International Orthopaedics (SICOT) DOI 10.1007/s00264-015-2785-8
REVIEW ARTICLE
Efficacy of mesenchymal stem cells injection for the management of knee osteoarthritis: a systematic review and meta-analysis Peng Xia 1 & Xiaoju Wang 1 & Qiang Lin 1 & Xueping Li 1
Received: 22 March 2015 / Accepted: 29 March 2015 # SICOT aisbl 2015
Abstract Objective The purpose of this study was to access the efficacy of mesenchymal stem cells (MSCs) injection in the treatment of knee osteoarthritis (OA). Methods Studies were identified from databases (Pubmed, Embase, Cochrane Library, Biosis Previews, ClincalTrials.gov, CBMdisc) searched to December 2014 using a battery of keywords. We included randomized controlled and controlled clinical trials of people with knee OA comparing the outcomes of pain and function for those receiving MSCs injection with those receiving no MSCs injection. Two reviewers independently selected studies, extracted relevant data and assessed study quality. Data were pooled and meta-analyses were performed. Results Seven randomized controlled and controlled clinical trials, studying a total of 314 participants with a diagnosis of knee OA were included. Overall, MSCs injection has no significant effect on pain [weighted mean difference (WMD) (95 % confidence interval (CI)) [−1.33(−3.08, 0.41), P=0.13], and tends to improve self-reported physical function [standardized mean difference (SMD) (CI)=2.35(0.92, 3.77), P=0.001] at the last follow-up. But results from two high quality trials (94 patients) show a positive effect of MSCs injection on pain [WMD(CI)=−0.49 (−0.79, −0.19), P=0.001]. Heterogeneity observed between studies regarding the effect of MSCs injection on pain and function was explained by the difference of follow-up time, outcome measures, control group, the source and dose of MSCs. The Peng Xia and Xiaoju Wang contributed equally to this work. * Xueping Li [email protected] 1
Department of Rehabilitation Medicine, Nanjing First Hospital, Nanjing Medical University, Nanjing 210006, China
quality of evidence supporting these effect estimates was rated as low. Conclusion MSCs injection could be potentially efficacious for decreasing pain and may improve physical function in patients with knee OA. The findings of this review should be confirmed using methodologically rigorous and adequately powered clinical trials. Keywords Mesenchymal stem cells (MSCs) . Osteoarthritis (OA) . Knee . Meta-analysis
Introduction Osteoarthritis (OA) is the most common clinical chronic joint disease with cartilage degeneration, damage and bone hyperplasia. Knee OA has an estimated prevalence between 12 % and 35 % in the general population and is considered the leading cause of musculoskeletal disability in the elderly population worldwide [1–3]. Current treatment options for articular cartilage injury and OA itself aim to relieve inflammation and pain, but they do little to delay disease progression [4]. The American College of Rheumatology (ACR) guidelines for the treatment of knee OA include non-pharmacological and pharmacological therapies [5]. Non-pharmacological approaches such as exercise and lifestyle modification are often associated with poor compliance [6]. Pharmacological therapies including analgesics, non-steroid and steroid antiinflammatory drugs and corticosteroid injections provide only temporary benefit and often have side effects [7–9]. Mesenchymal stem cells (MSCs) have attracted attention for clinical use because of their multilineage potential, limited immunogenicity, immunosuppressive activities, and relative ease of growth in culture. MSCs are found in all human tissues and have the capacity to differentiate into many different cell
International Orthopaedics (SICOT)
types including bone and cartilage [10]. MSCs can be isolated from bone marrow aspirate or adipose tissue and are used for autologous injection in knee joint as a form of treatment for chronic injury or degenerative conditions [11, 12]. Therefore, MSCs have been suggested for use in the cell-based treatment of OA. Because several case series have shown favourable results of MSCs injections in patients with cartilage damage and knee OA [13–15], an update on the current evidence is required. Due to these previous studies, we performed a systematic search of the literature to assess the efficacy of MSCs injection as a method of improving pain and physical function in patients with knee OA.
Methods Literature search The electronic databases of Pubmed, Embase, Cochrane library, Biosis Previews, and China Biology Medicine disc (CBMdisc) were searched through the combination of a series of logic keywords and text words related to OA, interventions of interest, randomized controlled trials (RCT) and controlled clinical trials (CCT) (Appendix 1). The most recent electronic search was conducted in December 2014. The reference lists of retrieved articles and reviews were identified. In addition, the following website was searched manually to retrieve unpublished and ongoing studies: ClinicalTrials.gov (http://www.clinicaltrials.gov/).
Outcome measures The primary goal of this study was to identify the effectiveness of pain management and function improvement by applying MSCs injection therapy. For pain, we identified pain on a visual analogue scale (VAS). In order to standardize the pain outcomes of different studies, all pain scales were converted into a scale of 0–10 and the pain scores were recalculated. If a study reported multiple pain scales, the highest one on the hierarchy of pain scale-related outcomes was selected, as described by Jüni et al. [17]. The Lysholm score, Lequesne index, osteoarthritis outcome score (OAOS) and Western Ontario and McMaster Universities osteoarthritis index (WOMAC) were the preferred measures for function in a meta-analysis. If a study reported outcomes at multiple time points after treatment, only the data from the final follow-up time point was extracted for analysis. The measures of the relative treatment effect were the change pain and function scores at follow-up time point. Statistical analyses The results of the studies were analysed using RevMan 5.2. If the data were sufficiently homogeneous (clinical and
Study selection and eligibility criteria First, two researchers £(PX and XW) reviewed all the retrieved abstracts and full texts independently. If any disagreement was raised, it would be resolved by discussion and consultation with another researcher. Those papers meeting the following criteria were included in the analysis: (1) RCTs or CCTs; (2) studies in patients with knee OA; (3) studies reporting the pain and physical function outcome of patients. The trials with additional modality (e.g., education or exercise) unbalanced between the experimental group and the control group were excluded. Quality assessment Two researchers (QL and XL) evaluated the methodological quality of the included studies separately. The modified Jadad score [16], a scale ranging from 0 to 7 according to the descriptions of randomization, concealment allocation, blinding method and reporting of participant withdrawals, was used to measure the methodological quality of all studies. The scores range from 1 to 3 was low quality and 5 to 7 was high quality.
Fig. 1 Flowchart of study selection
RCT
CCT
RCT
RCT
RCT
RCT
Varma et al. [27]
Koh et al. [29]
Koh et al. [30]
Wong et al. [28]
Tan et al. [33]
Gan et al. [32]
Bone marrow
Cold blood
Bone marrow
Bone marrow
Adipose tissue
Sample size of subjects whose data were included in the present meta analysis
30 patients with knee OA grade II, III and IV of Kellgren and Lawrence
12 patients with OA meet the knee OA classification standards of the American College of Rheumatology
56 patients with medialcompartment OA and genu varum, aged younger than 55 years, diagnosed either arthroscopically and /or radiologically, normal lateral joint space, no fixed flexion deformity of the knee, and no collateral ligament instability 72 patients with OA were aged younger than 70 years, Kellgren-Law scale grade I and II
44 patients were aged younger than 60 years; radiographs showing grade III or lower Kellgren-Lawrence symptomatic isolated medial knee compartment OA
50 patients were aged ≥ 30 years Adipose tissue with idiopathic or secondary knee OA
na =25 After arthroscopic debridement, 2 ml MSCs were injected into the knee joint n=25, 8 male, 17 female, age: 54.2±9.3 years After arthroscopic debridement, MSCs plus 3 ml PRP were injected into the knee joint n=21, 5 male, 16 female, age: 54.2±2.9 years, BMI: 25.7±2.9 kg/m² After arthroscopic surgery, MSCs plus 3 ml PRP were injected into the knee joint, then HTO was performed n=28, 15 male, 13 female, median age: 53 (range, 36–54) years, BMI: 23.81±2.17 After arthroscopic surgery followed by HTO and postoperative rehabilitation program, MSCs plus 2 ml HA were injected into the knee joint n=36, 10 male, 26 female, age: 53.37±6.94 years After arthroscopic debridement, 2 ml MSCs were injected into the knee joint n=6, 1 male, 5 female, age: 56.27±7.52 years, course:(6.21±2.34) years 3 ml MSCs were injected into the knee joint n=15, 8 male, 7 female, age: 56.60±9.57 years, 40 million MSCs were injected into the knee joint
Source of MSCs Study Group
50 patients with mild to moderate Bone marrow OA of knee
Participants
n=6, 2 male, 4 female, age: 55.96±6.93 years, course: (5.97±2.65) years 25 mg HA were injected into the knee joint n=15, 5 male, 10 female, age: 57.33±9.41 years 60 mg of HA were injected into the knee joint
n=36, 9 male, 27 female, age: 53.76±5.68 Arthroscopic debridement alone
n=25, 8 male, 17 female, age: 54.4±11.3 years After arthroscopic debridement, 3 ml PRP injected into the knee joint alone n=23, 6 male, 17 female, age: 52.3±4.9 years, BMI: 24.7±3.3 kg/m² After arthroscopic surgery, 3 ml PRP were injected into the knee joint alone, then HTO was performed n=28, 14 male, 14 female, median age: 49 (range, 24– 54), BMI: 23.89±3.20 After HTO and postoperative rehabilitation program, 2 ml HA were injected into the knee joint alone
n=25 Arthroscopic debridement alone
Control group
1. VAS score 2. Lequesne index 3. WOMAC 4. SF-12 Scores 5. MRI
1. Lequesne index 2. WOMAC
1. Lequene index 2. MRI
1. Lysholm score 2. Tegner activity scale 3. IKDC 4. MOKART scoring system
1. VAS score 2. Lysholm score 3. KOOS
1. VAS score 2. Lysholm score 3. Tegner activity scale
1. VAS score 2. OAOS
Outcome measures
3, 6, 12 months
3, 12 months
7, 12 months
6, 12, 24 months
3, 12, 24–25 months
12–18 months
1, 2, 3, 6 months
Follow-up
Low
Low
Low
High
High
High
Low
Quality
RCT randomized controlled trial, CCT controlled clinical trial, MSCs mesenchymal stem cells, VAS visual analogue scale, OAOS osteoarthritis outcome score, KOOS knee injury and osteoarthritis outcome score, HTO high tibial osteotomy, PRP platelet-rich plasma, HA hyaluronic acid, MRI magnetic resonance imaging
a
Orozco et al. RCT [31]
Design
Characteristics of included studies
Source
Table 1
International Orthopaedics (SICOT)
International Orthopaedics (SICOT)
statistical), we summarised these in a meta-analysis. Continuous outcomes were calculated and expressed as the weighted mean difference (WMD) (MD in RevMan 5.2) or as the standardised MD (SMD) with 95 % confidence intervals (CI) depending on the similarity of the scales used. To measure heterogeneity between studies, we used the χ2 (p value less than 0.10 indicates heterogeneity) and I2 statistic (a value of less than 40 % represents low heterogeneity and a value of 75 % or more indicates high heterogeneity) [18]. The outcomes were pooled using random-effects models. All p-values
REVIEW ARTICLE
Efficacy of mesenchymal stem cells injection for the management of knee osteoarthritis: a systematic review and meta-analysis Peng Xia 1 & Xiaoju Wang 1 & Qiang Lin 1 & Xueping Li 1
Received: 22 March 2015 / Accepted: 29 March 2015 # SICOT aisbl 2015
Abstract Objective The purpose of this study was to access the efficacy of mesenchymal stem cells (MSCs) injection in the treatment of knee osteoarthritis (OA). Methods Studies were identified from databases (Pubmed, Embase, Cochrane Library, Biosis Previews, ClincalTrials.gov, CBMdisc) searched to December 2014 using a battery of keywords. We included randomized controlled and controlled clinical trials of people with knee OA comparing the outcomes of pain and function for those receiving MSCs injection with those receiving no MSCs injection. Two reviewers independently selected studies, extracted relevant data and assessed study quality. Data were pooled and meta-analyses were performed. Results Seven randomized controlled and controlled clinical trials, studying a total of 314 participants with a diagnosis of knee OA were included. Overall, MSCs injection has no significant effect on pain [weighted mean difference (WMD) (95 % confidence interval (CI)) [−1.33(−3.08, 0.41), P=0.13], and tends to improve self-reported physical function [standardized mean difference (SMD) (CI)=2.35(0.92, 3.77), P=0.001] at the last follow-up. But results from two high quality trials (94 patients) show a positive effect of MSCs injection on pain [WMD(CI)=−0.49 (−0.79, −0.19), P=0.001]. Heterogeneity observed between studies regarding the effect of MSCs injection on pain and function was explained by the difference of follow-up time, outcome measures, control group, the source and dose of MSCs. The Peng Xia and Xiaoju Wang contributed equally to this work. * Xueping Li [email protected] 1
Department of Rehabilitation Medicine, Nanjing First Hospital, Nanjing Medical University, Nanjing 210006, China
quality of evidence supporting these effect estimates was rated as low. Conclusion MSCs injection could be potentially efficacious for decreasing pain and may improve physical function in patients with knee OA. The findings of this review should be confirmed using methodologically rigorous and adequately powered clinical trials. Keywords Mesenchymal stem cells (MSCs) . Osteoarthritis (OA) . Knee . Meta-analysis
Introduction Osteoarthritis (OA) is the most common clinical chronic joint disease with cartilage degeneration, damage and bone hyperplasia. Knee OA has an estimated prevalence between 12 % and 35 % in the general population and is considered the leading cause of musculoskeletal disability in the elderly population worldwide [1–3]. Current treatment options for articular cartilage injury and OA itself aim to relieve inflammation and pain, but they do little to delay disease progression [4]. The American College of Rheumatology (ACR) guidelines for the treatment of knee OA include non-pharmacological and pharmacological therapies [5]. Non-pharmacological approaches such as exercise and lifestyle modification are often associated with poor compliance [6]. Pharmacological therapies including analgesics, non-steroid and steroid antiinflammatory drugs and corticosteroid injections provide only temporary benefit and often have side effects [7–9]. Mesenchymal stem cells (MSCs) have attracted attention for clinical use because of their multilineage potential, limited immunogenicity, immunosuppressive activities, and relative ease of growth in culture. MSCs are found in all human tissues and have the capacity to differentiate into many different cell
International Orthopaedics (SICOT)
types including bone and cartilage [10]. MSCs can be isolated from bone marrow aspirate or adipose tissue and are used for autologous injection in knee joint as a form of treatment for chronic injury or degenerative conditions [11, 12]. Therefore, MSCs have been suggested for use in the cell-based treatment of OA. Because several case series have shown favourable results of MSCs injections in patients with cartilage damage and knee OA [13–15], an update on the current evidence is required. Due to these previous studies, we performed a systematic search of the literature to assess the efficacy of MSCs injection as a method of improving pain and physical function in patients with knee OA.
Methods Literature search The electronic databases of Pubmed, Embase, Cochrane library, Biosis Previews, and China Biology Medicine disc (CBMdisc) were searched through the combination of a series of logic keywords and text words related to OA, interventions of interest, randomized controlled trials (RCT) and controlled clinical trials (CCT) (Appendix 1). The most recent electronic search was conducted in December 2014. The reference lists of retrieved articles and reviews were identified. In addition, the following website was searched manually to retrieve unpublished and ongoing studies: ClinicalTrials.gov (http://www.clinicaltrials.gov/).
Outcome measures The primary goal of this study was to identify the effectiveness of pain management and function improvement by applying MSCs injection therapy. For pain, we identified pain on a visual analogue scale (VAS). In order to standardize the pain outcomes of different studies, all pain scales were converted into a scale of 0–10 and the pain scores were recalculated. If a study reported multiple pain scales, the highest one on the hierarchy of pain scale-related outcomes was selected, as described by Jüni et al. [17]. The Lysholm score, Lequesne index, osteoarthritis outcome score (OAOS) and Western Ontario and McMaster Universities osteoarthritis index (WOMAC) were the preferred measures for function in a meta-analysis. If a study reported outcomes at multiple time points after treatment, only the data from the final follow-up time point was extracted for analysis. The measures of the relative treatment effect were the change pain and function scores at follow-up time point. Statistical analyses The results of the studies were analysed using RevMan 5.2. If the data were sufficiently homogeneous (clinical and
Study selection and eligibility criteria First, two researchers £(PX and XW) reviewed all the retrieved abstracts and full texts independently. If any disagreement was raised, it would be resolved by discussion and consultation with another researcher. Those papers meeting the following criteria were included in the analysis: (1) RCTs or CCTs; (2) studies in patients with knee OA; (3) studies reporting the pain and physical function outcome of patients. The trials with additional modality (e.g., education or exercise) unbalanced between the experimental group and the control group were excluded. Quality assessment Two researchers (QL and XL) evaluated the methodological quality of the included studies separately. The modified Jadad score [16], a scale ranging from 0 to 7 according to the descriptions of randomization, concealment allocation, blinding method and reporting of participant withdrawals, was used to measure the methodological quality of all studies. The scores range from 1 to 3 was low quality and 5 to 7 was high quality.
Fig. 1 Flowchart of study selection
RCT
CCT
RCT
RCT
RCT
RCT
Varma et al. [27]
Koh et al. [29]
Koh et al. [30]
Wong et al. [28]
Tan et al. [33]
Gan et al. [32]
Bone marrow
Cold blood
Bone marrow
Bone marrow
Adipose tissue
Sample size of subjects whose data were included in the present meta analysis
30 patients with knee OA grade II, III and IV of Kellgren and Lawrence
12 patients with OA meet the knee OA classification standards of the American College of Rheumatology
56 patients with medialcompartment OA and genu varum, aged younger than 55 years, diagnosed either arthroscopically and /or radiologically, normal lateral joint space, no fixed flexion deformity of the knee, and no collateral ligament instability 72 patients with OA were aged younger than 70 years, Kellgren-Law scale grade I and II
44 patients were aged younger than 60 years; radiographs showing grade III or lower Kellgren-Lawrence symptomatic isolated medial knee compartment OA
50 patients were aged ≥ 30 years Adipose tissue with idiopathic or secondary knee OA
na =25 After arthroscopic debridement, 2 ml MSCs were injected into the knee joint n=25, 8 male, 17 female, age: 54.2±9.3 years After arthroscopic debridement, MSCs plus 3 ml PRP were injected into the knee joint n=21, 5 male, 16 female, age: 54.2±2.9 years, BMI: 25.7±2.9 kg/m² After arthroscopic surgery, MSCs plus 3 ml PRP were injected into the knee joint, then HTO was performed n=28, 15 male, 13 female, median age: 53 (range, 36–54) years, BMI: 23.81±2.17 After arthroscopic surgery followed by HTO and postoperative rehabilitation program, MSCs plus 2 ml HA were injected into the knee joint n=36, 10 male, 26 female, age: 53.37±6.94 years After arthroscopic debridement, 2 ml MSCs were injected into the knee joint n=6, 1 male, 5 female, age: 56.27±7.52 years, course:(6.21±2.34) years 3 ml MSCs were injected into the knee joint n=15, 8 male, 7 female, age: 56.60±9.57 years, 40 million MSCs were injected into the knee joint
Source of MSCs Study Group
50 patients with mild to moderate Bone marrow OA of knee
Participants
n=6, 2 male, 4 female, age: 55.96±6.93 years, course: (5.97±2.65) years 25 mg HA were injected into the knee joint n=15, 5 male, 10 female, age: 57.33±9.41 years 60 mg of HA were injected into the knee joint
n=36, 9 male, 27 female, age: 53.76±5.68 Arthroscopic debridement alone
n=25, 8 male, 17 female, age: 54.4±11.3 years After arthroscopic debridement, 3 ml PRP injected into the knee joint alone n=23, 6 male, 17 female, age: 52.3±4.9 years, BMI: 24.7±3.3 kg/m² After arthroscopic surgery, 3 ml PRP were injected into the knee joint alone, then HTO was performed n=28, 14 male, 14 female, median age: 49 (range, 24– 54), BMI: 23.89±3.20 After HTO and postoperative rehabilitation program, 2 ml HA were injected into the knee joint alone
n=25 Arthroscopic debridement alone
Control group
1. VAS score 2. Lequesne index 3. WOMAC 4. SF-12 Scores 5. MRI
1. Lequesne index 2. WOMAC
1. Lequene index 2. MRI
1. Lysholm score 2. Tegner activity scale 3. IKDC 4. MOKART scoring system
1. VAS score 2. Lysholm score 3. KOOS
1. VAS score 2. Lysholm score 3. Tegner activity scale
1. VAS score 2. OAOS
Outcome measures
3, 6, 12 months
3, 12 months
7, 12 months
6, 12, 24 months
3, 12, 24–25 months
12–18 months
1, 2, 3, 6 months
Follow-up
Low
Low
Low
High
High
High
Low
Quality
RCT randomized controlled trial, CCT controlled clinical trial, MSCs mesenchymal stem cells, VAS visual analogue scale, OAOS osteoarthritis outcome score, KOOS knee injury and osteoarthritis outcome score, HTO high tibial osteotomy, PRP platelet-rich plasma, HA hyaluronic acid, MRI magnetic resonance imaging
a
Orozco et al. RCT [31]
Design
Characteristics of included studies
Source
Table 1
International Orthopaedics (SICOT)
International Orthopaedics (SICOT)
statistical), we summarised these in a meta-analysis. Continuous outcomes were calculated and expressed as the weighted mean difference (WMD) (MD in RevMan 5.2) or as the standardised MD (SMD) with 95 % confidence intervals (CI) depending on the similarity of the scales used. To measure heterogeneity between studies, we used the χ2 (p value less than 0.10 indicates heterogeneity) and I2 statistic (a value of less than 40 % represents low heterogeneity and a value of 75 % or more indicates high heterogeneity) [18]. The outcomes were pooled using random-effects models. All p-values
