Comparative Analysis of Short-Term and Long-Term Clinical Efficacy of Mesenchymal Stem Cells from Different Sources in Knee Osteoarthritis: A Network Meta-Analysis

Background Joint articular injection of mesenchymal stem cells (MSCs) has emerged as a novel treatment approach for osteoarthritis (OA). However, the effectiveness of MSCs derived from different sources in treating OA patients remains unclear. Therefore, this study aimed to explore the differences between the effectiveness and safety of different sources of MSCs. Materials and Methods For inclusion consideration, we searched trial registries and published databases, including PubMed, Cochrane Library, Embase, and Web of Science databases. Revman (V5.3), STATA (V16.0), and R (V4.0) were utilized for conducting data analysis, while the Cochrane Risk of Bias Tool was employed for assessing the quality of the studies. We derived outcome measures at 6 and 12 months based on the duration of study follow-up, including visual analog scale (VAS) score, WOMAC score, WOMAC pain, WOMAC Functional Limitation, and WOMAC stiffness. The evaluation time for short-term effectiveness is set at 6 months, while 12 months is utilized as the longest follow-up time for most studies to assess long-term effectiveness. Results The evaluation of literature quality showed that the included studies had excellent methodological quality. A meta-analysis revealed that different sources of MSCs improved knee function and pain more effectively among patients suffering from knee OA (KOA) than controls. The results of the network meta-analysis showed the following: short-term functional improvement (the indexes were evaluated after 6 months of follow-up) (WOMAC total score: bone marrow-derived MSC (BMMSC) vs. adipose-derived MSC (ADMSC) (mean difference (MD) = −20.12, 95% confidence interval (CI) −125.24 to 42.88), umbilical cord-derived MSC (UCMSC) (MD = −7.81, 95% CI −158.13 to 74.99); WOMAC stiffness: BMMSC vs. ADMSC (MD = −0.51, 95% CI −7.27 to 4.29), UCMSC (MD = −0.75, 95% CI −9.74 to 6.63); WOMAC functional limitation: BMMSC vs. ADMSC (MD = −12.22, 95% CI −35.05 to 18.86), UCMSC (MD = −9.31, 95% CI −44.26 to 35.27)). Long-term functional improvement (the indexes were evaluated after 12 months of follow-up) (WOMAC total: BMMSC vs. ADMSC (MD = −176.77, 95% CI −757.1 to 378.25), UCMSC (MD = −181.55, 95% CI −937.83 to 541.13); WOMAC stiffness: BMMSC vs. ADMSC (MD = −0.5, 95% CI −26.05 to 18.61), UCMSC (MD = −1.03, 95% CI −30.44 to 21.69); WOMAC functional limitation: BMMSC vs. ADMSC (MD = −5.18, 95% CI −316.72 to 177.1), UCMSC (MD = −8.33, 95% CI −358.78 to 218.76)). Short-term pain relief (the indexes were evaluated after 6 months of follow-up) (VAS score: UCMSC vs. BMMSC (MD = −10.92, 95% CI −31.79 to 12.03), ADMSC (MD = −14.02, 95% CI −36.01 to 9.81), PLMSC (MD = −17.09, 95% CI −46.31 to 13.17); WOMAC pain relief: BMMSC vs. ADMSC (MD = −11.42, 95% CI −39.52 to 11.77), UCMSC (MD = −6.73, 95% CI −47.36 to 29.15)). Long-term pain relief (the indexes were evaluated after 12 months of follow-up) (VAS score: BMMSC vs. UCMSC (MD = −4.33, 95% CI −36.81 to 27.08), ADMSC (MD = −11.43, 95% CI −37.5 to 13.42); WOMAC pain relief: UCMSC vs. ADMSC (MD = 0.23, 95% CI −37.87 to 38.11), BMMSC (MD = 5.89, 95% CI −25.39 to 51.41)). According to the GRADE scoring system, WOMAC, VAS, and AE scores were of low quality. Conclusion Meta-analysis suggests MSCs can effectively treat KOA by improving pain and knee function compared to control groups. In terms of functional improvement in KOA patients, both short-term (6-month follow-up) and long-term (12-month follow-up) results indicated that while the differences between most treatments were not statistically significant, bone marrow-derived MSCs may have some advantages over other sources of MSCs. Additionally, BM-MSCs and UC-MSCs may offer certain benefits over ADMSCs in terms of pain relief for KOA patients, although the variances between most studies were not statistically significant. Therefore, this study suggests that BM-MSCs may present clinical advantages over other sources of MSCs.


Introduction
Osteoarthritis (OA) is a common chronic degenerative joint disease that mainly affects articular cartilage and surrounding tissues, leading to joint space narrowing and osteophyte formation.Weight-bearing joints, such as ankles, knees, and hips, are especially affected.During the initial phases, it presents itself as localized pain and rigidity, progressively resulting in a decline in physical abilities and impairment of function [1,2].In the past, OA was considered a simple wear-and-tear condition; however, it is now understood to be a complex multifactorial process involving inflammation and metabolic factors [3,4,5,6,7].Knee OA (KOA) is a common type of OA.According to the Global Burden of Disease Study 2015, approximately 85% of the burden of OA is attributed to KOA, with an estimated prevalence of 10% in men aged > 60 years and 13% in women aged > 60 [8,9].Current clinical treatments for KOA include nonsteroidal anti-inflammatory drugs (NSAIDs), total knee replacement surgery, and intraarticular injections [10,11,12,13].However, NSAIDs provide pain relief but do not alleviate joint damage.While total knee replacement surgery has proven effective in treating KOA, it is expensive and burdensome, with potential complications such as infection, thrombosis, and the need for revision surgery, which increases the risk of mortality in elderly patients [14].Recently, stem cell transplantation has emerged as a promising treatment option for OA, with significant advantages.Mesenchymal stem cells (MSCs), a type of multipotent adult stem cells, possess anti-inflammatory, immune-modulatory, and regenerative properties and have been widely used in cell therapy for various diseases.MSCs can be derived from bone marrow, adipose tissue, and umbilical cord and are referred to as bone marrow-derived MSCs (BMMSCs), adipose-derived MSCs (ADMSCs), and umbilical cord-derived MSCs (UCMSCs), respectively [15,16,17].The main markers expressed by MSCs in the human body are CD90, CD73, and CD105, although slight differences exist in marker expression between stem cells from different tissues [12,18].MSCs have been used in the clinical treatment of OA with good results.Multiple studies have demonstrated the significant advantages of stem cells for pain relief, joint function improvement, and cartilage regeneration in patients with OA [19,20,21].Most of the studies included patients with Kellgren-Lawrence classifications ranging from I to III, while some studies included patients with IV classifications and showed good efficacy in improving patients' symptoms [21,22,23].However, owing to the differences in their origin and therapeutic effects, MSCs from different sources exhibit variations in clinical efficacy for treating KOA.The clinical efficacy of different sources of MSCs for the treatment of OA has been evaluated differently in previous studies, and the dosage of bone marrow, adipose, and UCMSCs commonly used in clinical practice and the selection of different cells are the current problems.Therefore, we reviewed the latest research reports on the treatment of KOA with MSCs from different sources and compared them with the control group, and indirectly compared MSCs from different sources through network meta-analysis, using the control group as an intermediate bridge to compare the treatment groups with different MSCs.So, the aim of this study was to evaluate the efficacy of MSCs from different sources and the differences between them, as well as to provide a more reliable and tailored reference for the therapeutic injection therapy of KOA.

Materials and Methods
2.1.Registration.The present study was a systematic review of published studies, which does not require the consent of patients or ethical approval [24].According to the guidelines for preferred reporting items and previously published protocols, this review was based on systematic reviews and metaanalyses [25].This systematic evaluation has been registered in PROSPERO (reference number: CRD 42023448217) as a detailed protocol.

Data Sources and Retrieval
Strategy.We conducted a comprehensive search of trial registries and published databases, including PubMed, Embase, the Cochrane Library, and Web of Science.The search encompassed all randomized controlled trials (RCTs) on MSCs and OA published in these databases until February 15, 2024.A combination of subject terms and free text were used to construct the search model, which was then applied to the four aforementioned databases.Then a search was conducted in the database based on the following search terms: ("Mesenchymal Stem Cells" or "Stem Cell, Mesenchymal" or "Mesenchymal Stem Cell" or "Bone Marrow Mesenchymal Stem Cells" or "Adipose-Derived Mesenchymal Stem Cells" or "Wharton Jelly Cells" and "Osteoarthritis" or "Osteoarthritides" or "Osteoarthrosis" or "Osteoarthroses" or "Knee Osteoarthritides" or "Osteoarthritis of Knee."The complete search formula is in Supplementary 1.

Inclusion and Exclusion
Criteria.The inclusion and exclusion process for articles followed the principles of PICOS (Participants, Interventions, Comparisons, Outcomes, and Studies).Inclusion criteria were as follows: (i) studies with an RCT design, (ii) patients with KOA, (iii) experimental groups receiving at least one intra-articular injection of MSCs from any source, and (iv) include at least one of the following efficacy parameters, such as the University of Western Ontario and McMaster University Osteoarthritis (WOMAC) score or its subscale or visual analog scale (VAS) score.The exclusion criteria were as follows: (i) studies involving vascular stromal cells, (ii) studies with only abstracts available (insufficient data), and (iii) studies for which full-text access was not available.In each included study, two independent evaluators evaluated the methodological quality according to the Cochrane Collaboration guidelines.As a result of this tool, selection bias, allocation bias, implementation bias, reporting bias, follow-up bias, and other biases were assigned a "low," "unclear," or "high" risk.
2.5.Outcome Indicators.The majority of the studies under review primarily assessed outcomes at 6 and 12 months.Consequently, the 12-month outcomes were employed to assess long-term efficacy, while the 6-month outcomes were utilized to gauge short-term efficacy.Functional improvement was evaluated using WOMAC total, WOMAC stiffness, and WOMAC functional limitation scores, whereas pain relief was assessed using the VAS and WOMAC pain scores.
2.6.Statistical Analysis.In this study, Revman (V5.3),STATA (V16.0), and R (V4.0) were used for data analysis.First, traditional meta-analysis and network diagram plotting were performed using STATA (Stata is a statistical analysis tool with menu-driven operations that cannot be performed with conventional meta-analysis software).Second, the included studies were analyzed using a random-effects model and network meta-analysis using the GeMTC package in R (R is a statistical analysis software based on the Bayesian model).
The level of study heterogeneity is quantified by i 2 .If i 2 exceeds 50%, a random effects model will be employed.When i 2 surpasses 75%, indicating substantial heterogeneity, sensitivity analysis will be conducted to identify the source of heterogeneity and assess the stability of the results, as well as compare the impact of various clinical factors.In cases where no underlying cause for heterogeneity is identified, descriptive analyses will be carried out.None of the studies formed a closed loop.Therefore, a consistency model is used, and no inconsistency testing is required.Finally, we utilize GRADE profiler software (https:// gradeprofifiler.software.informer.com/)according to the index and analysis results of this study to evaluate the quality of the evidence, and using Stata software (version 16.0, http://www.stata.com) Construct a funnel plot to determine publication bias.

Characteristics of Research.
A total of 1,917 papers were found in four databases: 721 in PubMed, 243 in Cochrane Library, 717 in Embase, and 236 in Web of Science.After removing duplicates, 1,437 articles remained.Among these, 85 articles were selected for full-text reading based on their titles and abstracts.Of these, 66 were excluded (the intervention modality is not compliant: included vascular interstitial fraction, bone marrow aspirate, and adipose tissue-derived fraction), leaving 19 studies for the final analysis (Figure 1).These included eight studies on BMMSCs, seven on ADMSCs, three on UCMSCs, and one on placenta-derived MSCs (PLMSCs) (Table 1).

Literature Quality Evaluation Results
. A tool for evaluating the quality of RCTs, the Cochrane Risk of Bias Tool, was used to assess the quality of the 19 studies included in this analysis.Among these, four studies were assessed as having a low risk of bias.Regarding randomization, 16 studies described the random allocation methods employed in the research, whereas three studies did not provide detailed explanations of their specific randomization methods.Regarding data integrity, 17 studies explained the follow-up duration and excluded cases.The literature quality evaluation revealed that the included studies had excellent methodological quality in general.The risk of bias in each study is illustrated in Figure 2.  [19,20,22,23,26,27,28,29,30,31,32,33,34,35,36] at 6 months of follow-up included 813 patients reporting WOMAC total scores between MSCs of different origins (including bone marrow, fat, and umbilical cord) and the control group (Figure 3(a)).The results of the network meta-analysis, categorized by treatment modality, are presented in Supplementary 2. Although the analysis revealed no significant differences among the treatment modalities, the efficacy of BMMSC was superior to that of ADMSC and UCMSC (BMMSC vs. the Saline (MD = −37.64,95% CI −143.18 4(a)).The Bayesian model ranks the methods described in the paragraph as follows: PRP, BMMSC, UCMSC, ADMSC, HA, CT, and Saline (Figure 5(a)).In addition, the ranking of areas under the SUCRA curve is UCMSC, BMMSC, PRP, ADMSC, HA, CT, and Saline (Figure 6(a)) (Supplementary 2).
The results of 12 months of follow-up showed that the MSCs (including bone marrow, fat, umbilical cord MSCs) and the control group were (BMMSC vs. the HA and PRP: MD = −1.18,95% CI −1.52 to −0.84, P ¼ 0:001; MD = −0.43,95% CI − 0.99 to 0.   The network meta-analysis evaluated the efficacy of different MSC sources, extracting VAS and WOMAC pain ratings for short-term (6 months) and long-term (12 months) efficacy.
3.5.Sensitivity Analysis and Publication Bias.Publication bias was evaluated using Egger's test for WOMAC, VAS, and AE, which showed no bias.Heterogeneity was significantly reduced after removing the study by Kim et al. [31], by excluding the included studies one by one using sensitivity analysis.However, the Kim et al. [31] study did not affect the pooled results, which were stable.Stem Cells International

Discussion
OA is a common disease globally that negatively impacts individuals' quality of life.The main treatment approaches for OA include conservative treatment and surgical interventions such as total knee arthroplasty and total hip arthroplasty [40].Surgical treatment can partially relieve joint deformities and pain in patients with end-stage OA; however, the surgical risks and associated complications cannot be overlooked.This is particularly important for elderly patients with a higher risk of complications, such as infections, thrombosis, and the need for secondary surgeries [41].
For conservative treatment, pharmacological therapies, especially NSAIDs, are widely used in clinical practice.However, their primary role is pain control with no reparative effects on cartilage [42].Therefore, cell transplantation has attracted considerable attention as a novel therapeutic approach.Among them, MSCs have emerged as a popular choice for treating OA because of their wide availability (from bone marrow, adipose tissue, muscle, periosteum, and synovium), ease of accessibility, safety, and potential to differentiate into cartilage [43,44,45].These have demonstrated promising results in clinical settings.
In an RCT, Lee et al. [20] treated 12 patients with OA by intra-articular injection of autologous ADMSCs.The MSCs maintained the morphological changes in the cartilage and effectively relieved pain in patients.In a study by Matas et al. [23], UCMSCs were injected intra-articularly, and patients were followed up for 12 months.Clinical scores and magnetic resonance imaging findings were compared, and the results revealed significant improvements in pain and function, with no evidence of cartilage damage on magnetic resonance imaging.Moreover, Lamo-Espinosa et al. [33] injected different doses of BMMSCs into the joint cavity of patients with OA, resulting in reduced pain and improved functional activities.In a four-arm RCT, researchers used autologous bone marrow aspirate concentrate, umbilical cord MSCs, autologous adipose stromal vascular fraction, and corticosteroids in a comparison, and the results showed that umbilical cord MSCs did not have a clear advantage [39].MSCs from various sources have demonstrated significant clinical efficacy in OA treatment, improving pain relief and functional activities in patients with OA.However, differences have not been reported in the therapeutic efficacy of different types of MSCs for OA.
Therefore, we discussed the comparative efficacy of different sources of MSCs (including MSMSCs, ADMSCs, PLMSCs, and UCMSCs) and control groups in treating KOA.Traditional and network meta-analyses were conducted based on a Bayesian framework.The results indicated good heterogeneity between the MSC and control groups.Regarding functional improvement, MSCs from different sources demonstrated improvement in the WOMAC total score for patients with OA compared to the PRP and HA control groups; however, the results were not significant.Numerous studies have demonstrated the clinical efficacy of PRP for KOA.However, the results vary owing to the differences in the preparation process.For example, Park et al. [46] conducted a randomized trial involving patients with KOA who received intra-articular injections of PRP or HA.PRP improved both function and pain, with PRP rich in high concentrations of growth factors demonstrating better efficacy than the other groups.However, some studies have revealed that PRP does not improve ankle OA [47].Overall, PRP and HA have good clinical effects on KOA.MSCs from various sources revealed significant differences compared with the conservative treatment or Saline control groups, demonstrating significant improvement.Moreover, different stem cell sources improved significantly in the long-term follow-up than the no-cell therapy control group.A further reticulation meta-analysis showed that the differences between all treatment modalities were not statistically significant in the WOMAC score metrics at 6 and 12 months of follow-up, and since all treatment modalities can improve patients' symptoms clinically, we observed the reticulation meta-analysis results to find that BMMSC may be advantageous among MSCs of different origins.In addition, in the WOMAC subscales (including WOMAC stiffness and WOMAC functional limitation), although only the WOMAC stiffness rating scale was statistically significant for each treatment modality compared to CT, observing the results of the reticulation meta-analysis also revealed that BMMSC may be advantageous among different sources of MSCs. of MSCs.The advantage of BMMSCs in functional improvement has not been confirmed in previous meta-analyses; however, multiple clinical studies have demonstrated their treatment advantages [21,33].This may be attributed to the inherent self-renewal ability of BMMSCs and the potential for growth factor production and differentiation into chondrocytes, adipocytes, and osteoblasts [15,48].Regarding pain relief, traditional meta-analyses demonstrated good heterogeneity, indicating that MSCs from different sources have significant analgesic effects in patients with OA.Further network meta-analysis of different sources of MSCs showed that only the results of the differences between BMMSC and HA were statistically significant in terms of improvement in VAS scores at 6 and 12 months of follow-up and that the differences between the other different treatment modalities were not statistically significant.However, in the results of the network meta-analysis, it was found that BMMSC may still be advantageous in different MSCs.Also, in the WOMAC pain score, although the differences between treatment modalities were not statistically significant, BMMSC and UCMSC may still have an advantage over ADMSC.A meta-analysis also demonstrated that UCMSC was advantageous in pain relief in comparisons involving various doses [49].Additionally, in a clinical study, repeated injections were more effective than single injections [23].These studies indicate that injection dosage may also improve pain in patients with KOA, although no recommended dosage has been established.MSCs significantly inhibit bone loss in the posterior medial femur area in patients with KOA after intra-articular injection of synovial MSCs [50].This indicates that the reparative effect of MSCs on cartilage may be the key to improving function and relieving pain in patients with OA.Conventional treatments such as physical therapy, NSAIDs, and surgery can eliminate inflammation, swelling, joint adhesion, and stiffness but do not 20 Stem Cells International improve cartilage damage [51].In the phase I/IIa trial, Chahal et al. [52] injected BMMSCs into the knee joint at different concentrations and discovered a significant decrease in cartilage degradation biomarkers and magnetic resonance imaging synovitis.Furthermore, Zhang et al. [53] demonstrated that UCMSCs downregulated the expression of matrix metalloproteinase-5 and upregulated the expression of type II collagen and Ki67 in articular cartilage in an iodine-acetic acid-induced OA rat model.Additionally, ADMSCs alleviated inflammatory reactions and cartilage degeneration in an OA mouse model and increased glycosaminoglycan production, inducing endogenous cartilage formation.In summary, studies have shown that MSCs play an important role in cartilage regeneration and repair, but the injection of MSCs in the clinic faces many problems, mainly including the lack of a strict standard for the injection dose.
Recently, a study conducted a net meta-analysis on the injection dose of MSCs, and the results showed that the clinical efficacy of the high-dose group was better than that of the low-dose group, but there was an increase in the number of adverse events, which is also a real problem [54].For the safety of MSCs, we performed a network meta-analysis of the adverse events associated with the injections, and the results showed that the safety of PLMSCs differed significantly from other treatment modalities, but only one study was included for PLMSCs, so more clinical studies may be needed to further confirm the safety of PLMSCs.There was no statistically significant difference in AE between different sources of MSCs, suggesting that there may be no difference in safety compared to other therapeutic modalities or between different sources of MSCs, and therefore, safety should be recognized.Overall, our study conducted a meta-analysis of MSCs from various sources.Although the comparisons among the included MSC studies were indirect, which could be considered a limitation, we systematically compared long-term and shortterm outcome metrics based on the follow-up duration.We designated 6 months as the timeframe for short-term efficacy evaluation and 12 months for long-term efficacy evaluation.This suggests that future efficacy studies should include longer follow-up periods.Our study had some limitations.MSCs play a crucial role in cartilage regeneration and repair.However, owing to the lack of quantitative evaluation methods for cartilage assessment, we did not analyze the cartilage.Second, the lack of clinical research on synovial MSCs and other clinical studies is another reason they were not included in this study.Lastly, no recommended dosage has been established for MSC injections in different studies; thus, we used the lowest dosage in this study.More clinical trials are needed to evaluate the dosage, clinical efficacy, and safety of MSCs from other sources in patients with OA.Therefore, more in-depth research is needed for the clinical application of MSCs, not only in terms of clinical efficacy but also in terms of the ethical issues involved in the collection of cells, which may include the informed consent of the donor, the distribution of benefits in the case of commercial utilization, the fairness of the subjects, and the privacy of the subjects.

Conclusion
In conclusion, we used a network meta-analysis to evaluate the short-term and long-term efficacy differences between different sources of MSCs (ADMSCs, BMMSCs, and UCMSCs) in patients with KOA.Our findings suggest that the traditional meta-analysis showed that MSCS from different sources were able to improve KOA compared to the control group.However, the difference in efficacy between the various MSCS was not significant in the network meta-analysis.Compared with MSCs from other sources, BMMSCs may improve the short-term and long-term knee function of KOA patients more significantly, while UCMSCs and ADMSCs are relatively weak.In addition, UCMSCs and BMMSCs may be superior to ADMSCs in shortterm and long-term pain relief in KOA patients.Our findings demonstrate the potential benefits of MSCs, which will lead to better recommendations for clinical and scientific research in the treatment of OA and other diseases with MSCs and may provide good suggestions for more in-depth research on MSCs in the future.However, owing to the limitations of this study, more accurate efficacy differences between MSCs from different sources, larger clinical samples, and long-term follow-up data are required.
sequence generation (selection bias) Allocation concealment (selection bias) Blinding of participants and personnel (performance bias) Blinding of outcome assessment (detection bias) Incomplete outcome data (attrition bias) Selective reporting (reporting bias) risk of bias Unclear risk of bias

FIGURE 2 :
FIGURE 2: Risk of bias summary of RCTs.RCT, randomized controlled experiment.

2
Stem Cells International 2.4.Data Extraction and Literature Quality Evaluation.Two researchers (DQX and WX) extracted information from the included studies, including the publication authors, publication year, sample size, average age, intervention measures, outcome parameters, and follow-up time.For the efficacy indicators, outcome indices at 6 and 12 months of follow-up, including WOMAC total, WOMAC pain, WOMAC stiffness, WOMAC functional limitation, and VAS score.Data not included in the original document were extracted using specialized extraction tools.

TABLE 1 :
Basic information of the included studies.