Equine Veterinary Journal ISSN 0425-1644 4 DOI: 10.1111/evj.12334

Caudal lumbar vertebral fractures in California Quarter Horse and Thoroughbred racehorses E. M. COLLAR*†‡, R. ZAVODOVSKAYA†§, M. SPRIET‡, P. L. HITCHENS†¶, T. WISNER†, F. A. UZAL** and S. M. STOVER†,†† †

J.D. Wheat Veterinary Orthopedic Research Laboratory, School of Veterinary Medicine, University of California, Davis, USA Department of Surgical and Radiological Sciences, School of Veterinary Medicine, University of California, Davis, USA § Comparative Pathology Graduate Group, School of Veterinary Medicine, University of California, Davis, USA ¶ One Health Institute, School of Veterinary Medicine, University of California, Davis, USA **California Animal Health and Food Safety Laboratory, School of Veterinary Medicine, University of California - Davis, San Bernardino, USA †† Department of Anatomy, Physiology and Cell Biology, School of Veterinary Medicine, University of California, Davis, USA. ‡

*Correspondence email: [email protected]; Received: 19.03.14; Accepted: 08.08.14

Summary Reasons for performing study: To gain insight into the pathophysiology of equine lumbar vertebral fractures in racehorses. Objectives: To characterise equine lumbar vertebral fractures in California racehorses. Study design: Retrospective case series and prospective case–control study. Methods: Racehorse post mortem reports and jockey injury reports were retrospectively reviewed. Vertebral specimens from 6 racehorses affected with lumbar vertebral fractures and 4 control racehorses subjected to euthanasia for nonspinal fracture were assessed using visual, radiographic, computed tomography and histological examinations. Results: Lumbar vertebral fractures occurred in 38 Quarter Horse and 29 Thoroughbred racehorses over a 22 year period, primarily involving the 5th and/or 6th lumbar vertebrae (L5–L6; 87% of Quarter Horses and 48% of Thoroughbreds). Lumbar vertebral fractures were the third most common musculoskeletal cause of death in Quarter Horses and frequently involved a jockey injury. Lumbar vertebral specimens contained anatomical variations in the number of vertebrae, dorsal spinous processes and intertransverse articulations. Lumbar vertebral fractures examined in 6 racehorse specimens (5 Quarter Horses and one Thoroughbred) coursed obliquely in a cranioventral to caudodorsal direction across the adjacent L5–L6 vertebral endplates and intervertebral disc, although one case involved only one endplate. All cases had evidence of abnormalities on the ventral aspect of the vertebral bodies consistent with pre-existing, maladaptive pathology. Conclusions: Lumbar vertebral fractures occur in racehorses with pre-existing pathology at the L5–L6 vertebral junction that is likely predisposes horses to catastrophic fracture. Knowledge of these findings should encourage assessment of the lumbar vertebrae, therefore increasing detection of mild vertebral injuries and preventing catastrophic racehorse and associated jockey injuries. Keywords: horse; racehorse; lumbar; spine; fracture; vertebra; spondylosis

Introduction The racing industry is adversely affected by musculoskeletal injuries of racehorses. Morbidity and mortality have detrimental effects on equine welfare, the livelihood of people supported by racing, the betting public and the public view of the racing industry [1,2]. Catastrophic breakdowns increase the risk of injury for jockeys and exercise riders [3]. Spinal pathology in the thoracolumbar and lumbosacral regions is a source of back pain and poor performance in horses [4–8]. Due to the inaccessibility of the lumbar spine, the ability of the veterinarian to recognise and manage horses with lumbar injury has been significantly restricted [7]. Vertebral fractures occur during racing, most commonly in the lumbar region [9,10]. Given that lumbar vertebral fractures are not typically related to external trauma, and mild repetitive overuse injuries often predispose to catastrophic injuries in racehorses [10], it is possible that lumbar vertebral fracture occurs secondary to pre-existing abnormalities, as has been previously theorised [11]. The objectives of this study were to document the frequency and nature of catastrophic lumbar vertebral fractures in California Quarter Horse and Thoroughbred racehorses, the frequency of associated jockey injury, and evidence of chronic pathological changes in fractured lumbar vertebrae.

Materials and methods Retrospective review From 1 January 1990 until 31 December 2012, horses that died on racetracks under the jurisdiction of the California Horse Racing Board (CHRB) were necropsied by California Animal Health and Food Safety Laboratory System (CAHFS) veterinary pathologists as part of the CHRB Equine Veterinary Journal 47 (2015) 573–579 © 2014 EVJ Ltd

Postmortem Program. After 1 July 2011, musculoskeletal specimens from necropsied horses that died because of a musculoskeletal injury were also examined in detail in the J.D. Wheat Veterinary Orthopedic Research Laboratory (VORL) at the University of California, Davis for the CHRB Racing Injury Prevention Program. Race dataa and necropsy data obtained retrospectively from CAHFS necropsy reports (1990–2012) for horses that had a diagnosis of lumbar vertebral fracture and VORL reports (July 2011 to Dec 2012) were examined retrospectively by one author (E.M.C.). For jockeys who fell due to catastrophic injury of the horse in association with a race, jockey injury data were acquired from jockey accident reports submitted to the CHRB by racetrack stewards from 1 January 2007 until 31 December 2012. A jockey injury was defined as an incident where stewards declared the jockey as unfit to ride or the jockey was transported to the hospital. Race videosb were viewed for horses that incurred fracture during a race after 1 December 2007.

Prospective study of lumbar vertebral fracture cases and controls Detailed examinations were performed on lumbosacral specimens from racehorses that incurred a primary lumbar vertebral fracture between 1 July 2011 and 31 December 2012 (cases) and from a convenience sample of racehorses that died during the same time frame for reasons other than vertebral fracture (controls). Specimens were manually cleaned of soft tissue and assessed using visual, radiographic (Eklin EDR-3 unitc; OsiriX software v.5.6d), computed tomographic (CT) and histological (decalcified, haematoxylin- and eosin- stained sections) evaluations. Specimens were imaged with a multislice CT scanner (GE LightSpeed 16e; 120 kV and 200 or 250 mA, 0.6 mm contiguous slices), reconstructed using a bone algorithm, reviewed (OsiriX software v.5.6d) by 2 authors (E.M.C. and M.S.) and quantitative data collected by one author (E.M.C.).

573

E. M. Collar et al.

Caudal lumbar vertebral fractures in racehorses

Vertebral body (L4, L5 and L6) heights and widths were measured at one-half the measured craniocaudal length. Height was measured at the midline, and width was measured at one-half the measured height. Vertebral body average bone density (in Hounsfield units, HU) was measured in dorsal, central and ventral trabecular bone regions of interest from a single transverse slice (Supplementary Item 1) taken in 3 standardised locations within L4, L5 and L6: 1 cm caudal to the cranial aspect of the cranial endplate, at one half the measured length, and 1 cm cranial to the caudal aspect of the caudal endplate. Boundaries of the regions of interest were drawn manually using discrete landmarks, identified through training on a subset of images, so that the largest region of trabecular bone within the vertebral body could be captured, while excluding cortical bone tissue, discrete vascular channels, physes and fracture lines.

Analysis of data from the case–control study The effects of horse status (case or control), vertebra (L4, L5 or L6) and region of interest (9 regions within a matrix of dorsal, central and ventral; cranial, middle and caudal sites) on bone density and on vertebral body dimensions were assessed using general linear mixed models (SAS 9.3, PROC MIXED)f. The most caudal 3 lumbar vertebrae were designated L4, L5 and L6 in horses with 5 or 6 total lumbar vertebrae for the purposes of comparison. Horse status, vertebra and region of interest were treated as fixed effects. Horse was treated as a random effect. Horse body weight and interaction effects were also examined, but excluded due to statistical nonsignificance (P>0.05). Least-squares means and pairwise differences were assessed. A P value of ≤0.05 was considered statistically significant for all tests.

Results Retrospective review of lumbar fracture case frequency and signalment From 1 January 1990 until 31 December 2012, 38 Quarter Horse and 29 Thoroughbred racehorses that died on racetracks under the jurisdiction of the CHRB were diagnosed with a lumbar vertebral fracture, some previously reported [9]. Of these horses, 33 Quarter Horses (87%; 15 2-year-olds, 13 3-year-olds, four 4-year-olds and one 5-year-old; 11 geldings, one colt and 21 females) and 14 Thoroughbreds (48%; two 2-year-olds, eight 3-year-olds and four 4-year-olds; 2 geldings, 3 colts and 9 females) died from a primary fracture affecting one or both of the most caudal 2 lumbar vertebrae (hereafter referred to as L5–L6, although the vertebrae would be L4 and L5 in horses with 5 lumbar vertebrae). Of the remaining 15 Thoroughbreds, 7 horses died from a primary fracture of other lumbar vertebrae, for a total of 21 (72%) Thoroughbred lumbar vertebral fractures resulting in death. The remaining 8 horses had a lumbar vertebral fracture secondary to another musculoskeletal cause of death. Of the 33 Quarter Horses and 14 Thoroughbreds that died from a primary L5–L6 fracture, 30 Quarter Horse (91%) and 6 Thoroughbred (43%) fractures occurred in a race. The number of starts ranged from zero to 19 for Quarter Horses and from zero to 21 for Thoroughbreds. From 1 July 2011 until 31 December 2012, 45 Quarter Horse and 229 Thoroughbred racehorses died from musculoskeletal reasons. Six Quarter Horses (13%) and 2 Thoroughbreds (1%) died from lumbar vertebral fractures (6 at L5–L6). Lumbar fracture was the third greatest musculoskeletal cause of Quarter Horse deaths (Table 1).

Retrospective review of jockey data In a 6 year period (2007–2012) when jockey injuries were recorded, 18 Quarter Horses and 11 Thoroughbreds sustained a primary lumbar vertebral fracture, 16 Quarter Horses (89%) and 8 Thoroughbreds (73%) at L5–L6, some of which have previously been reported [3]. Of the L5–L6 fracture incidents, 14 Quarter Horses (88%) and 4 Thoroughbreds (50%) sustained the fracture during a race. For race-day incidents, 10 Quarter Horse (71%) and 2 Thoroughbred (50%) incidents involved the jockey falling or coming down with the horse. Eight Quarter Horse jockey falls (57% of L5–L6 race-related Quarter Horse fractures) resulted in jockey injury.

574

TABLE 1: The regional frequencies (expressed as percentages) of catastrophic musculoskeletal failure for the 45 Quarter Horse racehorses that died between 1 July 2011 and 31 December 2012 on racetracks under the jurisdiction of the California Horse Racing Board Musculoskeletal cause of death

Percentage of Quarter Horse deaths

Fetlock Hoof/laminitis Lumbar vertebrae Carpus Forelimb proximal to the carpus Soft tissue Third metacarpal bone (not involving the fetlock) Pastern luxation Tarsus Pelvis

24% 22% 13% 11% 7% 7% 4% 4% 4% 2%

Race-associated fracture incidents were visible in 8 videos. Five horses fell and somersaulted, 2 lost use of their hind end and threw their riders, and one horse pulled up uneventfully.

Case–control anatomical findings Lumbosacral specimens were examined from 5 Quarter Horses (four 2-year-olds and one 4-year-old; 3 geldings and 2 females) and one Thoroughbred (3-year-old, female) that incurred a primary L5–L6 fracture (cases) and from 3 Quarter Horses (one 2-year-old and two 3-year-olds; one gelding and 2 females) and one Thoroughbred (2-year-old, female) that died for reasons unrelated to vertebral fracture (controls). The number of vertebrae in the lumbar region could be determined in only 2 Quarter Horse fracture cases and 3 Quarter Horse control specimens because the cranial lumbar vertebrae were not submitted for examination with the fractured lumbar segments for the other case and control specimens. One control specimen had only 5 lumbar vertebrae (Fig 1). The dorsal spinous processes of L5 and L6 were absent or incomplete in 2 fracture cases. The case with an incomplete L5 dorsal spinous process had a dorsal portion that was held in anatomical position by an abnormal bone attachment to the dorsal spinous process of L4, with only a soft tissue attachment to a short protrusion of the L5 dorsal spinous process where a normal dorsal spinous process would originate on the dorsal lamina (Fig 1). Intertransverse articulations were present in all specimens between L5–L6 and L6–S1, but in only 3 fracture specimens and one control specimen between L4–L5. The L6 transverse processes were peripherally blunted and the transverse processes of L5 and L6 were fused in one control (Supplementary Item 2). Subjectively, all Quarter Horse specimens appeared to have wider, more transitional (lumbar vertebral body shape more like a sacral body shape) caudal lumbar vertebrae than the Thoroughbred specimens. Vertebral height decreased from L4 to L6 in cases and controls (L4 to L5 P

Caudal lumbar vertebral fractures in California Quarter Horse and Thoroughbred racehorses.

To gain insight into the pathophysiology of equine lumbar vertebral fractures in racehorses...
23MB Sizes 1 Downloads 13 Views