A challenging case: A collie with acute neurologic signs

A 3-year-old 77-lb (35-kg) neutered male collie was presented to the Veterinary Teaching Hospital at the Virginia-Maryland Regional College of Veterinary Medicine for evaluation of an acute onset of progressive ataxia, tetraparesis, a single tonic-clonic seizure, vomiting, and ptyalism. The owner reported that the signs had started as intermittent periods of excessive anxiety and dementia and had rapidly progressed over the previous four days despite supportive care (140 ml/hr intravenous Normosol-R [Abbott Laboratories]; no oral food or water) administered by the referring veterinarian. The results of a complete blood count, serum chemistry profile, and abdominal radiographic examination performed before referral had been unremarkable. Pertinent history included administration of heartworm prophylaxis (ivermectin with pyrantel) orally the day before the onset of the clinical signs. The dog had received the monthly preventive since 6 months of age. In addition, the dog had been allowed to roam unsupervised and had had access to carrion the week before the clinical signs developed. Its vaccination status was current.

Vital Stats

Physical and neurologic examinations

On presentation, the dog was obtunded and in lateral recumbency with marked extensor rigidity involving all four limbs. In addition, the patient demonstrated continuous ptyalism and persistent bradycardia (heart rate = 40 to 50 beats/min).

We detected several abnormalities on neurologic examination. When allowed to ambulate without assistance, the dog circled to the right and had severe truncal ataxia and asymmetric tetraparesis. We noted postural reaction deficits in all four limbs, but the right side was more severely affected. The cranial nerve examination revealed a right head tilt and bilateral mydriasis with normal pupillary light responses. Pathological, positional rotary nystagmus was present when the dog was placed in dorsal recumbency, as was a ventrolateral strabismus in the right eye. Segmental spinal reflexes were normal in all four limbs. Because of these neurologic examination findings, we suspected that a lesion was affecting the central portion of the vestibular system, although after considering the historical evidence of seizure activity, we could not exclude multifocal intracranial disease.

Differential diagnoses

The patient's initial problems included ptyalism, bradycardia, vomiting, and central vestibular (or possibly multifocal intracranial) disease. Our differential diagnoses for the ptyalism included metabolic disorders, neurologic diseases, toxicoses, gastrointestinal disorders, and salivary gland diseases. Potential causes of bradycardia include excess vagal tone secondary to gastrointestinal or respiratory disease, hyperkalemia, increased intracranial pressure, endocrinopathies, sick sinus syndrome, and a drug reaction. Acute vomiting can be associated with many systemic diseases, neurologic disorders, toxin exposures, and gastrointestinal diseases. Our primary differential diagnoses for the central nervous system signs included toxin exposure, infectious meningoencephalitis, and neoplasia.

After considering the cumulative clinical signs, we thought polysystemic metabolic (hepatic disease, hypoadrenocorticism), toxic (bacterial toxins, ivermectin, lead, organophosphates), and infectious conditions (canine distemper, rickettsial diseases, rabies) or multiple concurrent diseases were the most likely differential diagnoses.

Initial diagnostic tests and therapy

At admission, we performed a complete blood count, serum chemistry profile, and urinalysis. The results of the complete blood count and urinalysis were unremarkable (Table 1). Hypokalemia may have resulted from the ptyalism, vomiting, inappetence, previous administration of potassium-deficient fluids, or a combination of these. We attributed the mild elevations in protein to dehydration.

Table 1 : Results of Initial Laboratory Tests

We submitted additional blood samples for a tick-borne disease panel and measurement of whole blood cholinesterase activity and lead concentration as well as plasma ivermectin concentration. The results of six-lead electrocardiography confirmed a sinus bradycardia (heart rate = 42 beats/min). We also performed thoracic radiographic and abdominal ultrasonographic examinations, and the results were unremarkable.

Initial therapy included intravenous administration of lactated Ringer's solution supplemented with 40 mEq/L of potassium chloride (90 ml/kg/day) to maintain the dog's hydration, provide for ongoing losses, and correct the hypokalemia. Atropine sulfate administration (0.02 mg/kg intravenously and 0.02 mg/kg subcutaneously) resolved the sinus bradycardia, which suggested that the cardiac arrhythmia was caused by excess vagal tone. We also administered doxycycline (5.5 mg/kg intravenously b.i.d.) and metoclopramide hydrochloride (0.2 mg/kg subcutaneously q.i.d.) and withheld food and water for 24 hours.

Follow-up tests

Twelve hours after admission, the patient showed marked clinical improvement. The vomiting and nystagmus resolved, the ptyalism improved, and the animal was able to ambulate, although it remained mildly tetraparetic and ataxic. We performed further diagnostic tests, consisting of a fasting plasma ammonia concentration, sulfobromophthalein clearance test, and an ACTH stimulation test, to rule out biochemically occult hepatic disease and atypical hypoadrenocorticism. The plasma ammonia and pre-ACTH and one-hour post-ACTH stimulation cortisol concentrations were within the reference ranges, but the sulfobromophthalein retention was abnormal (Table 2).1,2 Results of the tick-borne disease panel were normal except for elevated Lyme IgM and IgG antibody concentrations (Table 2). It was later confirmed that the dog had been vaccinated annually for Lyme disease, which could potentially explain the elevated titers, but a Western blot analysis was not performed. The blood cholinesterase activity and lead and ivermectin concentrations were also within reference ranges (Table 2).

Table 2 : Results of Initial Laboratory Tests

Based on these results, hepatic encephalopathy became the primary differential diagnosis, although not all of the dog's clinical signs were consistent with classic manifestations of a metabolic encephalopathy.3 We planned a laparoscopic exploratory examination and hepatic biopsy. The results of prothrombin time and activated partial thromboplastin time tests performed before surgery were normal (Table 2). The patient was premedicated with hydromorphone, and anesthesia was induced with propofol. Isoflurane and oxygen were given to sustain anesthesia. We performed a second abdominal ultrasonographic examination while the dog was anesthetized to look for anomalous portosystemic vasculature, but none were found. Gross laparoscopic findings were largely unremarkable, although the liver subjectively appeared slightly small. No evidence of anomalous vasculature was present. We submitted five hepatic biopsy samples obtained from the right lateral, right medial, and caudate liver lobes for histologic examination, quantitative copper analysis (Table 2), and aerobic and anaerobic bacterial culture and antimicrobial sensitivity testing.

We began administering lactulose (0.5 ml/kg orally t.i.d.) after recovery from anesthesia to reduce ammonia production and facilitate ammonia elimination from the gastrointestinal system. Lactulose accomplishes this by acidifying intestinal contents, which allows colonic trapping of ammonium ions and decreased systemic absorption of ammonia. It also provides a nonprotein substrate for colonic bacteria and causes osmotic diarrhea.4 The hepatic culture results were negative for both aerobic and anaerobic bacterial growth, but the doxycycline administration makes interpretation of the culture results difficult. Histologic examination of the hepatic biopsy samples revealed some portal areas with a normal complement of artery, vein, and bile duct, but most fields had no discernible portal or central areas. Instead, small vessels and bile ductules were randomly positioned throughout the sections. The histopathologic diagnosis was hepatic microvascular dysplasia (Figure 1). The sections were reviewed by three board-certified anatomic pathologists, all of whom concurred with the diagnosis.

Figure 1 : A photomicrograph of a hepatic biopsy sample of the dog in this case. Note the absence of normal portal and central areas of the hepatic lobule and random placement of small vessels and bile ductules, which are consistent with hepatic microvascular dysplasia (hematoxylin-eosin stain;bar = 15 μm).

Outcome

Over the next three days, the patient's clinical signs and hypokalemia resolved, and parenteral fluid therapy was tapered off gradually. By the fourth day of hospitalization, the dog was clinically normal and eating and drinking. Outpatient therapy consisted of a diet specifically formulated for dogs with hepatic dysfunction (Prescription Diet Canine l/d—Hill's Pet Nutrition), oral doxycycline (6.1 mg/kg b.i.d. for 10 days), and lactulose (0.5 ml/kg orally t.i.d.).

In a telephone interview about four weeks later, the owner reported that the dog seemed normal and that the lactulose dose had been reduced to 0.15 ml/kg given orally three times a day. About two months later, the dog experienced a recurrence of its neurologic signs. The dog was still receiving Prescription Diet Canine l/d and lactulose at the dosage mentioned above. An exploratory celiotomy with portal scintigraphy to confirm the absence of a shunt was offered to the owner but was declined. The patient was euthanized after failing to improve after 48 hours of aggressive supportive care provided by a local veterinary hospital.

Necropsy and definitive diagnosis

A necropsy examination was performed. The only gross abnormalities consisted of several punctate gastric ulcers and mucosal hemorrhages. No evidence of anomalous portosystemic vasculature was present, and the liver appeared normal. The findings of a gross examination of the central nervous system were unremarkable. Histologic examination of serial sections of the brain and cervical spinal cord revealed bilaterally symmetric areas of polymicrocavitation of the white matter of the cerebral cortices, pons, caudal cerebellar peduncle, medulla oblongata, and cervical spinal cord. Alzheimer type II astrocytes were observed in the basal nuclei and hippocampal areas. A focal area of neuronal necrosis accompanied by gliosis was also noted in the hippocampus. Histologic changes noted in the liver were similar to those described previously, although several foci of regenerative nodular change were also noted. The only other clinically relevant finding was multifocal lymphocytic-plasmacytic erosive gastritis. The histologic diagnoses were primary hepatic microvascular dysplasia with changes in the brain consistent with hepatic encephalopathy.5

Discussion

Hepatic microvascular dysplasia has been described only recently in veterinary medicine. Atypical liver morphology associated with macroscopic portosystemic shunting has been well-reported in the veterinary literature.1,3-6 However, the diagnosis of hepatic microvascular dysplasia without any macroscopic anomalies did not surface until 1996 when it was described as a congenital disorder of hepatic vasculature in a kindred of Cairn terriers.7 Hepatic microvascular dysplasia is defined as an abnormal liver with apposed hepatic venous and portal vessels that communicate through random small-caliber or juvenile-like interconnecting microvascular channels, thus bypassing the sinusoids.7,8 Hepatic microvascular dysplasia can develop as an isolated disease or in conjunction with portosystemic shunting.9 Although uncommon, dogs with hepatic microvascular dysplasia may manifest with any of the constellation of clinical abnormalities associated with hepatic encephalopathy. However, the disease remains clinically occult in many cases.7

It has been hypothesized that the natural remodeling of embryogenic vitelline veins is disrupted, resulting in the persistence of functional vitelline veins, and that this is the most likely cause of hepatic microvascular dysplasia.8 Vitelline veins normally break down and form sinusoids during the terminal stages of embryogenesis. As a result, blood is shunted from the portal system to the central vein and, subsequently, to the systemic venous circulation.10 Another proposed theory describes ultrastructural changes in the sinusoidal capillaries that result in reduced endothelial permeability and limited access to hepatocellular surfaces.7,10

The incidence of hepatic microvascular dysplasia is still unknown because of the difficulty and controversy of the diagnosis. Classic congenital and acquired vascular anomalies can often be presumptively diagnosed based on the clinical history, signalment, and the results of routine serum chemistry profile testing, liver function testing (e.g. bile acids assay), hepatic ultrasonography, and nuclear scintigraphic scanning.7 However, dogs with hepatic microvascular dysplasia may have no clinical signs or marked serum chemistry abnormalities. Even when hepatic microvascular dysplasia has clinical manifestations, the extreme variability of clinical presentations can make the diagnosis difficult. Also, as demonstrated in the patient described here, the results of routine clinicopathologic tests can be normal or nonspecific. The variation in the severity of disease is theorized to be a direct reflection of the number of persistent vitelline veins.8

A definitive diagnosis of hepatic microvascular dysplasia is based on histologic examination of hepatic tissue. Histologic lesions can vary among liver lobes, with some appearing normal and others being completely abnormal. For this reason, it is best to obtain biopsy samples from more than one liver lobe.7 Although not performed in this case, additional diagnostic tests to rule out macroscopic portosystemic vascular anomalies could have included mesenteric angiography, splenic portography, or transcolonic scintigraphy.10 Each of these imaging techniques has limitations because a shunt distal to the point of contrast injection (as with mesenteric portography) or a shunt located close to the liver (as with transcolonic scintigraphy) could be overlooked.8 A nuclear scan would have been helpful in this case but was not performed after the biopsy samples were obtained because we thought that the isolation required at our institution after radioisotope administration was not in the patient's best interest. Although unlikely given the work-up performed, this patient may have had hepatic microvascular dysplasia and a macroscopic shunt.

Dogs that present with clinical signs of hepatic microvascular dysplasia are often older (2 or 3 years old), can have relatively unremarkable blood test results, and have lower preprandial and postprandial serum bile acid concentrations when compared with dogs having macroscopic anomalies.8,9 Like most hepatic vascular diseases, hepatic microvascular dysplasia has been historically identified in toy and small-breed dogs.7-11 According to a literature review, this case is only the second reported in a large-breed dog (> 30 kg).8 Females have been overrepresented in reported studies, making up 70% of the documented cases.10 As with portosystemic shunting, neurologic and gastrointestinal signs are the most common presenting clinical signs. This dog's vomiting may have resulted from gastric ulceration and inflammation. The mechanisms by which hepatic disease causes gastrointestinal inflammation and hemorrhage are complex and poorly understood but may result from reduced mucosal microcirculation and reduced mucosal renewal.12 Gastrointestinal hemorrhage and hypokalemia, which were both present in the case, are known to exacerbate hepatic encephalopathy13 and may have contributed to this dog's acute decompensation. However, the dog's neurologic signs were atypical because the disease presented primarily with focal signs referable to the central vestibular system as opposed to the global and diffuse cerebral cortical signs typically seen with hepatic encephalopathy.

Common neurologic abnormalities in dogs with hepatic encephalopathy are usually suggestive of a diffuse cerebral disease process and include central blindness, alterations in mentation or behavior, head-pressing, and, less commonly, seizures. Ptyalism can be a feature of hepatic encephalopathy, but it occurs more frequently in cats.3 Hepatic encephalopathy is typically seen as a result of decreased functional liver mass in combination with portosystemic shunting of blood.3 So hepatic encephalopathy is often seen in dogs with either congenital or acquired portosystemic shunts. Acquired shunts can often result from chronic hepatitis, cirrhosis, and portal vein hypoplasia.

Although not proven, it is possible that a concurrent disease such as an unidentified toxicosis (e.g. from possible carrion consumption) or infection could have caused this patient's clinical signs. Plasma ivermectin concentrations do not reliably correlate with brain ivermectin concentrations. Brain ivermectin concentrations may be more related to defective blood-brain barrier protective mechanisms than to plasma concentrations of the drug.14 Even though serum, not central nervous system samples, was tested for ivermectin concentrations, ivermectin toxicosis from the type of heartworm preventive administered would be extremely rare.14,15 Both antemortem and postmortem diagnostic findings were most consistent with hepatic microvascular dysplasia and hepatic encephalopathy.

An antibiotic-responsive encephalitic syndrome, although poorly documented in the literature, is anecdotally incriminated by clinicians of many disciplines as a common cause of acute intracranial disease. In many cases, rickettsial organisms are often speculated as the etiologic agent. The neurologic manifestations of Rocky Mountain spotted fever, ehrlichiosis, and borreliosis have been well-described in animals and people and are consistent with this dog's neurologic signs.16,17 After considering that this dog had elevated serum antibodies to Borrelia burgdorferi, that new infectious diseases are constantly emerging with the advent of sophisticated diagnostic testing methods, and that intoxications can be extremely difficult to definitively diagnose, we cannot totally discount the possibility that this dog had a doxycycline-responsive infectious disease or a toxicosis and that the hepatic microvascular dysplasia was an incidental but important finding.16,17 Although rare, tetracycline has been associated with adverse reactions and can be hepatotoxic.15

Conclusion

The overall incidence of hepatic microvascular dysplasia will likely increase with acceptance of diagnostic criteria, and the disease should be included in the differential diagnoses of animals presenting with signs of hepatic encephalopathy or gastrointestinal disease. Generally, the long-term prognosis for dogs with hepatic microvascular dysplasia appears to be better with medical management when compared with that for dogs with a portosystemic shunt.10 Dogs in which the disease is diagnosed at 1 year of age or less tend to remain in excellent or good condition with medical treatment.10 As has been described in the literature and in this case report, the prognosis for dogs with hepatic microvascular dysplasia diagnosed later in life (2 or 3 years of age) is guarded to poor because of medically intractable, recurrent clinical signs.10 Although neurologic signs often respond well to dietary management (i.e. low protein, highly digestible diets fed long-term), as demonstrated in this case, they, like the gastrointestinal manifestations previously reported in the literature, may have a variable response to treatment, with some patients having refractory clinical signs resulting in marked morbidity.10

Peter J. Lotsikas, DVM*

John H. Rossmeisl Jr., DVM, MS, DACVIM (internal medicine and neurology)

Department of Small Animal Clinical Sciences

Virginia-Maryland Regional College of Veterinary Medicine

Virginia Tech

Blacksburg, VA 24061

*Current address: Veterinary Teaching Hospital, College of Veterinary Medicine, Iowa State University, Ames, IA 50011-1250

REFERENCES

1. Center SA. Liver function tests in the diagnosis of portosystemic vascular anomalies. Semin Vet Med Surg Small Anim 1990;5:94-99.

2. Flatland B, Leib MS, Warnick LD, et al. Evaluation of the bromosulfophthalein 30-minute retention test for the diagnosis of hepatic disease in dogs. J Vet Intern Med 2000;14:560-568.

3. Rothuizen J, Meyer HP. History, physical examination, and signs of liver disease. In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine. Philadelphia, Pa: WB Saunders Co, 2000;1272-1275.

4. Nelson RW. Treatment of complicated hepatic failure. In: Small animal internal medicine. Baltimore, Md: Mosby, 1998;549-550.

5. Summers BA, Cummings JF, de Lahunta A. Degenerative diseases of the central nervous system. In: Veterinary neuropathology. St. Louis, Mo: Mosby, 1995;208-210.

6. Zawie DA, Gilbertson SR. Interpretation of canine liver biopsy. A clinician's perspective. Vet Clin North Am Small Anim Pract 1985;15:67-76.

7. Schermerhorn T, Center SA, Dykes NL, et al. Characterization of hepatoportal microvascular dysplasia in a kindred of cairn terriers. J Vet Intern Med 1996;10:219-230.

8. Phillips L, Tappe J, Lyman R, et al. Hepatic microvascular dysplasia in dogs. Prog Vet Neurol 1996;7:88-96.

9. Allen L, Stobie D, Mauldin GN, et al. Clinicopathologic features of dogs with hepatic microvascular dysplasia with and without portosystemic shunts: 42 cases (1991-1996). J Am Vet Med Assoc 1999;214:218-220.

10. Christiansen JS, Hottinger HA, Allen L, et al. Hepatic microvascular dysplasia in dogs: A retrospective study of 24 cases (1987-1995). J Am Anim Hosp Assoc 2000;36:385-390.

11. Baer KE, Patnik AK, MacDonald JM. Hepatic vascular dysplasia in dogs and cats (105 cases), in Proceedings. 42nd Ann Meet Am Coll Vet Pathol 1991;71.

12. Albillos A, Colombato LA, Enriquez R, et al. Sequence of morphological and hemodynamic changes of gastric microvessels in portal hypertension. Gastroenterology 1992;102:2066-2070.

13. Center SA. Pathophysiology of liver disease: Normal and abnormal features. In: Strombeck's small animal gastroenterology. 3rd ed. Philadelphia, Pa: WB Saunders Co, 1996;573-574.

14. Hopper K, Aldrich J, Haskins SC. Ivermectin toxicity in 17 collies. J Vet Intern Med 2002;16:89-94.

15. Plumb D. Veterinary drug handbook. 3rd ed. Ames: Iowa State University Press, 1999;230:358-361.

16. Mandel NS, Schneider EM, Bosler EM, et al. Intrathecal production of Borrelia burgdorferi-specific antibodies in a dog with central nervous system Lyme borreliosis. Compend Cont Ed Pract Vet 1993;15:581-589.

17. Bleck TP. Central nervous system involvement in Rickettsial diseases. Neurol Clin 1999;17:801-812.