Introduction
Methods
Mesenchymal stem cells
Mesenchymal stem cell-based cartilage regeneration from bench to bedside
In vitro studies
Ex vivo studies
In vivo animal studies
Scientific publication | Animal model | Simulated defect characteristics | Implanted/injected construct | Follow-up period | Key findings |
---|---|---|---|---|---|
Guo et al. (2004) [65] | 28 sheep | Medial femoral condyle osteochondral defects; cylindrical (8 mm diameter) | Implantation of isolated BM-derived MSCs seeded on a TCP scaffold; compared to cell-free scaffolds and empty defects | 6 months | Macroscopic: smooth, integrated tissue in MSC group. Histologic: proteoglycan and type II collagen consistent with hyaline cartilage in MSC group, compared with fibrocartilage in cell-free group; subchondral osseous regeneration. Biochemical: GAG quantity in MSC group was 89% of native cartilage |
Wayne et al. (2005) [66] | 10 dogs | Medial and lateral femoral condyle osteochondral defects; cylindrical (6 mm diameter) | Implantation of isolated BM-derived MSCs suspended in alginate and seeded on a PLA scaffold; precultivated for 3 wk; compared to cell-free scaffolds | 1.5 months | Macroscopic: improved coverage of defects in MSC group. Histologic: mixture of hyaline and fibrocartilage integrated with surrounding tissue; higher quality tissue in MSC group compared with cell-free group; no mineralization noted within osseous defects. Mechanical: lower resistance to compression than native cartilage |
Ando et al. (2007) [67] | 9 piglets | Medial femoral condyle chondral defects; cylindrical (8.5 mm diameter) | Implantation of isolated, allogeneic synovial tissue MSCs derived from piglets and cultured in a three-dimensional scaffold-free TEC; compared to empty defects | 6 months | Macroscopic: greater defect coverage in TEC group; subchondral erosion in the empty defects. Histologic: smooth, integrated tissue containing proteoglycans and type II collagen in the TEC group; empty defects showed signs of OA; higher ICRS scores in the TEC group. Mechanical: similar viscoelastic properties between TEC and native cartilage |
Lee et al. (2007) [68] | 27 mini-pigs | Medial femoral condyle chondral defects; cylindrical (8.5 mm diameter) | Injection of isolated BM-derived MSCs with HA (Synvisc) followed by HA weekly × 2 wk; compared to HA alone | 3 months | Macroscopic: greater defect coverage in the MSC + HA group. Histologic: hyaline-like cartilage noted in MSC + HA group; minimal defect filling in HA group; improvement in Wakitani histologic score with MSCs |
Saw et al. (2009) [69] | 15 goats | Femoral trochlea chondral defects; cylindrical (4 mm diameter) | Injection of BMDC collection with HA (Hyalgan) weekly for 3 wk starting 1 wk after subchondral drilling; compared to drilling with or without HA | 6 months | Macroscopic: greater defect coverage in the BMDC + HA group. Histologic: HA group had some proteoglycans and type II collagen mixed with type I collagen; BMDC + HA group had superior proteoglycan and type II collagen content; cell morphology was improved in the BMDC + HA group |
Zscharnack et al. (2010) [38] | 10 sheep | Medial femoral condyle osteochondral defects; cylindrical (7 mm diameter) | Implantation of isolated BM-derived MSCs in type I (rat) collagen gel either immediately following seeding or after 2 wk of precultivation | 6 months | Macroscopic: precultivation group produced more homogenous hyaline-like cartilage. Histologic: significantly better O’Driscoll and ICRS scores in the precultivation group compared with non-precultivated group, specifically with respect to surface features, integration, cell distribution, and mineralization. Mechanical: precultivated tissue was firm |
Shimomura et al. (2010) [70] | 7 pigs, 6 piglets | Medial femoral condyle chondral defects; cylindrical (8.5 mm diameter) | Implantation of isolated synovial tissue MSCs derived from piglets and cultured in a three-dimensional scaffold-free TEC; compared to empty defects | 6 months | Macroscopic: greater defect coverage in TEC group. Histologic: good integration of tissue that stained well for proteoglycans in the TEC group versus signs of OA in empty defects; higher ICRS scores in the TEC group. Mechanical: similar properties between TEC and native tissue |
Wegener et al. (2010) [71] | 9 sheep | Medial femoral condyle chondral defects; cylindrical (8 mm diameter) | Implantation of BM cells in fibrin glue seeded on a PGA scaffold; secured to subchondral bone by PLGA darts; compared to cell-free scaffolds | 3 months | Macroscopic: BM-seeded scaffolds had improved regeneration compared with cell-free scaffolds. Histologic: variation noted with fibrous tissue in some and hyaline-like cartilage in other BM cell-seeded scaffolds; O’Driscoll score was similar between cell-free and cell-seeded scaffolds |
Marquass et al. (2011) [72] | 9 sheep | Medial femoral condyle osteochondral defects; cylindrical (7 mm diameter) | Implantation of isolated BM-derived MSCs in type I (rat) collagen gel implanted either immediately following seeding or after 2 wk of precultivation; compared to MACI | 12 months | Macroscopic/histologic: significantly better O’Driscoll and ICRS scores with precultivated MSCs compared with both non-precultivated MSCs and MACI, specifically with respect to surface quality, matrix quality and integration; type II collagen content was superior in precultivated group. MRI: precultivated MSCs were similar to MACI but significantly better than non-precultivated MSCs on the MOCART score |
McIlwraith et al. (2011) [73] | 10 horses | Medial femoral condyle chondral defects (1 cm2) | Injection of isolated BM-derived MSCs with HA (Hyvisc) into the knee joint 1 month after MFX; compared to cell-free HA injection and MFX | 12 months | Macroscopic: greater repair tissue area with MSCs, but no difference in volume. Histologic: no difference in surface, structure, integration, cellular architecture, and subchondral regeneration; contradictory proteoglycan and aggrecan staining. Biochemical: equivalent GAG. Mechanical: tissue derived from MSCs was firmer. MRI: no difference |
Ando et al. (2012) [74] | 6 piglets | Medial femoral condyle chondral defects; cylindrical (8.5 mm diameter) | Implantation of isolated, allogeneic synovial MSCs and cultured in a three-dimensional scaffold-free TEC; compared to empty defects | 6 months | Histologic: tissue containing proteoglycans in the TEC group; empty defects were partially covered with fibrous tissue and showed signs of OA; higher O’Driscoll scores in the TEC group. Mechanical: similar properties between TEC and native cartilage |
Zhang et al. (2012) [75] | 20 mini-pigs | Femoral trochlea chondral defects; cylindrical (6 mm diameter) | Implantation of BMDCs or isolated, expanded BM-derived MSCs in type II collagen (porcine) hydrogel; compared to cell-free gels | 2 months | Macroscopic: good defect filling with both MSCs and BMDCs; irregularity with cell-free gels. Histologic: hyaline-like cartilage with both MSCs and BMDCs; O’Driscoll score was greater in the MSC group at 4 wk, but equivalent between the BMDC and MSC groups at 8 wk |
Bekkers et al. (2013) [76] | 8 goats | Medial femoral condyle chondral defects; cylindrical (5 mm diameter) | Implantation of chondrons and BM-derived MSCs suspended in fibrin glue; compared to MFX | 6 months | Macroscopic: improved defect filling with MSC + chondrons in comparison to MFX. Histologic: O’Driscoll score was significantly higher in the MSC + chondron group. Biochemical: GAG content and GAG/DNA in the repair tissue was greater in the MSC + chondron group than the MFX group |
Kamei et al. (2013) [77] | 16 mini-pigs | Patella chondral defects; cylindrical (6 mm diameter) | Magnetic accumulation of injected ferumoxide labeled MSCs; compared to gravity-focused MSCs | 3 months | Arthroscopic: improved smoothness and integration with magnetic accumulation. Histologic: superior integration and type II collagen content with magnetic accumulation; improved scoring on the Wakitani scale |
Nam et al. (2013) [78] | 18 goats | Medial femoral condyle chondral defects; cylindrical (5 mm diameter) | Injection of isolated BM-derived MSCs weekly (×3 wk) starting 2 wk after subchondral drilling; compared to drilling alone | 6 months | Macroscopic: smooth, integrated tissue with MSCs versus partial, irregular filling with drilling alone. Histologic: O’Driscoll score was significantly higher in the MSC group; improved proteoglycan and type II collagen content with MSCs. Biochemical: higher GAG quantity with MSCs |
Clinical studies
Scientific publication | Study type | Subject details | Defect characteristics | Implanted/injected construct | Follow-up period | Key findings |
---|---|---|---|---|---|---|
Kuroda et al. (2007) [82] | Case report: level IV evidence | 1 M (age 31 y) | 1 medial femoral condyle chondral defect (6.0 cm2) from trauma | Implantation of isolated BM-derived MSCs within porcine type I collagen gel on a collagen scaffold; covered by a periosteal flap | 12 months | Arthroscopic: firm, smooth repair tissue. Histologic: hyaline-like cartilage covered superficially by fibrous tissue. MRI: focal chondral and subchondral irregularities. Clinical: return to previous level of activity |
Wakitani et al. (2007) [88] | Case series: level IV evidence | 3: 2 M, 1 F (age 32-45 y) | 5 femoral trochlea (0.7-4.2 cm2) and 4 patella chondral defects (1.0-1.7 cm2); defects in 2/3 participants from trauma | Implantation of isolated BM-derived MSCs within bovine type I collagen gel on a porcine collagen scaffold; covered by a periosteal flap or synovium; adjunctive subchondral drilling | 18 months | Arthroscopic: firm, smooth tissue. Histologic: atypical cartilage. MRI: complete coverage of defects but quality unclear. Clinical: improvement of symptoms and return to work; IKDC improvement |
Giannini et al. (2009) [80] | Case series: level IV evidence | 48: 27 M, 21 F (mean age 28.5 ± 9.5 y) | 48 talar dome osteochondral defects (2.07 ± 0.48 cm2); 35 from trauma; previous MFX, debridement or ACI in 15 | Implantation of BMDCs suspended within collagen/platelet paste or seeded on HA (Hyaff-11) scaffold | 24-35 months | Arthroscopic: smooth tissue in some, hypertrophic in others; all integrated with firmness of native cartilage. Histologic: mixed with some hyaline quality. MRI: newly formed tissue in all lesions. Clinical: improvement in AOFAS scores with time and return to sports with no difference between scaffold types; worse outcomes with previous surgery |
Buda et al. (2010) [81] | Case series: level IV evidence | 20: 12 M, 8 F (mean age 28.5 ± 9.5 y) | 16 medial femoral condyle and 6 lateral condyle osteochondral defects (no area provided); 18 traumatic and 2 OCD defects | Implantation of BMDCs seeded on a HA (Hyalofast) scaffold supplemented with platelet-rich fibrin; adjunctive meniscus repair or debridement, ACL-R, or HTO | 29 ± 4.1 months | Histologic: collagen II noted throughout repair tissue with focal proteoglycan content consistent with hyaline-like cartilage. MRI: variable signal intensity that correlated with KOOS score. Clinical: improvement in IKDC and KOOS scores post-operatively |
Giannini et al. (2010) [90] | Prospective comparative study: level III evidence | 81: 47 M, 34 F (mean age 30 ± 8 y); 25 BMDC; 10 ACI; 46 MACI | 81 talar dome osteochondral defects (>1.5 cm2) from trauma | Implantation of BMDCs seeded on a HA (Hyaff-11) scaffold supplemented with platelet-rich fibrin | 59.5 ± 26.5 months | Arthroscopic: good defect coverage. Histologic: hyaline-like cartilage noted. MRI: complete integration in 76% and homogenous tissue in 82% of all cases with hypertrophy in 3 BMDC and 2 ACI patients. Clinical: improvement in AOFAS scores after surgery with no difference between BMDC-scaffold implants, ACI and MACI; lower overall cost for BMDC transplantation compared to ACI/MACI |
Haleem et al. (2010) [83] | Case series: level IV evidence | 5: 4 M, 1 F (mean age 25.4 y) | 5 femoral condyle chondral defects (3-12 cm2); 2 traumatic, and 3 OCD defects (1 OA from neglected OCD) | Implantation of isolated BM-derived MSCs within platelet-rich fibrin glue; covered by a periosteal flap | 12 months | Arthroscopic: smooth tissue. MRI: complete defect filling with good congruity in 3/5 patients. Clinical: improvement in Lysholm and RHSSK scores with return to sports; worse outcomes in 1 patient with pre-operative OA |
Nejadnik et al. (2010) [84] | Prospective comparative study: level III evidence | 72: 38 M, 34 F (mean age 44.0 ± 11.4 y), 36 MSCs; 36 ACI | 13 patella, 4 femoral trochlea, 12 femoral condyle, and 7 multiple knee chondral defects (4.6 ± 3.5 cm2); 14 traumatic, 20 OA and 2 other defects | Implantation of isolated BM-derived MSCs; covered by a periosteal flap; adjunctive partial meniscectomy, patellar realignment, ACL-R, or HTO | 24 months | Arthroscopic: smooth tissue in most cases. Histologic: aggrecan and collagen II content consistent with hyaline cartilage. Clinical: greater improvement in SF-36 Physical Role Functioning in MSCs versus chondrocytes; equivalent IKDC, Tegner and Lysholm score improvement following both MSC and chondrocyte transplantation; superior outcomes in males versus females |
Gobbi et al. (2011) [89] | Case series: level IV evidence | 15: 10 M, 5 F (mean age 48 y, range 32-58 y) | 7 patella, 6 femoral trochlea, 4 medial tibial plateau, 6 medial femoral condyle, and 1 lateral condyle chondral defects (9.2 ± 6.3 cm2); all defects from trauma; 6 patients had multiple defects | Implantation of BMDCs mixed with batroxobin (Plateltex Act) to produce a clot; covered by a type I/III collagen matrix (Chondro-Gide); adjunctive ACL-R, HTO, patellar realignment | 24-38 months | Arthroscopic: smooth, integrated tissue in all cases; no hypertrophy. Histologic: variability with properties of hyaline and fibrocartilage. MRI: complete defect filling in 80%, integration in 93%, and no hypertrophy in all patients. Clinical: improvement in all scores (VAS, KOOS, Tegner, Marx, IKDC and Lysholm) following surgery; patients with single lesions and smaller lesions had better outcomes |
Kasemkijwa-ttana et al. (2011) [85] | Case series: level IV evidence | 2 M (age 24-25 y) | 2 lateral femoral condyle chondral (2.2-2.5 cm2) | Implantation of isolated BM-derived MSCs seeded on a type I collagen scaffold supplemented with fibrin glue; covered by a periosteal flap; adjunctive ACL-R, meniscal repair | 30-31 months | Arthroscopic: good defect fill, integration and firmness. Clinical: significant improvement in IKDC score and KOOS post-operatively |
Saw et al. (2011) [98] | Case series: level IV evidence | 5: 1 M, 4 F (mean age 39.4 y, range 19-52 y) | 3 focal defects: 1 lateral femoral condyle (2 cm2), 1 patella (8.8 cm2), 1 femoral trochlea (0.5 cm2); 2 OA defects | Injection of peripheral blood-derived MSCs with HA weekly (×5) starting 1 wk after subchondral drilling; adjunctive HTO or lateral patellar release; pre-injection GCSF | 10-26 months | Arthroscopic: good filling in focal defects; range from devoid areas to smooth repair tissue in OA defects. Histologic: intense proteoglycan staining; type I collagen in superficial area with predominance to type II collagen in deep area; chondrocytes in subchondral drill holes |
Gigante et al. (2012) [86] | Case report: level IV evidence | 1 M (age 37 y) | 1 medial femoral condyle chondral defect (3 cm2) from trauma | Implantation of BMDCs within fibrin glue (Tisseel) and coverage with a collagen membrane (MeRG) after arthroscopic MFX (CMBMC) | 24 months | MRI: good defect filling with tissue that was isointense relative to native cartilage; no signs of bone edema. Clinical: return to activity and asymptomatic |
Enea et al. (2013) [87] | Case series: level IV evidence | 9: 5 M, 4 F (mean age 48 ± 9 y) | 6 medial femoral condyle and 3 lateral condyle chondral defects (2.6 ± 0.5 cm2); previous meniscectomy, debridement or ACL-R | Implantation of BMDCs within fibrin glue and coverage with a PGA-HA membrane (Chondro-tissue) after arthroscopic MFX (CMBMC); adjunctive meniscectomy, osteochondral fixation, or trochlea resurfacing | 22 ± 2 months | Arthroscopic: 1 normal, 3 nearly normal and 1 abnormal on ICRS CRA. Histologic: hyaline-like cartilage repair tissue. MRI: complete defect filling in all; mild subchondral irregularities in all; hypertrophy in 1 patient. Clinical: improvement in IKDC and Lysholm scores compared with pre-operative scores; no change in Tegner score from pre-injury; one failure |
Giannini et al. (2013) [91] | Case series: level IV evidence | 49: 27 M, 22 F (mean age 28.1 ± 9.5 y) | 49 talar dome osteochondral defects (2.2 ± 1.2 cm2); 36 traumatic defects with unknown etiology in others; previous debridement, MFX, ACI, or BMDCs in 17 | Implantation of BMDCs within collagen/platelet paste or seeded on HA (Hyaff-11) scaffold supplemented with platelet gel | 48 months | MRI: complete defect filling in 45%, hypertrophy in 45%, integration in 65%, subchondral disruption in 65% of cases; 78% of repair area had hyaline quality. Clinical: improvement in AOFAS scores - maximal value at 24 months; decreased at 36-48 months; decreased AOFAS associated with fibrocartilage quality; return to pre-injury sports in 78% |
Saw et al. (2013) [29] | RCT: level II evidence | 49: 17 M, 32 F (mean age 38 ± 7 y); 25 MSC + HA; 24 HA | 49 chondral defects of the knee (57% patella, 29% trochlea, 12% femoral condyle, and 8% tibial plateau) | Injection of peripheral blood-derived MSCs and HA weekly (×5) starting 1 wk after subchondral drilling and then weekly (×3) at 6 months; pre-injection GCSF | 24 months | Arthroscopic: smooth defect filling. Histologic: ICRS II score was significantly better in MSC + HA group. MRI: improved cartilage morphology, defect filling and integration in MSC + HA group. Clinical: improvement in IKDC scores with no difference between MSC + HA and HA |
Construct | Transplantation protocol | Advantages | Disadvantages |
---|---|---|---|
Bone marrow aspiration, separation of nucleated cell population (BMDCs) by centrifugation, scaffold seeding, and implantation of BMDC-scaffold construct into the AC defect site | Accessory cells/GFs create a natural microenvironment | Low number of MSCs | |
One step procedure with aspiration and implantation in the same surgery | Cells other than MSCs could promote immunorejection in allogeneic transplantation | ||
Bone marrow aspiration, in vitro MSC isolation by adherence to plastic flasks, in vitro expansion of MSCs, scaffold seeding with MSCs, and implantation of MSC-scaffold construct into the AC defect site | High MSC numbers are available due to expansion | In vitro expansion may increase the risk of contamination | |
Isolation allows for purification of MSCs and potentially reduced likelihood of rejection in allogeneic transplant | MSCs have the capacity to become bone without in vitro cueing prior to implantation (bone may be beneficial in osteochondral lesions) | ||
Mid-range time consumption | |||
Bone marrow aspiration, MSC isolation by adherence to plastic flasks, expansion of MSCs in vitro, scaffold seeding with MSCs, in vitro precultivation in medium promoting chondrogenesis, and implantation of a cartilage tissue construct into the AC defect site | High MSC numbers are available due to expansion | In vitro expansion and cultivation may increase the risk of contamination | |
Chondrogenesis is stimulated | Highest time and resource consumption | ||
Increased mechanical stability of the implanted construct | No clinical assessment to date | ||
Early neo-tissue remodeling occurs in vitro and may be accounted for at the time of implantation |
Optimizing technique in mesenchymal stem cell transplantation
Collection of mesenchymal stem cell-containing tissue
MSC source | In vivoassessment of focal AC defect treatment | Advantages | Disadvantages |
---|---|---|---|
Bone marrow
| Most rigorous investigation and strongest supporting evidence | Propensity to form osseous tissue (could be beneficial for osseous regeneration in osteochondral lesions) | |
Ease of collection by needle | |||
Long-term safety reported | |||
Peripheral blood
| Ease of collection by needle | Paucity of literature comparing this source to others | |
Synovial tissue
| Greatest chondrogenic capacity noted based on in vitro study | Clinical assessment is lacking | |
Periosteum
| Equivalent chondrogenic capacity to bone marrow | Propensity to form osseous tissue | |
Clinical assessment is lacking | |||
Adipose tissue
| Abundance of tissue | Reduced chondrogenic capacity | |
Widespread anatomic availability | Clinical assessment is lacking |