Clinical and histopathological features in horses with neuroaxonal degeneration: 100 cases (2017-2021)

2.1 Necropsy evaluation

Necropsy examinations, including gross evaluation of the axial musculoskeletal and central nervous systems, were performed by board-certified anatomic pathologists at PADLS New Bolton Center. Tissue samples, including brain and spinal cord, were fixed in 10% formalin, sectioned transversely, and examined for macroscopic abnormalities. Selected tissue sections were routinely processed, embedded in paraffin, sectioned at 5 μm and stained with hematoxylin and eosin (H&E) for microscopic evaluation. Specific sections evaluated varied by case and supervising pathologist but consistently included ≥1 sections of medulla oblongata (at or just rostral to the obex) and multiple sections of spinal cord (including sites of vertebral canal stenosis noted grossly). For most cases, sections of cerebrum and cerebellum also were evaluated. Before final inclusion in the study, a single pathologist (SJB) reviewed the central nervous system (CNS) sections from all horses diagnosed with eNAD/EDM during the designated time period to confirm the presence of eNAD/EDM-like lesions as previously described in the literature.^{1, 2, 6}

Lesions compatible with eNAD included bilaterally symmetrical axonal degeneration in brainstem and spinal cord gray matter (predominantly lateral cuneate nuclei with lesser involvement of medial cuneate and gracile nuclei of the brainstem and thoracic nuclei of the spinal cord) characterized by axonal spheroids with or without vacuolation, chromatolysis, gliosis and lipofuscin pigment. Lesions compatible with EDM included the gray matter lesions described above along with spinal cord white matter degeneration, characterized by dilated myelin sheaths with scattered myelin-laden macrophages (digestion chambers) or spheroids or both in ascending and descending (particularly dorsolateral and ventromedial) white matter tracts of multiple spinal cord segments, most notably in the mid to caudal thoracic spinal cord. Tissue sections also were assessed for other noninfectious and infectious conditions that might otherwise explain the clinical signs, including cervical vertebral stenotic myelopathy (CVSM) and protozoal myeloencephalitis (EPM). Cases were excluded if insufficient tissue was available for examination (eg, brainstem sections lacking both left and right sides or nuclei of interest). In addition, cases with other microscopic lesions (eg, compressive myelopathy) that might otherwise explain the neurologic signs were not included in the study.

2.2 Statistical analysis

Descriptive statistics were performed using Excel (Microsoft 365, Version 16.62).

3.1 Signalment and history

Histologic evaluation of 137 horses with suspected eNAD/EDM was performed. Thirty-seven horses were excluded because of presence of other pathologic findings that could have contributed to the horse's clinical signs or the presence of exceedingly mild degenerative changes. Therefore, 100 horses met the inclusion criteria for the study (Table 1). Median age was 8 years (range, 1-22 years), 80 were geldings, 18 were mares, and 2 were stallions. Warmbloods were overrepresented (72). Other breeds included Thoroughbreds (15), Quarter Horses (5), and 8 horses of other breeds (Table 1).

Seventy-three horses had a history of ataxia. Sixty-nine horses had a history of abnormal behavior under saddle (54) or on the ground (51). Thirty-five horses were performing poorly, and 27 horses had a history of lameness. For most horses (68), the primary presenting complaint was behavioral changes, 37 of which had concurrent features suggestive of ataxia (frequent tripping, episodes of falling). Behavioral changes included unpredictable spooking, dangerous or aggressive behavior toward humans and other horses, unusual excitability or anxiety, explosiveness, abnormal dropping of the penis, unwillingness under saddle, rearing and bolting, periods of disorientation, and vacant expression or unusual calm despite stressful stimuli. Other presenting complaints included ataxia alone (30), behavioral changes alone (25), lameness alone (4), poor performance (3) and uveitis (1). Duration of clinical signs was known for 33 horses, with a median duration of 90 days (range, 1-1068 days).

3.2 Clinical evaluation

Eighty-three horses were presented to the internal medicine service for neurologic evaluation. Seventeen horses were presented to other specialty services for evaluation of poor performance or other complaints. Twenty-eight horses had mentation changes, and 26 horses had abnormal behavior on presentation, characterized by dull mentation, descriptions of vacant expression, aggression, excessive spooking, anxiety, and hyper-reactivity. Horses often alternated between atypical dullness and hyper-reactivity. Six horses had abnormal cranial nerve examination, with delayed or inconsistent menace response (1 horse), weak tongue tone (2 horses), weak tongue tone with muzzle deviation and head tilt (1 horse), and strabismus and head tilt (1 horse). All horses showed signs of general proprioceptive ataxia at presentation. For all horses, gait examination was consistent with cervical or diffuse myelopathy with general proprioceptive ataxia and upper motor neuron paresis of all 4 limbs. One horse also displayed vestibular ataxia, and 1 had multifocal neurolocalization, with a combination of cranial nerve deficits and general proprioceptive ataxia. Horses displayed mild to moderate general proprioceptive ataxia, with an average grade of 2/5 (range, 0.75-3.5/5) on the modified scale.⁹ Median grade of ataxia was higher in the pelvic limbs (2/5; range, 0.75-3.5/5) than in the thoracic limbs (1.75/5; range, 0-3.5/5). Twenty-six horses were lame on admission. Cervical range of motion was decreased in 11 horses. One of 20 horses that underwent ophthalmic evaluation had pigment retinopathy (lipofuscin deposits) on retinal examination.

3.3 Diagnostic imaging

Survey radiographs of the CVC were performed at New Bolton Center in 86 horses, 57 of which had abnormal findings. Of those horses, 50 had osteoarthritis of the cervical articular process joints (CAPJs), and 7 had stenosis of the cervical vertebral canal based on inter- or intravertebral minimum sagittal diameter ratios. One or more intervertebral minimum sagittal diameter ratios were <50% in 6 horses, and ≥1 intravertebral minimum sagittal diameter ratios were <50% in 4 horses. A cervical myelogram was performed in 58 horses, and 14 had narrowing of the dorsal contrast column of >50% with corresponding decrease of the ventral contrast column, suggestive of spinal cord compression (5 horses in a neutral position, 9 horses in extension, and 6 in flexion). Fourteen horses underwent standing CBCT of the CVC, 9 of which were performed postmyelogram. Findings on CBCT included osteoarthritis of the CAPJs in 8 horses, intervertebral foraminal stenosis in 5 horses, and intervertebral disc mineralization in 1 horse. Findings on CBCT postmyelogram suggested spinal cord compression in 3 horses.

3.4 Clinical pathology

Serum amyloid A (SAA) concentrations were measured in 3 horses and were 0 μg/mL in all. Plasma fibrinogen concentrations were measured in 66 horses, with a mean concentration of 353 mg/dL (range, 198-611 mg/dL). Twenty-three horses had plasma fibrinogen concentrations above the normal range (150-375 mg/dL). Serum vitamin E concentrations were evaluated in 60 horses. Mean concentration of serum vitamin E was 4.1 ppm (range, 1.5-9.1 ppm). Six horses had vitamin E concentrations below the acceptable range (2-10 ppm). Cerebrospinal fluid cytology was performed in 74 horses. Median red cell count, nucleated cell count, and total protein concentration were 1 cell (range, 0-22 200 cells), 1 cell (range, 0-34 cells), and 64 mg/dL (range, 27-178 mg/dL), respectively. All samples were clear and colorless except 2, 1 of which was pink and cloudy (because of blood contamination), and 1 of which was xanthochromic and cloudy. Cytologic abnormalities in CSF were detected in 28 samples, including albuminocytologic dissociation in 22 samples, blood contamination in 3, neutrophilic inflammation in 2 and lymphocytic pleocytosis in 1. Serum or CSF or both tests for EPM (S. neurona and N. hughesi) were performed using SnSAG2/4/3 and Neospora ELISA (Equine Diagnostic Solutions, Lexington, KY) in 71 cases. Serum titers (range, <1 : 250-1 : 16 000) for S. neurona were indicative of immunoreactivity in 48 cases. Cerebrospinal fluid titers (range, <1 : 2.5-1 : 160) for S. neurona were indicative of immunoreactivity in 21 cases. Serum to CSF titer ratios for S. neurona were inconsistent with a diagnosis of EPM in all cases (ratio of <100 consistent with active infection). Serum titers (range, <1 : 500-1 : 1000) for N. hughesi were indicative of immunoreactivity in 9 cases. Cerebrospinal fluid titers (range, <1 : 5-1 : 5) for N. hughesi were indicative of immunoreactivity in 1 case. Serum to CSF titer ratios for N. hughesi were inconsistent with a diagnosis of EPM in all cases. Serum or CSF or both median fluorescent intensities (MFIs) for B. burgdorferi were performed using Lyme Multiplex assay (Animal Health Diagnostic Center, Cornell University, Ithaca, NY) in 66 cases. Median serum MFI for outer surface protein A (OspA), outer surface protein C (OspC) and outer surface protein F (OspF) were 291 (range, 34-4099), 104 (range, 4-1605) and 238 (range, 33-6847), respectively. Median CSF MFI for OspA, OspC and OspF were 390 (range, 25-12 402), 123 (range, 7-5718) and 343 (range, 28-17 811), respectively. Serum MFIs were positive for OspA in 5 cases, for OspC in 6 cases, and for OspF in 12 cases. Borrelia burgdorferi antibody results, in combination with CSF cytology results, were inconsistent with a diagnosis of Lyme neuroborreliosis in all cases.¹⁰ Phosphorylated neurofilament heavy (pNF-H) concentration was evaluated on serum and CSF in 27 horses. Serum pNF-H concentrations (mean, 0.37 ng/mL; range, <0.07-1.36 ng/mL; normal, <1.0 ng/mL) were above the normal range in 2 horses, and CSF concentrations (mean, 2.7 ng/mL; range, <0.07-24.42 ng/mL; normal, <3.0 ng/mL) were above the normal range in 8 horses.¹¹ In 1 horse, both serum and CSF results were above the normal range.

3.5 Treatment and outcome

Sixteen horses underwent treatment for EPM before presentation (10 treated with ponazuril, 2 treated with ponazuril and pyrimethamine sulfadiazine and 4 treated with an unknown antiprotozoal). Three horses were treated for suspected neuroborreliosis; 2 with PO minocycline and 1 had no treatment details. Six horses underwent intra-articular treatment of the CAPJs. Six horses were supplemented with vitamin E. Sixty-two horses were euthanized on the first visit to New Bolton Center. For horses that were reevaluated at New Bolton Center, median number of repeat evaluations was 1, and median duration of time between first and last visits was 42 days (range, 5-395 days). Median time from first evaluation to euthanasia was 7 days (range, 1-395 days). For the 37 horses that were reevaluated, 21 horses had progression of ataxia, with a median degree of progression of 0.5 grades (range, 0.5-2) on the modified scale. Twenty horses had progression of behavioral changes. Median grade of ataxia at euthanasia was 2 (range, 0.75-4/5) on the modified scale.

3.6 Necropsy findings

Hematoxylin and eosin-stained sections of brainstem (1-8 sections per case) and spinal cord (6-38 sections per case) were reviewed by a single pathologist (SJB) for the 100 horses included in the study. All 100 horses had bilaterally symmetrical axonal degeneration in the gray matter of the brainstem. Lateral cuneate nuclei were most consistently and extensively affected, with lesser involvement of the medial cuneate, gracile, olivary and lateral vestibular nuclei or reticular formation or both. Gray matter lesions consisted of axonal spheroids that varied in size, number, staining intensity, and presence or absence of clear vacuoles (Figure 1), accompanied in some cases by chromatolytic or necrotic neurons, gliosis, lipofuscin pigment or some combination of these. Twenty-four horses had mild concurrent axonal degeneration in ascending and descending white matter tracts of the spinal cord that was most apparent in the dorsolateral and ventromedial white matter of mid- to caudal thoracic segments. Three horses had gross evidence of cervical vertebral canal stenosis (1 at C4-C5, 1 at C6-C7, and 1 at both C3-C4 and C4-C5), but spinal cord sections adjacent to these sites did not have a pattern of white matter degeneration typical of compressive myelopathy. In the horse with vestibular ataxia (age, 12 years), necropsy evaluation of the middle and inner ear was unremarkable, as were antemortem otoscopic evaluation, endoscopy of the guttural pouches, and repeated computed tomography of the head.

4.1 Signalment

Horses included in our study were a median age of 8 years and were predominantly Warmbloods, representing a different signalment from those horses originally identified with eNAD/EDM. The first report of EDM described young horses of several breeds, but Arabians were overrepresented.^{1, 7} Affected horses displayed ataxia as weanlings or yearlings. A report of Morgan horses with eNAD stated that the majority showed clinical signs by 6 months of age, but some did not display clinical signs until after 3 years of age.² Factors contributing to the overrepresentation of young to middle-aged Warmbloods in our patient population are unknown. Genetic predilection may trigger a later onset of neurodegeneration and hence clinical signs. Previous studies identified antioxidant (vitamin E) deficiency and oxidative stress as risk factors for disease development, and therefore our patients might have undergone a period of oxidative imbalance at a later age than horses in earlier reports.^{12, 13} Finally, horses in our case population might represent mild examples of the neurodegenerative disease spectrum, and clinical signs of mild proprioceptive ataxia and erratic behavior may mistakenly be considered normal for a young horse. Early in life, affected horses might compensate for mild deficits, and be able to train and compete successfully. However, as exercise demands and complexity increase, affected horses may no longer be able to compensate for loss of proprioception, and deficits may become more noticeable.

The overrepresentation of geldings in our population is likely secondary to sampling bias associated with the use primarily of hunters, jumpers, and dressage horses, for which the temperament of geldings often is desired. This bias is reflected in a similar distribution of sex encountered in horses presenting for other performance-related issues such as musculoskeletal disease at New Bolton Center. Although a predisposition for males has not been previously encountered in descriptions of eNAD/EDM, the potential for sex predilection cannot be completely dismissed.³

4.2 Clinical presentation

Unlike the original descriptions of neuroaxonal degeneration in horses, in which there were no reports of aggression or increased reactivity, the horses described in our study commonly displayed these behavioral abnormalities. In fact, 68% of horses were presented for evaluation because of a primary complaint of behavioral changes, and 37% were presented for behavioral abnormalities only. Because of the retrospective nature of our study, it is possible that a higher percentage of horses had behavior changes, but the descriptions were not captured in our medical records. In addition, ataxia typically was mild, with a median grade of 2/5 on the modified scale, and not necessarily recognized by owners. Common descriptions of behavioral abnormalities included unpredictable spooking and hypervigilance, combined with periods of dullness. Bad behavior under saddle was described commonly, including bolting, bucking, rearing, spinning, refusing fences, or being difficult at the mounting block or when entering the ring. Other recent reports of horses with neuroaxonal degeneration also have reported clinical signs that were not apparent in earlier literature. A 5-month-old Pony of the Americas colt was described to have atypical behavior, including a lack of fear or flight response when confronted with loud objects such as farm equipment.¹⁴ The foal also had cerebellar dysfunction, including variable head tilt, wide head and neck excursions, head tremors, and pendular nystagmus. In an investigation of 148 Quarter Horses on a breeding farm, 88 of which were suspected of having eNAD/EDM, horses showed variable degrees of obtundation as well as inconsistent or absent menace responses with normal vision. Lack of a fear or flight response also was noted.³ In a different report, 6/15 young Lusitano horses on a breeding farm were clinically abnormal, and 5 of 8 tested Lusitanos had inconsistent menace responses, but no ophthalmologic lesions were present to explain this finding.¹⁵

The definitive cause for the behavioral abnormalities in horses with eNAD/EDM is unknown. None of the horses in our report had detectable histologic abnormalities in areas of the brain thought to influence behavior. Although some of the described behaviors, in particular alterations in sensation and somnolence, could be explained by abnormalities of the reticular formation, this area was not consistently affected in our population of horses. Subtle histologic changes in the brain could have been overlooked, because standard necropsy techniques utilized only H&E staining, or relevant areas of the brain may not have been examined. However, our results are consistent with other reports in the literature. Horses in other reports that demonstrated atypical signs had typical histologic lesions of eNAD/EDM without additional histologic lesions to describe their abnormal behavior, quiet to obtunded mental status, inconsistent to absent menace responses or cerebellar dysfunction.^{3, 14} In the absence of detectable structural (histologic) abnormalities to explain these abnormal behaviors, functional causes must be considered, as previously postulated.^{4, 14} Such functional disturbances could be occurring on a neurotransmitter level, as suggested by murine models of vitamin E deficiency.¹⁶ Mice with an inadequate supply of vitamin E during growth developed alterations in brain glutamate concentrations and demonstrated increased anxiety as adults compared to mice with adequate vitamin E status.

Importantly, behavioral abnormalities are common in human patients with a variety of neurodegenerative disorders, including Parkinson disease, Alzheimer disease, and frontotemporal dementia. Approximately 75% of humans with Parkinson disease describe experiencing psychosis, which includes minor and major hallucinations and delusions.^{17, 18} Some of these patients describe passage hallucinations (glimpsing a shadow of a person or animal), visual illusions and presence hallucinations (experiencing the sense that someone is nearby). One could expect that should horses with neurodegenerative disease be having these types of experiences, behaviors such as unpredictable spooking might naturally be the consequence. The underlying pathophysiology of neuropsychiatric symptoms in these diseases remains incompletely understood and is likely complex and multistructural, with multiple neurotransmitter systems implicated.¹⁹ Additionally, people with neurodegenerative disorders report neuropathic pain, and horses with eNAD/EDM might demonstrate aberrant behavior secondary to unpredictable or unaddressed pain.

4.3 Etiology of disease

Evidence suggests that oxidative stress contributes to neurodegeneration and likely underlies eNAD/EDM. Oxidative stress can result from antioxidant deficiency or excessive oxidative stressors. In addition, familial predisposition to eNAD/EDM has been suggested in the literature.⁷ Despite the predominance of Warmbloods in our study, many different lines and breeds were affected, decreasing the likelihood of a simple genetic cause. Similarly, efforts to identify a genetic cause of EDM have repeatedly failed.²⁰ However, it seems likely that genetic factors are important, and recent work suggests that horses with eNAD might metabolize vitamin E abnormally, increasing susceptibility to disease.²¹ Although horses in our study frequently had vitamin E concentrations within the normal range, it is possible that the horses had a period of early development during which inadequate dietary vitamin E was ingested. Interestingly, many of the affected horses in our report were imported from Europe. The exact number could not be ascertained, but our impression is that most affected Warmbloods were imported from various European countries. Anecdotally, many large breeding farms in Europe do not have adequate pasture turnout, relying on other sources of forage, which could result in vitamin E deficiency.

Insecticides, wood preservative or sealers and neurotoxins chemically related to anthelmintics used in horses also have been linked to eNAD/EDM.^{1, 12} The possibility that these environmental toxins have contributed to the increased prevalence of eNAD/EDM has been raised. A recent study documented a cluster of foals in Pennsylvania with dysphagia and altered mentation attributed to environmental chemical exposure because of proximity to hydraulic fracturing (fracking).²² Anecdotally, some horses in our report originated from the same barn or the same trainer. Whether this finding represents environmental factors contributing to disease development or selection bias is uncertain. Many horses were described as unusually calm before the onset of behavior changes, which might increase the likelihood of being purchased for certain types of equestrian programs.

4.4 Diagnostic testing

Similar to previous studies, our study found no antemortem diagnostic result consistently associated with a diagnosis of eNAD/EDM. Serum vitamin E concentrations were mostly within normal limits, with only 10% of horses tested being deficient. Because serum vitamin E concentrations do not reflect past periods of vitamin E deficiency and do not always correlate well with tissue vitamin E concentrations, it is possible that another method of testing, such as muscle biopsy, may be a more consistent marker of vitamin E deficiency. Only 22 of 74 horses (28%) with CSF testing had albuminocytologic dissociation, a non-specific indicator of neurologic disease with altered blood-CSF barrier. A single horse had mild inflammation on CSF analysis (lymphocytic pleocytosis) and mild chronic focal inflammation of the leptomeninges of unknown clinical relevance. Two horses had very mild neutrophilic inflammation with no corresponding pathologic cause on necropsy evaluation, 1 of which was likely to have had secondary blood contamination and blood-associated leukocytes. Of 27 horses for which pNF-H was evaluated, only 2 had increased serum concentrations of pNF-H, and 8 had increased CSF concentrations of pNF-H, consistent with previous reports of low sensitivity of this diagnostic test.¹¹ Currently, definitive diagnosis requires necropsy evaluation, and clinical diagnosis of neurodegenerative disease continues to be based on clinical signs and exclusion of other potential disease processes. However, at our institution, horses with ataxia that also have notable behavior changes, with or without low serum vitamin E concentrations, are much more likely to have a final necropsy diagnosis of eNAD/EDM than CVSM or EPM.

Cervical vertebral comorbidities including CVSM were common in horses with neuroaxonal degeneration. Six horses were excluded from analysis because histologic lesions were consistent with both eNAD/EDM and CVSM. Of the 100 horses that had only histologic lesions consistent with eNAD/EDM, the majority (57%) had radiographic abnormalities in the cervical region of the spinal cord. Fourteen had cervical myelograms that were suggestive of spinal cord compression but did not have evidence of compression histologically. The prevalence of cervical vertebral comorbidities in horses with neurodegenerative disease complicates decision-making for horses with clinical signs of a cervical or diffuse myelopathy. Even if cervical spinal cord disease is suspected and cervical imaging is suggestive of CVSM, eNAD/EDM might be the underlying cause of ataxia. Because of the prevalence of neurodegenerative disease in our patient population, we are increasingly hesitant to recommend cervical fusion surgery for middle-aged Warmbloods with behavior changes, even if imaging results suggest spinal cord compression. In addition, the recognition of horses with an imaging diagnosis of CVSM but pathologic diagnosis of eNAD/EDM might explain why some horses thought to have CVSM fail to show improvement after cervical fusion.

4.5 Limitations

The primary limitation of our study is the lack of a control group of horses free from ataxia with which to compare histopathologic findings in the CNS. Accumulation of spheroids in the CNS is considered an aging change in other species such as humans and dogs and it has been postulated for horses as well, although clinical correlation is lacking. A study documenting histopathologic changes in the brains of horses reported the presence of spheroids in almost all horses (91/100) regardless of age, but horses with multiple spheroids (>10) per anatomical area were older on average than those with single spheroids.²³ Whether these changes are associated with cumulative oxidative stress over many years causing neurodegeneration or are incidental lesions has yet to be determined. In our study population, 64% of the horses were <10 years of age and 94% were <15 years old. Although age-related degeneration cannot be definitively ruled out, given the absence of other lesions to explain the neurologic signs in these horses, we are hesitant to dismiss the degenerative lesions as simply age-related or incidental without further investigation and comparison to an adequate control group known to be free of behavioral abnormalities and mild proprioceptive ataxia (which was difficult to find in our case population). The possibility exists that the eNAD/EDM-like lesions are not the primary cause of the clinical signs in these horses but are simply the only lesions detected during routine necropsy evaluation. Neurodegenerative diseases in people such as Alzheimer disease, Parkinson disease, and amyotrophic lateral sclerosis generally are characterized by accumulation of misfolded proteins into insoluble aggregates in the CNS, accompanied by progressive loss of neurons in the affected regions.²⁴ Assuming similar pathogenesis in horses, advanced immunohistochemical or other imaging techniques might be required to identify the full extent of degeneration. These horses might also have peripheral neuropathies that have not yet been described or characterized. A recent report describes ganglioneuritis as the potential cause of dangerous behavior in a similar population of horses (young to middle-aged sport horses).²⁵

Other limitations include the use of serum alpha-tocopheral concentrations for quantification of vitamin E deficiency instead of tissue samples. Serum concentrations of alpha-tocopherol represent a small percentage of total body stores, and tissue concentrations might be more reflective of body alpha-tocopherol concentrations.²⁶ However, studies indicate that serum and liver alpha-tocopheral concentrations respond similarly to supplementation in horses, and therefore serum is likely a reasonable approximation of tissue concentrations.²⁷ Additionally, blinded review of all the pathologic samples included in our study was not performed, and therefore bias of a pathologist aware of the horse's clinical history and clinician's opinion cannot be overlooked. Because knowledge of the horse's history and the clinician's differential diagnosis are critical for complete and targeted necropsy evaluation, a completely unbiased evaluation was not achieved.

In conclusion, we have provided clinical and histopathologic information on a large number of horses with neurodegenerative disease and with signalment and clinical presentation distinct from those described earlier. Neurodegenerative disease remains an important cause of neurologic dysfunction in the horse. It is unclear whether a unifying genetic abnormality, risk factor, or disease process will be discovered for these cases or, as others have suggested, diverse metabolic, structural and functional abnormalities are being manifested with the same histologic lesions.¹⁴ Until a definitive antemortem test is identified, we must identify methods of increasing antemortem diagnostic certainty. The increasing prevalence with which these horses are diagnosed and the safety risk posed to owners and handlers emphasize the continued necessity for focused research on neurodegenerative disease in horses.