFK index = −2.12), showing considerable publication bias.

Figure 2.

Forest plot for (a) variceal rebleeding and (b) mortality.

Subgroup analyses based on geographical locations, severity of liver disease, and type of NSBB (Figure 2) did not reduce the heterogeneity. Sensitivity analysis by excluding the oldest study, i.e., Sauer 2002, reduced the heterogeneity to 0%. Additionally, excluding García-Pagán 2010 reduced the heterogeneity to 0% as well. Subgroup and sensitivity analysis are shown in Supplementary Figure 2.

We assessed the certainty of evidence to be low. The outcome was graded down twice for imprecision and publication bias.

Mortality

All eight studies reported mortality, with 336 participants in the TIPS group and 293 in the EBL plus NSBB group. Our analysis revealed that TIPS was not significantly associated with reduced mortality as compared to EBL plus NSBB (OR 0.67, 95% CI 0.42–1.07, P = 0.10). We observed mild heterogeneity among the studies (I² = 30%, Figure 2). We did not detect publication bias, as the Doi plot showed no asymmetry (LFK index = 0.22).

Subgroup analysis based on severity of liver disease revealed significant results favoring TIPS for the moderate to severe disease subgroup (P = 0.02), and heterogeneity was reduced to 0%. Subgroup analyses based on geographical locations and type of NSBB did not affect the heterogeneity. Sensitivity analysis by excluding Lv 2018, i.e., a high risk of bias study, showed results significantly favoring TIPS (P = 0.008), and the heterogeneity was reduced to 0%.

We assessed the certainty of evidence to be moderate. The evidence was graded down once for imprecision. Subgroup and sensitivity analysis are shown in Supplementary Figure 3.

Rebleeding episodes

Four out of eight studies included in our analysis reported total episodes of rebleeding per participant, with 146 participants in the TIPS group and 142 in the EBL plus NSBB group. Our analysis showed a significant improvement in rebleeding episodes in the TIPS group as compared to EBL plus NSBB (OR 0.14, 95% CI 0.06–0.36, P < 0.0001). We found moderate heterogeneity among the studies (I² = 46%, Figure 3). We also observed substantial publication bias, with major asymmetry in the Doi plot (LFK index = −2.78). None of the subgroup analyses or sensitivity analyses were able to explain the heterogeneity.

Figure 3.

Forest plot for (a) rebleeding episodes and (b) all-cause rebleeding.

The certainty of evidence was assessed to be very low due to downgrading three times for inconsistency, imprecision, and publication bias. Subgroup and sensitivity analysis are shown in Supplementary Figure 4.

All-cause rebleeding

Seven out of the eight included studies reported all-cause bleeding, with 299 participants in the TIPS group and 257 in the EBL plus NSBB group. Our pooled analysis revealed that TIPS is associated with significantly lower all-cause rebleeding compared with EBL plus NSBB (OR 0.29, 95% CI 0.15–0.54, P < 0.0001). We observed moderate heterogeneity among the studies (I² = 44%, Figure 3). We found substantial publication bias, as the Doi plot showed major asymmetry (LFK index = −3.16). Subgroup analyses and sensitivity analyses did not affect the heterogeneity.

We assessed the certainty of evidence to be low. The outcome was graded down twice for imprecision and publication bias. Subgroup and sensitivity analysis are shown in Supplementary Figure 5.

Hepatic encephalopathy

All eight studies reported data on HE, with 115 patients in the TIPS group and 75 patients in the EBL plus NSBB group. Our analysis showed no significant difference in the occurrence of HE between the two groups (OR 1.52, 95% CI 0.95–2.43, P = 0.08). A moderate level of heterogeneity was observed (I² = 40%, Figure 4). We found no publication bias, as no asymmetry was observed in the Doi plot (LFK index = −0.18).

Figure 4.

Forest plot for (a) hepatic encephalopathy and (b) ascites.

Subgroup analyses based on geographical location and severity of liver disease were able to explain the heterogeneity, since it was reduced to zero in all subgroups. The outcome was significant for the Europe subgroup (P = 0.001). Sensitivity analysis by removing the second oldest study, i.e., García-Pagán 2010, revealed a significant result (P = 0.02). Subgroup and sensitivity analysis are shown in Supplementary Figure 6.

The certainty of the evidence was assessed to be moderate, with downgrading once in the domain of imprecision.

Ascites

Six studies reported data on new or worsening ascites among 256 participants in the TIPS group and 215 in the EBL plus NSBB group. Our pooled analysis showed a significant decrease in the incidence and worsening of ascites in the TIPS group (OR 0.33, 95% CI 0.20–0.56, P < 0.0001). We found no heterogeneity among the studies (I² = 0%, Figure 4).

We found substantial publication bias, since major asymmetry was found in the Doi plot (LFK index = −2.27). Subgroup and sensitivity analysis are shown in Supplementary Figure 7.

The certainty of evidence was assessed to be low after grading down twice due to imprecision and publication bias.

Peritonitis

Six studies included in our meta-analysis provided data on peritonitis. There were 256 participants in the TIPS group and 215 participants in the EBL plus NSBB group. The results showed a significant decrease in the incidence of peritonitis in patients in the TIPS groups versus those in the EBL plus NSBB group (OR 0.23, 95% CI 0.07–0.70, P = 0.01). No heterogeneity was observed in the studies (I² = 0%, Figure 5). We found some publication bias, since minor asymmetry was found in the Doi plot (LFK index = 1.09). Subgroup and sensitivity analysis are shown in Supplementary Figure 8.

Figure 5.

Forest plot for (a) peritonitis, (b) sepsis, (c) liver transplant, and (d) hepatocellular carcinoma.

The certainty of the evidence was assessed to be low. The outcome was downgraded once due to imprecision and once due to publication bias.

Sepsis/pneumonia

Seven studies reported data on sepsis and/or pneumonia. Among 568 patients, 11 in the TIPS group and 10 in the EBL plus NSBB group developed sepsis/pneumonia. No significant difference was observed in the incidence of sepsis/pneumonia between the two groups (OR 0.90, 95% CI 0.37–2.16, P = 0.81). No heterogeneity was observed (I² = 0%, Figure 5). We found no publication bias, as the Doi plot showed no asymmetry (LFK index = 0.51). Subgroup and sensitivity analysis are shown in Supplementary Figure 9.

The certainty of evidence was assessed to be moderate, with downgrading occurring once for imprecision only.

Liver transplant

Six studies provided data on liver transplantation. Among 471 patients, 19 underwent liver transplantation, with 10 patients in the TIPS group and 9 patients in the EBL plus NSBB group. Our analysis showed no significant difference in the incidence of liver transplant between the two groups (OR 1.02, 95% CI 0.41–2.57, P = 0.96). We found no heterogeneity among the studies (I² = 0%, Figure 5). We found substantial publication bias, as major asymmetry was observed in the Doi plot (LFK index = −3.70). Subgroup and sensitivity analysis are shown in Supplementary Figure 10.

The certainty of the evidence was assessed to be low because of downgrading once in the domains of imprecision and publication bias.

Hepatocellular carcinoma

Five out of eight studies reported data on the incidence of hepatocellular carcinoma among 191 participants in the TIPS group and 188 in the EBL plus NSBB group. We found no significant difference in the incidence of hepatocellular carcinoma between the two groups (OR 0.53, 95% CI 0.24–1.19, P = 0.12). We observed no heterogeneity among the studies (I² = 0%, Figure 5). We found substantial publication bias revealed by major asymmetry in the Doi plot (LFK index = 3.03).

We assessed the certainty of evidence to be low. The outcome was graded down twice for imprecision and publication bias. Subgroup and sensitivity analysis are shown in Supplementary Figure 11.

DISCUSSION

In this meta-analysis, we assessed the efficacy of TIPS compared to EBL plus NSBB in preventing rebleeding from esophageal varices. Our analysis comprised 8 RCTs involving 629 patients (336 in the TIPS group and 293 in the EBL plus NSBB group). We observed a significantly lower rate of variceal rebleeding, all-cause rebleeding, peritonitis, or new/worsening ascites among patients receiving TIPS. Additionally, the overall mortality rate was lower in the TIPS group. However, there were no significant differences between the interventions concerning HE, sepsis/pneumonia, or chances of liver transplant.

TIPS has been used for managing variceal bleeding since 1982²⁶ and is currently recommended as a second-line treatment for preventing rebleeding and when the primary first-line therapy, i.e., EBL plus NSBB, fails.²⁷ Our analysis corroborates recent studies showcasing the superiority of TIPS in decreasing variceal rebleeding compared to EBL plus NSBB therapy.^6,28–31 Most of these analyses involved both EBL and sclerotherapy or combinations of both without beta-blockers, while others drew conclusions from indirect comparisons rather than a direct comparison—making our study one of the few that directly compared TIPS and EBL plus NSBB specifically.

Additionally, our subgroup analysis suggests potential similar effectiveness between TIPS and EBL combined with carvedilol in reducing rebleeding. Although carvedilol has replaced propranolol as the primary NSBB of choice in variceal bleeding, we noted limited available data for this combination.^27,32 Our findings highlight the need for further robust RCTs to evaluate the efficacy of carvedilol and EBL compared to TIPS. We also found TIPS to be effective in patients with Child-Pugh stage A liver failure, even though it is only recommended in high-risk patients (patients in Child-Pugh stage B with active bleeding on endoscopy, patients with Child-Pugh stage C). This finding suggests a possible extension of TIPS benefits beyond high-risk patients, warranting in-depth investigation into this patient subset.

Portal hypertensive bleeding, especially variceal rebleeding, is linked to startlingly higher mortality, especially within the first 6 weeks after the index bleed.³³ While our initial analysis did not unveil a distinct difference in all-cause mortality in TIPS and EBL plus NSBB, sensitivity analysis excluding studies with a high risk of bias demonstrated significantly lower overall mortality within the TIPS group. We removed the study of Lv 2018, which may have higher bias, as it included patients with portal vein thrombosis as well as esophageal variceal bleeding due to liver cirrhosis. A recent meta-analysis analyzing the efficacy of TIPS vs EBL plus propranolol exclusively in patients with portal vein thrombosis, thereby including the studies of Lv 2018 and Luo 2015 that were also analyzed by our study, also showed nonsignificant differences in mortality.³⁴ This might mean that patients with no portal vein thrombosis may show lower mortality. The effect of TIPS on mortality in variceal bleeding remains controversial, with some studies echoing our findings^31,35,36 and others having conflicting results.^28,30,37 This highlights the need for further research.

Our findings indicate that TIPS placement contributes to reduced rates of ascites and spontaneous bacterial peritonitis (SBP) compared to EBL plus NSBB. The efficacy of TIPS compared to other treatment modalities in reducing ascites and preventing SBP has also been reported by similar studies, highlighting its importance as a secondary treatment for refractory ascites.^29,31,38 There may be a correlation between the lower incidence of ascites and peritonitis due to ascites potentially disposing individuals to developing SBP.³⁹ Prevention of these events is imperative because they lead to severe complications such as hepatorenal syndrome–induced acute kidney injury.^40,41

HE has also remained a potential concern in TIPS placement and is a contributor to the gross underuse of TIPS in real-world situations.^42,43 TIPS diverts blood away from the liver, so toxins normally metabolized by the liver accumulate in the systemic circulation and damage the cerebral cortex.⁴⁴ Our analysis unveiled no significant association between TIPS and the development of HE, although a tendency was observed, aligning with one of our included studies, Sauer 2002.²¹ Later insertion and larger stent size—which are known risk factors of post-TIPS HE—may be why.⁴⁵ Other studies utilizing early small-diameter TIPS placement (<72 hours using 8 mm stent) all showed comparatively lower rates of HE development. Thus, fear of HE should not necessarily impede the consideration of TIPS placement.^6,30,37,42 However, further investigations are necessary to better understand the variables influencing HE occurrences post-TIPS placement.

Our results showed no improvement in infectious outcomes such as sepsis or pneumonia. However, a recent meta-analysis reported a lower incidence of sepsis in anticipant TIPS compared with conventional TIPS or standard therapy.²⁹ Another study reported lower readmission rates due to sepsis after diameter-controlled expansion TIPS, suggesting a role of size control in postprocedural complications, which warrants further exploration.⁴⁶

We found that TIPS did not reduce the chances of liver transplant as compared to standard therapy. Recent studies have reported a conflicting incidence of the need for liver transplantation after TIPS, ranging from minimal^47–49 to frequent.⁵⁰ Currently, TIPS is considered beneficial for improving outcomes in cirrhosis until transplant.⁵¹

Our results showed that TIPS did not reduce the occurrence of hepatocellular carcinoma. Currently, there is scanty evidence in favor of TIPS. However, recent studies have reported that TIPS is not associated with a higher risk or worse prognosis of hepatocellular carcinoma.^52,53

The main strength of our study is its focus on TIPS and the current primary therapy, i.e., EBL plus NSBB, with regards to important clinical outcomes. To our knowledge, no previous meta-analysis compared these two interventions exclusively, as stated above. To maintain the quality of evidence, we included only RCTs in the analysis. However, the small number of studies used to pool the data might diminish the generalizability of these outcomes. Furthermore, half of our included studies were from China and had variation in follow-up time, which may also decrease the generalizability of the findings.

Overall, our results have important clinical implications. While alterations in current standards of therapy would require additional undisputed evidence, our analysis helps to establish several key facts. First, TIPS is associated with significantly lower rebleeding events. Second, considering recent RCTs, TIPS might correspond with lower mortality. Third, various postprocedural complications are lower in the TIPS group. Lastly, our results disprove harmful effects of TIPS in terms of hepatocellular carcinoma, which corroborates similar newer evidence and has worthwhile implications in the treatment of patients with cirrhosis due to hepatocellular carcinoma.

Our meta-analysis highlights the need for further well‐designed RCTs with larger sample sizes. Further spotlight on important outcomes is required to establish conclusive results and dispel the existing conflicts in evidence, as shown in this review. Specific focus is needed on clinical outcomes—as pointed out beforehand—which might have pivotal implications in improving current standards of therapy.

In conclusion, the use of TIPS for treatment of esophageal variceal bleeding is beneficial for ameliorating rebleeding risk and might have some mortality benefits. Many complications were found to be lower after TIPS as compared to the standard of care, i.e., EBL plus NSBB. However, current evidence is insufficient to establish the decisive superiority of one procedure over the other. Further research is needed to obtain higher-quality evidence in this domain to maximize patient outcomes and quality of life.

Disclosure statement/Funding

No potential conflict of interest or funding was reported by the author(s).