Stem Cell Considerations for the Clinician

Summary

While there is ample evidence that beneficial results can be obtained from the use of MSC, several questions regarding their use remain to be answered. A better definition is needed for the mechanism/s by which the MSC are eliciting the positive outcome of their use whether it lies in the paracrine factors that are made by the MSC that affect the innate or adaptive immunity or impact the metabolism of resident cells, or their contribution as precursor cells giving rise to more differentiated reparative cells. If paracrine effects are desirable, commercial sources of allogeneic MSC could provide an easily obtainable, well-characterized reagent for the clinician. In their function as reparative cells, the longevity and fate of allogeneic MSC needs to be further evaluated on longer-term basis. The use of autologous cells has both advantages and its own inherent disadvantages involving additional procedures for collection, preparation and characterization of the MSC. It is apparent that a critical number of MSC are necessary for an optimal outcome for implantation. It remains to be determined if tissues such as adipose tissue provide more MSC than bone marrow and if the capability for differentiation is equivalent. A better understanding is also needed for how individual patient factors such as age, sex, drug regimens and co-morbidities influence the MSC population. For many of these questions, preclinical models will be helpful, but the task of evaluating and implementing these findings for orthopaedic patients falls onto the shoulders of clinical researchers. Evaluation of these questions is a daunting, but such a challenge fits the concept of personalized medicine in today’s medicine.

Introduction

Over the past 60 years, evidence has accumulated supporting the existence of a multipotent adult stem cell population in the body that has the potential to differentiate into bone, cartilage, tendon, ligament, adipocytes, dermis, muscle and connective tissue. These cells are now collectively grouped under the term mesenchymal stem cells (MSC) or multipotent mesenchymal stromal cells. A large proportion of the studies on MSC have involved the role of these cells in the development and repair of bone and cartilage heightening interest in the clinical orthopaedic community. In 1966, intraperitoneal diffusion chambers implanted with mouse bone marrow cells demonstrated that undifferentiated “stem” cells were present and resulted in osteogenic foci of cells producing alkaline phosphatase and fibroblasts while haematopoietic cells were lost (1). Interestingly, orthopaedic surgeons were already using viable cancellous bone chips containing these cells in fracture repair. In the first edition of Campbell’s Operative Orthopaedics (1939), a recommended treatment of non-unions included a combination of stable fixation and packing of cancellous bone chips from the proximal tibia (2). Even though the concept of stem cells as we know today was unknown at the time, early orthopaedic surgeons recognized the osteogenic effect of cancellous bone and bone marrow.

Today it is clear that stem cell supplementation offers a valuable tool for correcting some of the clinical challenges in treatment of musculoskeletal diseases and injury. Between 2006 and 2012, there was a 3-fold increase in the number of MSC product Investigational New Drug submissions to the FDA resulting in clinical trials initiated worldwide (246 trials; source: http://www.clinicaltrials.gov). Although much of the initial research focused on bone marrow-derived mesenchymal stem cells (bmMSC) with umbilical or placental sources serving as a secondary source, an increasing trend of adipose-derived MSC-based product INDs has occurred since 2011 (3). Many of these new IND deal with MSC destined for allogeneic use with over 80% using cryopreservation for storage of the MSC products to facilitate transport to the clinical site where they are used. Cell banking of cultured MSC (35%) in these endeavors has also been denoted; however, one report showed reduced immunomodulatory function of thawed cryopreserved MSC immediately after thawing that was recovered after subsequent in vitro culture (4). The bioactivity of these products in the IND is variably described with molecular markers such as “secreted factors or expression of proteins on the surface of either the MSC or target cells (e.g., T cells) that may be related to a given biological activity ” (3).

Mesenchymal stem cell definition

What is the definition of this stem cell population and what are the differences between stem cell populations isolated from different tissues (bone marrow, adipose tissues, cord blood, muscle, synovium, dental pulp, muscle and others)? Caplan has stated that “All or most MSC arise in vivo from perivascular cells (pericytes) that are released from the damaged or inflamed blood vessels at the site of injury ” (5, 6). If this is so, then tissues that have a poor vascular supply would heal poorly or not at all and this is supported by clinical observations of the mid to inner portions of the meniscal and articular cartilages. However, healing by fibroblast proliferation and scar formation and the regrowth of differentiated cells originating from stem cells that are specific to the function of the damage tissue are separate issues. Are stem cells from different tissues equivalent?

Review of the published literature on this question gives evidence to support shared characteristics of MSC from different tissues, but also support that variations are present as well. Perivascular MSC from both bone marrow (BM) and dental pulp (DP) tissues localized immunohistochemically and isolated by immunoselection document that these cells do show expression of STRO1 an early marker of stem cells and CD146 an endothelial marker, but that a 3G5 antigen marker of pericytes was predominantly in the DP population and only in a small portion of the BM cells (7). Another report observed that the MSC population derived from veins, artery, perivascular cells, and fibroblasts showed similarity of cell morphology and phenotypes established with 22 markers with heterogeneous expression of genes related to angiogenesis (8). According to criteria set up by the Mesenchymal and Tissue Stem Cell Committee of the International Society for Cellular Therapy, MSC are adherent cells that have the capacity to differentiate in vitro into osteoblasts, chondrocytes and adipocytes and must express CD105, CD73 and CD90 and lack expression of CD45, CD34, CD14, CD11b, CD79α, CD19 and HLA-DR surface molecules (9). A joint statement of the International Federation for Adipose Therapeutics and Science and the International Society for Cellular Therapy, identified CD90⁺, CD73⁺, CD105⁺, CD44⁺ as adipose tissue stem cell markers with absence of CD45⁻, and CD31⁻. The adipose-derived MSC cells are different from bone marrow MSC in being positive for CD36 and negative for CD106 (10).

FDA regulatory guidelines

The harvesting of bone marrow or adipose tissue from the patient for autologous transplantation results in a heterogenous mixture of cells including the desired stem cell population. Differential centrifugation and erythrocyte lysis accomplish some concentration of the stem cells, but this population still includes haematopoietic stem cells, endothelial cells, erythrocytes, fibroblasts, lymphocytes, monocyte/macrophages and pericytes, among other nucleated cells. The relatively unpurified cells from bone marrow can meet in many cases the “minimally manipulated” classification as 361 tissue with exemption from premarket review and regulation by the US Food and Drug Administration (FDA) regulations. For adipose tissue, this stromal vascular fraction (SVF) has been estimated to contain 15–30% stromal cells and 3–5% pericytes, with hematopoietic- and endothelial-related cells making up 20–45% and 10–20% respectively of the remainder (10). However, enzymatic harvesting or mechanical release of the cells from adipose tissue currently disqualifies them from the 361 classification (11). Expansion of the adherent stem cells in culture in vitro allows another purification step and also permits differentiation of the cells, but may also be considered disqualification from 361 status dependent upon culture conditions (12).

Harvesting and characterization of MSC from different tissues

Review of the literature for the use of human MSC identified four major harvest sites from placental, adipose, bone marrow and umbilical cord tissues (11). Although other tissues have been used, reports of these tissues are not as common. Comparisons of the yield of MSC from the major sites were difficult due to the large variations among different techniques, but adipose tissue consistently yielded higher levels of MSC ranging from 4.7 × 10³ to 1.5 × 10⁶ cells/ml of tissue compared to bone marrow ranges of 1–30 cells/ml to 3.2 × 10⁵ cells/ml and umbilical cord tissues yielding 1.0 ×10⁴ to 4.7 × 10⁶. The number of MSC from placental tissues and synovium ranged from 1.0–30.0 × 10³ cells per ml. Quantitation of the connective-tissue progenitor yield were conducted in the primary tissue isolate principally by limited dilution fibroblast colony forming unit (cfu) assay and enumeration of the colonies (ranging from 20–50 cells per colony) forming in vitro. This allows comparison from various studies or of the yields from different tissues, but does not evaluate the capability of these cells to undergo other types of differentiation. Some investigations have reported the findings with respect to the number of phenotypic cfu when cultured under conditions whereby multi-lineage differentiation occurs (13). Information for selective differentiation of cultures is reported as Alizarin Red staining cfu for osteogenesis, Alcian Blue-positive staining cfu for chondrogenesis or Oil Red O staining cfu for adipogenesis (Figure 1). These numbers are more relevant to the MSC population than the nucleated cell counts of the primary cells isolated as they make up such a minority of the total cells. It has been estimated that there is a mean of fifty-five osteoblastic progenitors per 10⁶ nucleated cells from the bone-marrow (14) with a cautionary note for risk of variation due to blood contamination (11).

Figure 1.

In vitro differentiation of porcine MSC derived from subcutaneous adipose tissue. The adipose-derived MSC (ADSCs) were isolated from dorsal subcutaneous fat. Following 24 h in culture the non-adherent hematopoietic cells were removed and the cells differentiated for 14 days. The Alcian Blue staining was performed for (B) chondrogenic cultures (StemPro Chondrogenesis Differentiation Kit, Thermo Fisher Scientific, CA), Alizarin Red staining was performed for (C) osteogenic differentiation (StemPro Osteogenesis kit, CA) and the Oil Red O Staining was performed for (D) adipogenic cultures (StemPro Adipogenesis kit, CA).

Not only did MSC vary dependent upon the tissue source, the anatomical location was important. The MSC from subcutaneous fat proliferate in vitro at a higher rate and differentiate better than those from visceral fat (15). Human MSC from different adipose anatomical sites showed that that the yield was site dependent with higher frequencies of MSC from the abdomen than the hip or thigh. In this study, abdominal adipose tissue was determined to contain 2.6–10.2 × 10⁶ MSC from 100 g of adipose tissue. The anatomic site of bone marrow harvesting also associated with significant variation with up to two-fold difference in the yield. Pierini et al. analyzed MSC yield from the posterior or anterior iliac crest and the subchondral knee sites and concluded that the posterior iliac crest was preferable although no differences were found for viability, phenotype, expansion kinetics or multilineage differentiation (13). Hernigou studied the risk of complications when harvesting from the iliac crest and determined that higher risks such as trocar contact with the external iliac artery were observed with aspiration in the four most anterior sections with a higher risk in women (16). Placement of the trocar deeper than 6 cm in the posterior iliac crest was at risk for sciatic nerve and gluteal vessel damage.

Pittenger et al. showed only 0.001–0.01% of the cells in the iliac crest aspirate of normal adults exhibited the capacity for colony forming units in vitro. This low number mandates larger collection volumes or culture expansion (17). While the iliac crest has been the traditional site of harvest for bone marrow, the intramedullary cavities of the long bone offers another resource with high numbers of stem cells present in the bone marrow (18). The number of stem cells that can be obtained has been enormously increased by the adaptation of a Reamer-Irrigator-Aspirator (RIA) (Synthes, Westchester, Pennsylvania) consisting of a cutting head on a drive shaft with an attached irrigating and aspirating system (19). Comparisons were made for MSC harvested using a traditional 10 ml iliac crest bone marrow aspiration (ICBMA) and MSC collected during the reaming of the femur to remove 1.5 mm at the isthmus. The material obtained during the reaming was divided into liquid (RIA liquid) and that which was strained out of the liquid through a cell strainer (RIA solid) and further digested with collagenase. Manual white cell count and in vitro culture of the cells under conditions of differentiation showed that showed colony formation was equivalent in ICBMA and RIA liquid fractions although the volume of the RIA liquid was 70 times that of the ICBMA. The IRA solid had almost 4 times more colonies than the ICBMA sample and had ~2.5 times the volume. The ability to collect such high amounts of these stem cells suggest that this methodology can provide 50,000 MSC in a bone marrow concentrate, a number that has been effective in nonunions, without culture expansion (20).

Effects of age and gender on MSC

Other variables to be considered for autologous stem cell supplementation concern the age and gender of the patient. What is the impact of age on stem cell number in an individual? A decrease in stem cell number seen with aging could contribute to delayed or incomplete repair in older individuals and would be a factor for consideration of supplementing with allogeneic vs autologous MSC. The data are somewhat controversial for an age-related change in the number of MSC in bone marrow, but the bulk of the evidence supports that aging decreases MSC numbers in the bone marrow and increases markers of senescence. In studies of normal volunteers, both the number and the proliferative capacity of bmMSC from aged donors (> 40 years) decreased compared to young or adult individuals (21). Cells from older individuals showed increased levels of radical oxygen species and senescent markers such as p53 and p21 while alkaline phosphatase-positive colony forming units were reduced. These data agree with a similar study in which comparisons of MSC harvested from young (18 years) and older (59–75 years) donors showed that the number of alkaline phosphatase-positive colonies were decreased in older individuals as well as their proliferation. The decreased proliferation of the older cells was associated with a shorter telomere restriction fragment. Telomere length of a cell positively correlates with replicative ability (22).

A study of mesenchymal stem cells (STRO-1⁺ cells) taken from bone marrow from discarded femoral tissues from 57 subjects, ages ranging from 17–90 year-old, undergoing hip arthroplasty for osteoarthritis showed that the number of cells STRO-1⁺ in the marrow did not change with age. However, significant age-associated effects were observed such as delayed proliferation due to prolongation of the duration of cell cycling, increased apoptosis, a 4-fold greater senescence-associated β-galactosidase (SA-β-gal) in the group ≥ 55 years, and osteoblast differentiation in terms of alkaline phosphatase activity and in the number of alkaline phosphatase (AlkP)-positive cells (23). The levels of p53, a negative regulator of osteoblast differentiation, showed an age-related increase in p53. In contrast, an earlier study of bone marrow aspirates in orthopaedic patients ranging in age from 13 to 83 years showed a significant decline of nucleated cells with age, but only the female patients showed a significant decline in AlkP-positive colony forming units with age, possibly relevant to osteoporosis in older women (14).

Tissue-dependent effects were obtained comparing the MSC of old and young rabbits. The bone marrow-derived stem cells from old rabbits showed reduced proliferation and chondrogenic response whereas muscle-derived stem and adipose-derived stem cells exhibited no negative effects although the yield from all tissues were lower in the older animals (24). While bone marrow MSC show detrimental changes with aging, this may not be the case for adipose-derived MSC. The number of MSC derived from adipose tissue or their differentiation capacity did not correlate with age of the donor in several studies (15)) (25), although the androgenic status of the pre-adipocytes at different locations influences the proliferation and differentiation of these cells (25). Another factor that has been shown to affect the stem cell population is previous treatment with steroids. Corticosteroid treatment results in decreased stem cells in the bone marrow of the iliac crest (26).

Mode of action of MSC supplementation

While much of the initial enthusiasm for the use of MSC pertained to their function as a precursor population for a more differentiated phenotype and their plasticity and multipotency in giving rise to different reparative cell types, it is apparent that these cells can also function to impact local resident cells and to decrease the immune response of both innate and adaptive immune cells and reduce inflammation (27). The MSC that are present in an initial injury inhibit the immune system early in the reparative process possibly as an attempt to forestall autoimmune reactions to components of the damaged tissue (28). In addition, they also produce molecules that stimulate replication of progenitor cells, stimulate formation of new blood vessels and inhibit the apoptosis of cells due to ischemia. Building upon this hypothesis, investigations of the downregulation of the immune reponse by MSC have been undertaken in a variety of autoimmune diseases. This immunomodulatory influence of MSC has been utilized in the treatment of lupus erythematosus (29), multiple sclerosis (30), systemic sclerosis (31) and graft versus host disease (32). Many of the beneficial effects upon autoimmunity and the reparative function are elicited by production of soluble mediators and through cell-to-cell interaction (Table I). Co-cultures of MSC suppress proliferation of stimulated mononuclear cells (33) or T-lymphocytes (34) (35) and Fas ligand-mediated apoptosis of T-cells (36). The interaction of T1 or T17 lymphocytes upregulates adenosine production by MSC that suppresses these lymphocyte populations (37) (38) and leads to new blood vessel formation through stimulation of vascular endothelial growth factor (39). The interaction of MSC and monocytes results in their production of IL-10 and an anti-inflammatory phenotype (40).

Table 1.

Immunomodulatory Molecules Produced by MSC

Source	Factor	Action	Reference
Mouse bmMSC	IL-1ra	Decreased inflammatory response to bleomycin in the lung	(64)
Human bmMSC	Soluble TNF receptor-1	Attenuated systemic inflammatory response to intraperitoneal injection of lipopolysaccharide	(65)
Human bmMSC	TNF-α stimulated gene/protein 6	Reduced early inflammatory response to permanent ligation of the anterior descending coronary artery and size of the myocardial infarcts	(66)
Human MSC	TNF-α stimulated gene/protein 6	Reduced inflammation in a model of sterile cornea injury	(67)
Human bmMSC	TNF-α stimulated gene/protein 6	Reduction of neutrophils and macrophages in model of zymosan-induced peritonitis	(68)
Mouse bmMSC	PGE2	Mouse model of sepsis induced by cecal ligation and puncture	(69)
Human bmMSC	PGE2	Increased numbers of anti-inflammatory M2 phenotype in human macrophages in vitro co-cultures	(70)
Mouse MSC	Adenosine	Suppression of T-cell proliferation	(37)
Human bmMSC	Adenosine	Suppresses immune responses of Th17 cells	(38)
Human bmMSC	IL-6	Skew monocytes to an IL-10 producing phenotype	(40)

In addition to the immune suppressive properties of MSC, their ability to give rise to more differentiated cell types needed for damaged tissues such as cartilage, bone and cardiac muscle are equally important. Interestingly, these cells have been reported to express HLA Class I but very little HLA class II molecules. This suggests that the cells can be used in an allogeneic as well as an autologous setting, an attractive alternative supplying an available off-the-shelf supply of well-characterized, healthier MSC. However, recently other studies have shown MHC II expression presents following stimulation with interferon gamma (41), after chondrogenic differentiation (42) (43) (44) or osteogenic differentiation (45). Others have found that immunosuppressive therapy is needed for differentiated MSC to survive and evade infiltrating immune cells (46, 47). This has raised concern for the longevity of allogeneic cells in the donor recipient (48) particularly for therapies where the differentiated cell type is important such as cartilage replacement in OA.

Indeed, monitoring the in vivo survival of implanted autologous MSC that undergo differentiation in vivo is difficult as the implanted cells do not differ from the adjacent resident cells and do not have identifying markers. One way of marking stem cells is by labeling them with fluorescent dyes, bioluminescent reporters, radionucleotides or paramagnetic particles before implantation (for review see (49)). We have used fluorescently labeled rabbit bone marrow-derived MSC for implantation into a surgically created partial-defect in the physis in young growing rabbits (50). Fluorescent microscopy of histological sections of the implanted cells in the rabbit physis shows their persistence in the tissue at 3-weeks following implantation (Figure 2) and their phenotypic appearance in histological stained sections as newly formed chondrocytes organized as proliferating isogenous groups that were contiguous with the remaining physis (Figure 3). If left untreated in this model, the animals develop an angular deformity due to the unequal growth across the physis. Angular deformity was significantly different from the untreated control group I in all groups receiving MSC with the greatest reduction in angular deformity in the groups that received MSC cultured on Gelfoam with transforming growth factor beta 3 (TGF-β3), a growth factor that enhances the chondrogenic differentiation of MSC.

Figure 2.

Adipose derived MSC transplanted into a physeal defect in the rabbit hind limb. Allogeneic ADSC were labeled with the fluorescent dye, dialkylcarbocyanine (DiI) before transplantation into a surgically created partial defect in the epiphyseal growth plate of young rabbits. Histological sections of implanted cells in the growth plate after a 3-week interval. (A) Without fluorescent imaging, (B) with fluorescent imaging, (C) merged images. (reprinted by permission from Ahn JI, Canale TS, Butler SD, Hasty KA. Stem cell repair of physeal cartilage. J Orthop Res 2004; 22:1215–21).

Figure 3.

Histology of the surgically damaged site of a growth plate in a rabbit treated with cells cultured on Gelfoam in medium containing 10 ng/ml of TGF-β3 in group C analyzed in Figure 4. Images at A. 200X and B. 400X magnification show differentiated chondrocytes and isogenous columns irregularly arranged in the matrix and contiguous with the damaged growth plate. (H–E stain) (reprinted by permission from Ahn JI, Canale TS, Butler SD, Hasty KA. Stem cell repair of physeal cartilage. J Orthop Res 2004; 22:1215–21).

Use of MSC for orthopaedic problems

The current expectations are very high for the benefits of using stem cell therapy for diverse diseases and traumatic injuries. The largest body of data on the clinical use of these cells is within the orthopaedic clinical setting although the use of these cells within the U.S. has been under the more restrictive guidelines of the FDA than those outside the U.S. Coverage of trials worldwide is beyond the scope of this report, but the most common applications are:

Fracture repair and non-unions. Nonunions following stabilization and treatment of a fracture of a long bone are problematic and can be exacerbated by previous drug treatment and ongoing co-morbidities of the patient. Many of these fractures occur in aging patients with osteoporosis with compromised fracture repair. Treatment with autologous bmMSC was evaluated in patients of varying ages with atrophic tibial diaphyseal non-unions (20). Percutaneous grafting with bone marrow concentrates harvested from the iliac crest showed a positive correlation between the volume of the mineralized callus at 4 months and the number and concentration of the fibroblast colony-forming units that were injected. Sixty patients with non-unions, defined as a failure of the fracture to heal within 6 months and considered to be atrophic due to minimal callus formation, were treated with bone marrow concentrates averaging a total of 50,000 colony-forming units. Fifty-three of the grafted patients obtained bone union with radiographic evidence of callus formation on average at 12 weeks. The seven patients who did not achieve union received lower concentrations of progenitor cells.

Osteonecrosis of the hip. Osteonecrosis of the femoral head was one of the first conditions of the hip where cellular therapies showed good clinical outcomes for review see (51). Many of these patients have been previously treated with steroids, a drug that results in adipocyte differentiation of bone marrow stem cells. Extensive osteocyte death and the observation that reduced MSC numbers are seen in the bone marrow of the iliac crest (26) form a logical basis for this treatment and possibly recommend the use of allogeneic MSC. Another likely benefit may result from angiogenic cytokines secreted by the mononuclear bone marrow cells (51). Interestingly, the bone marrow concentrate can be percutaneously injected with a small trocar into the necrotic head (52). In some cases, autologous bmMSC were combined with beta-tricalcium phosphate ceramic composites (53) or mixed with allogeneic MSC before implantation. Recently, good results were shown with harvested osteoprogenitor cells for grafting secondary osteonecrosis of the knee (54). In another study of patients with sickle cell disease, decompression and autologous bone marrow grafting proved effective with a reduced incidence (only 12.5%) of collapse after 5 years of follow-up compared to 87% collapse seen historically. Two recent reviews of the literature on core decompression with autologous stem cell injection determined that addition of stem cells was superior to core decompression alone in osteonecrotic femoral heads by different investigations several with long-term follow-up (55). Approximation of the number of stem cells needed for loading in a osteonecrotic femoral head was calculated by evaluation of the number of MSC in a normal femoral head whereby a total of 35,000 bmMSC is considered as a useful approximation of the number of MSC present in a femoral head (26, 56).

Cartilage repair. One of the most difficult challenges in the future for tissue engineering with MSC will be to undertake the re-construction of articular cartilage. Although this tissue is comprised of only one cell type and lacks innervation and vascularization, the chondrocytes and their matrix components are stratified in their phenotype from the articular surface to the subchondral underpinning in apparent response to the ratio of the different types of mechanical stresses exerted upon the layers. It is clear that MSC cultured in vitro under chondrogenic conditions express differentiated cartilage components, but do not display the stratified appearance of articular chondrocytes. Even MSC differentiating at the articular surface with microfracture; the drilling through the osteochondral boundary and the subsequent upgrowth and chondrogenic differentiation of the stem cells from the marrow cavity, do not achieve a mature hyaline cartilage phenotype (57)57). The new cartilage that forms under this condition initially looks promising, but gradually progresses to a fibrocartilage phenotype with an unknown half-life (58). As it is known that the expressive phenotype of the chondrocyte is responsive to mechanical stress, it is clear that more information is needed regarding the impact of physical rehabilitation and the long-term response to weight bearing on newly formed or implanted cartilage tissue.

The phenotype of articular cartilage in the joint often modifies with aging displaying hypertrophic markers such as type 10 collagen and matrix metalloproteinase expression typical of the hypertrophic zone of the physis. These markers are thought to signal the onset of OA (for a review see (59)). It is unknown if these changes are inherent in the chondrocytes themselves or arise from the reponse of the chondrocyte to changes in their milieu whether it be to soluble factors or mechanical forces. If an external factor is the primary impetus eliciting the hypertrophic changes, would newly implanted chondrocytes be resistant to such changes or would the same sequela develop over time? If hypertrophy develops due to inherent properties of the aging chondrocytes, would autologous MSC used for tissue engineering of the cartilage replacement be subject to early degeneration as well? Alternatively, do MSC exert a beneficial modification of the inflammatory milieu through paracrine effects in addition to their role as chondrocyte precursors? Clinical trials going forward in this area should incorporate some of the guidelines for “detailed methodological recommendations…developed for the statistical study design, patient recruitment, control group considerations, study endpoint definition, documentation of results, use of validated patient-reported outcome instruments, and inclusion and exclusion criteria for the design and conduct of scientifically sound cartilage repair study protocols” developed by the International Cartilage Repair Society (60).

Homologous use of MSC is their implantation in a tissue for a condition in which the cells would normally participate. An example of this is the use of bone marrow-derived MSC for non-union of bone fractures. The employment of bmMSC or adipose-derived MSC for cartilage repair is considered non-homologous use by the FDA a classification inhibiting clinical application of their use in the US; however, clinical investigations outside the US show promising results with these cells. Patients with unilateral osteoarthritis have been treated with autologous bone marrow stem cells expanded in autologous serum for 3 weeks and supplemented by intra-articular injection into the knees of fifty-six patients (<55 years of age) who had previously undergone microfracture and medial opening-wedge high tibial osteotomy (61). Those patients receiving MSC showed more improvement in short-term clinical and longer-term MRI findings at 1 year than patients similarly treated who did not receive the cells. However, the authors stated that a limitation of this study was that the injected MSC could not be followed to the desired regeneration site raising the question if the beneficial effects were due to paracrine factors from the MSC or their direct participation as chondrogenic precursor cells. In another study, Jo et al. injected enzymatically isolated, culture expanded adipose-derived MSC into the knee joint of eighteen patients with arthroscopically graded OA lesions (62). Significantly improved WOMAC scores and decreased knee pain were observed compared to baseline in only patients receiving the highest number of 1 × 10⁸ cells. MRI evaluations at 3 months showed a thick layer of articular cartilage that was thickened at 6 months. Arthroscopic measurement of the original cartilage lesion in the lateral femoral and tibial condyles at 6 months post-injection showed a significant decrease in the high dose group, but not the other groups. Histological evaluations of biopsies at six months showed a hyaline-like cartilage in the mid to deep zones, but the upper half of the mid zone and the superficial zone demonstrated that type I collagen positive fibrocartilage was present. Although this study injected the cells in the solution, implantation of such cells or autologous chondrocytes in various scaffolds is now second generation for cartilage replacement engineering. However, utilization of a scaffold requires two invasive procedures: one for harvesting of the autologous stem cells from tissues with possible expansion in culture or cell/scaffold preparation and another surgery for implantation of this tissue. Comparison of treatments with autologous bone marrow stem cells and autologous chondrocytes showed equivalent results although the stem cell treatments were less expensive (63).

Summary/Discussion

While there is ample evidence that beneficial results can be obtained from the use of MSC, several questions regarding their use remain to be answered. Several questions are in regard to the mechanism/s by which the MSC are eliciting the positive outcome of their use whether it be paracrine factors that are made by the MSC that affect the innate or adaptive immunity or impact the metabolism of resident cells or by their contribution as precursor cells giving rise to reparative cells. If paracrine factors are the paramount impetus, then commercial sources of allogeneic MSC could possibly provide an easily obtainable, well-characterized reagent for the clinician if such can be suitably standardized. Tissue engineering of MSC contained within a suitable scaffold also has value for stabilizing and localizing these cells within the implantation site. However, in their function as reparative cells, the longevity and fate of allogeneic MSC need to be further evaluated on longer-term basis. Does the fact that allogeneic cells have the potential to express their own inherent HLA protein under certain stimuli impair their desirability for serving as reparative cells for cartilage, ligaments and bone; tissues that may be subject on occasion to inflammation or the activity of innate and acquired immune cells?

Alternatively, the use of autologous cells has both advantages and its own inherent disadvantages involving additional procedures for collection, preparation and characterization of the MSC. It is apparent that a critical number of MSC are necessary for implantation to effect an optimal outcome. This underscores the importance of deciding if tissues such as adipose tissue provide more MSC than bone marrow and if the capability for differentiation is the same for MSC isolated from different tissue sources. If autologous MSC are limited, what are the conditions for optimal culture expansion and are there benefits for initiating differentiation pathways before re-implantation? Also we need to better understand how individual patient factors such as age, sex, drug regimens and co-morbidities influence the MSC population. Do some conditions presuppose for administration of allogeneic cells? For many of these questions, preclinical models will be helpful, but in the final analysis the task of evaluating and implementing these findings for orthopaedic patients falls onto the shoulders of clinical researchers.

Figure 4.

Differences in angular deformity in the treated and untreated rabbits. Deformity measured as the angle formed between perpendiculars from the distal femoral and distal tibial articular surfaces in radiographs. No cells were implanted into the control groups, group I (4 rabbits) the physeal defect was left untreated and group II (3 rabbits) the physeal defect was treated with Gelfoam. Experimental groups with approximately 2–4 million MSC cells per implant: Group A (5 rabbits), the defect was filled with 5% gelatin containing cultured MSC, Group B (3 rabbits), the defect was filled with 10% gelatin containing the same cells and a Gelfoam resorbable sponge plug placed at the periphery of physeal defect to prevent leakage of the cells, Group C (5 rabbits), the defect was filled with cells cultured on Gelfoam in medium containing 10 ng/ml of TGF-β3 to enhance chondrogenesis. Angular deformity was significantly different (*) from control group I in all animal groups receiving MSC with the most reduction of (p < 0.001) in deformity seen in group C (MSC + Gelfoam + TGFβ3). (reprinted by permission from Ahn JI, Canale TS, Butler SD, Hasty KA. Stem cell repair of physeal cartilage. J Orthop Res 2004;22:1215–21).

Key Points.

1. Mesenchymal stem cells isolated from different tissues share many characteristics such as the capability for multilineage differentiation, absence of HLA-DR expression and possession of the markers CD105, CD73 and CD90, but may vary in expression of other markers such as CD36 and CD106.

2. In the U.S., unpurified stem cells from bone marrow and other tissues can meet minimally manipulated classification as 361 tissue with exemption from FDA premarket review and regulation; however, enzymatic harvesting and culture expansion of MSC are exclusive of 361 classification.

3. Harvesting of MSC from different anatomical sites yields varying cell numbers and proliferation rates as exemplified by comparison of MSC isolated from bone marrow and adipose tissues. Age and sex differences were observed with decreased proliferation and differentiation and increased senescent markers.

4. Paracrine effects of soluble mediators and cell-to-cell interactions of MSC affect innate and adaptive immunity and decrease inflammation.