Retrospective analysis of oblique and straight distal sesamoidean ligament desmitis in 52 horses.

BACKGROUND
Injuries to the oblique (ODSL) or straight (SDSL) distal sesamoidean ligaments are a recognised cause of distal limb lameness in the horse. However, there are only limited publications addressing common diagnostic features and prognosis.


OBJECTIVES
1) Report findings on ultrasonography and standing low-field magnetic resonance imaging (sMRI) in horses with ODSL or SDSL injury; 2) Identify clinical variables associated with lesion type; 3) Identify factors associated with return to soundness in horses with ODSL or SDSL injury.


STUDY DESIGN
Retrospective case series.


METHODS
Horses with a primary diagnosis of ODSL or SDSL injury confirmed with a combination of diagnostic analgesia and detection of a lesion on imaging (ultrasonographic +/- advanced imaging) were included. Return to soundness and performance follow-up data were obtained.


RESULTS
Fifty-one horses were included. SDSL injuries were more common in the forelimb (13/21, 62%) while ODSL injuries had equal frequency in fore- (15/30, 50%) and hindlimbs (15/30, 50%). ODSL injuries were more likely than SDSL injuries in the proximal third of the ligament (OR 13, 95% CI 2.3-74.3; p=0.004) and often presented with periligamentar swelling (20/30;67%) and focal pain (22/28;79%). Lesions were frequently detected using ultrasonography (35/42;83%) and sMRI examination (18/25;72%). Only 27/49 cases (55%) returned to soundness, with only 15/49 (31%) returning to intended use. There were no significant associations between outcome and clinical features or treatment.


MAIN LIMITATIONS
There was no comparative "gold standard" to validate lesions such as high-field MRI or histopathology.


CONCLUSIONS
In contrast to previous studies, ODSL and SDSL injuries were readily identified ultrasonographically using appropriate views, and with sMRI. Given the limited availability, cost and general anaesthetic risks associated with high-field MRI, more focus should be placed on optimising the ultrasonographic examination. Owners of affected horses should be informed of the guarded prognosis for return to full use.


| INTRODUC TI ON
The distal sesamoidean ligaments (DSL) consist of three paired (2/27) and 20% (9/45) of cases, respectively, with the remainder diagnosed under general anaesthesia with the use of high-field (1-1.5T) MRI. [1][2] However, both of these studies failed to describe specific ultrasonographic imaging planes optimised for detecting DSL injury. [1][2] Given the risks of general anaesthesia, limited availability and high costs of high-field MRI, sMRI would be the more clinically applicable, but only one study reports the use of low-field (0.27T) sMRI to diagnosis ODSL desmitis in just three racehorses. 4 Furthermore, the prognosis for these injuries is also unclear, as the previously reported small case series have provided a highly variable (40%-76%) prognosis for a return to full athletic function. 1,3,4 Previous studies have not attempted to identify factors associated with return to soundness or clinical variables associated with lesion type.
The objectives of this exploratory study were therefore to: 1) Report findings (lesion detection rate and common imaging abnormalities) on ultrasonography and sMRI in horses with ODSL and SDSL injury; 2) Identify clinical variables associated with lesion type (ODSL or SDSL) and 3) Identify factors associated with return to soundness in horses with ODSL and SDSL injury.

| MATERIAL S AND ME THODS
Case records from a single referral hospital over a 17-year period (2002-2018) were reviewed. Only horses with a primary diagnosis of unilateral ODSL or SDSL injury confirmed with a combination of diagnostic analgesia and detection of a lesion on imaging (ultrasonographic ± advanced imaging) were included in the study.
All clinical assessments were performed by one of three experienced clinicians (ACVS/ECVS diplomat) and sMRI examinations were analysed by a diagnostic imaging specialist (ECVDI diplomat/associate). Lameness was subjectively assessed using a standard 0-10 scale with 0 indicating the horse is sound and 10 indicating non-weight bearing lameness. 7 More recent select cases (after 2012) were additionally examined using objective gait analysis (MTw; Xsens and The Mathworks). All horses underwent a combination of regional, intrathecal and intra-articular anaesthesia unique to each individual case. Treatment was categorised as conservative (rest and exercise management only) or including additional therapies (extracorporeal shockwave therapy, regenerative therapies and tenoscopic debridement). Imaging protocol differed slightly between clinicians but involved subdividing the pastern region into four anatomical levels; proximal third (P1a), intermediate third (P1b) and distal third (P1c) of the proximal phalanx together with the proximal aspect of the middle phalanx (P2). 8 The SDSL was examined both in transverse and longitudinal views using a palmar/plantar approach. Care was taken to avoid misinterpreting the normally present hypoechoic region seen within the SDSL on midline near its insertion on P2 as pathology, which is variable between animals and also visible on a number of MRI sequences. 6,[8][9][10] The ODSLs were also examined in transverse and longitudinal section using a palmaro/plantarolateral/medial approach. The most proximal images were obtained immediately distal to the proximal sesamoid bones (PSBs) from the palmaro/plantaro medial/lateral aspects of the limb with the transducer angled upwards so that the ultrasound beam was orthogonal to the DSL fibres. Longitudinal sections were aligned in an oblique abaxial (proximal) to axial (distal) from the same palmaro/ plantaro-lateral/medial orientation ( Figure 1). Doppler examination, when performed, was undertaken with the limb non-weight Owners of affected horses should be informed of the guarded prognosis for return to full use.

K E Y W O R D S
horse, lameness, distal sesamoidean ligament, ultrasonography, tenoscopy bearing. Lesion location was reported as involving the proximal third, distal two thirds (along its attachment) or full length of the respective structure. In all cases, comparative images of the contralateral limb were obtained.

| Standing low-field MRI
A 0.27-Tesla dedicated equine sMRI system (Equine Limb Scanner; Hallmarq Veterinary Imaging, Ltd.) was used to obtain images of the pastern region in select cases since sMRI instalment in 2009. Imaging protocol differed between cases but usually included a combination of T1-weighted gradient recall echo (GRE), T2/T2*-weighted fast spin echo (FSE)/GRE and short tau inversion recovery (STIR) FSE sequences were acquired in the sagittal, frontal and transverse planes.
Exclusion of artefacts, such as "magic angle" effect seen with imaging the proximal aspect of the ODSL, 11 was minimised by use of T2weighted FSE sequences where possible.

| Other imaging modalities
Radiography of the distal limb (including contrast tenography), highfield MRI under general anaesthesia, nuclear scintigraphy and computed tomography were also performed in select cases.

| Rest and rehabilitation protocol
Typically, this consisted of 30-60 days of strict box rest, followed by a further 30-60 days of in-hand (controlled) walking that was gradually increased to 45-60 minutes daily, depending on clinical progress. Following this, a ridden exercise program was completed increasing from walk, to trot and canter. Small paddock turnout (10m x 10m) was advised 4-6 months after the diagnosis was made. Serial clinical and ultrasonographic re-examinations were advised every 6-8 weeks throughout rehabilitation to adapt the program accordingly.

| Additional (non-surgical) treatments
In addition to rest and rehabilitation, some horses were treated with extracorporeal pressure wave therapy (EPWT), intralesional biological therapies (platelet-rich plasma (Veterinary Platelet Enhancement Therapy) (VBS Direct Limited) or mesenchymal stem cells (MSCs) or with intrathecal corticosteroid medication, depending on individual clinician's preference (Table S1).

| Surgery
Tenoscopic debridement was performed in cases where communication of the lesion with the digital flexor tendon sheath (DFTS) was suspected ( Figure 2). The surgical approach, as described by Nixon (1990), 12 was similar in most cases, however, the main difference being some cases were positioned in dorsal rather than lateral recumbency (clinician-dependent). 13 In all cases, tenoscopy was performed using a 4mm 25° forward oblique arthroscope

| Outcome
Outcome data relating to return to soundness and performance was One case (forelimb SDSL) was bilaterally affected and, because determination of a return to soundness would be problematical, this case was excluded from the study for the purpose of statistical analysis.

| Factors associated with lesion type
Univariable screening identified four variables with potential as-    (Table S3). No significant association was found between any of these variables with return to soundness in the final multivariable model (Table 3), although lesion location justifies consideration given that the small numbers leads to low precision in the estimated OR.

| D ISCUSS I ON
In agreement with Sampson et al. (2007), 1 lesions were more commonly seen in the ODSL than in the SDSL, but this difference was less evident in this study (58% vs. 67%). 1 In contrast, Smith et al. of medial and lateral ODSL injuries were also similar to that reported previously; the medial ODSL was affected in the forelimb in 60% of cases compared to 87% described previously and 87% affected the lateral ODSL in the hindlimb compared to 71% reported previously. 1 Biomechanical asymmetry during loading may explain the cause of medial over-representation in the forelimbs and lateral in the hindlimbs as this reflects the loading patterns in these limbs. This is supported by the reported higher frequency of medial suspensory ligament branch injuries in the forelimbs in racehorses, [14][15] although a study in Warmbloods showed no significant association with laterality. 16 Unlike previous studies, palpable abnormalities (periligamentar swelling ± effusion) and focal pain on palpation were associated with ODSL injury, likely due to their abaxial and more superficial location in comparison to the SDSL which is located deep to the DDFT. [1][2][3] Due to the high prevalence shown in our study, swelling and pain in the palpable groove located biaxially immediately distal to the basilar aspect of each proximal sesamoid bone and immediately palmar/ plantar to the medial and lateral edges of the proximal phalanx appears to be a key diagnostic feature of ODSL injuries. This is best ap-  (Figure 2) or diffuses locally to anaesthetise the regional nerves as described for foot diagnostic analgesic techniques. 17 Only 1/26 (4%) cases was fully positive to a palmar/plantar digital nerve block although approximately a third of cases showed some response (30%). Abaxial sesamoid and low 4-point nerve blocks most frequently abolished the lameness, consistent with previous studies. [1][2] As with all regional diagnostic analgesia, a negative result to these blocks can be the result to ineffectively performed blocks.
Ultrasonography was a useful diagnostic tool in this study with  It is not possible from these data to conclusively determine whether sMRI or ultrasonography is superior for detecting DSL pathology as both imaging modalities were not used in every case. In addition, there were unavoidable complicating factors such as nonavailability of sMRI in the first half of the study timeframe and also improvements in our ability to differentiate pathology from artefacts in both sMRI and ultrasonography over the years. However, the high frequency at which DSL pathology was detected using specific ultrasonographic techniques supports our conclusion that the first line imaging modality in cases of DSL injury should be ultrasonography, but if findings are inconclusive or non-diagnostic, sMRI should be considered. However, it is also recognised that not all DSL injuries will be visible with sMRI, in part due to potential susceptibility to magic angle effect within the proximal aspect of the ODSL and lower resolution images acquired with sMRI which may underestimate the extent of the injury. 11 In addition, movement (particularly with hindlimb sMRI scans) and individual horse temperament issues can interfere with the overall quality of image acquisition. Thus, high-field MRI should also be considered for some cases. 18 There was no significant difference in outcome between conservative management alone (51%) and additional therapies (64%).
Although the relatively small case numbers reduce the power of the study to predict more effective treatments, the limited difference suggests more intense treatments do not impact outcome in a major way.
Unlike in previous studies, 1,3 intrathecal corticosteroid medication and ligament splitting was not frequently performed in our cases.
Clinicians should be aware of the possibility of extension of palmar/plantar injuries into the DFTS cavity due to the ligaments' close proximity to the DFTS in this region ( Figure 2) which thereby offer the prospect of surgical debridement. We were able to identify this opening into the sheath cavity pre-operatively in two of our surgical cases using a combination of ultrasonography and contrast tenography. Only a limited number of cases underwent surgery, which prevents any valid statistical analysis, but the success rate was high (80%) if the two horses who did not survive for other reasons are excluded, which encourages further evaluation of this therapeutic approach for such cases.
Limitations of this study include its retrospective nature and insufficient subject numbers to determine which treatments were the most effective. A 'gold standard' was lacking for validating the presence of injury (such as high-field MRI or histopathology), although we believe that the combination of clinical and diagnostic features made the diagnoses sufficiently robust. Potential reasons for the lack of significant association with return to soundness following multivariable analysis are explained by the relatively small study number and low statistical power to detect such potential associations.
There is a possibility of misclassification of return to soundness due to the inherent subjectivity of lameness assessment performed by various referring veterinarians (greater interobserver variation) with only four cases assessed with concurrent objective gait analysis both at the time of diagnosis and at follow-up. In addition, as mentioned previously, although certain ORs are of a high magnitude, the wide confidence intervals obtained in the multivariable analyses indicate low precision of the estimated ORs. Consequently, statistical conclusions drawn from this observational study must be interpreted with caution because of the small sample size. It could be argued that the varied outcomes seen between the current and largest previous study 1 are a result of differences in imaging modalities, where the previous study used a more sensitive imaging modality (high-field MRI) and diagnosed less severe injuries which translated to improved outcomes.y 1 We believe that the prognosis for a return to soundness calculated in this study can be used reliably because it is based on the largest case numbers to date. We acknowledge that this is limited to the population of horses within the referral practice area and shortcomings of imaging modalities used. Owners of horses with SDSL or ODSL injuries should therefore be advised of the guarded prognosis to return to full athletic function.

ACK N OWLED G EM ENTS
We thank all the referring veterinary surgeons for their case contributions, Ms Nancy Brown (nee Ellis) for her work on ultrasonographic assessment of the cross-sectional area of oblique and straight distal sesamoidean ligaments and Dr Ruby Chang for her statistical advice.

CO N FLI C T O F I NTE R E S T S
No competing interests have been declared.

AUTH O R CO NTR I B UTI O N S
A. Hawkins was the principal author and contributed to study design, data collection and analysis and manuscript preparation.
L. O'Leary contributed to study design and data collection. D.
Bolt and A. Fiske-Jackson contributed cases to the study and revision of the manuscript. D. Berner contributed to preparation and revision of the manuscript. R. Smith was the senior author and contributed to overall study design, data analysis and preparation of the manuscript. All authors gave their final approval to the manuscript.

E TH I C A L A N I M A L R E S E A RCH
Research ethics committee oversight not required by this journal: retrospective analysis of clinical data.

I N FO R M E D CO N S E NT
Explicit owner informed consent for inclusion of animals in this study was not stated.

PE E R R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1111/evj.13438.

DATA ACCE SS I B I LIT Y S TATE M E NT
The data that support the findings of this study are available from the corresponding author upon reasonable request.