Competition, regardless of discipline, exposes the modern equine athlete to the risk of injury, which contributes significantly to days lost from competition and training as well as wastage of horses (Stover, 2003). Tendon injuries are commonplace in the equine athlete (Dyson, 2002), are reported to be one of the most prevalent forms of musculoskeletal injuries occurring in horses competing in all disciplines (O'Sullivan, 2007;Reardon et al, 2012) and remain a significant cause of wastage in racehorses in particular (Williams et al, 2001; Dyson, 2004;Reardon et al, 2012). Tendonitis is defined as non-traumatic injury to a tendon, but in the horse tendinopathy or tendon disease have been adopted by the veterinary profession as they incorporate the inflammatory cascade responses commonly associated with tendon injury (Smith and McIlwraith, 2012).
Risk factors associated with impending tendon injury are variable. Contributory factors include:
Additionally, injury is also thought most likely to occur when the horse is performing at its maximal level (O'Meara et al, 2010).
Location of tendon injuries
Some tendons are more prone to injury than others; most tendon injuries (97–9%) are reported to occur to the forelimb tendons (Kasashima et al, 2004; Lam et al, 2007), with the superficial digital flexor tendon (SDFT) being the most commonly injured tendon (O'Meara et al, 2010; Reardon et al, 2012). A higher susceptibility to injury occurs in disciplines that involve galloping and jumping, as this places increased strain on tendon units in the distal limb (Williams et al, 2001; Pinchbeck et al, 2004; Thorpe et al, 2010). Tendon injuries in racehorses vary between race type; a 46% tendon injury prevalence has been reported in fat racehorses (Williams et al, 2001; Ely et al, 2004), which increases to 53% in national hunt horses where increased loading occurs as a result of jumping (Pinchbeck et al, 2004). Tendon injury also accounts for 43% of injuries that occur to event horses in training, with 33% of these occurring to the SDFT (Singer et al, 2008).
Level of tendon injury
Tendon injuries can occur at all levels of the tendon, from the insertion to the musculotendinous junction, and may affect the tendon core or peripheral regions (O'Sullivan, 2007). Research suggests that lesions typically occur at or just below the mid-metacarpal level in the ‘core’ or central region of the SDFT, as this area appears to be preferentially loaded and is thought to degenerate more with age and exercise than other areas within the tendon (O'Sullivan, 2007;Patterson-Kane and Firth, 2009). A suggested reason for this increased susceptibility to injury include the small cross-sectional area of the SDFT in the mid-metacarpal region, relative to other regions, which can result in a deficient blood supply to this region experienced during exercise (Patterson-Kane and Firth, 2009).
Diagnosis of tendon lesions
Ultrasonography is the most commonly used technique to diagnose tendonitis and monitor healing as it provides an affordable and non-invasive method that enables repeatable practical and objective assessment of lesions over time (Smith and McIlwraith, 2012). Although ultrasonography is useful in recognition of the site, extent and overall nature of the tendon lesion, evaluation of the lesion prognosis is still difficult and recurrences are frequent. The entirety of the central hypoechoic lesion can be difficult; damage can be hard to delineate as adjacent areas of the damaged tendon will not show the same level of hypoechogenicity (Smith and McIlwraith, 2012). A wide level of variance is observed naturally in equine tendon morphology; therefore, to aid assessment a reference is often sought via ultrasonic evaluation of the contralateral tendon or an area of the affected tendon that has no pathological changes present (Smith and McIlwraith, 2012).
When imaging the distal limb for the purpose of an accurate diagnosis, the region is often divided into specific topographical locations (Ross and Dyson, 2003; Smith and McIlwraith, 2012). One approach is to divide the metacarpus region (comprising the SDFT) into seven zones from the accessory carpal bone to the ergot: 1A, 1B, 2A, 2B, 3A, 3B, 3C respectively, where sections 2A, 2B and 2C represent the mid-metacarpal region (Ross and Dyson, 2003; Figure 1.).
Ultrasonography and histology findings in tendo-nitis and throughout tendon healing have been found to be closely correlated (Smith and McIlwraith, 2012), demonstrating that both are beneficial when assessing tendon damage and monitoring tendon repair. The commonly measured variables to assess the severity of injury include tendon and lesion cross-sectional area, lesion type and location, and assessment of fibre alignment (Dowling and Dart, 2005; Smith and McIlwraith, 2012). An increase in tendon cross-sectional area is reportedly the most sensitive indicator of fibre damage (Genovese et al, 1996; Dowling and Dart, 2005) and for detecting re-injury during the recovery period, when the reintroduction of exercise may have been too excessive (Dowling et al, 2000). This is because the cross-sectional area of the tendon is inversely proportional to collagen content and tensile strength; thus injured tendons present with an increased cross-sectional area on examination (Dowling and Dart, 2005). A good fibre alignment after healing is believed to be indicative of a successful outcome (Dowling et al, 2000; Smith and McIlwraith, 2012). Additional measurements, such as changes in echogenicity, estimation of the cross-sectional area of the damaged region and the summation of the area affected, can be combined to provide a comparative severity score or rating; an example of one system used is the Havemeyer grading system (Table 1; Smith and McIlwraith, 2012).
| Havemeyer grade 1: mild | Havemeyer grade 2: moderate | Havemeyer grade 3: severe | |
|---|---|---|---|
| Tendon ‘volume’ affected* | 0–15% | 16–25% | >25% |
| Lesion size at maximum injury zone | 0<10% | 10–40% | >40% |
| Maximum tendon cross-sectional area (CSA) | <2cm2 | 2–5cm2 | >5cm2 |
*Volume affected = (summation of lesion’s CSA for 7 levels) (summation of tendon CSA for 7 levels) x 100%
Rehabilitation of tendon injuries can vary between 9 and 18 months depending on the severity of the initial damage (Davis and Smith, 2006). As remodelled tendons exhibit reduced elasticity, it is suggested that ‘healed’ tendons will require an increased muscular effort to compensate for this reduction; therefore, a decrease from pre-injury performance is to be expected and this may contribute to the high levels of re-injury observed (O'Meara et al, 2010).
In normal tendons, echogenicity, the relative brightness of the ultrasonographic image, is mainly related to the density of collagen and is caused by the bright specular refection that occurs from the fbres’ interfaces (Van Schie and Bakker, 2000). Fibre disruption as a result of injury causes disorganization of the acoustic interfaces and loss in collagen density; the inflammatory process that accompanies this during the early stages of injury, such as haemorrhage, oedema and cellular infiltration, determine a loss of echogenicity, anechogenicity or hypoechogenicity. During the healing process, fibroplasia and fibre misalignment continue to reduce echogenicity, whereas progressive replacement of granulation tissue by collagen fibres will progressively re-orientate along the lines of stress and will induce a gradual recovery of echogenicity, indicating healing.
Aims of the study
The study aimed to identify if any relationship existed between topographical location of tendon lesions and lesion severity, assessed via ultrasonogra-phy. Lesion location and severity were also examined to evaluate if the discipline the horses competed in before injury or the laterality of the forelimb injury were related.
Methodology
Data collection
Secondary data were collated retrospectively from ultrasound scans of the forelimb SDFT of horses presenting with primary ‘core’ lesions provided by an equine veterinary hospital in the south west of England. To maintain anonymity, the names of the clients were removed and each limb was allocated a number for identification purposes. The sample consisted of 100 SDFT lesions, each deriving from a different individual horse. The ultrasound scans were obtained by multiple veterinarians using a 7.5– 10 MHz linear array transducer. Both transverse and longitudinal scan views were available in the majority of tendons; however some tendons were only scanned in the transverse plane. Thus, for the purpose of this study, data were recorded from scans in the transverse plane only. For each individual limb (n =100) a number of printed scan pictures were provided relating to the specific topographical locations within the distal limb. One method of imaging the distal limb is to divide the limb into zones; this is useful as it provides norms for comparison. The metacarpus, from the accessory carpal bone to the ergot, was therefore divided into seven zones — 1A, 1B, 2A, 2B, 3A, 3B and 3C, respectively — each approximately 4 cm in length in accordance with the methodology described by Ross and Dyson (2003). Each scan represented one of the seven regions within the SDFT for each individual limb to enable the core lesion to be subsequently located. Some of the scan pictures (often those corresponding to 1A or 3C) did not present with any lesion; this was because the lesion did not span across these particular regions, as not all lesions spanned the whole length of the tendon from insertion to the musculotendinous junction.
The seven scan pictures from each limb were analyzed and all regions containing a lesion were recorded. This enabled the region in which the lesion originated and terminated to be identified, indicating the length of the lesion. The limb in which the lesion occurred was also noted — left or right — along with the discipline that each particular horse predominantly competed within. The lesion within each scan picture, corresponding to a specific location, was then allocated a score corresponding to the estimated percentage of the SDFT that the lesion covered within that particular topographical region — 0%, 25%, 50%, 75% and 100%, respectively. Thus each tendon was assigned seven percentage scores, one referring to each region. This scoring system was established in accordance with that used by the veterinarians at the equine hospital as a means of estimating the extent of fibres disrupted to quantify any morphological deviation from normal, defined as 0%.
In addition to this scoring system, each lesion at each location was also allocated a score for echogenic-ity (Table 2 ). Lesions vary in echogenicity depending on the morphological consistency at the time of examination, and the implementation of a scoring system has the potential to improve objectivity when assessing the severity of injury.
| Echogenicity score | Description |
|---|---|
| 0 | Isoechogenic (normal) |
| 1 | Slightly hypoechogenic, mostly echogenic |
| 2 | Mixed echogenicity (50% echogenic and 50% anechogenic) |
| 3 | Mostly anechogenic or totally anechogenic |
Each lesion was then assigned an overall severity score; this was calculated by allocating each of the seven regions of the SDFT affected another score representing the estimated percentage of that region of the SDFT affected by the lesion (Table 3). The sum of the scores for each region was then calculated to provide a comparative value that could be used to evaluate the span of the lesion along the SDFT. For example, a small lesion may only span two regions and score up to a maximum of 6, while an extensive lesion could span all seven regions and could obtain a maximum score of up to 21.
| 0 | 0 |
|---|---|
| 25 | 1 |
| 50 | 2 |
| 75 | 3 |
Data analysis
Descriptive statistics identified the prevalence of each lesion’s size and echogenicity for each of the topographical regions. Chi-squared tests ascertained if differences existed between the prevalence of lesions observed in the right and left forelimbs of the total sample population; analysis of variance (ANOVA) tested if significant differences were present between the average lesion severity score within the three disciplines. Subjects were also subgrouped for analysis according to the discipline they competed in: racing (n =65); eventing (n =13); and ‘other’ (n =22). A series of paired t -tests (two-tailed) compared the average lesion severity score between the left and right limbs in each of the discipline categories.
Results
Lesion distribution and ‘percentage’ affected
The topographical location of SDFT lesions appeared to be related to their severity (Table 4). The proximal and distal regions of the SDFT were largely unaffect-ed by lesions, with most being present in the mid-metacarpal regions. The prevalence of large lesions, where 75% of the SDFT was covered by the lesion, were also predominately located in the mid-metacar-pal regions.
| Topographical location of the SDFT | Number of cases that occurred within each estimated percentage region | ||||
|---|---|---|---|---|---|
| 0% | 25% | 50% | 75% | ||
| 1A | 65 | 25 | 2 | 8 | |
| 1B | 40 | 30 | 12 | 18 | |
| 2A | 22 | 30 | 22 | 26 | |
| 2B | 16 | 39 | 22 | 23 | |
| 3A | 26 | 39 | 22 | 13 | |
| 3B | 55 | 28 | 8 | 9 | |
| 3C | 79 | 12 | 5 | 4 | |
SDFT, superficial digital fexor tendon
Lesion distribution and echogenicity
For the proximal and distal regions of the limb, 1A and 3C, a large number of tendons showed no sign of lesion development and recorded an echogenicity score of ‘0’, indicating isoechogenic or normal tendon tissue. However, the mid-metacarpal regions, 2A, 2B and 3A, demonstrated a high prevalence in limbs presenting with type three (anechogenic) or type two (mixed echogenicity) lesions recorded (Figure 2. )
Right versus left forelimb lesion distribution
No significant difference was found for this population with regards to the distrubution of lesions between right and left forelimbs (chi-squared test, p >0.05); this pattern was repeated when the discipline subgroups were analyzed (t -test, p >0.05).
Severity of lesions
Interestingly in this population, event horses presented with a trend for SDFT lesions of increased severity when compared with the racing and ‘other’ categories (Figure 3. ), although this relationship was subsequently found to be non significant (ANOVA,>0.05).
Discussion
Tendon injuries within this population span across the entirety of the SDFT and are not concentrated solely in one topographical location. However, the severity of the lesions in relation to size and echo-genicity do differ throughout the seven topographical locations surveyed; less severe lesions occurred in the proximal and distal regions, and more severe lesions within the mid-metacarpal region of the SDFT. Echo-genicity scores for lesions within the mid-metacarpal region were also highest in these areas and thus displayed the greatest amount of fibre disruption. This is consistent with research (Dowling and Dart, 2005; Patterson-Kane and Firth, 2009) and adds weight to the hypotheses that increased loading combined with a reduced cross-sectional area, hypoxia and hyperthermia related to poor perfusion in the mid-metacarpal regions of the SDFT contribute to enhanced fibril damage and produce a more severe and extensive lesion, relative to other regions. However, it should be noted that fibril disruption did occur at all topographical levels in the scans examined, and therefore during diagnosis care should be taken by practitioners to ensure that the entirety of the tendon is examined ultrasonographically.
In this population of horses, no apparent differ-ences existed between the severity score of tendon lesions within disciplines or between the right and left leg scans. This may be because of the convenience sampling method undertaken as all horses in the study had to present with clinical tendon pathology to be included, thus negating the influence of their main discipline as a positive bias is present. However, 65% of the convenience sample assessed were racehorses, which could also suggest that horses competing in this discipline are generally at a higher risk of SDFT injury. In this population, 78% of subjects were engaged in racing or eventing before injury; these disciplines regularly include galloping and jumping during training, which have been previously identified as increasing strain within the tendon unit and are thought to act as causal factors for injury to the SDFT (O'Meara et al, 2010; Reardon et al, 2012). Training regimens for horses within the ‘other’ group were unknown and also may have included known risk factors for tendon injury; perhaps similarities in work undertaken may explain the lack of difference between disciplines reported here.
Why are tendon lesions more prevalent and more severe in the mid-metacarpal regions of the SDFT?
Birch and colleagues (2002) stated that the total amount of collagen in the mid-metacarpal level is similar to that of most other regions of the SDFT, which implies that it is not necessarily a weaker area. This led researchers to find alternative explanations for the reported increase in susceptibility to injury of the mid-metacarpal region. The hierarchical structure of the tendon is such that, during a loading cycle, some fibrils may experience higher strains and therefore be at higher risk of damage than others (Dowling and Dart, 2005; Smith and McIlwraith, 2012). Physiological strains usually result in tendon adaptation; however, unusually high strains may cause deformation and damage to the fibril, resulting in tendonitis (Dowling and Dart, 2005; Patterson-Kane and Firth, 2009; Smith and McIlwraith, 2012). The mid-metacarpal level in the ‘core’ or central region of the SDFT appears to be preferentially loaded during exercise, and this region is also thought to degenerate more with age and repeated exercise than other areas within the tendon (Patterson-Kane and Firth, 2009). Fatigue and the associated increase in asymmetrical loading in the limbs (Smith and McIlwraith, 2012) or the impact of distal limb conformation such as hoof balance and resultant changes in both SDFT and deep digital flexor tendon loading (Jorgensen and Genovese, 2003) also exert the potential to increase the risk of injury in this region within the competition horse.
The relatively smaller cross-sectional area of the mid-metacarpal regions and associated poor capillari-zation have been postulated to predispose the region to injury (Kraus-Hansen et al, 1992; Thorpe et al, 2010). The poor blood supply within the tendon is thought to result in localized hyperthermia and ischaemia; generated heat is dissipated relatively slowly, exposing the tendon cells to a high temperature over a longer period (Thorpe et al, 2010); it should also be noted that many competition horses wear protective boots or leg wraps that could enhance these temperatures. While these temperature rises do not always result in teno-cyte lysis, they may still induce matrix damage (Dowl-ing et al, 2000; Patterson-Kane and Firth, 2009); at temperatures above 45°C equine tenocytes produce more pro-inflammatory cytokines, thus increasing the production of matrix-degrading enzymes.
Conclusions
This work aimed to review a sample population of equine SDFT lesion pathology to identify if relationships between topography and echogenicity of lesions could measure severity. Higher cumulative scores did provide an objective assessment of tendon lesion severity; lesions here predominately presented with the highest severity primarily in the mid-meta-carpal regions. However, lesions also extended to the proximal and distal regions of the SDFT, and within these areas variable echogenicity, and thus severity, was observed. This indicates that the entire length of the SDFT should be subjected to ultrasonographic examination during initial diagnosis and throughout rehabilitation. No statistical differences were found between tendon lesion severity within the equestrian disciplines or between right and left SDFT injury; however, further research in a wider population is warranted to confirm these findings.