Tendon and ligament injuries (TLI) continue to be a major cause of morbidity and premature retirement in racehorses. There have been comprehensive reviews published relating to TLI and tendon biology in the horse that provide much useful background information relating to tendon biology, aetiopathogenesis of injury and treatment [1–3]. In the last decade there has been substantial progress in the understanding of many aspects of equine tendon biology relating to the racehorse much of which has been achieved through financial support from the Horserace Betting Levy Board.
Tendon and ligament injuries continue to be one of the most frequent causes of orthopaedic injury and early retirement in the Thoroughbred racehorse. In the UK, the Horserace Betting Levy Board has invested considerable research funding in the last decade in order to investigate aspects of tendon epidemiology, biology and therapy, in order to decrease both the welfare and economic consequence of such injuries. This review summarises key aspects of this research effort and identifies specific research directions for the future.
There have been a number of studies in the UK that have detailed the epidemiology of TLI in racehorses, particularly identifying the incidence and risk factors for these conditions. Internationally, there has also been much work detailing epidemiological data relating to both frequency and risk factors for TLI injuries in racing, although the data may not always be directly transferable to other racing jurisdictions due to differences in horse demographics, surfaces and both racing and training practices [4–7].
Analysis of injuries sustained on UK racecourses in the 1996–1998 racing season identified that 46% of injuries were TLIs with the incidence of such injuries being greater in steeplechase than hurdle races even though age-specific rates of tendon injury were higher in hurdle races than in chases . The importance of TLIs in racing was further highlighted in a cohort study undertaken at 6 racecourses in 2000 and 2001. This study identified that TLIs were the most frequent cause of injury during racing with a frequency of 6.9/1000 starts. Of the TLIs reported in this study, 90% occurred within the superficial digital flexor tendon (SDFT) and 20% of the TLIs necessitated euthanasia on the racecourse .
An investigation of the epidemiology of TLIs during National Hunt (NH) horse training identified the TLI incidence rate to be 1.9/100 horse months and varied significantly by trainer (P<0.05) and with increasing age (P<0.001) but not by gender or background. The majority of TLI injuries occurred in training with 41% associated with racing and 59% associated with training. Ex-store NH horses were significantly more likely to have a TLI on the racecourse than ex-flat horses (P<0.01). Superficial digital flexor tendon injuries accounted for 89% of all TLIs, the remainder being suspensory ligament injuries. Of the injuries that occurred during racing 7% were subjected to euthanasia compared with 1% of injuries sustained during training .
In a study to determine the prevalence of SDFT injuries and to improve methods of predicting injury in NH racehorses routine ultrasonographic assessment of the palmar metacarpal soft tissues of 263 NH racehorses was performed on up to 6 occasions over 2 NH racing seasons. The study identified the prevalence of SDFT pathology detected using ultrasonography in this cohort of horses in training was 24% (n = 148), with a nonsignificant variation between different trainers of 10–40%. Older horses had a significantly higher prevalence of SDFT pathology compared with younger horses and horses with a previous history of tendinopathy were more likely to suffer an acute injury compared with horses with no evidence of previous pathology. This study concluded that SDFT tendinopathy is common in N horses, with substantial variation between training yards . A case-control study compared 401 horses with a first occurrence of SDFT tendonitis with control horses with no history of a tendon injury of same age, sex and trainer. This study identified that 53% of the case horses reinjured a tendon within 3 years of first treatment, although an equal number of case and control horses (80%) raced again after the date of injury and raced at least 3 times. However, control horses were significantly more likely to complete 5 races post treatment date than case horses. The authors concluded that assessment of the outcome of horses with an SDF injury using the number of races post injury requires a minimum of 5 races post injury to be a useful indicator .
More recently, a large epidemiological study of SDFT injuries that occurred in UK hurdle races and were diagnosed at the racecourse, identified 20 variables associated with altered risk for tendon injuries. Variables found to be associated with increased odds of SDFT injury included: firmer going, increased horse age at first race, having had a previous SDFT injury and racing in the summer compared with other seasons. Variables found to be associated with decreased odds included being trained by a more successful trainer and having raced more frequently in the preceding year .
Early recognition and prediction of TLI
There has been much interest in developing methods to identify both horses at risk of TLI, as well as early recognition of injury prior to potential catastrophic tendon rupture. Using data from a cohort study of 263 NH racehorses, no changes in SDFT cross-sectional area (CSA) or ultrasonographic appearance were detected prior to injury. Cross-sectional area of normal horses did not vary significantly with age, limb or over 2 racing seasons but did with sex and whether the horses originated from NH store background or whether it was an ex-flat horse. It was concluded that ultrasonography at 3-month intervals was unable to predict acute SDFT injuries and there was a need to develop other methods of early disease prediction .
As in the case of joint disease [14,15] there has been interest in the development of fluid biomarkers to identify early or preclinical TLI. Studies have assessed the use of biomarkers of both type I collagen synthesis (PICP) and degradation (ICTP/CTX-MMP) and cartilage oligomeric matrix protein (COMP), to determine whether these assay could detect changes in connective tissue remodelling associated with injury to the SDFT. Data from all these investigations is preliminary; however, there is no indication that such assays will become validated for clinical usefulness in the foreseeable future [16–18].
The use of in vivo limb biomechanical measures to determine the mechanical properties of injured tendons does show some promise as a monitoring method in the horse. A study demonstrated significant correlation between an in vivo limb stiffness index and in vitro SDFT stiffness. This correlation means that it is possible to relate in vivo biomechanical measurements to tendon stiffness, which can be used to monitor tendon healing. Further work is required to translate this finding into a useful clinical monitoring tool to decide when a horse is able to return to athletic work or its usefulness as a tool for assessing tendon therapies .
Tendon cell and matrix biology
Tendons show a complex response to development, training, ageing and injury. Data have demonstrated a complex proteinase response in tendon cells at differing stages of pathology and identified that tendon cells differentiated into a cartilage-like phenotype in chronic tendon disease [20,21]. It has also been demonstrated that tendons themselves contained a small population of mesenchymal progenitor cells that could be differentiated into a number of mesenchymal lineages . A number of studies have compared the matrix and cell biology of the frequently injured high strain SDFT, with the rarely injured low strain common digital extensor tendon (CDET). An initial study assessed differences in matrix turnover and cell phenotype between these tendons and, somewhat surprisingly, discovered that there was in fact much lower collagen metabolism in the SDFT in comparison to the CDET and concluded that reduced or inhibited collagen turnover in the SDFT may account for the high level of degeneration and subsequent injury compared with the CDET . The same researchers extended their investigations to determine how ageing affects matrix turnover and cell phenotype in SDFT and CDET and from their data proposed that the increased susceptibility to injury in older individuals results from an inability to remove partially degraded collagen from the matrix leading to reduced mechanical competence .
Cells in tendons exist as large networks that communicate with each other in part by intercellular communication sites known as gap junctions. Gap junction intercellular communication (GJIC) is necessary for strain-induced collagen synthesis by tendon cells. It was hypothesised that gap junction proteins, connexin 43 and 32 (Cx43; Cx32), GJIC and associated collagen expression levels differ in the SDFT and CDET and differentially alter with development in these 2 tendons. It was determined that both cellularity and gap junctions decreased significantly with skeletal maturity in both the SDFT and CDET. In the adult SDFT there was significantly less collagen produced in comparison with the immature tendon, although in the CDET collagen production conversely increased subsequent to skeletal maturity. It was concluded that in the commonly injured energy storing SDFT that whilst the cellularity of the SDFT decreased following skeletal maturity, a failure of the remaining tenocytes to show a compensatory increase in GJ expression and collagen synthesis may explain why cell populations are not able to respond to exercise and to repair microdamage in some adult athletes . Expression of the Cx43 and Cx32 proteins per tenocyte were found to be significantly higher in a fetal group compared with all other age groups in both CDET and SDFT. Higher levels of cellularity and gap junction protein expression in immature tendons were thought to relate to requirements for tissue remodelling and growth during development .
There has been continuing interest relating to the role of exercise in improving tendon quality through development and the potential role of exercise in hastening tendon degeneration and ultimately failure in the skeletally mature animal. In particular, the subject of early conditioning exercise prior to skeletal maturity and the inability of skeletally mature tendon tissue to respond positively to training has been a popular hypothesis in recent years [26,27].
One study addressed the question whether high-intensity exercise induced degenerative change in the injury-prone SDFT but not in the CDET. Tendons from trained and untrained horses were compared following an 18-month training programme. High-intensity training resulted in a significant decrease in the glycosaminoglycan content in the SDFT, but no change in collagen content, despite a decrease in collagen fibril diameters. There were no signs of degeneration or change in mechanical properties of the SDFT with training. The CDET had lower water content following training and a higher elastic modulus. The authors concluded that long-term training in skeletally mature individuals results in changes that suggest accelerated ageing in the injury-prone SDFT and adaptation in the CDET. However, the training imposed on these individuals did not result in any active pathology or loss of mechanical strength .
There have been a number of investigations into the role of exercise on the developing skeleton. A controlled study of exercise in foals aged 2–15 months determined that the exercised foals had larger tendons when the tendon CSA was determined by ultrasound at some but not all the time points. There was also a significantly greater rate of increase in tendon CSA as a function of time in the exercise compared with the control group. These data suggest that tendon development could be modulated by exercise during growth in the horse . Further data from these authors investigated the effect of exercise in development on CDET and SDFT mechanical properties, molecular composition and morphology. Whilst there were matrix alterations identified in the CDET in the exercised group, matrix properties and mechanical properties were not significantly changed by the conditioning exercise programme in the SDFT by exercise. It was concluded that exercise in immature horses failed to augment the development of the SDFT over and above that induced by normal pasture exercise .
A large experimental study in New Zealand aimed to investigate whether conditioning exercise during development had an effect on skeletal development and the workload and clinical injury rate in the animals' subsequent 2- and 3-year-old racing careers. Data from the study determined that those animals that received conditioning exercise as foals had no differences in orthopaedic disease compared with the conventionally trained animals. There was a tendency for the conventionally trained animals to show signs of orthopaedic injury sooner than the early trained animals. The authors concluded that subjecting Thoroughbred foals to conditioning exercise early in life did not affect the animals' subsequent 2- and 3-year-old racing careers . This study specifically looked at tendon development on its own in the 2 groups of early vs. conventionally trained animals and it was concluded that there was no conclusive evidence for a structural adaptive hypertrophy of the SDFT, probably because the regimen was insufficiently rigorous or because spontaneous pasture exercise may induce maximal development of energy storing tendons .
In the last decade or so, there has been a huge amount of clinical and scientific interest in the application of regenerative-based therapies to improvement treatments for TLIs. The rationale of this has been to improve the quality of repair in injured tendons more towards biologically relevant tendon tissue and away from formation of scar tissue within the tendon that leads to both an inferior athlete and a tendon that has a higher propensity to injury. Such therapies have had a rapid rise in clinical popularity and whilst having an excellent scientific rationale, clinical evidence supporting their use has frequently been weak .
Techniques have used either cell therapy  or the administration of regenerative growth factors, which have either been sourced endogenously or exogenously [35,36]. A variety of sources of cells have been used including bone marrow derived, fat derived and attempts to derive embryonic stem cells and induced pleuripotent stem cells (iPS) for such purposes [37–45] (Figs 1, 2).
A preliminary study used a small number of experimental horses to determine whether autologous and allogenic mesenchymal progenitor cells would survive after injection into a tendon and whether they would be incorporated into tendon tissue without causing an inflammatory response. This study demonstrated that they could determine the fate of exogenous cells once they had been injected into the SDFT. The data indicated the cells could both remain localised to a surgically created lesion but could also integrate into healthy tendon. Furthermore, it appears that both autologous and allogeneic cells could be used without stimulating an undesirable cell mediated immune response from the host . There has been recent publication of a clinical study of the use of bone marrow derived mesenchymal stromal cells in TLI that compares the effects, particularly relating to re-injury rates with historic control populations. The data do indicate an apparent improvement in re-injury rates in cell treated horses in comparison to the historic control populations, which would support the use of such therapies . However a recent experimental study using cell-based therapy failed to identify a positive response relating to matrix synthesis with treatment .
Whilst over the last decade there has been considerable advances in understanding the biology of TLIs in the Thoroughbred racehorse and much of these data have had a great deal of clinical relevance, tendon injuries continue to be one of the most significant economic and welfare issues facing racing. Many more questions need to be answered relating to such injuries. In particular, the role of specific training regimens and training and racing surfaces, both during development and later in an animal's racing career need much further research. Several studies have shown a major effect of different trainers on the risk of injury and this effect certainly demands further scrutiny. Whilst the effect of training type and intensity both during development and ageing are undoubtedly important, novel methods of investigating these effects are required as controlled experimental training studies are expensive and frequently lack sufficient power. Standardised outcome measures are required to assist with clinical trials to assess novel therapies in a practical and cost-effective manner, as well as in the development of better clinical diagnostics. There is still much lacking in our knowledge of basic tendon cell and mechano-biology relating to development and ageing and whilst such mechanistic investigations may not be immediately clinically relevant, such research can underpin translational developments. Undoubtedly the next decade is going to be an exciting time both for researchers interested in tendon biology, as well as for clinicians dealing with the injured racehorse, as there are likely to be major scientific and clinical advances relating to tendons in the short to medium term.
Conflicts of interest
Horserace Betting Levy Board's (HBLB's) Veterinary Advisory Committee commissioned and sponsored this article as part of a series summarising progress made in areas relating totheir priorities for research funding. The author and other workers at his university hold current and previous research grants funded by the HBLB.
The EVJ is delighted to publish HBLB's Advances in Equine Veterinary Science and Practice Review Series in recognition of the major contribution that HBLB research and educational funding has made to the health and welfare of the Thoroughbred.
Source of funding
The author would like to thank the HBLB and the Horse Trust for generously funding some of the authors research cited in this review.
The author would like to thank all colleagues, veterinary surgeons, scientists and trainers who contributed to the cited studies.