Mesenchymal stem cell therapy in horses: useful beyond orthopedic injuries?

Abstract

In the past decade, mesenchymal stem cells (MSC) have received much attention in equine veterinary medicine. The first therapeutic use of equine MSC was reported in 2003. Since then, the clinical application of MSC has been exploding with thousands of horses now treated worldwide. At present, MSC are mainly used in veterinary medicine to treat musculoskeletal diseases based on their ability to differentiate into various tissues of mesodermal origin. This is in marked contrast to human medicine, where MSC therapies are primarily focused on immune-mediated, inflammatory, and ischemic diseases. In this review, both orthopedic as well as non-orthopedic clinical applications of equine MSC are discussed. A brief overview is provided on the potential of MSC for non-orthopedic injuries with emphasis on those diseases, which occur in both humans and horses.

Keywords:

• equine
• horse
• mesenchymal stem cells
• orthopedic applications
• non-orthopedic applications
• stem cell therapy

1. Introduction

The considerable therapeutic potential of equine mesenchymal stem cells (MSC) in regenerative medicine is based on both their simplicity of isolation and their ability to promote tissue regeneration, prevent pathological scar formation, modulate immune responses, and regulate inflammation (Stewart & Stewart 2011). So far, clinical studies have been reported using equine MSC derived from bone marrow (BM), adipose tissue, and umbilical cord blood (UCB). In equine medicine, MSC are currently mainly used to treat musculoskeletal diseases, while in human medicine, MSC therapies are primarily focused on immune-mediated, inflammatory, and ischemic diseases (Borjesson & Peroni 2011). The public clinical trials database lists almost 300 clinical trials in which MSC are used for a variety of therapeutic applications such as ulcerative colitis, type 1 diabetes mellitus, liver cirrhosis, spinal cord injury, graft versus host disease, osteoarthritis, myocardial infarction, and ischemic stroke (http://www.clinicaltrials.gov).

Interestingly, several infectious, allergic/atopic, developmental, and autoimmune diseases have a similar pathogenic etiology in humans and in domestic animals (Gershwin 2007). Therefore, the use of MSC in equine medicine might be considerably expanded (Figure 1). As such, equine medicine could learn from human medicine while on the other hand, horses could prove to be a valuable model to test the efficacy and safety of novel treatments for these diseases. In this concise review, a summary of the orthopedic applications of equine MSC is followed by a discussion on the potential non-orthopedic applications based on the use of human MSC in either experimental models or clinical trials. Hereby, diseases with a similar pathogenesis in both humans and horses are emphasized.

Figure 1. Schematic overview of possible applications of equine mesenchymal stromal cells in veterinary medicine.

2. How do MSC function to heal tissues?

Initially, the use of MSC for primary tissue regeneration was advocated based on their ability to differentiate into various tissue types. As such, the regeneration of damaged tissues would be directly stimulated since injected MSC colonize the injury site, differentiate into the appropriate mesenchymal tissue type, and affect repair (Stewart & Stewart 2011). However, some controversies still exist on the issue whether MSC primarily contribute to lesion healing by integrating into the injured tissue or indirectly by secreting immunomodulatory and bioactive trophic factors (Koch et al. 2009; Fortier & Travis 2011). First, the in vitro differentiation capacities of MSC toward different lineages were confirmed by assessing some qualitative aspects of tissue formation using histological stainings. However, it is obvious that the in vivo tissue-level complexities of e.g. mineralized bone or articular cartilage are far from realized in vitro (Stewart & Stewart 2011). Thus, an evidence of in vitro MSC differentiation does not guarantee their in vivo clinical usefulness. Along this line, it is noticed that the synthesis and deposition of matrix proteins by MSC are generally far less distinct in comparison to the activities of differentiated chondrocytes. Nevertheless, the MSC chondrocyte-specific mRNA expression is often similar or even higher than that of intrinsic articular chondrocytes (Stewart & Stewart 2011). Furthermore, there is evidence that MSC do not remain at the injury site after injection. In a study of Quintavalla et al. (2002), fluorescently labeled MSC on a gelatin scaffold were implanted in full-thickness cartilage defects in goats and 14 days later, an extensive loss of the implanted MSC throughout the defect was observed. Fluorescent MSC were detected in the deeper regions of the defect as well as in the subchondral bone spaces, suggesting a migration of cells. On the other hand, it has been reported that MSC are guided by chemokines to migrate from their niche and home to sites of damaged tissue, although the exact nature of their signaling factors remains unknown (Kode et al. 2009). Guest et al. (2008) demonstrated that fluorescent-labeled mesenchymal progenitor cells, which were injected into the superficial digital flexor tendon, mainly remained localized within the lesions although some labeled cells were present in healthy tendon surrounding the lesions, again indicating that migration does occur.

3. Clinical applications of equine MSC for orthopedic injuries

Complementary to the excellent recent reviews on this topic (Alves et al. 2011; Fortier & Travis 2011), we here provide an updated overview of the current state of knowledge, emphasizing the remaining controversies, gaps in knowledge, and possible focus for future research.

3.1. Tendon

Injuries to the tendons situated at the palmar/plantar side of the equine distal limb are very common in competition horses subjected to high-intensity exercise. The poor success with conventional therapy supported the need to search for novel treatments, which should aim at restoring functionality and regenerating a tissue as close to the tendon as possible (Richardson et al. 2007). An overview of the in vivo studies performed to date, using equine MSC to treat tendon injuries, is given in Table 1. It must be noted that the efficacy of equine MSC therapy is difficult to evaluate since the use of appropriate control groups is not always included. Furthermore, this treatment is often combined with other biological factors such as BM supernatant, autologous serum, and platelet-rich plasma (Koch et al. 2009). Moreover, the current clinical literature frequently relies on study designs that do not respond to the gold standard of evidence-based medicine, i.e. blinded randomized control trials. Although some studies are controlled, the experimental power is often lacking because of the limited sample size in horse-based studies and the interanimal variability of the pathological conditions (Clegg & Pinchbeck 2011). Nevertheless, a positive therapeutic effect of MSC upon treatment of tendon lesions has been suggested (Table 1). Both in studies with experimentally induced tendonitis and in studies with naturally occurring lesions, the re-injury rate was shown to be significantly lower in horses treated with MSC compared to conservative therapy (Table 1).

Table 1. Overview of clinical studies using equine MSC to treat musculoskeletal diseases in the horse. Where applicable, the efficacy of the MSC treatment, i.e. effect as assessed by diagnostic tools or percentage (%) of horses re-injured after MSC treatment versus the control group, is indicated.

Study	N horses	Control group	Nature of the lesion	Diagnosis	Follow-up period	Efficacy MSC (re-injury rate treatment vs. control)
Tendon
Smith et al. (2003)	1	NI	Natural	CE + U	6 weeks	NA
Pacini et al. (2007)	11	Conservative treatment	Natural	CE + U	2 years	18.2% vs. 100%
Del Bue et al. (2008)	16	NI	Natural	CE + U	240 days	12.5% vs. NI
Guest et al. (2008)	2	NI	Induced	IHC	10 and 34 days	No significant effect
Nixon et al. (2008)	8	PBS	Induced	U + H + IHC+ PCR + BA	6 weeks	Positive effect
Smith et al. (2008)	500	NI	Natural	CE + U	>1 year	13–36%^* vs. NI
Ferris et al. (2009a)	97	NI	Natural	CE	7–39 months	15–27%
Leppänen et al. (2009)	58	NI	Natural	PI	18–24 months	26.4% vs. NI
Schnabel et al. (2009)	12	Control limb: PBS	Induced	U + MT + H+ PCR + BA	8 weeks	Positive effect
Crovace et al. (2010)	6	Saline	Induced	CE + U + H + IHC	21 weeks	Positive effect
de Mattos Carvalho et al. (2011)	8	Control limb: no MSC	Induced	CE+ U+ H+ IHC	8–21 weeks	Positive effect
Caniglia et al. (2012)	6	Control limb: BM supernatant	Induced	HM	12 weeks	No significant effect
Godwin et al. (2012)	141	NI	Natural	CE + U + S + H	3 years	27.4% vs. NI
Marfe et al. (2012a)	3	Conservative treatment	Natural	U	3 years	0% vs. 100%
Cartilage
Wilke et al. (2007)	6	Control limb: PBS	Induced	A + H + (I)HC + ISH	8 months	No significant effect
Ferris et al. (2009b)	40	NI	Natural	PI	21 months	28% vs. NI
Frisbie et al. (2009)	24	Placebo	Induced	CE + RX + H + punction	70 days	No significant effect
McIlwraith et al. (2011)	10	Control limb: hyaluron alone	Induced	CE + RX + A + MRI + H + HM + IHC + BA	12 months	No significant effect

Note: A - arthroscopy; BA - biochemical analysis; CE - clinical examination; H - histopathology; HM - histomorphometry; IHC - immunohistochemistry; ISH - in situ hybridization; MRI - magnetic resonance imaging; MT - mechanical testing; NA - not applicable; NI - not included; PBS - phosphate buffered saline; PCR - polymerase chain reaction; PI - phone interview; RX - radiography; S - scintigraphy; U - ultrasonography; * depending on discipline.

3.2. Cartilage

Due to the hypocellular and avascular nature of articular cartilage, the ability to obtain effective repair is limited (Frisbie & Stewart 2011). Full-thickness cartilage defects in horses heal with fibrous tissue that might become fibrocartilage, which has inferior biomechanical properties compared to articular hyaline cartilage (Taylor et al. 2007). Giving the ability of MSC to undergo chondrogenic differentiation, much of the recent research on cartilage resurfacing in the horse has focused on the use of these stem cells (Figure 1) (Frisbie & Stewart 2011). Although MSC differentiation into cartilage has been demonstrated in vitro, a successful in vivo use of such cell-based therapies might be hampered by the compressive load exerted on the injected cells and scaffolds (Koch et al. 2009). Indeed, only a scarce number of equine in vivo studies have been performed so far with little success (Table 1).

Even more challenging is the treatment of osteoarthritis since the articular cartilage damage is often more diffuse and not only the corresponding surfaces, but also the periarticular tissues such as the synovial membrane, the joint capsule, ligaments, menisci, and subchondral bone, can be impaired (Frisbie & Stewart 2011). Nevertheless, a beneficial effect of MSC on cartilage morphology and histology has been demonstrated in various osteoarthritis animal models (Murphy et al. 2003; Agung et al. 2006; Alfaqeh et al. 2008; Grigolo et al. 2009; Matsumoto et al. 2009). These beneficial effects, however, are most likely not due to the formation of new cartilage as it has been demonstrated in cell tracking studies that only a limited number of the injected MSC differentiated toward chondrocytes (Murphy et al. 2003; Matsumoto et al. 2009). Interestingly, it has been demonstrated in a recent study of van Buul et al. (2012) that MSC secrete soluble factors, which cause multiple anti-inflammatory effects and influence the matrix turnover in synovium and cartilage explants. So far, 21 clinical studies were conducted using BM-derived MSC in patients with osteoarthritis. As most of these studies are still ongoing, a positive effect of the MSC treatment has been suggested based on the outcome of only eight studies (Gupta et al. 2012).

3.3. Bone

In contrast to tendon and cartilage repair, bone fractures usually regenerate with similar biochemical and biomechanical properties as the original tissue (Taylor et al. 2007). However, when large quantities of bone need to be regenerated, it may be required to stimulate the natural processes of bone repair (Kraus & Kirker-Head 2006). Examples include substantial loss of host bone from trauma or tumor resection, arthrodesis, spinal fusion, non- or delayed unions, osseous cyst-like lesions, metabolic disease, arthroplasty, or insufficient healing potential of the host because of local or systemic disease or old age (Taylor et al. 2007). Based on large animal model studies, it has been demonstrated that MSC transplantations not only significantly increase bone formation, but also that the treated bones retain the same strength as the uninjured control bones (Kon et al. 2000; Liu et al. 2008). In a case study of Quarto et al. (2001), three patients with non-union fractures in limbs with 4–7 cm critical size defects, regained limb function with no adverse effects 15–27 months after MSC treatment. Callus formation and integration of the graft were already evident at 2 months post transplantation, whereas this recovery would have taken 12–18 months with traditional treatment. Also after a long-term follow-up of 7 years, long bone defects were healed with complete fusion and integration of the graft by an average of 7 months (Marcacci et al. 2007).

In horses, a few preliminary experiments have been performed in which a pastern joint arthrodesis was supported by a combined therapy of stem cells and a bone replacement material, resulting in an adequate development of bone fusion (Brehm et al. 2012). However, controlled clinical studies on the application of MSC in bone regeneration in horses have not yet been reported.

4. Clinical applications of equine MSC for non-orthopedic injuries

4.1. Immune-mediated and inflammatory diseases

MSC are known to modulate local inflammatory responses and recruit local autologous stem cells inside injured tissues to stimulate cell survival and tissue repair (Stewart & Stewart 2011). Therefore, in humans, MSC have been proposed useful in cases of organ transplantation, inflammatory and autoimmune diseases. In addition, the immunomodulatory effects of MSC appear to be restricted to inflamed tissues since there is no evidence of systemic immunosuppression or increased risk of infections upon administration of MSC to immune-competent patients (Sensebe et al. 2010). A number of studies using animal models demonstrated the efficacy of human or rodent MSC as a tool for immunomodulation in the protection against allograft rejection, autoimmune encephalomyelitis, collagen-induced arthritis, sepsis, and autoimmune myocarditis (Augello et al. 2005; Einstein et al. 2007; Ohnishi et al. 2007; Gonzalez et al. 2009; Sordi & Piemonti 2011).

Also in horses, there are several equine autoimmune pathologies for which stem cell therapy could likely be interesting. Equine recurrent uveitis, an organ-specific, T-cell-mediated autoimmune disease of high prevalence (i.e. 10%) in horses, is characterized by active inflammatory episodes in the eye followed by periods of minimal ocular inflammation (Deeg et al. 2008; Gilger & Deeg 2011). As horses are the only animal species, which develop this disease with clinical and immunopathological aspects that are very similar to human autoimmune uveitis, equine recurrent uveitis could serve as a reliable spontaneous model to study the histopathological changes in uveitis and to better understand the course of inflammation (Deeg et al. 2002).

Horses, just like humans, can also suffer from systemic lupus erythematosus (SLE). This is an autoimmune disease characterized by the accumulation of autoreactive lymphocytes and immune complexes, leading to the destruction of targeted organs including renal, cardiovascular, neural, musculoskeletal, and cutaneous involvements (Gu et al. 2010). As MSC have immunomodulating capacities, the imbalance of cytokine homeostasis, which is a prominent feature of SLE, might be modulated by the soluble factors produced by MSC upon their activation (Liang et al. 2010). Indeed, treatment-refractory human SLE patients showed clinical and hematological improvements after allogeneic MSC transplantation (Sun et al. 2009; Shi et al. 2012). A similar approach as in these human clinical trials might prove beneficiary in horses suffering from SLE.

4.2. Ischemic diseases

Stem cell therapy could also be used for the treatment of ischemic diseases, which cause oxygen deprivation, cell injury, and related organ dysfunction (Chen et al. 2006). Although ischemic injuries are usually local in nature, they are often part of disorders with a highly complex pathophysiology involving numerous biochemical changes in several cell types (Lange et al. 2005). The multidifferentiation and immunomodulatory abilities of MSC provide an attractive perspective of using them in the treatment of a variety of human diseases such as stroke, ischemic retinopathy, myocardial infarction, ischemic diseases of the liver, ischemic renal failure, and ischemic limb dysfunction, as demonstrated in different rodent models (Li et al. 2002; Terai et al. 2003; Tang et al. 2005; Tögel et al. 2005; Li et al. 2009; Zhu et al. 2011). Since MSC appear to migrate specifically to ischemic regions, the functional recovery of the damaged tissue is supported by circulating stem cells (Chen et al. 2006). In line with this, it has been demonstrated that MSC, intravenously administered to rats after myocardial infarction, are localized in the infarct region and improve ventricular function, whereas MSC delivered to non-infarcted rats home to the BM (Saito et al. 2002).

In horses, laminitis is a multifactorial disease of the equine foot with various initiating causes including local ischemia (Engiles 2010). Therefore, it has been proposed that stem cell therapy might improve the current variably successful treatment options (Koch et al. 2009). Another ischemic disease is the perinatal asphyxia syndrome seen in foals, which usually arises from the combination of ischemia and hypoxemia and affects many organs (Galvin & Collins 2004). It can be hypothesized that a fast intervention using e.g. UCB-derived MSC obtained immediately after birth, could help to increase the foals’ survival rate. The isolation and characterization of equine UCB-derived MSC is now well established (De Schauwer et al. 2011), although autologous use might be hampered due to the long initial culture period needed before enough MSC can be collected. An allogeneic approach could alternatively be used, providing a readily available, off-the-shelf product without the inherent lag period associated with the expansion of autologous MSC (Alves et al. 2011). Consequently, it might be considered to start a bank of allogeneic UCB-derived equine MSC in analogy with humans.

4.3. Wound repair

It has been recently demonstrated that intravenously administered murine MSC at the site of a skin lesion in a mouse model are able to accumulate and differentiate into multiple skin cell types (Sasaki et al. 2008). In a preliminary clinical study of Badiavas and Falanga (2003) on human patients, autologous BM-derived MSC were used to treat chronic wounds (i.e. older than 1 year) which were nonresponsive to other treatments. Hereby, a complete closure and evidence of dermal rebuilding was observed in all three patients treated with MSC. However, a study including more patients is warranted to further substantiate these findings.

As horses are predisposed to traumatic skin wounds that can be labor intensive and expensive to heal, equine MSC could play an important role in wound repair considering the abovementioned potential of MSC to improve the healing of skin defects (Figure 1) (Theoret 2009). Indeed, a retrospective study of 422 horses with traumatic wounds showed that primary closure was obtained in as little as 24% of the wounds (Wilmink et al. 2002). The historical gold standard to replace lost skin is an autologous skin graft but unfortunately, graft failure is relatively common in equine patients due to infection, inflammation, fluid accumulation beneath the graft, and motion (Theoret 2009). Besides, full-thickness autografting is limited to relatively small wounds since the horse lacks redundant donor skin (Theoret 2009). To date, only one study describes the use of equine MSC at the site of a surgically repaired soft palate defect (Carstanjen et al. 2006). Labeled autologous BM-derived MSC were implanted into the repaired defect at surgery and the horse was euthanized 14 days later. Microscopic examination revealed that the MSC were oriented and integrated along the axis of the skeletal myocytes under the epithelium, which is indicative for a successful engraftment (Carstanjen et al. 2006). Taken together, these preliminary data suggest that MSC might indeed contribute to wound healing in horses.

4.4. Ophthalmology

Ma et al. (2006) evaluated in a rat model whether or not MSC could be used to treat corneal disorders. For this purpose, human MSC were grown and expanded on amniotic membranes, and subsequently transplanted into rat corneas 7 days after chemical burns. The corneal surface was successfully reconstructed 4 weeks later. This therapeutic effect was likely associated with the inhibition of both inflammation and angiogenesis after MSC transplantation, rather than with the epithelial differentiation of the transplanted MSC (Ma et al. 2006).

In equine ophthalmology, only one recent study describes the use of equine MSC in four chronic cases of corneal ulcer and one case of retinal detachment, all of which were nonresponsive to conventional treatment. Within 3 months after MSC treatment, consisting of local as well as systemic injections, all four patients showed a significant improvement of the inflammatory reaction as well as the restoration of the epithelial surface of the central cornea (Marfe et al. 2012a). It must be mentioned that in the latter study, not only the isolated stem cells from peripheral blood expressed known MSC markers like CD90, CD105, and CD117, but also hematopoietic stem cell markers such as CD34, or embryonic stem cell markers such as Sox2, Oct3-4, and Nanog.

4.5. Neurological disorders

As MSC have been shown to differentiate in vitro into neurogenic progenitors that express specific neuronal markers, the potential efficacy of MSC for functional repair of nervous tissues has been studied with great interest (Jamnig & Lepperdinger 2012). Two possible approaches in the treatment of human neurological diseases exist: either the MSC can be cultured and differentiated before implantation toward the desired differentiated neuronal cell type, or the MSC can be implanted immediately after which differentiation could be induced by endogenous factors (Nery et al. 2013). An important question, however, is whether the neuronal cells differentiated from MSC become functional neurons or not. Most studies confirm neuronal differentiation using morphological features and/or neuronal marker gene expression analyses at the mRNA or protein level (Franco Lambert et al. 2009). In the study of Wenisch et al. (2006), the differentiation toward mature neuron-like cells was suggested based on the immunocytochemistry results but this finding was not confirmed ultrastructurally and electrophysiologically. As such, it was demonstrated that the expression of neuronal marker proteins does not prove that these cells are truly mature functional neurons. A normal mature central nervous system is only unambiguously defined by its excitability and its ability to fire action potentials and to communicate with other cells by releasing neurotransmitters (Lu & Tuszynski 2005). To our knowledge, this was confirmed only recently for MSC (Ma et al. 2011).

In the clinical trial of Ra et al. (2011), eight male human patients with a spinal cord injury were intravenously treated with autologous adipose tissue-derived MSC. Both the safety of the culture-expanded MSC and the efficacy were monitored during a 3-month follow-up. Some restorations of function were observed such as the improvement of the motor score in three patients and the overall improved pin prick scores, which is indicative for a better prognosis to regain functional ambulation. Nevertheless, no resolute conclusions about efficacy of MSC therapy could be drawn due to the limited size of the clinical trial, the short follow-up period and the lack of a control group (Ra et al. 2011).

Since most neurodegenerative diseases have common underlying pathological processes, a specific therapeutic agent like MSC could improve the symptoms of several neurodegenerative disorders, based on their ability to replace damaged cells or secrete trophic factors and immunomodulating cytokines (Sadan et al. 2009). Equine myeloencephalopathy and equine motor neuron disease are examples of neurodegenerative disorders for which MSC therapy might be interesting. Also for laryngeal hemiplegia, caused by a progressive idiopathic paralysis of the intrinsic laryngeal muscles, equine MSC might contribute to improve proper treatment (Figure 1). However, as for humans, it still remains to be determined whether MSC display sufficient neurogenic differentiation capacity in vivo, and whether they survive well after being transplanted (Jamnig & Lepperdinger 2012).

5. Conclusions

In conclusion, the use of MSC for the treatment of equine injuries and diseases holds an immense potential and should expand from the current very valuable albeit exclusive treatment of orthopedic lesions into the evaluation of future treatments of ischemic, inflammatory, and neurological disorders. As such, equine medicine could benefit from findings in human medicine, while on the other hand, horses also could prove a very valuable model to test the efficacy and safety of innovative treatments for these nonorthopedic diseases.

Still, more in-depth research and well-designed prospective clinical trials are mandatory to fully understand the potential of equine MSC and to optimize the equine MSC-based therapies. However, the true efficacy of equine MSC therapy is sometimes difficult to evaluate mainly due to the absence of appropriate control groups. Therefore, it should be a collective goal of the equine veterinary community to use clinical trials, which include sufficient similar cases and a consistent and standardized panel of objective outcome measures (Stewart 2011).

Acknowledgements

Prof A. Martens, Prof P. Simoens, Dr E. Raes and Dr L. Lefère are acknowledged for providing the necessary pictures to compose Figure 1.