Introduction

A disorder known as border collie collapse (BCC) occurs in border collies throughout North America, Europe, and Australia.^1–4 Dogs with BCC are normal at rest and with moderate exercise but sometimes develop an abnormal gait after prolonged strenuous exercise. Severe episodes can result in collapse and inability to ambulate. This disorder is recognized as a common and important problem within the breed, but there have been very few reports of veterinary evaluation of affected dogs.^3,4 Veterinarians evaluating border collies with a history of abnormal episodes must consider a wide range of established causes for exercise intolerance including hypoglycemia, hypoadrenocorticism, cardiac arrhythmias, pulmonary hypertension, malignant hyperthermia, myasthenia gravis, metabolic myopathies, and episodic neurologic disorders such as dynamin 1-associated exercise-induced collapse (dEIC), epilepsy, paroxysmal dyskinesia, and narcolepsy/cataplexy.^5–10 Clinical and laboratory examination at rest and during collapse episodes in a limited number of dogs with BCC has failed to demonstrate a metabolic, cardiovascular, respiratory, or neuromuscular cause for their collapse.^2–4 Trigger activities and factors important in inducing episodes in affected dogs have not been established. There is no test for BCC, so a tentative diagnosis can only be made by recognizing the characteristic clinical features of BCC episodes and eliminating other causes for exercise intolerance through diagnostic evaluation. The characteristic clinical features of BCC episodes have not, however, been well described.

This study reports the result of surveys completed by owners of 165 border collies with presumed BCC and veterinary description of videotaped collapse episodes from 40 of these dogs. This is the first large-scale descriptive study of dogs presumed to have BCC and should provide veterinarians with the information needed to recognize typical BCC episodes and to advise owners regarding the management and prognosis of affected dogs.

Materials and Methods

Results

Descriptive Study

Completed surveys were obtained from the owners of 165 border collies meeting study criteria. Male (54%) and female (46%) dogs were approximately equally represented. Fifty-six percent of the male dogs were neutered and 74 % of the female dogs were spayed before their owners filled out the survey. Age at the first observed episode of collapse ranged from 4 mo to 7 yr (median 2 yr). When asked to comment on their dog's body condition, level of fitness, and temperament, all owners reported that their dogs were in good physical condition (85% perfect condition, 15% thin) and many (47 %) described them as excitable. Most owners (72%) indicated that they felt that their dog with BCC had always been less tolerant of high temperature and humidity than other dogs. Owners of 24/165 (15%) of the affected dogs were aware of dogs closely related to theirs that had experienced similar episodes while exercising.

Dogs had experienced from two to more than 100 (median six) observed episodes at the time the survey was completed. Twenty dogs (12%) had experienced more than 20 episodes. No dogs for which a survey was completed had died during an episode of collapse. Episodes occurred during or immediately following repetitive hand-thrown “fun” retrieves of balls, Frisbees, or toys in 112 dogs (68%) and after herding stock in 39 dogs (24%) (Table 1). A few dogs (7%) had experienced episodes both retrieving and after working stock. Other activities leading to episodes included excited running play with other dogs (13%), agility training (3%), and chasing or running after wildlife (rabbits, deer, kangaroos) or an all-terrain vehicle (3%). Dogs typically experienced an episode after 5–20 min of strenuous exercise (median 10 min).

Table 1 Activities Associated with Collapse in 165 Border Collies with Presumed BCC

View Fullscreen

Owners described clinical findings before, during, and after each episode in their own words and by selecting descriptors provided in the questions. None of the dogs were reported to exhibit abnormal behavior or systemic signs immediately prior to an episode. Most dogs (82%) were considered by their owners to be mentally abnormal during their BCC episodes and were described as unfocused, dazed, confused, or disoriented. Many of the dogs became unresponsive to verbal commands. It was common for a loss of mental focus on the task being performed (retrieving, herding) to be the first sign that an owner noticed at the onset of an episode. Squinty eyes and/or rapid blinking during some or all episodes were reported in nine dogs (5%). All of the dogs were considered by their owners to be mentally normal after recovering from an episode.

Owners described their dog's gait during episodes as wobbly or uncoordinated. Most (88%) reported that the gait in all four limbs was abnormal, with the pelvic limbs most severely affected. Only the pelvic limbs were reported to be abnormal in 20 dogs (12%). Dogs stumbled and staggered during episodes, and 61 dogs (37%) were reported to fall over or collapse and be unable to walk during one or more severe episodes. Owner description of the number of legs affected (all four legs) and mentation (altered) were in complete agreement (100%) with analysis of videotaped episodes in the 40 dogs for which both tools were available. Owner descriptions of the details of gait and muscle tone were often incomplete, making accurate description of the gait during episodes impossible based on the questionnaire alone. Owners described the recovery from episodes as gradual, with the time required for return to normal mentation and gait estimated at 5 to 45 min (median 10 min).

Owners were asked to speculate on factors that they felt contributed to causing an episode in their dog (Table 2). Excessive environmental heat (111/165 dogs, 67%) and excitement associated with a trigger activity (67/165 dogs, 41%) were the most commonly reported factors. Although many owners felt that excessive environmental heat increased the likelihood of collapse in their dogs, individual episodes were reported in a wide range of temperatures. Some owners (28%) indicated that they felt that over time their dogs were becoming more susceptible to having episodes when participating in trigger activities. None of the dogs with BCC developed significant systemic or neurologic abnormalities during the period of 3 mo to 13 yr (median 3 yr) between their first observed episode and completion of the survey.

Table 2 Factors Reported by Owners as Contributing to Collapse in 165 Border Collies with Presumed BCC

View Fullscreen

Veterinary Description of Videotaped Collapse Episodes

All of the videotaped episodes were similar, varying only in their degree of severity. Characteristics are summarized in Table 3 and can be viewed in Supplementary Videos I, II, III, and IV. All of the dogs appeared to be mentally abnormal. When the onset of an episode was observed (seven dogs, 17%), it was evident that there was a sudden loss in focus or interest in the task being performed. Dogs were poorly responsive or disoriented, and many (18 dogs, 45%) paced compulsively in wide circles or wandered aimlessly and paid no attention to owner commands or obstacles in their path. Squinting eyes and/or rapid blinking were apparent in seven dogs (17%). All dogs had symmetrical ataxia affecting all four limbs, with the pelvic limbs more obviously affected than the thoracic limbs. Increased extensor tone of the pelvic limbs was common (37 dogs, 92%), and affected dogs often stood with their pelvic limbs extended caudally (Figure 1). The protraction phase of the gait was prolonged or delayed in all dogs, resulting in scuffing, toe dragging, and occasional knuckling of the pelvic and sometimes the thoracic limbs. Limbs often crossed when the dogs walked in a circle. An exaggerated high-stepping or floating thoracic limb gait was noticed in some dogs (21 dogs, 52%). Truncal swaying was evident in most of the dogs (34 dogs, 83%) as they stood or walked, and this often resulted in the dogs staggering or falling to the side. Only one of the videotaped episodes analyzed showed a dog that collapsed and was unable to rise (Supplementary Video IV). A head tilt and nystagmus were not noted in any of the dogs.

Table 3 Clinical Features of Videotaped Border Collie Collapse Episodes in 40 Dogs

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

Discussion

This report describes the typical signalment and history of dogs with presumed BCC and owner perception of factors important in BCC, as well as summarizes the gait and mentation abnormalities evident during BCC episodes. Study criteria excluded dogs likely to be having brief losses of consciousness typical of syncope and also excluded dogs with generalized seizures due to epilepsy, which is a common disorder in the border collie breed.^5,6,11–13 Although it is possible that some of the individual dogs included in this study had a medical condition other than BCC causing their recurrent collapse, the episodic nature of their collapse, normal veterinary evaluation at rest, the consistent clinical description of episodes, and failure to develop other signs or progressive disease during the long follow-up period made the diagnosis of BCC most likely.^4–6

The investigators recently reported extensive veterinary evaluation at rest and during exercise-induced episodes in 13 dogs with BCC.⁴ Body temperature, blood glucose, serum electrolytes, arterial blood gas analysis, serum creatine kinase, and sequential lactate and pyruvate measurements were not different in these dogs from border collies without BCC performing the same exercise. Electromyography and muscle biopsies were also normal. Many metabolic, cardiac, and neuromuscular disorders were eliminated as the cause of BCC episodes. A disorder affecting the central nervous system was suspected, but the precise cause of BCC episodes could not be determined.⁴

BCC has been speculated to be a heat-related disease and is often referred to by owners and breeders as heat intolerance, exercise-induced hyperthermia, or malignant hyperthermia.^1–4 Body temperatures have been reported to be extremely elevated during collapse in dogs with BCC (>42°C), but not higher than body temperatures of normal exercise-tolerant border collies performing the same exercise.⁴ Body temperatures are similarly increased in normal dogs of many breeds participating in strenuous exercise.^14–17 Most of the owners participating in this study (66%) reported that high environmental temperatures increased the likelihood of their dog collapsing on a given day, and many (72%) felt that their dog had always been relatively intolerant to heat and humidity and panted or sought water or shade more quickly than other dogs they had owned or trained. Although these findings are extremely subjective, it is clear that owners of dogs with BCC believe that their dogs are relatively intolerant to heat and that overheating may be a trigger for BCC episodes.

Generating a description of typical BCC episodes is important in order to facilitate the recognition of BCC by owners and veterinarians. The videotaped episodes evaluated in this study revealed very consistent clinical characteristics during collapse, suggesting that a presumptive diagnosis of BCC can be made when a border collie's collapse episodes fit this pattern and other causes of exercise intolerance have been eliminated through diagnostic evaluation.

The videotaped dogs with BCC had an altered sensorium during episodes, with a loss of mental focus often being the first thing noticed, progressing to the dog being dull and disoriented. Abnormal compulsive activity was evident in many dogs, manifested as aimless wandering, pacing, or circling. Squinting eyes and rapid blinking were apparent in seven of the videotaped episodes (17%) and had been noticed by owners during some episodes in an additional seven dogs. Repetitive eyelid blinking and photosensitivity are a common automatism in several human non-convulsive (focal seizure) epileptic disorders and in children with hyperventilation-induced “blank spells” that are not considered to be a seizure disorder.^18–20

The gait deficit in these dogs was most consistent with upper motor neuron paresis and general proprioceptive ataxia that was most obvious in the pelvic limbs, but the thoracic limbs were also affected. The protraction phase of gait was prolonged or delayed, resulting in characteristic scuffing and knuckling of the limbs and the crossing of legs during turning. Involvement of cerebral and/or brainstem-associated functions (compulsive behavior, altered mentation) together with the presence of paresis and ataxia suggests that BCC is an episodic diffuse or multifocal central nervous system disorder.²¹

The clinical features and duration of BCC episodes are not typical of established paroxysmal disorders like cataplexy, epilepsy, or paroxysmal dyskinesia in dogs, but it is possible that BCC does represent an unusual manifestation of one of these disorders.^5,6,10–12 Cataplexy in dogs is characterized by a sudden loss of muscle tone triggered by strong positive emotions (food, play) or excitement.^22,23 Episodes are typically brief, lasting a few seconds or minutes, and mentation is not altered.^22,23 Partial seizures without secondary generalization to convulsions have been well described in dogs with epilepsy and may in some cases resemble the BCC episodes described here.^12,13,24 Features that would suggest that a BCC episode could be a manifestation of a partial seizure would include identification of an aura, the presence of autonomic signs (urination, defecation, hypersalivation) before or during an episode, postictal signs, demonstration of electroencephalography (EEG) abnormalities, progression to convulsive episodes, or response to antiepileptic treatment.²⁴ None of these features were identified in any of the dogs with presumed BCC reported here, but that does not completely eliminate the possibility that BCC episodes could be epileptic partial seizures.

Paroxysmal dyskinesias are episodes of abnormal involuntary movement, abnormal posture, or abnormal muscle tone in individuals with a normal neurologic examination between episodes.^10,25,26 Episodes are distinguished from partial seizures by normal mentation during episodes, lack of autonomic signs or postictal abnormalities, and lack of EEG abnormalities during and between episodes. Paroxysmal dyskinesias are rarely reported in dogs, but an inherited paroxysmal dyskinesia has been described in Chinook dogs, and another is suspected to be the cause of a disorder in border terriers known as canine epileptoid cramping syndrome.^10,25 Classification of human paroxysmal dyskinesia disorders is usually based on clinical characteristics, precipitating factors, and duration of attacks, but identification of the genetic cause of some familial forms has improved classification and understanding of the pathophysiology of these disorders.²⁶ Although mentation is usually normal during dyskinesia episodes, several mutations that cause episodic dyskinesia in humans are associated with a high prevalence of concurrent epilepsy.^27,28 Paroxysmal exercise-induced dystonia is a rare, inherited dyskinesia in which attacks are triggered by prolonged strenuous exercise, and episodes of abnormal mentation are common due to concurrent epilepsy with absence seizures.^27,28 It is impossible to say with certainty that BCC episodes are not a manifestation of a paroxysmal dyskinesia with or without concurrent epilepsy. EEG is an important tool in the differential diagnosis of paroxysmal disorders, but none of the dogs reported here had EEGs performed. The investigators have, however, fully evaluated one dog with presumptive BCC that had normal EEGs at rest and during an exercise-induced episode.

The clinical features of BCC episodes are very different from those reported in dogs with dEIC, the inherited episodic neurologic disorder affecting Labrador retrievers and other breeds.^7–9,29,30 Both BCC and dEIC cause collapse that is most likely to occur during participation in exciting trigger activities in hot weather, but the appearance of dogs during collapse is different. While dogs with dEIC remain alert and have relatively flaccid but uncoordinated pelvic limbs during collapse, dogs with BCC are mentally abnormal, are ataxic in all four limbs and have increased extensor tone in the rear limbs with scuffing and knuckling during walking.^7,29,30 Dogs with dEIC usually try to continue to run and retrieve while dragging their flexed and flaccid rear legs, while dogs with BCC tend to wander aimlessly or stagger and fall. The DNM1 mutation causing dEIC has not been previously identified in border collies and was not found in the 128 dogs tested here.⁹ This does not, of course, preclude a DNM1 problem in dogs with BCC caused by a different mutation resulting in a different phenotype.

Working border collie websites and organizations recognize BCC as a problem within the breed and speculate that it is a familial or heritable condition.^1–4 Most owners of the dogs in this study knew very little about the health of their dog's sire, dam, or siblings, but the owners of 24 dogs (15%) were aware of a similar or identical episodic disorder causing collapse in dogs closely related to theirs. Multiple border collie breed registry organizations exist, making it difficult to confirm the identity of dogs with similar or identical names, likely resulting in an underestimation of relatedness in the 93 dogs for which pedigrees were available. Analysis suggested that BCC is inherited, but mode of inheritance could not be determined.

Conclusion

At the present time, BCC is a presumptive diagnosis made by ruling out other disorders causing exercise intolerance and by observing characteristic clinical features in affected border collies. Reporting the results from this large study cohort of 165 dogs provides meaningful information that can help veterinarians to recognize and manage BCC. Data from this study suggest that BCC is a distinct episodic disorder that may be the result of diffuse central nervous system dysfunction.