Mesenchymal stem cell transplantation for severe acute pancreatitis-associated lung injury: A meta-analysis of animal studies

Abstract

BACKGROUND

This study is to evaluate the efficacy of mesenchymal stem cells (MSCs) in the treatment of acute respiratory distress syndrome-acute lung injury (ALI), so as to provide evidence-based reference for the clinical use of stem cells in the treatment of acute lung injury associated with pancreatitis in the future.

AIM

To systematically evaluate the efficacy of MSC transplantation in the treatment of severe acute pancreatitis (SAP)-ALI.

METHODS

Databases, including China National Knowledge Infrastructure, Wanfang Data, SinoMed, PubMed, and Web of Science, were electronically searched for animal experiments on MSC transplantation for SAP-ALI from inception of the databases to August 1, 2024. Two researchers independently retrieved the literature, extracted data, and assessed the risk of bias of individual studies. Meta-analysis was performed using the RevMan 5.4 software.

RESULTS

A total of 13 randomized controlled animal experiments were included, involving 248 rats, with 124 rats in the MSC group and 124 in the model control group. Meta-analysis indicated that MSC transplantation reduced the histopathological scores of pancreatic tissues and lung tissues [standardized mean difference (SMD) = -3.12, 95% confidence interval (CI): -3.99 to -2.25] in SAP rats. Additionally, MSC transplantation decreased lung wet-to-dry weight ratio (weighted mean difference = -0.50, 95%CI: -0.64 to -0.36), myeloperoxidase activity (SMD = -3.29, 95%CI: -4.40 to -2.19), and tumor necrosis factor-α (TNF-α) level (SMD = -3.57, 95%CI: -5.49 to -1.64) and its mRNA expression (SMD = -3.74, 95%CI: -5.27 to -2.20) as well as serum TNF-α levels (SMD = -3.51, 95%CI: -4.51 to -2.51). Compared with the control group, the MSC group had significantly lower serum amylase levels (SMD = -3.63, 95%CI: -4.54 to -2.72).

CONCLUSION

MSC transplantation can reduce the pulmonary inflammatory response and alleviate SAP-ALI in SAP rats. However, due to the limited number of included studies and sample size, further multicenter, large-scale randomized controlled trials are warranted to validate our findings.

Key Words: Severe acute pancreatitis; Mesenchymal stem cells; Severe acute pancreatitis-associated lung injury; Meta-analysis; Systematic review

Core Tip: Mesenchymal stem cell transplantation can effectively reduce the pulmonary inflammatory response and alleviate the pathological damage of lung tissue in rats with severe acute pancreatitis (SAP), thus playing a significant protective role in SAP-related lung injury. However, due to the limited number of included studies and heterogeneity among studies, the robustness of the existing conclusions is insufficient. Therefore, multi-center, large-sample, rigorously designed randomized controlled trials combined with standardized stem cell preparation and administration regimen are needed to further systematically evaluate its efficacy, safety and long-term prognosis, so as to provide a higher level of evidence for clinical application.

INTRODUCTION

Severe acute pancreatitis (SAP) is a common acute abdominal inflammatory disease with sudden onset and rapid progression. It is a systemic inflammatory response syndrome characterized by massive inflammatory cell infiltration, lobular necrosis, and hemorrhage[1]. As the most common complication of SAP, SAP-associated acute lung injury (SAP-ALI) is a severe systemic complication characterized clinically by progressive hypoxemia and respiratory distress, which can progress to acute respiratory distress syndrome in severe cases and is also one of the main causes of early death in SAP patients[2]. As the result of a significant systemic inflammatory response, SAP-ALI is accompanied by increased permeability of endothelial and epithelial barriers, which leads to the leakage of protein-rich exudates into the alveolar spaces and interstitial tissue, thereby impairing oxygenation and gas exchange[3,4]. Although the mechanisms underlying acute lung injury (ALI) are increasingly understood, current treatments for SAP-ALI are mainly limited to symptomatic and supportive therapies such as thoracentesis, control of pulmonary infection, improvement of respiratory function, and oxygen inhalation. There is still a lack of direct recommendations and effective methods for treating this disease.

In recent years, mesenchymal stem cells (MSCs) have gradually become a research hotspot for the treatment of ALI due to their advantages such as self-renewal, multidifferentiation, easy availability, and low immunogenicity[5,6]. Current evidence indicates that MSCs alleviate SAP-associated inflammatory lung tissue injury through mechanisms such as inhibiting iron deposition and lipid peroxidation in lung tissue, improving oxidative stress and inflammatory response levels in acute pancreatitis rats, and inducing the M2 polarization of lung macrophages[7-9]. Thus, MSC transplantation has gradually become one of the main potential therapeutic strategies for SAP and SAP-ALI.

In this study, we conducted a meta-analysis to evaluate the efficacy of MSCs in the treatment of SAP-ALI. We integrated scattered mechanistic clues and constructed a clearer and deeper knowledge framework, drawing an “evidence map”, which not only indicates the current research trends, but also accurately points out the evidence gaps to be filled for clinical translation.

MATERIALS AND METHODS

Literature search

English language databases (PubMed, Cochrane Library, Web of Science, and EMBASE) and Chinese databases [China National Knowledge Infrastructure (CNKI), CBM, VIP, and Wanfang Data] were searched for literature related to the application of stem cell therapy for SAP-ALI from inception of the databases to August 1, 2024. The Chinese search terms included “stem cells”, “mesenchymal stem cells”, “severe acute pancreatitis”, and “acute lung injury”, whereas the English search terms included “Mesenchymal Stem Cells”, “Stem Cells”, “severe acute pancreatitis”, “acute pancreatitis”, “lung injury”, and “acute lung injury”. The detailed search strategies for PubMed and CNKI are shown in Table 1.

Table 1 Search strategies for an English database (PubMed) and a Chinese database (CNKI).

	ID: Query
Search strategy for PubMed	#1 “Pancreatitis”[Mesh]
	#2 ((((((((Edematous Pancreatitis, Acute) OR (Acute Edematous Pancreatitis)) OR (Edematous Pancreatitis, Acute)) OR (Pancreatic Parenchymal Edema)) OR (Acute Pancreatitis)) OR (Peripancreatic Fat Necrosis)) OR (Parenchymal Edema, Pancreatic)) OR (Necrosis, Peripancreatic Fat)) OR (Pancreatitis, Acute)
	#3 “Pancreatitis, Acute Necrotizing”[Mesh]
	#4 ((((((Necrotizing Pancreatitis, Acute) OR (Pancreatitis Necrotising)) OR (Acute Necrotizing Pancreatitis)) OR (Pancreatic Necrosis)) OR (Hemorrhagic Necrotic Pancreatitis)) OR (Necrotic Pancreatitis, Hemorrhagic)) OR (Pancreatitis, Hemorrhagic Necrotic)
	#5 #1 OR #2 OR #3 OR #4
	#6 “Stem Cells”[Mesh]
	#7 ((((((((((((((Cell, Stem) OR (Cells, Stem)) OR (Stem Cell)) OR (Progenitor Cells)) OR (Cell, Progenitor)) OR (Cells, Progenitor)) OR (Progenitor Cell)) OR (Mother Cells)) OR (Cell, Mother)) OR (Cells, Mother)) OR (Mother Cell)) OR (Colony-Forming Unit)) OR (Colony Forming Unit)) OR (Colony-Forming Units)) OR (Colony Forming Units)
	#8 “Mesenchymal Stem Cells”[Mesh]
	#9 (((((((((((Stem Cell, Mesenchymal) OR (Bone Marrow Mesenchymal Stem Cells)) OR (Bone Marrow Stromal Cells)) OR (Multipotent Bone Marrow Stromal Cell)) OR (Adipose-Derived Mesenchymal Stem Cells)) OR (Mesenchymal Stem Cells, Adipose-Derived)) OR (Adipose Tissue Derived Mesenchymal Stem Cells)) OR (Adipose Tissue Derived Mesenchymal Stem Cells)) OR (Mesenchymal stem Cells)) OR (Multipotent Mesenchymal stem Cells)) OR (Progenitor Cell, Mesenchymal)) OR (Wharton Jelly Cells)
	#10 #6 OR #7 OR #8 OR #9
	#11 “Lung Injury”[Mesh]
	#12 ((Injuries, Lung) OR (Injuries, Pulmonary)) OR (Pulmonary Injuries)
	#13 “Acute Lung Injury”[Mesh]
	#14 ((Acute Lung Injuries) OR (Lung Injuries, Acute)) OR (Lung Injury, Acute)
	#15 #11 OR #12 OR #13 OR #14
	#16 #5 AND #10 AND #15
Search strategy for CNKI	#1 “acute pancreatitis”[MeSH Terms]
	#2 “severe acute pancreatitis”[MeSH Terms]
	#3 #1 OR #1
	#4 “stem cells”[MeSH Terms]
	#5 “mesenchymal stem cells”[MeSH Terms]
	#6 #4 OR #5
	#7 “lung injury”[MeSH Terms]
	#8 “acute lung injury”[MeSH Terms]
	#9 #7 OR #8
	#10 #3 AND #6 AND #9

Eligibility criteria

Study type: Randomized controlled animal studies.

Inclusion criteria: (1) Rat studies; (2) Successful establishment of SAP-ALI models; and (3) The intervention groups receiving stem cell therapy, while the control group receiving phosphate-buffered saline, normal saline, or culture medium.

Exclusion criteria: (1) Review articles, meta-analyses, editorials, dissertations, and academic conference abstracts; (2) Duplicate publications; (3) Articles for which the full text could not be obtained; and (4) Unavailable data.

Data extraction

Two researchers independently performed literature screening and data extraction, and cross-checked the results. In case of discrepancies, a third researcher was consulted and a decision was made through discussion. During literature screening, duplicates were removed. Preliminary screening was then conducted by reading titles and abstracts to exclude obviously irrelevant articles. Finally, full texts were read for secondary screening to determine whether the studies could be included.

The main data extracted included first author’s name, publication year, country, experimental subjects, SAP-ALI modeling method, sample sizes of the stem cell group and the control group, stem cell type, transplantation route, number of stem cells transplanted, control protocol, duration of transplantation, follow-up duration, and outcome measures.

Quality assessment of the included studies

The risk of bias in the included studies was assessed using the 10 items of the SYRCLE’s Risk of Bias tool for animal experiments[10], and the results were categorized as “yes”, “no”, and “unclear”, representing low, high, and unclear risk of bias, respectively.

Outcome measures

The primary outcome measures were the histopathological scores of pancreatic and lung tissues. The secondary outcome measures included lung wet-to-dry weight ratio, myeloperoxidase (MPO) activity, and tumor necrosis factor (TNF)-α level and its mRNA expression as well as serological markers [e.g., amylase, TNF-α, and interleukin (IL)-1β].

Statistical analysis

All analyses were conducted using the RevMan 5.4 software. For articles that reported data only in graphical form, two independent authors extracted the mean and SD values from the figures. For studies reporting results at multiple time points, the results were treated as different studies. Heterogeneity was tested using the Q-test and I² test. For continuous variables, the effect size was expressed as mean difference with 95% confidence interval (CI). If the measurement tools or methods for the same indicator differed, the standardized mean difference (SMD) was used as the pooled effect size. If P ≥ 0.10 and I² ≤ 50%, studies were regarded as having low heterogeneity, and a fixed-effects model was applied for the meta-analysis; I² > 50% or P < 0.10 indicated high heterogeneity among studies, for which a random-effects model was used, and subgroup analysis was conducted to further explore the sources of heterogeneity. Sensitivity analysis was performed to assess the stability of the pooled results. Egger’s test was used to assess potential publication bias. All tests were two-sided, and a P value of < 0.05 was considered statistically significant.

RESULTS

Literature screening process and results

A total of 343 relevant articles were identified in the initial search. After stepwise screening, 13 studies were finally included (Figure 1).

Figure 1  Flow diagram of the study.

General characteristics of the included studies and results of risk of bias assessment

The general characteristics of the included studies are shown in Table 2, and the results of the risk of bias assessment are shown in Table 3.

Table 2 General characteristics of the included studies.

Ref.	Country	Animals	SAP modeling	Sample size (T/C)	Type of stem cells	Route of transplantation	Number of stem cells transplanted	Control protocol	Duration of transplantation	Follow-up duration	Outcome measures
Yu and Song[7], 2022	China	SD rats (200 g)	Injection of 3% sodium taurocholate (1 mL/kg) into the biliopancreatic duct	6/6	BMSCs	Tail vein	1 × 107/rat	PBS	6 hours	72 hours	a, b, e, h, k
Li et al[31], 2020	China	SD rats (200-240 g)	Injection of 4% sodium taurocholate (1 mL/kg) into the biliopancreatic duct	12/12	PMSCs	Tail vein	1 × 107/kg	PBS	6 hours	12 hours, 24 hours	a, b, c, d, e, g, h, j, k
Song et al[32], 2019	China	SD rats (150-200 g)	Injection of 3% sodium taurocholate (1 mL/kg) into the biliopancreatic duct	6/6	BMSCs	Tail vein	1 × 107/kg	PBS	12 hours	72 hours	a, b, d, f, i, k
Yu et al[33], 2017	China	SD rats (180-220 g)	Injection of 1.5% sodium deoxycholate (1 mL/kg) into the biliopancreatic duct	10/10	BMSCs	Tail vein	3 × 106/kg	Blank	0	7 days	c
Chen et al[34], 2016	China	SD rats (200-250 g)	Injection of 4% sodium taurocholate (1 mL/kg) into the biliopancreatic duct	10/10	BMSCs	Tail vein	NR	NS	24 hours	24 hours	a, b, e, g, j
Zhao et al[35], 2016	China	SD rats (200-250 g)	Injection of 3% sodium taurocholate (1 mL/kg) into the biliopancreatic duct	10/10	BMSCs	Tail vein	(5-7) × 107/rat	PBS	24 hours	72 hours	a, g, j
Yang et al[36], 2013	China	SD rats (250-280 g)	Injection of 5% sodium taurocholate (1 mL/kg) into the biliopancreatic duct	12/12	UCMSCs	Tail vein	5 × 106/kg	NS	0, 1 hour, 6 hours, 12 hours	48 hours	a, b, e, k
Sheng et al[37], 2012	China	SD rats (200-250 g)	Injection of 5% sodium taurocholate (1 mL/kg) into the biliopancreatic duct	36/32	BMSCs	Tail vein	1 × 107/kg	NS	1 hours	3 hours, 6 hours, 12 hours, 24 hours	b, c, e, h, k
Wu et al[26], 2012	China	SD rats (200-230 g)	Injection of 5% sodium taurocholate (1 mL/kg) into the biliopancreatic duct	20/20	BMSCs	Femoral vein	(0.5-1) × 104/kg	Medium	1 hour	6 hours, 12 hours	a, b, c, e, h, k
Wang et al[12], 2012	China	SD rats (250-300 g)	Injection of 5% sodium taurocholate (1 mL/kg) into the biliopancreatic duct	25/25	BMSCs	Tail vein	1 × 106/rat	Medium	2 hours	1 hour, 3 hours, 6 hours, 12 hours, 24 hours	c, d, g, k
Li et al[38], 2011	China	SD rats (260-280 g)	Injection of 5% sodium taurocholate (1 mL/kg) into the biliopancreatic duct	20/20	BMSCs	Tail vein	2 × 105/kg	NS	0	12 hours	b, c, e, g
Li et al[39], 2011	China	SD rats (180-200 g)	Injection of 5% sodium taurocholate (1 mL/kg) into the biliopancreatic duct	30/30	BMSCs	Tail vein	1 × 107/kg	NS	0	3 hours, 6 hours, 12 hours	b, c, f, k
Lu et al[40], 2011	China	SD rats	Two intraperitoneal injections of L-arginine (2 g/kg)	48/48	BMSCs	Femoral vein	NR	Blank	6 hours	12 hours, 24 hours, 48 hours, 72 hours	d

a: Pancreatic histopathological score; b: Lung histopathological score; c: Lung wet/dry weight ratio; d: Lung myeloperoxidase level; e: Serum tumor necrosis factor-α (TNF-α) level; f: Lung TNF-α level; g: Lung TNF-α mRNA expression; h: Serum interleukin (IL)-1β level; i: Lung IL-1β level; j: Lung IL-1β mRNA expression; k: Serum amylase level. SAP: Severe acute pancreatitis; T: Treatment group; C: Control group; BMSCs: Bone marrow mesenchymal stem cells; PBS: Phosphate buffered saline; PMSCs: Placenta mesenchymal stem cells; NR: Not reported; NS: Normal saline; UCMSCs: Umbilical cord mesenchymal stem cells.

Table 3 Results of risk of bias assessment in the included studies.

Ref.	Generation of random sequences	Baseline characteristics	Allocation concealment	Randomized animal allocation	Blinding of animal caretakers and investigators	Assessment of random outcomes	Blinding of outcome assessors	Incomplete data reporting	Selective reporting of outcome	Other bias
Yu and Song[7], 2022	Yes	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear
Li et al[31], 2020	Yes	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear
Song et al[32], 2019	Unclear	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear
Yu et al[33], 2017	Unclear	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear
Chen et al[34], 2016	Unclear	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear
Zhao et al[35], 2016	Unclear	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear
Yang et al[36], 2013	Unclear	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear
Sheng et al[37], 2012	Unclear	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear
Wu et al[26], 2012	Unclear	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear
Wang et al[12], 2012	Unclear	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear
Li et al[38], 2011	Unclear	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear
Li et al[39], 2011	Unclear	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear
Lu et al[40], 2011	Unclear	Yes	Unclear	Unclear	No	Unclear	No	No	No	Unclear

Results of meta-analysis

Pancreatic histopathological scores: Among the included studies, seven articles reported a total of 12 outcomes regarding pancreatic histopathological scores. The heterogeneity test indicated significant heterogeneity among these studies (I² = 92.1%, P < 0.001); therefore, a random-effects model was employed. Meta-analysis showed that compared with the control group, stem cell therapy reduced pancreatic histopathological scores in SAP rats (SMD = -4.25, 95%CI: -5.80 to -2.70) (Figure 2A).

Figure 2 Meta-analysis of pancreatic and lung histopathological scores in severe acute pancreatitis rats: Stem cell therapy group vs control group. Letters after the study name (e.g., “a” and “b”) indicate experimental results at different time points within the same study. A: Pancreatic histopathological scores; B: Lung histopathological scores. SMD: Standardized mean difference; CI: Confidence interval.

Lung histopathological scores: Among the included studies, nine reported a total of 19 outcomes regarding lung histopathological scores. The heterogeneity test indicated significant heterogeneity among these studies (I² = 88.2%, P < 0.001); therefore, a random-effects model was employed. The results of the meta-analysis showed that compared with the control group, stem cell therapy reduced lung histopathological scores in SAP rats (SMD = -3.12, 95%CI: -3.99 to -2.25) (Figure 2B).

Lung wet-to-dry weight ratio: Among the included studies, seven reported a total of 18 outcomes regarding lung wet-to-dry weight ratio. The heterogeneity test indicated significant heterogeneity among these studies (I² = 88.4%, P < 0.001); therefore, a random-effects model was employed. Meta-analysis showed that compared with the control group, stem cell therapy reduced the lung wet-to-dry weight ratio in SAP rats (weighted mean difference = -0.50, 95%CI: -0.64 to -0.36) (Figure 3A).

Figure 3 Meta-analysis of lung wet-to-dry weight ratio and myeloperoxidase activity in severe acute pancreatitis rats: Stem cell therapy group vs control group. Letters after the study name (e.g., “a” and “b”) indicate experimental results at different time points within the same study. A: Wet-to-dry weight ratio; B: Myeloperoxidase. SMD: Standardized mean difference; CI: Confidence interval.

Lung MPO activity: Among the included studies, four reported a total of 12 outcomes regarding lung tissue MPO activity. The heterogeneity test indicated significant heterogeneity among these studies (I² = 80.4%, P < 0.001); therefore, a random-effects model was employed. Meta-analysis showed that compared with the control group, stem cell therapy reduced lung MPO activity in SAP rats (SMD = -3.29, 95%CI: -4.40 to -2.19) (Figure 3B).

Serum TNF-α levels, lung TNF-α levels, and lung TNF-α mRNA expression: Among the included studies, 11 reported a total of 15 outcomes regarding serum TNF-α levels, four regarding lung TNF-α levels, and six regarding TNF-α mRNA expression in lung tissue. The heterogeneity test indicated significant heterogeneity among these studies, with I² values of 86.8%, 82.8%, and 82.2%, respectively; therefore, a random-effects model was employed. Meta-analysis showed that compared with the control group, stem cell therapy reduced serum TNF-α levels (SMD = -3.51, 95%CI: -4.51 to -2.51), lung TNF-α levels (SMD = -3.57, 95%CI: -5.49 to -1.64), and lung TNF-α mRNA expression (SMD = -3.74, 95%CI: -5.27 to -2.20) in SAP rats (Figure 4).

Figure 4 Meta-analysis of serum tumor necrosis factor-α levels, lung tumor necrosis factor-α levels, and lung tumor necrosis factor-α mRNA expression in severe acute pancreatitis rats: Stem cell therapy group vs control group. Letters after the study name (e.g., “a” and “b”) indicate experimental results at different time points within the same study. SMD: Standardized mean difference; CI: Confidence interval.

Serum IL-1β levels, lung IL-1β levels, and lung IL-1β mRNA expression: Among the included studies, 11 reported a total of nine outcomes regarding serum IL-1β levels, one regarding lung IL-1β levels, and four regarding lung IL-1β mRNA expression. The heterogeneity test indicated significant heterogeneity among the studies for serum IL-1β levels and lung IL-1β mRNA expression, with I² values of 87.0% and 70.5%, respectively; therefore, a random-effects model was employed. Meta-analysis showed that compared with the control group, stem cell therapy reduced serum IL-1β levels (SMD = -3.93, 95%CI: -5.48 to -2.39), lung IL-1β levels (SMD = -3.46, 95%CI: -5.35 to -1.56), and lung IL-1β mRNA expression (SMD = -3.91, 95%CI: -5.83 to -2.35) in SAP rats (Figure 5).

Figure 5 Meta-analysis of serum (interleukin-1β levels, lung interleukin-1β levels, and lung interleukin-1β mRNA expression in severe acute pancreatitis rats: Stem cell therapy group vs control group. Letters after the study name (e.g., “a” and “b”) indicate experimental results at different time points within the same study. SMD: Standardized mean difference; CI: Confidence interval.

Serum amylase levels: Among the included studies, eight articles reported a total of 22 outcomes regarding serum amylase levels. The heterogeneity test indicated significant heterogeneity among these studies (I² = 85.7%, P < 0.001); therefore, a random-effects model was employed. Meta-analysis showed that compared with the control group, stem cell therapy reduced serum amylase levels in SAP rats (SMD = -3.63, 95%CI: -4.54 to -2.72) (Figure 6).

Figure 6 Meta-analysis of serum amylase levels in severe acute pancreatitis rats: Stem cell therapy group vs control group. Letters after the study name (e.g., “a” and “b”) indicate experimental results at different time points within the same study. SMD: Standardized mean difference; CI: Confidence interval.

Results of subgroup analysis

For outcome indicators with significant heterogeneity, including lung histopathological score, MPO activity, and wet-to-dry weight ratio, subgroup analyses were performed based on stem cell type, transplantation route, duration of transplantation, and follow-up duration. However, the sources of heterogeneity were not identified. Subgroup analysis showed that compared with the control group, stem cell therapy reduced lung histopathological scores, lung MPO activity, and lung wet-to-dry weight ratios regardless of changes in stem cell type [bone marrow MSCs (BMSCs) or non-BMSCs], transplantation route (tail vein or femoral vein), transplantation duration (< 6 hours or ≥ 6 hours), and follow-up duration (< 12 hours or ≥ 12 hours), and the differences were all statistically significant (all P < 0.05) (Table 4).

Table 4 Subgroup analysis.

Subgroup	Number of studies included	Test for heterogeneity		Effect model	SMD/WMD (95%CI)	P value
		I2 (%)	P value
Lung histopathology score
Stem cell type
BMSCs	13	86	< 0.001	Random	-2.95 (-3.96 to -1.93)	< 0.001
Non-BMSCs	6	89	< 0.001	Random	-2.87 (-4.46 to -1.29)	< 0.001
Transplantation route
Tail vein	17	83	< 0.001	Random	-2.37 (-3.12 to -1.62)	< 0.001
Femoral vein	2	79	0.03	Random	-8.09 (-12.90 to -3.28)	0.001
Transplantation duration
< 6 hours	12	89	< 0.001	Random	-3.34 (-4.50 to -2.18)	< 0.001
≥ 6 hours	7	79	< 0.001	Random	-2.15 (-3.29 to -1.00)	< 0.001
Follow-up duration
< 24 hours	10	88	< 0.001	Random	-3.21 (-4.46 to -1.95)	< 0.001
≥ 24 hours	9	86	< 0.001	Random	-2.62 (-3.79 to -1.46)	< 0.001
Lung MPO activity
Stem cell type
BMSCs	10	79	< 0.001	Random	-3.03 (-4.22 to -1.85)	< 0.001
Non-BMSCs	2	39	0.2	Random	-2.56 (-4.15 to -0.97)	0.002
Transplantation route
Tail vein	8	78	< 0.001	Random	-4.39 (-6.35 to -2.43)	< 0.001
Femoral vein	4	69	0.02	Random	-2.00 (-2.93 to -1.07)	< 0.001
Transplantation duration
< 6 hours	5	87	< 0.001	Random	-7.32 (-12.13 to -2.51)	0.003
≥ 6 hours	7	58	0.03	Random	-2.24 (-2.98 to -1.50)	< 0.001
Follow-up duration
< 24 hours	6	81	< 0.001	Random	-3.07 (-4.91 to -1.22)	0.001
≥ 24 hours	6	79	< 0.001	Random	-4.47 (-6.82 to -2.12)	< 0.001
Lung wet/dry weight ratio
Stem cell type
BMSCs	16	89	< 0.001	Random	-0.46 (-0.61 to -0.32)	< 0.001
Non-BMSCs	2	0	0.33	Random	-0.87 (-1.13 to -0.61)	< 0.001
Transplantation route
Tail vein	16	88	< 0.001	Random	-0.50 (-0.66 to -0.34)	< 0.001
Femoral vein	2	95	< 0.001	Random	-0.54 (-1.02 to -0.06)	0.03
Transplantation duration
< 6 hours	16	89	< 0.001	Random	-0.46 (-0.61 to -0.32)	< 0.001
≥ 6 hours	2	0	0.33	Random	-0.87 (-1.13 to -0.61)	< 0.001
Follow-up duration
< 24 hours	14	84	< 0.001	Random	-0.46 (-0.59 to -0.33)	< 0.001
≥ 24 hours	4	94	< 0.001	Random	-0.62 (-1.11 to -0.13)	0.01

SMD: Standardized mean difference; WMD: Weighted mean difference; CI: Confidence interval; BMSCs: Bone marrow mesenchymal stem cells; MPO: Myeloperoxidase.

Results of sensitivity analysis

Sensitivity analysis was performed on the pooled results for lung histopathological score, MPO activity, and wet-to-dry weight ratio by sequentially excluding individual studies. Among these indicators, no single study exerted excessive influence (Figure 7).

Figure 7 Sensitivity analysis results. A: Lung histopathological score; B: Lung myeloperoxidase activity; C: Lung wet-to-dry weight ratio. CI: Confidence interval; MPO: Myeloperoxidase.

Publication bias

Funnel plots were generated based on the lung histopathological score, lung MPO activity, and wet-to-dry weight ratio in SAP rats treated with stem cell therapy, and Egger’s test was used to assess publication bias. The funnel plots for lung histopathological score (Egger’s test: t = -8.95, P < 0.001) and lung MPO activity (Egger’s test: t = -6.71, P < 0.001) were asymmetrical, indicating possible publication bias. In contrast, the funnel plot for lung wet-to-dry weight ratio was basically symmetrically distributed, and Egger’s test (t = -1.67, P = 0.115) also indicated no obvious publication bias (Figure 8).

Figure 8 Funnel plots. A: Lung histopathological score; B: Lung wet-to-dry weight ratio; C: Lung myeloperoxidase activity. WMD: Weighted mean difference; MPO: Myeloperoxidase.

DISCUSSION

SAP-ALI is considered a complex and challenging clinical disease. Currently, the treatment for ALI mainly focuses on symptomatic and supportive therapies, such as thoracentesis and drainage, controlling pulmonary infection, correcting hypoalbuminemia, and oxygen inhalation, while effective therapeutic regimens are still lacking. In ALI models, the pathophysiological processes of the lung mainly include oxidative stress injury, inflammatory cytokine infiltration, and alveolar cell apoptosis[11]. However, the underlying mechanisms of SAP and how it leads to ALI are not fully understood. Mechanisms such as signal transduction pathways, activation of biomarkers, and anti-pyroptosis play important roles in the pathogenesis of ALI[3].

The present meta-analysis indicates that MSCs can alleviate ALI by downregulating MPO, TNF-α, and IL-1β, as well as reducing the lung wet/dry weight ratio. These results are consistent with the findings of Wang et al[12]. Being classic inflammatory cytokines, both IL-1β and TNF-α have been widely demonstrated to play extremely important roles in the different pathways of SAP-ALI. Various inflammatory mediators contribute to the aggregation of macrophages and neutrophils, subsequently triggering a cascade of pathological changes in the pulmonary microcirculation, leading to the occurrence and aggravation of SAP-associated lung injury[13,14]. The histopathological score of rats was selected as the primary outcome of observation as the score is objective and can more objectively show the degree of lung tissue injury in rats. In addition, the lung function and survival rate of rats can be affected by many factors. Xu et al[15] found that emodin might reduce neutrophil infiltration and improve SAP-ALI in rats by inhibiting NLRP3/IL-1β/CXCL1 signaling activated by cold-inducible RNA-binding protein. TNF-α is released by macrophages and monocytes in activated fixed tissues and plays an important role in the early occurrence and development of SAP-ALI[16]. It first binds to TNF receptor 1 (TNF-R1) and interacts with adaptor proteins such as TNF-R1-associated death domain protein and receptor-interacting protein; subsequently, it triggers an intracellular cascade reaction and ultimately activates nuclear factor kappa B (NF-κB)[17,18]. Activation of NF-κB can enhance transcription of the TNF-α gene, thereby forming a vicious cycle that amplifies early inflammatory signals and exacerbates the initial inflammatory effects[19].

The past few decades have witnessed increasing attention towards the biomedical applications of MSCs with the development of cell therapy. Although the mechanism of MSCs in the treatment of ALI is not yet fully elucidated, studies have demonstrated that MSCs can inhibit ALI through various pathways. MSCs can be easily isolated from tissues such as bone marrow, adipose tissue, and placenta, and can differentiate into various cell lineages according to the requirements of specific biomedical applications[20]. As MSCs do not significantly express major histocompatibility complex and immunostimulatory molecules, they will not cause severe graft rejection after implantation[21]. MSCs can differentiate into almost all terminal lineage cells, enabling them to engraft in specific microenvironments[22]. Their immunological properties, including anti-inflammatory, immunomodulatory, and immunosuppressive capabilities, contribute to their potential role as immunotolerant agents. Moreover, evidence indicates that MSCs exert anti-inflammatory and immunomodulatory effects via diverse pathways in a variety of inflammatory diseases[23-25]. Studies have shown that allogeneic BMSC transplantation can inhibit the expression of IL-1β and TNF-α by suppressing aquaporin-1 and aquaporin-5[26]. With regard to SAP-ALI, Yin et al[27] found that BMSCs may alleviate SAP and SAP-ALI by regulating the release of inflammatory cytokines through paracrine pathways. Focusing on immunoregulation, Huang et al[28] demonstrated that exogenous placental chorionic plate-derived MSCs tend to home to injured tissues (pancreas, lung, etc.) and reduce pancreatic injury and systemic inflammation in SAP rats by secreting TNF-α-stimulated gene 6 protein and inducing M2 polarization of macrophages. Pancreatitis-associated lung injury is characterized by diffuse alveolar damage, microvascular injury, type I pneumocyte necrosis, and influx of inflammatory cells, resulting in injury to endothelial and epithelial cells in lung tissue[29]. Evidence indicates that MSCs can exert a protective effect against oxidative stress-induced apoptosis via their antioxidant activity and by downregulating MPO activity[30].

Although this study systematically evaluated the role of MSCs in the treatment of SAP-ALI, several limitations exist. First, only Chinese and English databases were searched; as a result, relevant studies in other languages might have been missed, potentially leading to selection bias. Second, some included studies did not describe the randomization method, allocation concealment, or blinding implementation, which might result in selection and performance bias. Third, a certain degree of publication bias was observed for some outcome measures, which might affect the credibility of our conclusions. Finally, none of the included studies reported rejection after stem cell transplantation, and future research should pay more attention to the safety of stem cell therapy.

CONCLUSION

Taken together, the available evidence suggests that MSCs play an important role in the treatment of SAP and its complications. However, further well-designed preclinical and clinical studies are needed to confirm their efficacy. Due to the limitations in the number and quality of the included studies, high-quality studies are warranted to further validate our conclusions. In addition, differences in stem cell source, cell dose, route of transplantation, timing of treatment, and follow-up time may affect the observed therapeutic effect, and a large number of validated experiments are required to confirm these factors.