Alterations in the Gut Microbiota of Cirrhotic Patients With Sarcopenia and PH After TIPS

Abstract

INTRODUCTION:

Transjugular intrahepatic portosystemic shunt (TIPS) can effectively reduce portal hypertension (PH) and may be associated with improvement in muscle status; however, its impact on the gut microbiota and muscle condition remains unclear. Therefore, this prospective study aimed to evaluate gut microbiota alterations in patients with sarcopenia and PH after TIPS treatment.

METHODS:

Gut microbiota analysis was prospectively conducted in 20 cirrhotic patients with sarcopenia and PH receiving TIPS and their first-degree healthy family between December 2019 and July 2021 in our center. Fecal samples were collected before and 3 months after TIPS and analyzed by 16S rRNA sequencing. Microbial diversity and composition were assessed using standard bioinformatic pipelines. Sarcopenia was defined by L3 skeletal muscle index according to European Association for the Study of the Liver criteria.

RESULTS:

Compared with healthy controls, cirrhotic patients with sarcopenia and PH exhibited significantly reduced α-diversity (Chao1, Shannon, Simpson; all P < 0.001) and distinct β-diversity clustering, confirming gut dysbiosis. Although TIPS did not significantly alter overall α-diversity, a significant increase in the order Burkholderiales and a reduction in Pseudomonadales and Staphylococcales were detected after TIPS. Analysis of the family taxonomic rank revealed that TIPS was followed by an increase in a greater abundance of Ruminococcaceae and Sutterellaceae and decreased levels of Tannerellaceae, Marinifilaceae, and Pseudomonadaceae. At the genus level, after TIPS placement, increased levels of Lachnospiraceae_ND3007_group, Intestinibacter, Fusicatenibacter, and Faecalibacterium and lower abundances of Pseudomonas, Prevotella, Peptostreptococcus, Parabacteroides, Muribaculaceae, and Butyricimonas were detected. Sarcopenia reversal was observed in 25% of patients, characterized by enrichment of short-chain fatty acid-producing genera such as Faecalibacterium and Streptococcus, whereas patients without improvement showed high abundance of Fusicatenibacter and low levels of beneficial taxa.

DISCUSSION:

TIPS induces significant taxonomic shifts without changing overall diversity in cirrhotic patients with sarcopenia and PH, suggesting partial microbial restoration. Enrichment of short-chain fatty acid-producing bacteria may correlates with improved sarcopenia, whereas proinflammatory taxa may indicate poor recovery and higher risk of complications. These findings highlight the gut-muscle-liver axis and support microbiome-based biomarkers and interventions to optimize outcomes after TIPS.

INTRODUCTION

Portal hypertension (PH) is a major complication of liver cirrhosis and represents a critical determinant of morbidity and mortality in this population (1,2). Transjugular intrahepatic portosystemic shunt (TIPS) is an established intervention for reducing portal pressure and preventing variceal bleeding or refractory ascites, thereby improving survival in selected patients (3–7). Beyond hemodynamic benefits, emerging evidence suggests that TIPS may exert broader effects on systemic homeostasis, including nutritional and metabolic status. Sarcopenia, defined as the progressive loss of skeletal muscle mass and function, is highly prevalent among cirrhotic patients with PH and has been recognized as an independent predictor of poor outcomes (8), including increased risk of hepatic encephalopathy (HE), reduced quality of life, and impaired posttransplant survival (9,10). However, the mechanisms underlying sarcopenia in cirrhosis and the potential impact of TIPS on muscle recovery remain incompletely understood.

In recent years, the gut microbiota has been identified as a key modulator of the gut-liver-muscle axis (11). Cirrhosis-related gut dysbiosis is characterized by reduced microbial diversity, overgrowth of pathogenic taxa, impaired intestinal barrier function, and altered microbial metabolite profiles, all of which contribute to systemic inflammation, hyperammonemia, and sarcopenia progression (8,12). Beneficial commensals such as Faecalibacterium and Ruminococcaceae are known to produce short-chain fatty acids (SCFAs), which have been reported to exert anti-inflammatory effects, preserve gut barrier integrity, and are believed to promote muscle anabolism (13). Conversely, expansion of pathobionts including Pseudomonas and Peptostreptococcus has been associated with intestinal inflammation and muscle catabolism (12). Although previous studies have reported microbial shifts after TIPS placement, the relationship between these alterations, sarcopenia dynamics, and clinical outcomes has not been systematically evaluated.

To date, little is known about how TIPS influences gut microbial composition in cirrhotic patients with sarcopenia and PH whether these microbiota changes correlate with muscle recovery. Most existing research has focused on TIPS-related hemodynamic or neurological outcomes, whereas the interplay between microbiota and sarcopenia in this setting remains underexplored. Furthermore, although studies have highlighted the potential contribution of SCFA-producing taxa to muscle health, their specific role in sarcopenia reversal after TIPS has not been elucidated.

Therefore, in this prospective study, we analyzed gut microbiota composition in cirrhotic patients with sarcopenia and PH before and after TIPS using 16S rRNA sequencing, with first-degree relatives as healthy controls (HCs). We aimed to characterize microbial diversity and taxonomic alterations associated with TIPS, and to investigate their relationship with sarcopenia reversal. Our findings provide novel insights into the gut-muscle-liver axis, highlighting potential microbial biomarkers and therapeutic targets for optimizing outcomes after TIPS.

MATERIALS AND METHODS

Study design and patients

This prospective observational study received ethical approval from the Institutional Review Board of the First Affiliated Hospital, Zhejiang University School of Medicine. All participants provided written informed consent before enrollment, in accordance with the ethical guidelines outlined in the Declaration of Helsinki and relevant institutional policies. We prospectively enrolled cirrhotic patients (with any etiology) with sarcopenia and PH requiring secondary prophylaxis for variceal bleeding or refractory ascites at our center between December 2019 and July 2021. The inclusion criteria were as follows: diagnosis of cirrhosis confirmed by clinical history. The cutoffs for sarcopenia were determined on the basis of the European Association for the Study of the Liver clinical practice guidelines -L3 skeletal muscle index (SMI), 50 cm²/m² for men and,39 cm²/m² for women (10,14). histopathology, or imaging; TIPS indicated for prevention of variceal rebleeding and/or refractory ascites; age 18–80 years; and no systemic or nonabsorbable antibiotic exposure within 1 week before or after stool sample collection. The exclusion criteria included recurrent overt hepatic encephalopathy (OHE); comorbid hepatocellular carcinoma or extrahepatic malignancies; congestive heart failure (New York Heart Association Class III–IV) or severe valvular insufficiency; inaccessible fecal samples; history of pulmonary hypertension; hepatic failure (defined as thrombospondin index <35%, serum bilirubin >100 μmol/L, or Child‒Pugh score >12); serum creatinine >250 μmol/L; uncontrolled sepsis; albumin allergy; pregnancy or lactation; contrast agent hypersensitivity; investigator-determined conditions potentially confounding study outcomes; and patient refusal.

Fecal samples were collected from patients 1–3 days before TIPS and 3 months after TIPS, with additional samples obtained from first-degree relatives without cirrhosis, all first-degree relatives were healthy and lived in the same household as the patients, ensuring similar environmental exposures. All the samples were cryopreserved at −80°C within 2 hours of collection. Demographic, clinical, and laboratory parameters were systematically recorded at each sampling timepoint.

TIPS procedure and follow-up

The TIPS procedures were performed by a consistent team of interventional radiologists with over a decade of extensive surgical experience. Before TIPS placement, all patients underwent standardized preoperative evaluations by computed tomography angiography and magnetic resonance angiography to delineate the spatial relationship between the anatomical origin (hepatic veins) and target termination (portal veins). after successful puncture, a combination of bare-metal stents (Boston Scientific Corporation, Marlborough, MA) and self-expanding polytetrafluoroethylene-covered stent grafts (GORE VIABAHN, W. L. Gore & Associates, Newark, DE) was deployed, utilizing 8 mm diameter stents. The portosystemic pressure gradient (PPG) was quantified by measuring the differential pressure between the portal venous system and right atrial pressure. Balloon angioplasty was performed if postprocedural pressure monitoring demonstrated inadequate PPG reduction below the target threshold of 12 mm Hg. Post-TIPS embolization was based on post-TIPS portography using metallic coils (Cook Medical, Bloomington, IN), n-butyl-2-cyanoacrylate glue (Guangzhou Baiyun Medical Adhesive, Guangdong, China), or a combination therapy of metallic coils with adhesive agents.

Follow-up was conducted when patients were presented and scheduled for 3 months after TIPS, clinical, laboratory, and hepatic computed tomography angiography assessments were performed at presentation, patient stool samples were collected, and the occurrence of any liver-related complications that had occurred since presentation was collected by the clinical research coordinator.

DNA extraction and 16S rRNA gene sequencing

Total genomic DNA was extracted from the fecal samples via the MagPure Soil DNA LQ Kit (Magen Biotech, Guangzhou, Guangdong, China) according to the manufacturer's protocol. DNA concentration and purity were assessed via a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA), with integrity verified via 1% agarose gel electrophoresis. The genomic DNA served as a template for polymerase chain reaction amplification with barcoded primers and Tks Gflex DNA Polymerase (Takara Bio, Kusatsu, Shiga, Japan).

For bacterial diversity analysis, the V3–V4 hypervariable regions of the 16S rRNA gene were amplified by the universal primers 343F (5′-TACGGRAGGCAGCAG-3′) and 798R (5′-AGGGTATCTAATCCT-3′). Amplicons were visualized on a 2% agarose gel, purified with AMPure XP beads (Beckman Coulter, Brea, CA), and subjected to a second round of polymerase chain reaction amplification. After purification with AMPure XP beads, amplicon concentrations were quantified using a Qubit dsDNA HS Assay Kit (Thermo Fisher Scientific). Equimolar pools of amplicons were prepared for paired-end sequencing (2 × 250 bp) on an Illumina NovaSeq 6000 platform (Illumina, San Diego, CA).

Bioinformatic analysis

Raw sequencing data (FASTQ format) were processed by OE Biotech Co., Ltd. (Shanghai, China). Adapter sequences were trimmed by Cutadapt v3.4, followed by quality filtering, denoising, read merging, and chimera removal via DADA2 v1.24.0 within QIIME2 v2020.11 (15–17). Amplicon sequence variants (ASVs) were generated and taxonomically classified against the SILVA database v138 (18) via the q2-feature-classifier plugin with default parameters.

Alpha diversity metrics (Chao1 richness, Shannon diversity) and beta diversity based on unweighted UniFrac distances were calculated in QIIME2. Principal coordinate analysis (PCoA) was performed to visualize beta diversity patterns. Intergroup differences were assessed via permutational analysis of variance (PERMANOVA, 999 permutations), with post hoc pairwise comparisons by Wilcoxon rank-sum or Kruskal-Wallis tests. Linear discriminant analysis effect size (LEfSe) was used to identify differentially abundant taxa (linear discriminant analysis score >3.0) (19).

Statistical analysis

All the statistical analyses and graphical visualizations were conducted via SPSS version 26 (IBM Corp, Armonk, NY) and R version 4.2.2 (R Foundation for Statistical Computing, Austria). A 2-tailed P value < 0.05 was predefined as the threshold for statistical significance. Continuous variables are reported as medians with interquartile ranges and were analyzed by nonparametric tests: the Wilcoxon rank-sum test for independent groups, the Wilcoxon signed-rank test for paired comparisons, or the Kruskal-Wallis test for multigroup analyses, as appropriate. Categorical variables are presented as counts and were compared by the Pearson χ² (χ²) tests.

Beta diversity was assessed through PCoA based on unweighted UniFrac distances, complemented by analysis of similarities with 999 permutations to evaluate intergroup versus intragroup dissimilarities.

To address data heterogeneity in sequencing analyses, linear discriminant analysis effect size (LEfSe) was used to identify differentially abundant microbial taxa across pre-TIPS (n = 20), post-TIPS (n = 20), and HCs (n = 20) groups, applying a stringent linear discriminant analysis score cutoff of >3.0.

Ethical issues

This prospective observational study was reviewed and approved by the Ethics Committee of the First Affiliated Hospital, Zhejiang University School of Medicine. All patients provided written informed consent to participate in the study. In addition, explicit consent was obtained from each participant for the publication of potentially identifiable images or data contained in this article.

RESULTS

Patient characteristics

A total of 106 patients underwent TIPS creation at our center between December 2019 and July 2021, with 86 patients excluded on the basis of predefined criteria, leaving a total of 20 patients for inclusion in the study (Figure 1). Table 1 summarizes the baseline characteristics of the study cohort (n = 20) and their first-degree healthy controls (n = 20). The cohort comprised predominantly male patients (55%), with a mean age of 59.25 ± 11.31 years. Hepatitis B virus infection accounts for 60% of cirrhosis etiologies, no patient had experienced hepatic encephalopathy before TIPS and lactulose/rifaximin was not routinely administered postprocedure. Five (25%) patients showed reversal of sarcopenia after TIPS. Reversal of sarcopenia was defined as a transition from a sarcopenia to a nonsarcopenia state. A significant SMI increase, defined as an increase of more than 10% from baseline, was observed in 5 patients (25%), whereas the remaining 10 patients (50%) did not significant SMI improvement. Four patients (20%) had undergone splenectomy before TIPS placement. The mean Child-Pugh score was 7.05 ± 1.57, with the majority classified as Child-Pugh class B (65%). The model for end-stage liver disease score averaged 17.38 ± 3.94. The primary indication for TIPS was secondary prophylaxis for variceal hemorrhage (85%). Preprocedural laboratory parameters included aspartate aminotransferase, albumin, the international normalized ratio, the white blood cell count, the platelet count, hemoglobin (Hb), creatinine, total bilirubin, direct bilirubin, the prothrombin time, glucose, and serum ammonia levels.

Figure 1.

Flowchart of patient enrollment for gut microbiota analysis in cirrhosis patients undergoing TIPS. HCC, hepatocellular carcinoma; TIPS, transjugular intrahepatic portosystemic shunt.

Table 1.

Baseline demographic and clinical characteristics of patients

	Pre-TIPS	HCs
Male	11 (55%)	5 (25%)
Age (mean ± SD) (yr)	59.25 ± 11.31	50.12 ± 15.56
Height	1.62 ± 0.07	1.61 ± 0.08
Weight	56.67 ± 7.05	62.88 ± 8.71
BMI	21.66 ± 2.16	24.15 ± 2.6
Etiology of cirrhosis
Hepatitis B virus	12 (60%)
Hepatitis C virus	1 (5%)
Alcohol	6 (30%)
PBC/PSC	1 (5%)
Sarcopenia	20 (100%)
Episodes of HE	0 (0%)
Previous splenectomy	4 (20%)
Laboratory parameters (mean ± SD)
AST (IU/L)	40.15 ± 57.15
ALB (g/dL)	33.65 ± 5.08
INR	1.28 ± 0.19
WBC	3.84 ± 2.64
PLT	85.85 ± 67.89
HB	76.85 ± 15.91
Creatinine (mg/dL)	77.25 ± 21.27
TB (mg/dL)	20.40 ± 21.19
DB (mg/dL)	10.92 ± 15.36
PT	14.75 ± 2.16
Glucose	5.72 ± 1.88
Serum ammonia	37.78 ± 15.83
Child-Pugh score	7.05 ± 1.57
Child-Pugh class
A	7 (35%)
B	13 (65%)
C	0 (0%)
MELD	12.38 ± 3.94
Indication for TIPS
Refractory ascites	2 (10%)
Variceal bleeding	17 (85%)
Refractory ascites + variceal bleeding	1 (5%)
PPG before TIPS (mm Hg)	21.65 ± 5.07
PPG after TIPS (mm Hg)	7.85 ± 2.85

ALB, albumin; ALT, albumina; AST, aspartate aminotransferase; BMI, body mass index; DB, indirect bilirubin; HB, hemoglobin; HCs, healthy controls; HE, hepatic encephalopathy; INR, international normalized ratio; MELD, model for end-stage liver disease; PBC, primary biliary cirrhosis; PLT, platelet; PPG, portosystemic gradient; PSC, primary sclerosing cholangitis; PT, prothrombin time; TB, total bilirubin; TIPS, transjugular intrahepatic portosystemic shunt; WBC, white blood cell.

Procedure and clinical outcomes

Technical success, defined as successful shunt creation with PPG reduction to <12 mm Hg, was achieved in all patients (100%). A hybrid stent configuration consisting of an 8 mm bare metal stent overlapped with an 8 mm expanded polytetrafluoroethylene-covered stent (GORE VIABAHN) was deployed in all patients. Concurrent variceal embolization was performed in 15 patients (75%), predominantly by a combination of n-butyl-2-cyanoacrylate glue and detachable coils (60% of embolized patients).

No major procedure-related complications (Clavien-Dindo grade ≥ III) occurred during the perioperative period. Over a mean follow-up duration of 19.5 ± 12.8 months, 2 patients (10%) died, 1 from progressive hepatic failure and 1 from sepsis-induced multiorgan dysfunction. The cumulative incidence rates of recurrent variceal hemorrhage and refractory ascites were 10% and 5%, respectively. Post-TIPS OHE developed in 15% of the cohort. De novo hepatocellular carcinoma was diagnosed in 2 patients during surveillance. Hemodynamically, the PPG decreased significantly from 21.65 ± 5.07 mm Hg pre-TIPS to 7.85 ± 2.85 mm Hg postprocedure. Shunt dysfunction, which was diagnosed by Doppler ultrasound and PPG reelevation, occurred in 2 patients (10%) (Table 1).

The gut microbiota composition of patients compared with that of HCs

To assess differences in the gut microbial community between pre-TIPS patients and HCs, we first performed α diversity analysis. Significant disparities were observed across α diversity indices, including Chao1 richness (P = 4.1 × 10⁻⁷), Shannon diversity (P = 9 × 10⁻⁸), and Simpson evenness (P = 0.00019) (Figure 2a). β-Diversity analysis by unweighted UniFrac distances further confirmed distinct microbial clustering between groups (analysis of similarities, P = 0.001), as visualized by PCoA (Figure 2b). These findings collectively demonstrate marked gut dysbiosis in cirrhotic patients compared with HCs.

Figure 2.

(a) Alpha diversity indices in cirrhotic patients pre-TIPS vs HCs. Boxplots display alpha diversity indices (Chao1, Shannon, Simpson, PD_whole_tree, goods_coverage) of gut microbiota in pre-TIPS cirrhotic patients (n = 20) and paired HCs (n = 20). Significant differences were observed: Chao1 (P = 4.1 × 10⁻⁷), Shannon (P = 9 × 10⁻⁸), Simpson (P = 0.00019), PD_whole_tree (P < 0.001). Differences were assessed using the Wilcoxon signed-rank test; P < 0.05 indicates significance, with *** denoting P < 0.001. (b) Beta diversity analysis of gut microbiota in cirrhotic patients pre-TIPS vs HCs principal coordinates analysis plot based on unweighted UniFrac distances illustrates the beta diversity of gut microbiota in pre-TIPS cirrhotic patients (n = 20, green) and HCs (n = 20, purple). PC1 and PC2 explain 8.6% and 4.77% of the variance, respectively. A significant difference between groups was observed (PERMANOVA, P = 0.001), indicating distinct microbial clustering. HCs, healthy controls; PC, principal component; PD, phylogenetic diversity; TIPS, transjugular intrahepatic portosystemic shunt.

Diversity of the gut microbiota after TIPS placement

Longitudinal analysis of α diversity metrics (Chao1, Shannon, Simpson, phylogenetic diversity [PD_whole_tree], and Goods_coverage) revealed no significant differences between pre-TIPS and post-TIPS samples (Figure 3). Paired analysis of microbial abundance across taxonomic ranks revealed significant compositional changes post-TIPS. At the order level (Figure 4a), we observed a marked increase in Burkholderiales abundance alongside reductions in Pseudomonadales and Staphylococcales abundance (all P < 0.05, FDR-corrected). Analysis of the family taxonomic rank (Figure 4b) revealed that TIPS was followed by an increase in a greater abundance of Ruminococcaceae and Sutterellaceae and decreased levels of Tannerellaceae, Marinifilaceae and Pseudomonadaceae. At the genus level, after TIPS placement, increased levels of Lachnospiraceae_ND3007_group, Intestinibacter, Fusicatenibacter, and Faecalibacterium and lower abundances of Pseudomonas, Prevotella, Peptostreptococcus, Parabacteroides, Muribaculaceae, and Butyricimonas were detected (Figure 4c).

Figure 3.

Changes in alpha diversity indices of gut microbiota pre- and post-TIPS. Line plots depict individual patient trajectories for 5 alpha diversity metrics: (a) Chao1 richness, (b) Shannon diversity, (c) Simpson evenness, (d) phylogenetic diversity (PD_whole_tree), and (e) Goods_coverage, comparing pre-TIPS (n = 20) and post-TIPS (n = 20) samples. Each line represents 1 patient, with redlines indicating an increase and blue lines indicating a decrease in the corresponding diversity index after TIPS. P-values were calculated using the paired Wilcoxon signed-rank test. Among all indices, only the Simpson index showed a significant change after TIPS (P = 0.0484), whereas others remained statistically nonsignificant (P > 0.05). PD, phylogenetic diversity; TIPS, transjugular intrahepatic portosystemic shunt

Figure 4.

(a) Order-level taxa changes pre- and post-TIPS segment plots show order changes (pre-TIPS n = 20, post-TIPS n = 20). Lines link paired samples; red: up, blue: down. Numbers in top-left: decreased (red), increased (blue) counts. Plot titles report P-values. Plot titles report both Log2 fold change and shrunken Log2 fold change (according to the DEse2 function lftcshrink) between pre-TIPS and post-TlPS groups. (b) Family-level taxa changes pre- and post-TIPS segment plots show family changes (pre-TIPS n = 20, post-TIPS n = 20). Lines link paired samples; red: up, blue: down. Numbers in top-left: decreased (red), increased (blue) counts. Plot titles report P-values. Plot titles report both Log2 fold change and shrunken Log2 fold change (according to the DEse2 function lftcshrink) between pre-TIPS and post-TlPS groups. (c) Genus-level taxa changes pre-TIPS and Post-TIPS segment plots show genus changes (pre-TIPS n = 20, post-TIPS n = 20). Lines link paired samples; red: up, blue: down. Numbers in top-left: decreased (red), increased (blue) counts. Plot titles report P-values. Plot titles report both Log2 fold change and shrunkenLog2 fold change (according to the DEse2 function lftcshrink) between pre-TIPS and post-TlPS groups. TIPS, transjugular intrahepatic portosystemic shunt.

Microbiota analysis of sarcopenia alterations after TIPS

To evaluate overall differences in microbial abundance among the 3 groups—sarcopenia reversal group (SRG, n = 5), significant SMI increase group (SSIG, n = 5), and no significant SMI improvement group (NSIIG, n = 10) (Table 2)—we performed a Kruskal-Wallis test on key taxa at the order, family, and genus levels. The Kruskal-Wallis H statistic revealed no statistically significant differences in overall microbial composition across the 3 groups. Although the lack of significance may be attributed to the small sample size and inter-individual variability, We further conducted a correlation heatmap analysis (Figure 5) to compare the relative abundance of gut microbiota among 3 groups: SRG, SSIG, and NSIIG. The intensity of the color in the heatmap represents the correlation of bacterial abundance with the group average, where red indicates a positive correlation (high abundance) and blue indicates a negative correlation (low abundance). The results revealed distinct microbial composition patterns across the 3 groups, reflecting differences in muscle health status.

Table 2.

Baseline characteristics of patients according to post-TIPS SMI response

	SRG (n = 5)	SSIG (n = 5)	NSIIG (n = 10)	P value
Age	57	53	59.5	0.755
Child-Pugh score	5	6	8	0.098
MELD	17	14.5	18.85	0.181
PPG before TIPS (mm Hg)	23	23	18.9	0.303
PPG after TIPS (mm Hg)	7	6	6.5	0.847

MELD, model for end-stage liver disease; NSIIG, no significant SMI improvement group; PPG, portosystemic gradient; SMI, skeletal muscle index; SRG, sarcopenia reversal group; SSIG, significant SMI group.

Figure 5.

Relative abundance of key gut microbiota taxa across different sarcopenia—related groups heatmap displays the relative abundance of key gut microbiota taxa (at order, genus, and family levels) across 3 groups: sarcopenia reversal group (SRG, [n = 5]), SSIG (with SMI increase > 10%, [n = 5]), and NSIIG ([n = 10]). Color intensity represents the relative abundance level, with the color scale ranging from −1 (low abundance, blue) to 1 (high abundance, red). NSIIG, no significant SMI improvement group; SMI, skeletal muscle index; SRG, sarcopenia reversal group; SSIG, significant SMI increase group.

In the SRG, several taxa demonstrated markedly higher abundance (red), including Faecalibacterium, Streptococcus, members of the order Sphingomonadales, and Pseudomonadales. This suggests that these taxa may be positively associated with the reversal of sarcopenia. Conversely, genera such as Fusicatenibacter and Peptostreptococcus showed lower abundance (blue), indicating a potential negative association with sarcopenia reversal. In contrast, the SSIG displayed an opposite pattern, characterized by a higher abundance of Peptostreptococcus and Intestinibacter. Notably, taxa that were enriched in the SRG, such as Faecalibacterium and Pseudomonas, were present at low levels in the SSIG. The NSIIG demonstrated a unique microbiota signature, with a strikingly high abundance of Fusicatenibacter (deep red), whereas several beneficial taxa, including Faecalibacterium, were markedly reduced. This profile suggests that the enrichment of Fusicatenibacter may be linked to the absence of significant muscle improvement.

DISCUSSION

Our study provides valuable insights into the dynamic alterations in the gut microbiota in cirrhotic patients with sarcopenia and PH after TIPS placement. Our findings demonstrate that, compared with HCs, cirrhotic patients with sarcopenia and PH exhibited profound gut dysbiosis, characterized by reduced α-diversity and distinct β-diversity clustering. Although overall diversity indices remained unchanged after TIPS, significant taxonomic shifts were observed, including enrichment of beneficial SCFA-producing bacteria such as Faecalibacterium and Lachnospiraceae_ND3007_group, alongside a reduction in potentially pathogenic taxa such as Pseudomonas and Parabacteroides. Importantly, sarcopenia reversal was observed in a subset of patients and was associated with enrichment of SCFA-producing genera, whereas those without improvement demonstrated persistence of proinflammatory taxa such as Fusicatenibacter.

First, we found a clear distinction in the gut microbiota composition between cirrhotic patients with sarcopenia and PH and HCs. Alpha diversity indices, including the Chao1, Shannon, and Simpson indices, are significantly reduced in cirrhotic patients, which is indicative of the microbial dysbiosis commonly associated with liver cirrhosis with sarcopenia and PH (20–22). Beta diversity analyses, which are based on both unweighted and weighted UniFrac distances, further confirmed marked differences in community structure between the groups. These findings align with previous studies demonstrating that chronic liver injury and PH are associated with gut barrier dysfunction, bacterial overgrowth, and endotoxemia, all of which may fuel microbial disarray (21,22)^.

Second, although TIPS placement did not significantly impact overall alpha diversity, it induced specific taxonomic shifts at multiple phylogenetic levels. At the order level, there was a notable increase in Burkholderiales, whereas Pseudomonadales and Staphylococcales decreased post-TIPS. At the family level, Ruminococcaceae and Sutterellaceae increased in relative abundance after the procedure, whereas Tannerellaceae, Marinifilaceae, and Pseudomonadaceae decreased. These changes suggest a trend toward a more eubiotic microbial composition postintervention (21,22). At the genus level, beneficial commensals such as Faecalibacterium, Fusicatenibacter, Intestinibacter, and Lachnospiraceae_ND3007_group were elevated after TIPS, whereas potential pathobionts such as Pseudomonas, Prevotella, Parabacteroides, Peptostreptococcus, Muribaculaceae, and Butyricimonas were reduced. Notably, Faecalibacterium prausnitzii, a major member of the Faecalibacterium genus, is known for its anti-inflammatory properties and its ability to produce short-chain fatty acids (SCFAs), which maintain gut homeostasis and mucosal integrity (23). Among the altered taxa, Ruminococcaceae merits special attention. As a key SCFA-producing family, Ruminococcaceae contributes to intestinal barrier maintenance, pH regulation, and suppression of pathogenic bacteria (24). Several studies have reported reduced Ruminococcaceae abundance in cirrhotic patients with HE (25–28). However, the role of this family is complex. For example, Ruminococcus gnavus, a member of this family, has been implicated in increased phenylethylamine production, a neuroactive compound that may worsen HE by modulating neurotransmitter systems (29). This dualistic role highlights the functional heterogeneity within Ruminococcaceae, suggesting that although some strains are protective, others may exacerbate disease, especially in the context of TIPS-induced shunting. This insight opens new perspectives on post-TIPS HE. The potential involvement of Ruminococcaceae-derived metabolites in HE pathogenesis could provide early microbial biomarkers and therapeutic targets for post-TIPS complications. Previous studies have linked changes in microbial metabolites—such as ammonia, indoles, and SCFAs—with altered neurocognitive function and astrocyte swelling, highlighting the gut-brain-liver axis in cirrhosis (27,30–32).

Although our findings suggest partial microbial recovery after TIPS, these results should be interpreted with caution. The lack of significant changes in alpha diversity may reflect the persistent influence of cirrhosis-related factors such as impaired immunity, dysbiosis-driven inflammation, and altered bile acid metabolism. Furthermore, although the increase in beneficial taxa such as Faecalibacterium and Lachnospiraceae suggests improved gut health, the long-term stability and functional relevance of these changes remain unclear.

Third, our analysis revealed distinct gut microbiota signatures among patients with SRG,SSIG, and NSIIG after TIPS, supporting a role for microbial taxa and their metabolites in muscle recovery. In SRG group, the enrichment of Faecalibacterium and other short-chain fatty acid (SCFA)-producing taxa is notable. SCFAs, particularly butyrate, exert anti-inflammatory effects and regulate energy metabolism in skeletal muscle. Experimental studies demonstrate that SCFA supplementation or increased abundance of SCFA-producing bacteria can attenuate age-related muscle loss and improve muscle function (33,34). This mechanistic link may explain why higher Faecalibacterium abundance was associated with sarcopenia reversal in our cohort. By contrast, NSIIG group was characterized by high Fusicatenibacter and low Faecalibacterium. Although Fusicatenibacter has been observed in various conditions, its enrichment has been associated with unfavorable host metabolic signatures and low-grade inflammation (35). This pattern suggests an impaired SCFA-producing capacity or a pro-inflammatory environment that may hinder muscle recovery. SSIG group showed a different microbial profile, marked by enrichment of Peptostreptococcus and Intestinibacter. Peptostreptococcus species are increasingly recognized for their proinflammatory potential and their ability to exacerbate intestinal inflammation in animal models (36). Although SSIG patients exhibited measurable SMI gains, the accompanying microbiota may represent compensatory but metabolically less favorable adaptations compared with SCFA-rich profiles in SRG. These results are consistent with growing evidence linking gut dysbiosis to sarcopenia. Several clinical studies and systematic reviews have reported reduced microbial diversity and depletion of beneficial taxa in sarcopenic individuals (37,38). Collectively, this supports a gut-muscle axis where the microbiome influences muscle homeostasis through metabolic, immune, and endocrine pathways.

Clinically, our findings suggest that microbiota-directed interventions—such as prebiotics, probiotics, or fecal microbiota transplantation—may represent adjunctive approaches to improve sarcopenia outcomes after TIPS. Furthermore, the heterogeneity between SRG and SSIG highlights the need to integrate microbial and metabolic biomarkers, beyond muscle mass alone, in assessing sarcopenia recovery.

Although our follow-up period was limited to 3 months, this time point was selected to capture early biological adaptations after TIPS. Previous studies have demonstrated that skeletal muscle mass may improve within the first 3 months after TIPS placement. For example, Liu et al (10), reported significant increases in skeletal muscle index at 3 months postprocedure. In parallel, Gitto et al (23), observed significant gut microbiota remodeling at 3 months after TIPS, suggesting that this interval represents an active early phase of physiological and microbial reorganization. Nevertheless, the present study primarily reflects early post-TIPS alterations. Longer-term follow-up with repeated assessments is required to determine the durability and clinical relevance of microbiome and muscle mass changes beyond the initial 3-month period, and to clarify whether early improvements translate into sustained functional benefits.

This study has several limitations. First, it was conducted at a single center with a relatively small sample size (n = 20), which limits the statistical power and generalizability of our findings. In particular, subgroup analyses stratified by sarcopenia reversal (e.g., SRG, SSIG, NSIIG) should be interpreted with caution because of the limited number of patients in each group and the potential for type II error. Second, although the 3-month follow-up was selected to capture early post-TIPS microbiome remodeling and initial skeletal muscle changes reported in previous studies, this relatively short duration does not allow assessment of the long-term durability of microbial and muscle adaptations. Such extended observation will be necessary to determine whether the early changes we observed translate into sustained physiological adaptations. Third, although all first-degree relatives were healthy and lived with the patients, age, sex, body mass index, and alcohol intake were not strictly matched, and residual confounding cannot be excluded. Fourth, post-TIPS factors such as detailed diet and physical activity were not comprehensively assessed, which may represent unmeasured confounders. Future multicenter studies with larger cohorts, extended follow-up, and integrated multiomics approaches are warranted to validate and mechanistically clarify the role of the gut-muscle-liver axis in patients undergoing TIPS.

In conclusion, our findings reveal that TIPS placement induces notable taxonomic shifts in the gut microbiota of cirrhotic patients with sarcopenia and PH, even in the absence of significant changes in alpha diversity. These alterations involve an increased abundance of SCFA-producing and anti-inflammatory genera, suggesting a trend toward partial microbial restoration. The role of specific bacterial groups, particularly Ruminococcaceae, in the pathogenesis and potential prediction of post-TIPS HE warrants further mechanistic exploration. These findings highlight the complex interplay between sarcopenia, PH, the gut microbiota, and cirrhosis progression, warranting further investigations with larger cohorts and functional analyses to fully elucidate the clinical significance of these microbiome alterations.

CONFLICTS OF INTEREST

Guarantor of the article: Bin Xiong, MD.

Specific author contributions: Conceptualization: B.X., J.W. Data curation: H.W., Y.C., X.Z., H.W. Methodology: B.X., H.W. Resources: B.X. Supervision: B.X., J.W., T.Z., S.C., K.J. Writing—original draft: H.W., Y.C., X.Z. Writing—review and editing: B.X., J.W.

Financial support: None to report.

Potential competing interests: None to report.

Ethics statement: All study protocols were approved by the Ethics Committees of The First Affiliated Hospital of Zhejiang University School of Medicine and were conducted in accordance with the institutional standards of the Declaration of Helsinki (1964) and its subsequent amendments. All details that might reveal the identities of the subjects have been removed from the manuscript.

ABBREVIATIONS:

ALB

albumin

ANOVA

analysis of Variance

AST

aspartate aminotransferase

ASV

amplicon sequence variant

BMI

body mass index

DB

direct bilirubin

DNA

deoxyribonucleic acid

EASL

European Association for the Study of the Liver

FDR

false discovery rate

Hb

hemoglobin

HCs

healthy controls

HE

hepatic encephalopathy

INR

international normalized ratio

IQR

interquartile range

LEfSe

linear discriminant analysis effect size

MELD

model for end-stage liver disease

NSIIG

No Significant Skeletal Muscle Index Improvement Group

NYHA

New York Heart Association

OHE

overt hepatic encephalopathy

PC

principal component

PCoA

principal coordinate analysis

PCR

polymerase chain reaction

PD (PD_whole_tree)

phylogenetic diversity

PERMANOVA

permutational multivariate analysis of variance

PH

portal hypertension

PLT

platelet count

PPG

portosystemic pressure gradient

PT

prothrombin time

QIIME2

quantitative insights into microbial ecology 2

rRNA

ribosomal ribonucleic acid

SCFAs

short-chain fatty acids

SD

standard deviation

SMI

skeletal muscle index

SRG

sarcopenia reversal group

SSIG

significant skeletal muscle index increase group

TIPS

transjugular intrahepatic portosystemic shunt

WBC

white blood cell