Mesenchymal stem cell therapy for liver fibrosis need “partner”: Results based on a meta-analysis of preclinical studies

Abstract

BACKGROUND

The efficacy of mesenchymal stem cells (MSCs) in treating liver fibrosis has been demonstrated in several clinical studies. However, their low survival and liver implantation rates remain problematic. In recent years, a large number of studies in animal models of liver fibrosis have shown that MSCs combined with drugs can improve the efficacy of MSCs in the treatment of liver fibrosis alone and inhibit its progression to end-stage liver disease. This has inspired new ways of thinking about treating liver fibrosis.

AIM

To investigate the effectiveness and mechanisms of MSCs combined with drugs in treating liver fibrosis.

METHODS

Data sources included four electronic databases and were constructed until January 2024. The subjects, interventions, comparators, outcomes, and study design principle were used to screen the literature, and the quality of the literature was evaluated to assess the risk of bias. Relevant randomised controlled trials were selected, and the final 13 studies were included in the final study.

RESULTS

A total of 13 studies were included after screening. Pooled analysis showed that MSCs combined with drug therapy significantly improved liver function, promoted the repair of damaged liver tissues, reduced the level of liver fibrosis-related indexes, and effectively ameliorated hepatic fibrosis by modulating the hepatic inflammatory microenvironment, promoting the homing of MSCs, and regulating the relevant signaling pathways, and the treatment efficacy was superior to MSCs alone. However, the combined treatment statistics showed no ame-lioration in serum albumin levels (standardized mean difference = 0.77, 95% confidence interval: -0.13 to 1.68, P = 0.09).

CONCLUSION

In conclusion, MSCs combined with drugs for treating liver fibrosis effectively make up for the shortcomings of MSCs in their therapeutic effects. However, due to the different drugs, the treatment mechanism and effect also differ. Therefore, more randomized controlled trials are needed to compare the therapeutic efficacy of different drugs in combination with MSCs, aiming to select the “best companion” of MSCs in treating hepatic fibrosis.

Key Words: Mesenchymal stem cell; Stem cell therapy; Liver fibrosis; Combination of drugs; Cirrhosis; Meta-analysis

Core Tip: The efficacy of mesenchymal stem cells (MSCs) therapy for hepatic fibrosis is affected by liver colonization rate and low survival rate. In recent years, an increasing number of studies have shown that the combination of MSCs and drugs can improve the efficacy of MSCs alone in treating liver fibrosis. Our Meta-analysis summarizing the current studies and systematically assessed the efficacy of this combined strategy to ameliorate liver fibrosis.

INTRODUCTION

Liver fibrosis is a chronic pathological reaction process in the liver caused by various factors such as alcohol, drugs, inflammation, or cholestasis[1], which can lead to cirrhosis and even hepatocellular carcinoma[2]. It is characterized by a progressive accumulation of extracellular matrix, leading to an imbalance in production and degradation in the body, which disrupts the physiological structure of the liver[3]. If not effectively reversed or contained, it will progress to irreversible end-stage liver diseases such as cirrhosis and hepatocellular carcinoma. Currently, liver transplantation is an effective treatment for advanced liver disease, but it is difficult to be widely promoted in clinical practice due to its drawbacks such as high cost, lack of donors, immune rejection, and the need for lifelong immunosuppressive drugs[4]. Therefore, there is an urgent need for efficient treatments to ameliorate liver fibrosis in patients, which is important for preventing the progression of liver disease and the development of hepatocellular carcinoma.

Mesenchymal stem cells (MSCs) are multipotent stem cells with the capacity of self-renewal, homing, and low immunogenicity[5]. MSCs have been reported to migrate to the site of liver injury to reverse hepatic fibrosis through immunomodulation, liver-derived differentiation, and paracrine mechanisms[6]. With this unique advantage, MSCs are attractive candidates for treating liver fibrosis. However, MSCs alone have limited efficacy in treating hepatic fibrosis, and their transplantation may suffer from the disadvantages of low liver colonization rate, low survival rate, and short duration of action[7,8]. In addition, some researchers have found that MSC implantation promotes the activation of hepatic stellate cells and weakens the anti-fibrotic ability of MSCs, thus accelerating the process of hepatic fibrotic lesions[8,9]. Therefore, there is an urgent need to change the treatment strategy to improve the efficacy of MSCs in liver fibrosis.

In recent years, to compensate for the insufficiency of MSCs, transplantation of MSCs combined with drugs to ameliorate liver fibrosis has gradually become a novel therapeutic idea. In many studies of animal models of liver fibrosis, it has been confirmed that MSCs combined with drugs ameliorate the degree of liver fibrosis and the level of liver function more significantly than MSCs alone, which suggests that it could be used as a therapeutic approach for liver fibrosis[10]. In recent years, several studies have conducted meta-analyses on the use of MSCs to ameliorate liver disease[10,11]. Unfortunately, there is no scholarly Meta-analysis of trials, which improves the therapeutic efficiency of MSCs based on randomized controlled trials. Here, we screened and extracted data on MSC combination drugs for the treatment of hepatic fibrosis to critically assess the effectiveness of MSC combination drug transplantation for treating hepatic fibrosis and explore the therapeutic mechanisms through systematic evaluation. Overall, these results will provide a new direction for the treatment of liver disease with MSCs in the clinical setting.

MATERIALS AND METHODS

The systematic reviews and meta-analysis were interpreted and elaborated as per Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA).

Search strategies

The sources of the search were mainly from PubMed, Web of Science, Embase, and Cochrane Library published in English. From the database creation to January 2024, two authors (Xu Y and Wang XS) browsed through the titles, abstracts, and keywords to select literature that might meet the inclusion criteria, using the keywords “mesenchymal stem cells” and “hepatic fibrosis”. The full texts were then obtained after initially screening and reading to determine their acceptability. Detailed search methods are described in Supplementary Table 1.

Study selection

The complete article was retrieved according to the subjects, interventions, comparators, outcomes, and study design (PICOS) principal criteria. The details are as follows: Subjects (P): Animals successfully modeled as having hepatic fibrosis, irrespective of country, region, or ethnicity. Intervention (I): Intervention was MSCs combined with medication only. Comparison (C): MSCs treated alone. Outcome (O): Serum liver function [albumin (ALB), alanine aminotransferase (ALT), aspartate aminotransferase (AST), hydroxyproline], hepatic histopathology (Masson stain and Sirius stain), and expression of genetic markers related to hepatic fibrosis [alpha-smooth muscle actin (α-SMA) gene, Transforming growth factor beta 1 (TGF-β1) gene, and Type III collagen gene]. Study design (S): Only randomized controlled trials were included in this study.

Exclusion criteria

(1) Human subjects; (2) Use of immunodeficient animals; animal models of cancer or tumors; (3) The study is not experimental, but rather a review, conference, clinical case, or meta-analysis; and (4) Studies unrelated to the article topic (e.g., interventions using clinical studies and non-MSCs combined drug infusions) were excluded.

Required data extraction

Two reviewers (Zhou XL and Xu Y) extracted data from the incorporated literature into the table. When data were not available in the text, we used GetData Graph Digitizer version 2.25.0.32 software to extract data from the graphs.

The following information was extracted from the included literature: Study information (first author, year of publication, country), animals (species and modeling method), MSCs (cell type, cell dose, route of administration), drug type, drug dose, and key outcome indicators.

Quality assessment

Two reviewers (Lu WM and Tang XK) independently assessed the risk of bias in each included study according to the Cochrane Assessment Tool. If necessary, the two reviewers could discuss or consult with the third author (Wang XS) to resolve differences. The following items were assessed: Allocation scheme, blinding of investigators and outcome assessments, selective reporting, and other sources of bias. Each item was categorized as low risk, high risk, or unclear risk.

Statistical analysis

The data analysis of the effect of MSC therapy for hepatic fibrosis in combination with drug therapy was performed using Review Manager 5.4 software. The required data were extracted directly from the original literature or calculated indirectly from the original data by means of a conversion tool. Finally, the data were analyzed.

In this meta-analysis, standardized mean difference (SMD) was employed to present continuous results for studies where the same outcome was evaluated at several time periods. Heterogeneity was analyzed using the I² statistic. Values of I² < 50% were considered to indicate low or moderate heterogeneity and were meta-analyzed using a fixed-effects model. I² ≥ 50% represented significant heterogeneity. If the P value of the heterogeneity test is greater than or equal to 0.05, we use a fixed-effects model; Otherwise, we use a random-effects model. Sensitivity analysis or subgroup analyses were conducted to evaluate the results when there was a significant amount of heterogeneity between the two groups. P < 0.05 indicated statistically significant differences.

RESULTS

Study selection

A systematic search of the four databases retrieved a total of 2673 articles. After screening for duplicates, 1914 articles remained. Of these, 1272 articles lacked relevant topics, 79 were clinical studies, 174 were previous reviews and meta-analyses, and 62 were non-randomized controlled trials, resulting in 327 potentially eligible articles. However, on review of the full text, 69 articles were non-quality randomized controlled trials, 92 had no relevant results, and 153 were undated and full-text conference abstracts. Ultimately, 13 articles were included in our meta-analysis. The specific filtering process is presented in Figure 1.

Figure 1 Literature selection and inclusion process. RCT: Randomized control trial.

Characteristics of the included studies

Of the 13 included studies, four were from China, three from Japan, two from Egypt, one from Russia, one from Korea, one from Iran, and one from Pakistan. All studies were randomized controlled trials. Nine studies infused bone marrow MSCs (BM-MSCs), two studies infused placental MSCs (PD-MSCs), one study infused adipose MSCs (AT-MSCs), and one study infused Wharton’s jellied MSCs. All studies were published from 2010 to 2023. The main infusion routes were tail vein infusion and intrahepatic injection, with a total dose of 1 × 10⁶ to 1 × 10⁷ cells/kg. Supplementary Table 2 shows the detailed characteristics of the included studies.

Risk assessment of bias

Figure 2 show the assessment results of the risk of bias and methodological applicability of the included studies[12]. It was found that there was a high risk of randomized outcome generation in one study, a low risk of bias in nine studies[7,13-19], and no mention of randomized outcome generation in three studies[20-22] (unclear risk of bias). Seven studies[7,13-16,19,23] (low risk of bias) mentioned allocation concealment and 6 studies did not[12,17,18,20-22]. For blinding of outcome assessment, 1 retrieved study was low risk[15] and 12 studies were not reported. The majority of studies had complete information on outcome, and 2 studies had incomplete information on outcome[12,22] (high risk of bias). It was not clear whether there was selective reporting bias in the three studies[18,21,22], and there was no other bias in all the studies.

Figure 2 Risk of bias graph and bias summary. A: Review authors’ judgments about each risk of bias item presented as percentages across all included studies; B: Review authors’ judgments about each risk of bias item for each included study.

Primary outcome

AST: Ten studies reported serum AST levels. A random effects model was used as an effect indicator because of the wide variation in means across studies. Pooled analysis of the forest plot results showed that the differences were statistically significant [SMD = -1.31, 95% confidence interval (95%CI): -1.93 to -0.68, P < 0.001; test for heterogeneity P = 0.006, I² = 61%]. Serum AST levels were significantly lower in the MSCs combined with drugs group than in the MSCs alone treatment group (Figure 3A).

Figure 3 Forest plots. A: Aspartate aminotransferase (AST); B: Mesenchymal stem cells source subgroups of AST levels; C: Animal model subgroup of AST levels; D: Mechanism subgroup of AST levels; E: Alanine aminotransferase (ALT); F: Mechanism subgroup of ALT levels; G and H: Hydroxyproline; I: Forest plot of type III collagen; J: Mechanism subgroup of transforming growth factor beta gene levels. MSCs: Mesenchymal stem cells; 95%CI: 95% confidence interval; BM-MSCs: Bone marrow mesenchymal stem cells; AT-MSCs: Adipose mesenchymal stem cells; PD-MSCs: Placental mesenchymal stem cells.

To study the effect of various factors on the efficacy of MSCs combination drugs, subgroup analyses of AST levels were performed.

MSC sources mainly included bone marrow and placenta. Pooled analysis showed that the AST levels in the MSCs combined with drugs group were significantly lower than those in the MSCs alone treatment group (SMD = -1.31, 95%CI: -1.93 to -0.68, P < 0.001; heterogeneity test P = 0.006, I² = 61%). A subgroup analysis of PD-MSCs using a random effects model demonstrated a significant reduction in AST levels (SMD = -1.74, 95%CI: -2.58 to -0.89, P < 0.001; heterogeneity test P = 0.62, I² = 0%). AST levels were also significantly reduced in the BM-MSCs subgroup (SMD = -1.21, 95%CI: -2.12 to -0.30,P = 0.009; heterogeneity test P = 0.002, I² = 72%) (Figure 3B).

In the included literature, liver fibrosis models included mice and rats. Pooled analysis showed that the AST levels in the MSCs combined with drugs group were significantly lower than those in the MSCs alone treatment group (SMD = -1.31, 95%CI: -1.93 to -0.68, P < 0.001; heterogeneity test P = 0.006, I² = 61%). Subgroup analysis using a random-effects model showed a significant reduction in AST levels using mouse modeling (SMD = -1.68, 95%CI: -2.52 to -0.84, P < 0.0001; heterogeneity test P = 0.52; I² = 0%). AST levels were also significantly reduced using rat modeling (SMD = -1.23, 95%CI: -2.00 to -0.46, P = 0.002; heterogeneity test P = 0.003, I² = 67%) (Figure 3C).

In the included studies, according to the action mode of different drugs, the mechanisms of MSCs combined with drugs to improve liver fibrosis were classified into modulation of the inflammatory response, modulation of MSCs differentiation, and modulation of hepatic fibrosis-related signaling pathways. Pooled analysis showed a significant reduction in the AST levels in the MSCs combined with the drugs group compared with the MSCs alone treatment group (SMD = -1.31, 95%CI: -1.93 to -0.68, P < 0.001; test for heterogeneity P = 0.006, I² = 61%). Subgroup analysis using a random-effects model showed that the drug caused a significant reduction in the AST levels by modulating the signaling pathway (SMD = -1.30, 95%CI: -1.82 to -0.77, P < 0.00001; test for heterogeneity P = 0.43, I² = 0%). A significant reduction in the AST levels was likewise observed in the subgroup in which the drug modulated inflammation (SMD = -2.49, 95%CI: -4.80 to -0.18, P = 0.02; test for heterogeneity P = 0.03, I² = 81%). However, no statistically significant differences were observed in the subgroups in which drugs modulated the differentiation of MSCs (SMD = -0.68, 95%CI: -1.94 to 0.59, P = 0.29; test for heterogeneity P = 0.03, I² = 70%) (Figure 3D).

ALT: Ten studies reported serum ALT levels. Due to the wide variation in mean values across studies, a random effects model was used as an effect indicator. Pooled analysis of the forest plot results showed that the differences were statistically significant (SMD = -1.32, 95%CI: -2.18 to -0.47, P = 0.02; heterogeneity test P < 0.0001, I² = 77%) (Figure 3E). Serum ALT levels in the MSCs combined with the drugs group were significantly lower than those in the MSCs alone treatment group.

Then subgroup analyses of ALT levels were performed to study the effect of various factors on the efficacy of MSCs combination drugs.

MSC sources mainly included bone marrow and placenta. Pooled analysis showed that the ALT levels in the MSCs combined with drugs group were significantly lower than those in the MSCs alone treatment group (SMD = -1.32, 95%CI: -2.18 to -0.47, P = 0.02; heterogeneity test P < 0.0001, I² = 77%). A subgroup analysis of BM-MSCs using a random effects model demonstrated a significant reduction in the ALT levels (SMD = -1.27, 95%CI: -2.46 to -0.08, P = 0.04; heterogeneity test P < 0.0001, I² = 80%). However, no statistical difference was observed in the PD-MSCs subgroup (SMD = -1.20, 95%CI: -3.04 to 0.65, P = 0.20; heterogeneity test P = 0.02, I² = 81%) (Supplementary Figure 1).

Liver fibrosis models included mice and rats. Pooled analysis showed that the ALT levels in the MSCs combined with drugs group were significantly lower than those in the MSCs alone treatment group (SMD = -1.32, 95%CI: -2.18 to -0.47, P = 0.02; heterogeneity test P < 0.0001, I² = 77%). Subgroup analyses using a random-effects model showed a significant reduction in the AST levels of the mouse model subgroup (SMD = -2.21, 95%CI: -3.04 to -1.38, P < 0.00001; heterogeneity test P = 0.49, I² = 0%). However, no statistical difference was observed in the subgroup of rat models (SMD = -0.95, 95%CI: -2.00 to 0.09, P = 0.07; heterogeneity test P < 0.0001, I² = 80%) (Supplementary Figure 2).

According to the action mode of different drugs, the mechanisms of MSCs combined with drugs to improve liver fibrosis were classified into modulation of the inflammatory response, modulation of MSCs differentiation, and modulation of hepatic fibrosis-related signaling pathways. Pooled analyses showed that the MSCs-drugs combination therapy group showed a significant decrease in the ALT levels compared with the MSCs-only therapy group (SMD = -1.32, 95%CI: -2.18 to -0.47, P = 0.02; heterogeneity test P < 0.0001, I² = 77%). Subgroup analyses using a random-effects model showed that the drugs caused a significant reduction in the ALT levels by modulating signaling pathways (SMD = -1.26, 95%CI: -2.08 to -0.44, P = 0.003; test for heterogeneity P = 0.05, I² = 57%). The ALT levels were also significantly lower in the subgroup in which drugs modulated inflammation (SMD = -2.82, 95%CI: -4.03 to -1.60, P < 0.00001; test for heterogeneity P = 0.20, I² = 38%). However, no statistical differences were observed in the subgroups in which drugs modulated the differentiation of MSCs (SMD = 0.37, 95%CI: -0.91 to 1.65, P = 0.57; test for heterogeneity P = 0.13, I² = 57%) (Figure 3F).

HYP: Five studies reported serum hydroxyproline levels. Pooled analysis showed that MSCs combined with drugs significantly reduced hydroxyproline levels compared to MSCs alone (SMD = -1.71, 95%CI: -2.69 to -0.74, P = 0.0006; test for heterogeneity P = 0.01, I² = 68%). Serum hydroxyproline levels were analyzed in subgroups of MSCs source, MSCs transplantation method, species, and modeling method, but no source of heterogeneity was found. When two studies were excluded[17,23], the result of heterogeneity was P = 0.15, I² = 47% (Figure 3G and H).

Area percentage of collagen fibers (%)

In eight studies, histopathological examinations were performed to quantify the extent of liver injury. The studies reported the area of hepatic fibrosis, as determined by Masson staining and Sirius red staining. In the MSCs combined with the drugs group, there was a significant reduction in the area of hepatic fibrosis compared to the MSCs alone treatment group (Supplementary Figure 3).

MSCs combined with drugs significantly ameliorate liver fibrosis

Nine studies reported ratings of liver fibrosis-related markers, in which MSCs in combination with drugs significantly reduced Type III collagen gene, TGF-β1 gene, and α-SMA gene levels more effectively than MSCs treatment alone in animal models of liver fibrosis.

Four studies reported Type III collagen gene expression levels. As the mean differences between studies were not significant, a fixed-effects model was used as an effect indicator. Pooled analysis of the forest plot results showed that the differences between both groups were statistically significant (SMD = -1.34, 95%CI: -1.98 to -0.71, P < 0.0001; test for heterogeneity P = 0.16, I² = 46%) (Figure 3I).

Five studies reported TGF-β1 gene levels. Due to the large differences in mean values between the studies, a random effect model was used as an effect indicator. Pooled analysis of the forest plot results showed that the differences between both groups were statistically significant (SMD = -2.09, 95%CI: -3.95 to -0.59, P = 0.006; test for heterogeneity P < 0.0001, I² = 84%) (Supplementary Figure 4).

The included studies were analyzed, and the results showed that the TGF-β1 gene level in the liver of the MSCs combined with drug group was significantly lower than that of the MSCs alone treatment group (SMD = -2.09, 95%CI: -3.95 to -0.59, P = 0.006; test for heterogeneity P < 0.0001, I² = 84%). Based on the different action modes of the different drugs, the mechanism of MSCs combined with drugs to improve liver TGF-β1 gene level can be divided into regulating MSC differentiation and regulating liver fibrosis-related signaling pathways. Subgroup analyses using a random-effects model indicated that the drugs significantly reduced TGF-β1 gene levels by modulating signaling pathways (SMD = -3.33, 95%CI: -4.87 to -1.79, P < 0.0001; heterogeneity test P = 0.09, I² = 59%). However, no statistical differences were observed in the subgroups in which the drugs regulated MSCs differentiation (SMD = -0.42, 95%CI: -1.18 to 0.34, P = 0.28; test for heterogeneity P = 0.32, I² = 0%) (Figure 3J).

Five studies reported α-SMA gene levels. Due to the wide variation in means between the studies, a random effect model was used as an effect indicator. Pooled analysis of the forest plot results showed that the differences were statistically significant (SMD = -1.75, 95%CI: -2.96 to -0.18, P = 0.03; test for heterogeneity P < 0.0001, I² = 84%) (Supplementary Figure 5).

Sensitivity analyses and publication bias

Sensitivity analyses were performed on the primary outcome indicators to test the stability and reliability of the results of the meta-analyses. The results showed that none of the studies had a significant impact on any of the combined outcomes, and none of the statistically significant differences were altered (Figure 4), suggesting that the results are relatively robust and reliable.

Figure 4 Sensitivity analysis. A: Aspartate aminotransferase; B: Alanine aminotransferase; C: Hydroxyproline; D: Transforming growth factor-β; E: Type Ⅲ collagen; F: Masson staining.

The AST and ALT funnel plots were symmetrical. The results of Begg’s test and Egger’s test for AST and ALT confirmed that the publication bias was absent (Figure 5).

Figure 5 Funnel plot and Egger’s publication bias plot. A: Funnel plot with pseudo-95% confidence limits [aspartate aminotransferase (AST)]; B: Egger’s publication bias plot (AST); C: Funnel plot with pseudo-95% confidence limits [alanine aminotransferase (ALT)]; D: Egger’s publication bias plot (ALT). SMD: Standardized mean difference.

DISCUSSION

Liver fibrosis caused by viral or metabolic chronic liver disease is a major challenge for global health[24]. The level of fibrosis correlates with the progression of liver disease[2]. If the pathological changes of liver fibrosis are not effectively controlled, it will in turn lead to vascular abnormalities and structural destruction of the liver, resulting in the loss of compensatory capacity of liver function, which will eventually progress to end-stage liver diseases such as cirrhosis and hepatocellular carcinoma[25,26]. Although numerous studies have shown that some drugs have great potential to combat liver fibrosis in animal and clinical trials[27-29], no drugs have been developed to target the reversal of the level of liver fibrosis, and liver transplantation is still the treatment choice for end-stage liver diseases due to liver fibrosis[30]. However, liver transplantation is hampered by the high cost of surgery, lack of liver donors, and immune rejection[31]. Therefore, there is an urgent need to seek efficient therapeutic approaches to ameliorate liver fibrosis in patients.

Currently, clinical treatments for hepatic fibrosis mainly include eliminating the cause of the disease, inhibiting the inflammatory response, suppressing the activation of hematopoietic stem cells (HSCs), and maintaining the in vivo balance of extracellular matrix (ECM)[32]. In recent years, with the in-depth research on MSCs for liver disease, it has gradually entered the clinical trial stage and become a hotspot of attention in the regenerative medicine development[33,34]. Liver fibrosis, as a reversible pathological stage, can be effectively ameliorated by mechanisms such as hepatogenic differentiation[35], paracrine effects[36], and immunomodulation[37] of MSCs. It has also been confirmed in clinical trials that MSCs can significantly improve liver function, reverse hepatic fibrosis, and alleviate patients’ clinical symptoms[33,38]. Unfortunately, after isolation and culture in vitro, MSCs face the disadvantages of lack of nutrients, oxygen and external growth factors, as well as being affected by the in vivo microenvironment and inflammatory response after transplantation[39], which ultimately leads to the disadvantages of low colonization rate and low survival rate of effect after transplantation of MSCs into the liver. Therefore, MSCs therapy for liver diseases is not widely used in clinical practice[8]. With the deepening of related studies worldwide, MSCs combined with various drugs have shown better therapeutic effects in animal models of liver fibrosis, with efficacy superior to that of MSCs alone[40-42].

A number of traditional Chinese medicines have been reported to have been widely used in the treatment of liver disease by the advantages of low toxicity and adverse effects and the ability to target and regulate fibrosis-related signaling pathways[27,43-45]. A large number of animal experiments have confirmed that MSCs combined with drugs can improve the homing and survival rates of MSCs in the liver[41,46] and significantly ameliorate liver fibrosis by further inhibiting the activation of HSCs[15], regulating the microenvironment of the organism[7], and modulating the signaling pathways of fibrogenesis and development[12,13,19], which provides a new idea for the poor efficacy of MSCs alone in treating liver fibrosis. Table 1 summarizes the mechanisms involved in the combination of MSCs with different drugs to improve the efficacy of liver fibrosis treatment in the 13 studies included in this paper.

Table 1 The mechanism of mesenchymal stem cells combined with different drugs to improve the therapeutic effect of liver fibrosis.

Ref.	Type of MSCs	Combination drugs	Mechanism
Ouyang et al[17], 2010	BM-MSCs	Salidroside	Promoting MSCs differentiation; the expression of TGF-β1 was inhibited
Chai et al[7], 2013	BM-MSCs	Oxymatrine	Regulating the inflammatory microenvironment
Motawi et al[22], 2014	BM-MSCs	Simvastatin	Attributed to their effect on the MMPs/TIMPs balance
Hammam et al[16], 2016	WJ-MSCs	Praziquantel	Promoting the differentiation of MSCs into functional hepatocytes
Jang et al[19], 2018	BM-MSCs	Simvastatin	Inhibition of TGF-β/Smad signaling pathway
Fathy et al[13], 2020	AT-MSCs	Eugenol	Inhibition of TGF-β/Smad signaling pathway
Mazhari et al[12], 2020	BM-MSCs	Imatinib	Blocking PDGF signaling pathway
El-Monem et al[14], 2021	BM-MSCs	Nano curcumin	Reducing TGFβ1 levels and attenuating the expression of Smad 2,3 and collagen I, III genes
Qiao et al[23], 2011	PD-MSCs	Baicalin	Promoting the differentiation of MSCs into functional hepatocytes
Iwasawa et al[20], 2021	BM-MSCs	Juzentaihoto	Inducing the production of Treg cells and anti-inflammatory macrophages
Zhang et al[15], 2022	BM-MSCs	Ferulic acid	Inhibition of HSCs activation
Rafiq et al[21], 2022	BM-MSCs	Melaton	Promoting MSCs homing
Slautin et al[18], 2024	PD-MSCs	Fucoxanthin	Decreasing the proinflammatory cytokine levels, TGF-β level, α-SMA+, TIMP-1+ areas and increasing HGF level, MMP-13+, and MMP-9+ areas

MSCs: Mesenchymal stem cells; WJ-MSCs: Wharton’s jellied mesenchymal stem cells; BM-MSCs: Bone marrow mesenchymal stem cells; AT-MSCs: Adipose mesenchymal stem cells; PD-MSCs: Placental mesenchymal stem cells; HSC: Hematopoietic stem cell; PDGF: Platelet-derived growth factor; TGF-β: Transforming growth factor beta; α-SMA: Alpha-smooth muscle actin; TIMP-1: Tissue inhibitor of metalloproteinases-1; HGF: Hepatocyte growth factor; MMP: Matrix Metalloproteinase.

ALT and AST, mainly distributed in hepatocytes, are the most commonly used serum markers in the clinical diagnosis and treatment of patients with liver disease, so they were also the most important indicators in our included studies[47]. If the liver is damaged, ALT and AST in hepatocytes will enter the bloodstream to result in elevated blood levels of ALT and AST[48]. In our included studies, MSCs combined with drugs reduced serum levels of AST and ALT more significantly than MSCs alone in animal models of liver fibrosis, effectively improving liver function. Subgroup analyses were performed based on serum levels of AST and ALT. The results of the mechanism subgroup showed that MSCs in combination with drugs were more effective in ameliorating hepatic fibrosis through modulating signaling pathways than modulating the inflammatory response. However, there was no statistically significant difference between MSCs combined with drugs and treatment alone in this subgroup of regulating differentiation. A subgroup analysis was also carried out based on the level of TGF-β1. The results showed that MSCs combined with drugs were more effective in modulating signaling pathways than in regulating the differentiation of MSCs into hepatocyte-like cells in ameliorating hepatic fibrosis. Therefore, it was hypothesized that modulating signaling pathways might dominate in ameliorating liver fibrosis.

It was also found that Iwasawa et al[20] intervened in hepatic fibrosis using BM-MSCs in combination with Juzentaihoto, which resulted in the most significant reduction in the AST and ALT levels among all the included studies when compared to MSCs alone in treating hepatic fibrosis. The drug combined with MSCs effectively increased the NKp46/CD45 and CD4CD25/CD4 ratios by affecting the immune response, induced the production of Treg cells and anti-inflammatory macrophages, and attenuated liver injury. In addition, Juzentaihoto effectively normalized lipid levels, tricarboxylic acid cycle, and urea cycle, which in turn ameliorated hepatic fibrosis. Therefore, the combination strategy of MSCs with this drug may bring a boon to patients with liver diseases in the clinical practice, and this direction will be continuously followed.

BM-MSCs are currently the most widely used seed cells in animal and clinical trials for liver fibrosis[49]. Many investigators have found that ALT and AST levels and model for end-stage liver disease (MELD) scores of patients with hepatic fibrosis were improved by peripheral intravenous infusion of their BM-MSCs, ultimately enhancing their quality of life[50-52]. However, as the patients’ age increases, the differentiation ability of their BM-MSCs decreases, and studies have shown that combination therapy with Salidroside can promote the differentiation of BM-MSCs[17]. The combination of Salidroside can be considered to improve the efficiency of cellular therapy in elderly patients when their own BM-MSCs are used.

In addition to the two indicators of ALT and AST, ALB serves as an important indicator of liver reserve and transport function. The decrease of ALB often indicates the severity and late stage of liver injury[53]. The forest statistical analysis of ALB levels in the included studies found that a statistically significant difference was not observed for MSCs in combination with drugs over MSCs alone (Supplementary Figure 6). Given that MSCs had effectiveness against liver fibrosis, it was speculated that MSCs in combination with nano curcumin, baicalin, and fucoxanthin would affect serum levels of ALB.

The progression of various chronic liver diseases is usually due to an imbalance between fibrogenesis and degradation in the liver, which results in excessive collagen deposition in the liver[3,54]. Using Sirius red staining, collagen fibers are stained red by acidic magenta, while Masson’s staining method stains collagen fibers blue by aniline blue during staining[55]. Eight studies revealed a significant reduction in the area of hepatic fibrosis when MSCs were combined with drugs compared to treatment with MSCs alone, and the differences were all statistically significant, suggesting that the combination of MSCs with different drugs can degrade collagen deposition and alleviate hepatic fibrosis.

When the liver is damaged by stimuli such as physical, chemical, or biological factors, the damaged hepatocytes release various cytokines, chemokines, and reactive oxygen species and recruit Kupffer cells into the damaged lesions, which in turn produce inflammatory factors and growth factors. Examples include interleukins IL-1β, IL-6, and TGF-β[56,57] are important paracrine signals that promote the activation of HSCs[58]. Activated HSCs can also transdifferentiate into myofibroblast-like cells, which are capable of synthesizing ECM proteins[24,59]. Eventually, excessive deposition of ECM proteins impairs the function and structure of the liver, leading to the progression of liver fibrosis. According to the included studies, four studies reported Type III collagen gene expression levels, five studies reported TGF-β1 gene levels, and five studies reported α-SMA gene levels. The effect of MSCs combined with drugs on improving liver fibrosis is better than that of MSCs alone. Subgroup analyses of TGF-β1 and α-SMA gene levels were performed, but no source of heterogeneity was found. This may be related to the MSC infusion dose, which was not analyzed in subgroups due to the high variability of MSC infusion dose in the included literature. In addition, the variability between the combination of different drugs and the route of transplantation may also have an impact on the source of heterogeneity. However, we performed a sensitivity analysis of liver fibrosis-related genes and biochemical markers that respond to liver function and found that the statistical difference was not affected by any study exclusion, suggesting that the results were stable.

Although many in vivo animal experiments have demonstrated that MSCs in combination with drug therapy are more effective in improving liver fibrosis than MSCs alone, different modeling modalities may affect the efficacy of MSCs in combination with drugs. CCl4 modeling was used in 10 of the included studies. The results confirmed treatment differences between the two groups. However, when treating liver fibrosis animals induced by receiving porcine serum[17], it was found that there was no difference in serum levels of ALB and ALT between the two groups. In studies using other modalities to induce liver fibrosis in animals, the researchers did not measure serum markers associated with ALT, AST, ALB, and ALP[16,19]. It was discovered that all 10 of these studies used tail vein transplantation of MSCs to treat liver fibrosis. In other studies, the researchers implanted MSCs into hepatic fibrosis animals by intrahepatic graft injection. Since the results of ALT, AST, and ALP levels after intrahepatic transplantation of MSCs were not evaluated, the impact of the two transplantation routes on liver function could not be assessed. It has been reported that MSCs are not easily and uniformly distributed in the liver after intrahepatic infusion and their effects are limited[60]. However, more studies are needed to evaluate the effects of both transplantation routes on serum levels of ALT, AST, and ALP to confirm whether transplantation of MSCs via the caudal vein route has a better effect on liver function than intrahepatic transplantation.

In addition, there are some difficulties in clinical trial research, including fewer patients, large individual differences in clinical trials, lack of comparability, and poor reproducibility. Therefore, there are few reports of MSCs combined with drugs for treating liver diseases, and most of them focus on safety and efficacy assessment. In a study of MSCs combined with pioglitazone for the treatment of patients with compensated cirrhosis, the patient’s clinical status remained stable, with no major adverse reactions[61]. A transient improvement in the MELD scores was observed at month 3 post-infusion, culminating in a return to baseline levels at month 12 and the result showed that the combination treatment was significantly better than MSCs alone. It is suggested that the combination of pioglitazone and MSCs is safe and feasible. Although combination therapy is a new direction that can be adopted, the drug concentration, route of administration, frequency of treatment, and the route and quantity of MSCs transplanted in combination with drugs need to be further investigated.

This review comprehensively and systematically analyzed the preclinical efficacy of MSCs combined with drugs for treating liver diseases. Pooled results from Meta-analysis showed that MSCs combination drugs significantly enhanced liver functions (as reflected in ALB, ALT, AST, ALP, and hydroxyproline levels), repaired damaged liver tissues (as reflected in Masson trichrome or Sirius red staining), and improved liver fibrosis-related gene levels (as reflected in Type III collagen, TGF-β1, and α-SMA levels) (Table 1). In our included study, Chai et al[7] showed that the combination treatment was significantly better than BM-MSCs alone in an experiment of BM-MSCs synergized with Oxymatrine in treating liver fibrosis in rats, and the researchers found that the drug did not increase the colonization of BM-MSCs in the liver, but promoted the secretion of IL-4 and IL-10 by MSCs to ameliorate liver fibrosis. However, there was no systematic evaluation and meta-analysis of whether MSCs combined with drugs could improve liver fibrosis by increasing anti-inflammatory and less pro-inflammatory factors. This is a deficiency of the article because the data on multiple inflammatory indicators given in the included literature are insufficient. Many countries around the world are currently exploring the pretreatment of MSCs or a combination of drugs for liver disease intervention, but the safety and efficacy of the drugs have not been fully addressed, making these approaches difficult to access clinically and wasting a lot of resources for medical trials. In conclusion, our Meta-analysis systematically assessed the effects of MSCs combined with drugs on liver fibrosis by regulating inflammatory levels, regulating differentiation and liver fibrosis-related signaling pathways by summarizing the current research on MSCs combined with drugs in treating liver fibrosis (Figure 6). In addition, we explored the comparative efficiency of different transplantation routes and modeling modalities in improving liver histopathology and liver function. These provide important evidence and direction for the subsequent promotion of clinical transformation of related research.

Figure 6 Diagram of the mechanism of mesenchymal stem cell therapy for liver fibrosis. MSCs: Mesenchymal stem cells; HSCs: Hepatic stellate cells.

CONCLUSION

In animal models of liver fibrosis, MSCs combined with drugs can improve liver function and reverse liver fibrosis and promote liver tissue healing more effectively than MSCs alone. In the future, MSCs combined with different drugs are expected to be a new direction for treating liver diseases in clinical practice, providing effective and safe treatment for patients with liver fibrosis and improving the status of liver disease worldwide. However, before using it as a clinical approach, it is necessary to further study in depth the therapeutic effects and risks of combining MSCs with drugs, standardize the combination strategy, and ensure the safety of combination therapy.