Introduction

The intracranial extension of otitis media/interna is a rare but serious medical condition requiring urgent and precise therapeutic interventions. Only a few case reports describe its clinical signs, imaging features, and outcome in dogs and cats.^1,2 The incidence of otogenic intracranial complications in humans is 0.24–0.36%.^3,4 Common complications include meningitis, brain abscess, lateral sinus thrombosis, hydrocephalus, and suppurative disease involving adjacent areas. Multiple complications were reported in 11–45% of patients.^4,5 Despite advanced imaging techniques and the advent of effective antibiotics, the mortality rate of otogenic intracranial complications remains high, ranging from 5 to 14% in humans.^5,6

Cerebral ventriculitis is uncommon in central nervous system infections and is considered a late and lethal complication of meningitis.⁷ It is variably referred to as ependymitis, ventricular empyema, pyocephalus, and ventriculitis.^7–9 Few reports have documented ventriculitis in the veterinary literature, and most of them were associated with feline infectious peritonitis. Only three reports describe ventriculitis in dogs, which resulted from blastomycosis, foreign body migration, and suspected ruptured brain abscess.^8–10 All five dogs in those reports either died or were euthanized over the course of their diseases.

Ventriculitis-complicated otogenic meningoencephalitis, to the best of the authors' knowledge, has not been documented in the veterinary literature. In human medicine, ventriculitis is also a rare intracranial complication of otitis media/interna. Although rare, ventriculitis is a life-threatening complication due to increased intracranial pressure (ICP) and because only a limited amount of antibiotics can penetrate cerebrospinal fluid (CSF) when the ventricular system is blocked by suppurative debris. This report describes the clinical features, MRI findings, and the long-term outcome of a dog with cerebral ventriculitis and meningoencephalitis that resulted from the intracranial extension of otitis media/interna.

Case Report

A 7 yr old male mixed-breed dog was referred for rapidly progressive mentation changes, ataxia, and tetraparesis. Only mild skin disease and chronic otitis externa were reported as previous illnesses. An occasional, mild left head tilt was noted for many years, the severity of which remained unchanged. Two weeks before referral, one episode of ataxia lasting for only 2 min was noted, but the dog recovered afterwards. An increase in vocalization was observed 2 days prior to presentation. The patient deteriorated rapidly, became ataxic, and howled through the night before referral.

At the time of presentation, the dog was in lateral recumbency, unable to rise, had mild dehydration, and tachypnea. A neurological examination revealed severe stupor, disorientation, and nonambulatory tetraparesis with normal spinal reflexes. Abnormalities observed on cranial nerve examination included anisocoria (a miotic left pupil and mydriatic right pupil), bilaterally absent menace response and palpebral reflex, bilaterally delayed direct and consensual pupillary light response (PLR), spontaneous nystagmus with either a horizontal (quick phase to the right) or rotatory direction, bilaterally decreased facial sensation, and a weak gag reflex. Neuroanatomic localization was consistent with a diffuse brainstem lesion; however, forebrain involvement was also considered based on the presence of disorientation. The differential diagnoses included infectious diseases (e.g., viral, bacterial, fungal, protozoal, rickettsial), inflammatory diseases (e.g., granulomatous meningoencephalitis), neoplasia, or unknown toxin exposure.

Hematology and serum biochemical analysis detected a matured neutrophilic leukocytosis (WBCs, 27.1 × 10⁹/L; reference range, 6–17 × 10⁹/L; neutrophils, 22.2 × 10⁹/L; reference range, 3–11 × 10⁹/L ), monocytosis (2.7 × 10⁹/L ; reference range, 0.15–1.35 × 10⁹/L), and mild hyperglycemia (11.54 mmol/L; reference range, 4.11–7.94).

Within 2 hr of admission, the patient's neurological signs deteriorated drastically, and the dog became comatose with bilateral mydriasis and lacked a PLR. Mannitol^a (0.5 g/kg IV q 12 hr) was administered to address the suspected elevated ICP and IV fluid therapy was initiated. Mentation improved subtly following the mannitol treatment; however, the dog remained severely stuporous. Based on the differential diagnoses, an anti-inflammatory dose of methylprednisolone^b (1 mg/kg IV q 12 hr) was administered to further stabilize the dog prior to anesthesia for intracranial investigation.

The dog's neurological status ameliorated gradually but significantly a few hours after methylprednisolone administration. The following day, however, the dog remained stuporous and lacked a menace response. The PLR had returned bilaterally and a left miotic pupil was observed. The right palpebral reflex, gag reflex, and bilateral facial sensation improved, but left facial paralysis remained. Other findings included a marked left head tilt, left body rolling, and nonambulatory tetraparesis with postural reaction deficits in all limbs, especially on the left side. Hyperesthesia was elicited when palpating the cranial cervical region. A brainstem lesion affecting the left side more severely than the right was suspected. The cervical hyperesthesia was thought to indicate either a cervical spinal lesion or a sign secondary to brain disease. Radiographs of the cervical spine and the thorax revealed mild right-sided cardiomegaly. Echocardiography and electrocardiography were unremarkable.

Once the neurological status was stabilized, MRI of the brain was performed under anesthesia 48 hr after admission using a 0.2 Tesla MRI unit^c with a DPA coil. The following sequences were performed: T1-weighted [repetition time (TR), 560 msec; echo time (TE), 18 msec], T2-weighted (TR, 3000 msec; TE, 90 msec), fluid-attenuated inversion recovery (FLAIR; TR, 5090 msec; TE, 90 msec; inversion recovery, 960 msec), postcontrast T1-weighted and postcontrast FLAIR (TR, 4000 msec; TE, 90 msec; inversion recovery, 960 msec). Transverse, dorsal, and sagittal planes were obtained from T1- and T2-weighted images. Pre- and postcontrast FLAIR images were performed for the dorsal and transverse planes, respectively.

The left tympanic cavity was filled with materials that appeared iso- to hypointense on T1-weighted images (Figure 1A) and hyperintense on T2-weighted images compared with the grey matter in the brain. Additionally, a mass bulged from the dorsomedial aspect of the left tympanic cavity through the dehiscence of the petrous bone, extending into the subarachnoid space near the left lateral aperture. That mass was iso- to hypointense on T1-weighted images and iso- to mildly hyperintense on T2-weighted images. Structures representing the normal inner ear were not present on the left side. On the left side of the brainstem, there was a small triangular area with mild hyperintensity on T2-weighted and FLAIR images. That area was in close contact with the mass described above, suggesting focal brain edema associated with the mass. On T2-weighted and dorsal FLAIR images, a poorly demarcated hyperintensity in the white matter dorsolaterally to the lateral ventricles was evident in the caudal aspect of the cerebrum, indicating severe periventricular edema (Figures 2A, D, E). Following gadolinium^d administration (0.13 mmol/kg IV), T1-weighted images revealed enhancement of the mucosal lining of the left tympanic cavity and the adjacent mass. Subtle enhancement on the left side of the brainstem close to the mass was also noted (Figure 1C). Strikingly, ependymal contrast enhancement of all ventricles was observed together with strong meningeal enhancement from the brainstem to the cervical spine (Figures 1, 2). Given that more slices of the ventricular system could be demonstrated in the transverse plane, it was chosen for postcontrast FLAIR images in search of subtle changes near the ventricular system. This revealed a markedly high signal intensity for the CSF in all ventricles, suggesting a possible blood-CSF barrier breakdown and gadolinium leakage into the CSF during the inflammatory process (Figure 2F).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

A CSF tap was performed, and the sample from the cerebellomedullary cistern was xanthochromic and viscous. Only a small quantity of CSF was successfully obtained from that location due to the high viscosity of the fluid. A cell count analysis revealed marked pleocytosis (WBCs, 3672 × 10⁶/L; reference range, 0–5 × 10⁶/L). Analysis of the CSF obtained from the caudal lumbar region revealed markedly increased protein (>4 g/L; reference range, 0–0.45 g/L) and severe neutrophilic pleocytosis (WBCs, 8290 × 10⁶/L; reference range, 0–5 × 10⁶/L; 82% neutrophils, 6% lymphocytes, and 12% monocytes; reference range, neutrophils <1%, lymphocytes 60–70%, monocytes 30–40%). An endoscopic evaluation of the left ear canal revealed a ruptured tympanic membrane. The cytological evaluation of the materials from the left tympanic cavity detected large amounts of gram-positive bacilli, gram-negative bacilli, gram-positive cocci, and a small amount of yeast. At that stage, severe meningoencephalitis and ventriculitis associated with the intracranial extension of otitis media/interna were diagnosed.

The patient recovered from anesthesia uneventfully. Empirical antibiotic therapy including trimethoprim-sulfonamide^e (15 mg/kg subcutaneously q 12 hr) and metronidazole^f (10 mg/kg IV q 8 hr) were initiated directly following the investigations. The dog's neurological status further improved. On the 4th day of hospitalization, he was alert and ambulatory but tetraparetic with the left side exhibiting more postural reaction deficits. Mannitol was discontinued, and parenteral antibiotics were changed to their corresponding oral formulations. Methylprednisolone was substituted with an anti-inflammatory dose of prednisolone^g [0.5 mg/kg per os (PO) q 12 hr].

The CSF bacterial culture was negative, but the sample from the tympanic cavity yielded Staphylococcus aureus, Proteus mirabilis, and Enterococcus faecalis. According to antibiotic susceptibility testing, enrofloxacin^h (2.5 mg/kg PO q 12 hr) was added for thorough coverage of all three bacteria. Total ear canal ablation and a lateral bulla osteotomy were performed on the 9th day. At that time, except postural reaction deficits in the left limbs, left head tilt, and left facial paralysis, the remainder of the neurological exam was unremarkable. During surgery, the tympanic cavity filling with dark green gelatinous material was observed. Aerobic bacterial culture of the materials revealed the same three bacteria as the previous culture. The dog was discharged, and continuation of the antibiotics was instructed. Prednisolone was gradually withdrawn over 1 mo.

Re-examination on the 25th day revealed left facial paralysis and left head tilt without other neurological abnormalities. Forty-five days after the first examination, the dog was clinically stable with only a mild left head tilt. A lumbar CSF tap was repeated, which revealed an elevated protein level (0.86 g/L), lymphocytic pleocytosis (WBCs, 335 × 10⁶/L; 93% lymphocytes, 5% monocytes, 2% neutrophils), and an increased red blood cell count (2106 × 10⁶/L; reference range, 0 × 10⁶/L) due to blood contamination. Antibiotic therapy was continued for 1 mo. Follow-up CSF analysis was proposed but declined by the owner. After a 2.5 mo course of antibiotic therapy, treatment was discontinued based on the stable condition of the dog.

Seven months after surgery, the dog presented with acute left facial swelling. Neurological status was unchanged compared with the previous visit. Computed tomography suggested the recurrence of otitis media/interna and secondary abscess formation without an intracranial abnormality. An otoscope-assisted lateral bulla osteotomy was performed to create a larger defect on the wall of the left tympanic cavity for better visualization and complete removal of the squamous epithelium remnants. Aerobic bacterial culture of the epithelium isolated Proteus mirabilis. Amoxicillin trihydrate/clavulanate potassiumⁱ (20 mg/kg PO q 12 hr) and enrofloxacin (5 mg/kg PO q 24 hr) were prescribed for 1 mo. The dog recovered well from the second surgery. Three years after the initial presentation, the dog remained healthy with only a mild residual left head tilt.

Discussion

Intracranial extension of otitis media/interna is an uncommon disorder in both humans and animals. The infecting organism may gain access into the intracranial cavity through progressive thrombophlebitis, erosion of the bony walls of the middle ear, or extension along a pre-formed pathway (e.g., the round window, congenital dehiscent sutures, a skull fracture, and surgically induced bony dehiscence). Multiple factors, including bacterial virulence, anatomical defects, and altered host immunity, can influence whether otitis media/interna extends into the intracranial cavity. Once the intracranial extension of otitis media/interna takes place, two patterns of intracranial infection are commonly observed (1) focal meningitis adjacent to a parenchymal or subdural abscess near the cerebellomedullary angle or (2) generalized meningitis resulting from a direct infection extension into the subarachnoid space.¹¹ Bacterial meningitis initiates a cascade of intracranial inflammation, resulting not only in inflammation of the subarachnoid space but also vasculitis, cerebral edema, and injury to cortical and subcortical brain structures. In a previous report on otogenic intracranial infections in dogs and cats, severe and generalized meningitis was found only in animals with acute neurological dysfunction. That was in contrast to focal meningitis adjacent to intracranial lesions in animals with subacute and chronic neurological signs.¹ Generalized bacterial meningitis may be responsible for the acute and fulminating deterioration of neurological signs, which led to a relatively poor outcome in that study.

Ventriculitis occurs most commonly as a late and fatal complication of meningitis. An alteration of the blood-brain barrier and blood-CSF barrier permeability in bacterial meningitis results in vasogenic cerebral edema and allows leakage of serum protein and other molecules into the CSF, which contributes to the formation of a purulent exudate in the subarachnoid space and ventricular system.¹² The purulent exudate impairs the CSF flow and resorption, resulting in the blockade of outflow tracts and hydrocephalus. It is not uncommon to observe obstruction of the ventricular system at the mesencephalic aqueduct or in the lateral apertures in young farm animals with bacterial meningoencephalitis or in cats with feline infectious peritonitis.

Ventriculitis-complicated meningitis puts animals at a high risk of sudden deterioration of neurological signs associated with elevated ICP due to brain edema and/or the formation of hydrocephalus.¹³ In the patient described herein, it was suspected that the extension of the suppurative focus of the otitis media/interna close to the left lateral aperture facilitated the infection's entrance into the ventricular system. Although obstructive hydrocephalus was not detected, generalized meningitis and ventriculitis with periventricular edema was evident on this dog's MRI. The profound periventricular edema in this scenario was most likely associated with the diffuse ventriculitis. However, it is also a typical imaging finding of hypertensive hydrocephalus.¹⁴ The study authors hypothesized that the drastic deterioration of neurological status at the time of initial presentation was caused by the rapid development of severe brain edema and/or obstructive hydrocephalus due to the diffuse ventriculitis and meningitis. Following the steroid and mannitol treatments, the severe edema and/or hydrocephalus subsided. Subsequently, the neurological dysfunction improved remarkably over a short period of time.

Characteristic findings of ventriculitis on MRI include ventricular debris, periventricular subependymal signal abnormality, ependymal contrast enhancement, signs of meningitis, and hydrocephalus.¹⁵ In this dog, periventricular hyperintensity on FLAIR and T2-weighted images and ependymal enhancement on postcontrast T1-weighted images were the most prominent abnormalities, indicating the presence of ventriculitis and the periventricular inflammatory changes.¹⁵ Furthermore, gadolinium leakage into all ventricles was noted on the postcontrast FLAIR images. Similar findings were observed in human patients with lesions close to either the subarachnoid space or ventricles, as well as patients with significant disruption of the blood-brain barrier or blood-CSF barrier. That phenomenon is best detected on FLAIR images due to its extreme sensitivity to alternations of the fluid composition.¹⁶ Although studies have shown that gadolinium concentration in the CSF increases after IV contrast administration in healthy animals, the visible change of CSF signal intensity primarily depends on the underlying clinical conditions that may exacerbate the effect and technical factors such as contrast dose and/or timing of image acquisition. One study of dogs showed that a triple dose of contrast agent or a 6 hr delay in scanning exhibited an increased CSF signal on the postcontrast FLAIR images.¹⁷ The dog described herein received a standard dose of contrast agent, and the time lag between contrast administration and image acquisition was <0.5 hr. In addition, with the equivalent time lag and contrast dose, the gadolinium leakage into the ventricles was not observed in other animals scanned by the study authors' MRI unit. After ruling out technical factors, the compromised blood-CSF barrier at the choroid plexus or disruption of the blood-CSF interface due to the severe and diffuse ventriculitis was the most reasonable explanation for gadolinium leakage in this dog.

Steroid administration in the dog reported herein significantly improved the neurological signs and stabilized the dog's condition, which allowed further intracranial investigation to be performed with lower risks under general anesthesia. The use of steroids as an adjunct treatment to antibiotics in bacterial meningitis is controversial. The results from several experimental studies indicate that a high dose of dexamethasone aggravates neuronal damage in the hippocampal formation and may potentiate ischemic injury to neurons.^18,19 In contrast, others suggest it is advantageous to give an immunosuppressive dose of dexamethasone either before or with the first dose of antibiotics against acute bacterial meningitis. However, this should be carefully monitored throughout the therapy and repeating CSF analysis in 24 to 48 hr should be considered to evaluate the status of infection, as steroids may reduce the blood-brain permeability and consequently the penetration of antibiotics into the subarachnoid space.²⁰ One quantitative systemic review of five clinical trials involving 623 adults that met standards of randomization, double-blinding, and withdrawals/drop outs concluded that steroid therapy for bacterial meningitis could substantially reduce neurological sequelae and mortality.²¹ Although the regimes of steroid therapy varied, they were all higher than a usual anti-inflammatory dose. However, controversy surrounding the use of adjunctive steroid treatment remains, especially as the outcome seemed highly related to the causative pathogens, the patient's immune status, and the timing of medical attendance. In the veterinary literature, an anti-inflammatory dose of steroids was suggested in case reports of otogenic intracranial infection and streptococcal meningoencephalitis.^1,22 Further studies are still required to generate guidelines addressing steroid therapy in canine bacterial meningoencephalitis. Nonetheless, this case report suggested a beneficial effect of an anti-inflammatory dose of steroids in this dog with acute bacterial meningitis and ventriculitis.

The principles of treatment for ventriculitis are generally the same as those for acute bacterial meningitis with the exception that the intraventricular administration of antibiotics may be considered in refractory and shunt-associated ventriculitis in humans.¹³ The optimal duration of antibiotic therapy for acute meningitis, even for the most common pathogens, remains unclear. Traditionally, treatment duration of 7–10 days is recommended for meningococcal meningitis in human patients, and a longer course (10–21 days) is recommended for other pathogens.¹⁹ In this canine patient, abnormal CSF findings remained after 1.5 mo of treatment. That may indicate the difficulties, despite the resolution of most clinical signs, with eradicating infections/inflammation in a dog with ventriculitis and meningoencephalitis.

Conclusion

Ventriculitis-complicated otogenic meningoencephalitis can lead to an acute deterioration of neurological signs. The short-term use of an anti-inflammatory dose of steroids may provide initial stabilization for further investigation without jeopardizing either imaging or clinical pathology findings. MRI utilizing pre- and postcontrast sequences is very sensitive for the detection of ventriculitis. When compared with a postcontrast T1-weighted image, postcontrast FLAIR imaging proved to have better sensitivity for demonstrating a CSF change in this case. Longer antibiotic therapy may be required for ventriculitis-complicated meningoencephalitis despite the resolution of the majority of clinical signs.