External Hydrocephalus in a Dog With Suspected Bacterial Meningoencephalitis

Case Report

An approximately 12-week-old, male fox terrier presented to the Texas A&M University Veterinary Medical Teaching Hospital (TAMU-VMTH) with a 6-week history of progressively worsening signs of encephalopathy. At 6 weeks of age, the owners had noticed that the dog had become less responsive to environmental stimuli, disinterested in food, and less willing to ambulate. When the dog did walk, he would circle to the right. Other abnormalities noted included progressive enlargement and dome-shaped appearance of the dog’s calvarium and increasing episodes of aggressive behavior, especially when handled. The puppy had not yet received any vaccinations and had recently been started on a regimen of oral prednisone (1.0 mg/kg body weight, q 12 hours) by the referring veterinarian.

Upon presentation to the TAMU-VMTH, the puppy appeared obtunded and constantly circled to the right. The patient was in overall thin body condition and had a dome-shaped calvarium that seemed disproportionately large in relation to his body. Several open fontanelles were palpable. During the neurological examination, the dog exhibited frequent bouts of apparent rage when handled, vocalizing and biting at the examiner and then at his own limbs for several seconds after being released. In addition to altered mental status and circling, abnormal neurological examination findings included absent menace response bilaterally, decreased facial sensation (inner nare) on the left side, bilateral ventrolateral strabismus, bilateral rotary nystagmus, decreased proprioceptive positioning in the left thoracic limb, and absent proprioceptive positioning in the remaining limbs. The puppy seemed hyperesthetic when his head and neck were palpated. The neuroanatomical localization was multifocal/diffuse encephalopathy, based on the combination of both cerebral and brain-stem dysfunction.

Blood was procured for a complete blood count (CBC) and a serum biochemistry panel. A mature neutrophilic leukocytosis (62 × 10³ white blood cells [WBCs]/μL; reference range, 6 to 17 × 10³ WBCs/μL) was apparent on the CBC. Serum biochemistry panel abnormalities included mild elevation of alkaline phosphatase activity (162 U/L; reference range, 24 to 147 U/L), considered normal for a young puppy, as well as mild hyponatremia (126 mmol/L; reference range, 138 to 148 mmol/L) and mild hypochloridemia (91 mmol/L; reference range, 108 to 118 mmol/L). The puppy was anesthetized, and a computed tomography (CT) scan of the brain was performed. The brain appeared distorted and shifted axially on the CT images, with large accumulations of fluid on either side. The lateral ventricles appeared moderately enlarged [Figure 1]. Fluid accumulation was also apparent dorsal to the cerebellum on transaxial images of the caudal fossa [Figure 2]. The imaging diagnosis, based upon the location of the accumulated fluid, was external hydrocephalus. A cerebrospinal fluid (CSF) sample obtained from the cerebellomedullary cistern was consistent with a mixed-cell pleocytosis. The WBC count of the CSF sample was 283/μL (reference range, 0 to 5/μL), with 49% neutrophils and 51% mononuclear cells. The majority of the neutrophils appeared nondegenerate. The red blood cell count was 40/μL. The CSF protein level was 259 mg/dL (reference range, <30 mg/dL). No organisms were apparent on CSF cytopathological evaluation, but a sample of the CSF was submitted for culture and antibiotic sensitivity testing. The culture results were negative for the initial 48 hours following sample submission. Serology was negative for Toxoplasma, Neospora, and Cryptococcus.

The patient was placed on a combination of intravenous ampicillin (40 mg/kg body weight, q 6 hours), oral clindamycin (11 mg/kg body weight, q 12 hours), and oral trimethoprim-sulfonamide (TMS; 15 mg/kg body weight, q 12 hours) while awaiting results of serology and CSF culture/sensitivity. Because the nature of the dog’s encephalitic process was unknown, a broad-spectrum antibiotic regimen with activity against gram-positive and anaerobic bacteria, as well as protozoal organisms, was chosen. Oral prednisone was continued (1.0 mg/kg body weight, q 12 hours). Although the patient’s WBC count declined dramatically (22.7 × 10³ WBCs/μL) between 48 and 72 hours following institution of antibiotic therapy, his neurological status appeared to be slightly worsened. The dog was less able to walk, tending to fall frequently. He also appeared more obtunded and would vocalize repeatedly. After consulting with the dog’s owner, the decision was made to surgically place a ventriculoperitoneal shunt (VPS)^a in order to divert the excess fluid in the cranial vault to the peritoneal cavity. The VPS was placed [Figure 3], and the patient recovered well from the procedure. During surgery, fluid was obtained directly from the cranial vault and submitted for cytopathology and culture/antibiotic sensitivity testing. Grossly, the fluid appeared somewhat viscous, had a reddish-brown color (assumed to be hemorrhagic), and had small particulate matter floating in it. Cytopathologically, the fluid was hemorrhagic, with a predominance of degenerate neutrophils. No organisms were identified, and bacterial cultures were negative. A meningeal biopsy was also obtained during surgery. The meningeal tissue was grossly abnormal, exhibiting a brown discoloration; the histopathological diagnosis was fibrinopurulent meningitis.

The dog’s neurological status was unchanged after surgery and did not appear to worsen over the subsequent few days. Approximately 24 hours after placing the VPS, a Staphylococcus capitis organism was cultured from the previously submitted cisternal CSF sample. The organism was sensitive to several antibiotics, including penicillins, clindamycin, and TMS. Two days prior to hospital discharge, clindamycin and ampicillin were discontinued, and the antibiotic regimen was switched to oral amoxicillin/clavulanic acid (22 mg/kg body weight, q 12 hours). The TMS was continued. The dog was discharged from the hospital approximately 1 week after surgery. At this time, the oral prednisone dose was decreased to 0.5 mg/kg body weight, q 48 hours for 1 week, then discontinued. During the second week after discharge, a recheck CBC was performed by the referring veterinarian, and a marked mature leukocytosis was again apparent (51.62 × 10³ WBCs/μL). The dog’s corneas also appeared dry, and a Schirmer tear test was abnormally low in both eyes. The TMS and the amoxicillin/clavulanic acid were discontinued, and oral clindamycin was reinstituted. Approximately 1 month later, the WBC count was rechecked by the referring veterinarian and was within reference ranges (12.94 × 10³ WBCs/μL). Clindamycin was discontinued at this time.

Four months after initial presentation, the dog was eating and drinking voluntarily and gaining weight. The owners reported a decrease in frequency of rage episodes and that the dog would occasionally walk in a straight line. The blindness persisted. One year after initial presentation, the dog was doing well according to the owner (contacted via telephone). The dog circled only occasionally, often when tired, and usually walked in a straight line. He also responded to his name. Episodes of rage were infrequent and typically associated with stressful events. The dog was still blind at this last follow-up communication.

Discussion

Hydrocephalus refers to the accumulation of excessive CSF within the ventricular system of the brain.¹ There are numerous terms used to classify hydrocephalus, most based upon either suspected or known etiologies, or suspected or known anatomical locations of impaired CSF flow. The author prefers the term congenital hydrocephalus to describe cases of hydrocephalus in young animals for which no underlying cause is identified. It is likely that many cases of congenital hydrocephalus represent a developmental abnormality which leads to impaired CSF absorption at the arachnoid villi level. Despite suggestions that these animals develop hydrocephalus as a sequela to events, such as in-utero central nervous system (CNS) viral infections and perinatal intracranial hemorrhage, the absence of any evidence of a causative disorder in these dogs helps clinically define this form of hydrocephalus. Congenital hydrocephalus is the most common form of the disorder encountered in small animal practice, and it most often occurs in miniature and toy dog breeds (e.g., Chihuahuas, Yorkshire terriers, Maltese) within the first year of life.^1–3 Acquired hydrocephalus refers to cases in which an etiology for the excessive fluid accumulation is identified. Examples of disease states that may lead to acquired hydrocephalus include neoplasia, infectious/inflammatory disorders (e.g., meningoencephalitis), and intracranial hemorrhage. These various disorders can lead to hydrocephalus via direct mechanical obstruction to intraventricular CSF flow (e.g., tumor) or via interference with absorption of CSF at the arachnoid villi level (e.g., meningoencephalitis).^1–3

The characteristic appearance of hydrocephalus on CT or magnetic resonance (MR) imaging is dilatation of the ventricular system, most predominantly involving the lateral ventricles.¹ Occasionally, humans will develop a form of hydrocephalus in which the majority of excessive CSF is located in the subarachnoid space surrounding the cerebral convexities, with no or mild evidence of ventriculomegaly. This distinct CT or MR appearance is termed external hydrocephalus.^4–8 The vast majority of external hydrocephalus cases in humans are idiopathic and occur in infants. These infants typically have disproportionately enlarged craniums associated with the external hydrocephalus, referred to as macrocephaly or megalencephaly. This disorder is often clinically mild and self-limiting.^4–8 Recent evidence suggests that idiopathic macrocephaly and external hydrocephalus in infants may be associated with primary hypomagnesemia.⁹ External hydrocephalus has also been reported in humans as an acquired disorder, secondary to conditions such as head injury (with intracranial hemorrhage) and meningitis.⁴¹⁰

The fluid dynamics responsible for the development of external hydrocephalus are poorly understood. It has been postulated that resistance to CSF flow through the arachnoid villi into the venous sinuses may be an early event in the development of hydrocephalus. This resistance would favor initial accumulation of CSF in the subarachnoid space.⁸ Alternatively, there is evidence in a rat model of hydrocephalus that extra-axial accumulation of CSF occurs following rupture of the caudal pole of the occipital lobe as a consequence of severe internal hydrocephalus.¹¹

A rare form of acquired internal hydrocephalus was first reported by Higgins, et al., in 1977.¹² This idiopathic disorder, termed hydrocephalus with periventricular encephalitis (HPE), occurs in young dogs (6 weeks to 6 months of age) and is characterized by rapid onset and progression of encephalopathic signs. Hydrocephalus in this disorder is thought to be a sequela to a severe meningoencephalitis of unknown etiology.^12–16

Hydrocephalus with periventricular encephalitis appears to be a rapidly progressive, usually fatal disease. Cerebrospinal fluid from affected dogs is typically hemorrhagic or xanthochromic, with a mixed-cell pleocytosis and increased protein concentration. Free-floating aggregates of cellular debris have also been described in the CSF of dogs with HPE.^12–16 Intense inflammatory infiltrates and hemorrhagic encephalomalacia are characteristic postmortem findings in brain tissue from HPE patients.^12–16 Although several infectious agents have been implicated as the cause of HPE, the pathological changes are most consistent with a bacterial etiology.^12–16 In two cases, postmortem bacteriological culture results were positive but were considered of questionable clinical significance. Pasteurella multocida and Staphylococcus aureus were isolated from brain tissue in one dog, and a Pasteurella species was isolated from CSF fluid in another dog.¹²¹³

To the author’s knowledge, this is the first clinical report of external hydrocephalus in a dog. The dog described in this case report exhibited clinical and pathological features characteristic of HPE. As suggested in confirmed reports of HPE, it is possible that the positive CSF culture of Staphylococcus capitis obtained in this case represents a contaminant. The rapid and repeatable reduction of the dog’s peripheral WBC count shortly after instituting clindamycin therapy is evidence of a bacterial etiology. Prednisone was continued for a short time after obtaining the positive CSF culture, due to the severe inflammatory nature of the dog’s disease process. Although glucocorticoid use in the face of bacterial infection is usually contraindicated, there is evidence that transient glucocorticoid therapy at antiinflammatory doses improves outcome in human bacterial meningitis.¹⁷

The decision to place a VPS in the dog was made in advance of the positive CSF culture results and in the absence of any cytopathological evidence of bacterial infection. However, an infectious etiology was suspected before shunt placement. Preexisting infection is regarded as a strict contraindication to VPS placement, as this can lead to shunt failure (i.e., clogging of the shunt with inflammatory debris) as well as potentially life-threatening septic complications for the patient.¹⁸¹⁹ It was decided that there would be little hope for neurological improvement in this case if a sustained conduit for accumulated CSF was not provided in addition to medical therapy. The risks of placing a VPS in a potentially infected environment were weighed against the risks of not relieving the intracranial fluid accumulation in this particular patient. These risks were also carefully explained to the dog’s owners before the decision was made to place the VPS. Whether or not the VPS contributed to the dog’s neurological improvement is debatable, as no follow-up brain imaging was obtained. However, the dog appeared to be worsening neurologically despite antibiotic therapy and only exhibited sustained improvement in neurological function following VPS placement. Although the dog did exhibit improvement in his neurological status following treatment, substantial neurological impairment persisted. These neurological deficits may reflect permanent brain damage incurred prior to treatment initiation. Alternatively, the persistent neurological deficits may reflect malfunction of the VPS.