Feline meningoencephalomyelitis of unknown origin: A retrospective analysis of 16 cases

Abstract

This study aimed to describe the signalment, clinical signs, magnetic resonance imaging (MRI) findings, cerebrospinal fluid (CSF) analysis, treatment, and outcome of feline meningoencephalomyelitis of unknown origin (FMUO). Medical records from 16 cats meeting the inclusion criteria of CSF pleocytosis, negative CSF polymerase chain reaction (PCR)-infectious disease results, and characteristic MRI findings were retrospectively reviewed. Median age was 9.4 years. Clinical signs included ataxia, proprioceptive deficits, seizures, and spinal hyperesthesia. The CSF nucleated cell count was increased (median 70.7 cells/μL), with predominantly mixed pleocytosis and CSF protein concentration was increased in 15/16 cats. Magnetic resonance imaging showed intraparenchymal infiltrative ill-defined lesions in 13 cases. All cats received a corticosteroid-based treatment protocol; additional therapies included lomustine, cytarabine, and anticonvulsant medications. Mild neurological signs were recorded in 5/12 cats but 7/12 cats were neurologically normal at re-examination. This represents the first study of feline MUO, highlighting FMUO as an important differential diagnosis in cats with variable neurological presentation. Prognosis appears to be good with immunomodulatory therapy.

Introduction

Inflammatory central nervous system (CNS) disorders are one of the most common causes of neurological dysfunction in dogs and cats and can be divided into 2 broad classes: CNS inflammation due to an infectious origin and CNS inflammation without an identifiable infectious cause (1,2). The latter is commonly referred to as meningoencephalitis or meningoencephalomyelitis of unknown origin (MUO) (2,3). It is useful to consider disorders classed within MUO individually when possible, as clinical features and response to therapy tends to differ between sub-groups. However, as a histological diagnosis is often not possible, the term MUO is commonly used (2).

While MUO is extensively reported in dogs (2,4–7), few reports of non-infectious meningoencephalomyelitis have been described in cats (8–10). Bradshaw et al (10) reported that 11% of 286 cats with encephalitis or meningitis had no evidence of an infectious agent, while 5 cases of presumed feline MUO were classified in Singh et al (8) as either non-suppurative or steroid-responsive based on cerebrospinal fluid (CSF) characteristics and response to treatment. Although the underlying cause of MUO remains elusive, it appears to be associated with an aberrant immune response directed against the CNS (11) and consequently immunomodulatory therapy is the mainstay of canine MUO treatment. There are no reported treatment protocols for feline MUO (FMUO), although 2 cases were successfully treated with prednisolone (8). Response to therapy and outcome of canine MUO are highly variable while data on the prognosis of cats with the condition are currently unknown (2,9).

This report represents the largest study of presumed FMUO cases in the literature. The aim of this study was to describe the signalment, clinical signs, magnetic resonance imaging (MRI) findings, ancillary diagnostic testing, CSF analysis, treatment and outcome of FMUO, providing background information for further studies focusing on clinicopathological features and optimal treatment protocols.

Materials and methods

Selection criteria

Medical records (2008 to 2016) of cats presented to the Neurology and Neurosurgery Service at Dick White Referrals were retrospectively reviewed. Inclusion criteria comprised CSF pleocytosis, MR characteristics indicative of inflammatory lesions, and negative CSF reverse-transcription polymerase chain reaction (RT-PCR) infectious diseases results. Where neoplasia was suspected as a differential diagnosis, further imaging and laboratory tests were performed and reviewed. Additional diagnostic investigations, including Toxoplasma gondii serology, were conducted in some cases. Cats with a history of steroid administration before presentation were excluded. Signalment, duration of clinical signs, physical and neurological examination findings, neuroanatomical localization, CSF nucleated cell count and cytological characteristics, CSF protein concentration, diagnostic imaging results including MRI, treatment, neurological examination findings at re-examination(s), and occurrence and therapy at time of relapse were recorded. Clinical signs were deemed acute if present for < 2 wk, and chronic if reported for > 2 wk.

Cerebrospinal fluid analysis

Cerebrospinal fluid was collected from the cisterna magna in all cases. Analysis was performed within 1 h of collection. Cerebrospinal fluid nucleated cell counts were classified as: normal (< 5/μL), mildly (5 to 80/μL), moderately (81 to 500/μL), or markedly (> 500/μL) increased (8). Cerebrospinal fluid protein levels were classed as: normal (< 0.3 g/L), mildly (0.31 to 1.0 g/L), moderately (1.1 to 3.0 g/L), or markedly (> 3.0 g/L) increased (8). Nucleated cell population was classified as neutrophilic (> 50% neutrophils), mononuclear (> 80% mononuclear cells), eosinophilic (> 50% eosinophils), or mixed (no predominance of any 1 cell type) (6). The RT-PCR analysis on all CSF samples was carried out in all cases for the following infectious agents: feline calicivirus, feline herpesvirus, Chlamydophila felis, Toxoplasma gondii, Bornavirus, feline leukemia virus, feline immunodeficiency virus, Leishmania infantum, feline parvovirus (panleucopenia virus), and feline coronavirus.

Magnetic resonance imaging protocol

Magnetic resonance imaging was performed using a 0.4T permanent magnet scanner (Hitachi Aperto Lucent 0.4T; Hitachi Medical Systems, Wellingborough, UK). Pulse sequences varied but in all cats transverse T1WI, T2WI, FLAIR sagittal T2WI, and post-contrast gadoteric acid (Gadovist; Bayer Schering Pharma, Reading, UK), 0.1 mmol/kg body weight (BW), IV, transverse T1WI were acquired. Images were reviewed by board-certified radiologists who were blinded to the study. Images were assessed for the presence of lesions, lesion pattern (focal, diffuse or multifocal), and location. The presence of meningeal involvement was recorded. Lesion margins were described as well-defined, irregular, or infiltrative. Mass effect was indicated by effacement of sulci, shift of midline structures, cerebellar herniation (evaluated on sagittal midline T2WI) or ventricular system deviation/compression. Pre- and post-contrast T1WI were compared to assess for contrast enhancement and the pattern of contrast uptake was described.

Treatment

Treatment depended on clinician preference and individual case requirements. All cats received clindamycin (Antirobe; Zoetis, London, UK), 12.5 mg/kg BW, PO, q12h, while waiting for 5 to 7 d for CSF infectious disease PCR results. All cats were treated with a corticosteroid-based protocol (dexamethasone only, dexamethasone followed by prednisolone, or prednisolone only). Immunosuppressive steroid doses were initially used, including prednisolone (Prednicare; Animalcare, York, UK), 1 mg/kg BW, PO, q12h and dexamethasone (Dexamethasone; Aspen, Redditch, UK), 0.2 to 0.3 mg/kg BW, PO, q24h. Corticosteroid therapy was tapered by typically halving the dose or dosage every 2 to 3 wk unless a relapse was seen. Additional immunomodulatory therapies included lomustine (Lomustine; Nova laboratories, Wigston, UK), 10 mg/cat, PO, single dose on day 1, and cytarabine (Cytarabine; Pfizer, Sandwich, UK), 50 mg/m², SC, q12h for a total of 4 doses, at the clinician’s discretion. Anticonvulsant medication was used in all cats presented with seizures and included phenobarbitone (Phenoleptil; Animalcare), 1 mg/kg BW, PO, q12h, levetiracetam (Keppra; UCB Pharma, Brussels, Belgium), 10 mg/kg BW, PO, q8h, or diazepam (Diazemuls; Actavis, Barnstaple, UK), 0.5 mg/kg BW, IV, in case of seizure.

Follow-up and outcome

Re-examination was scheduled in all cats 2 to 3 wk after discharge. Clinical and neurological findings and any changes in ongoing medication protocols were recorded. Subsequent follow-up was at the clinicians’ and owners’ discretion.

Results

The clinical, neurological, and diagnostic findings are summarized in Table 1.

Table 1.

Signalment, clinical, and diagnostic findings, treatment and outcome of 16 cases of feline meningoencephalomyelitis

Case	Age	Gender	Breed	Neurological signs	Onset	Localization	CSF TNCC (cells/μL)	CSF protein (g/L)	CSF cytology	Treatment	Outcome
1	9 y	MN	DSH	Ataxia, decreased postural reactions, horizontal nystagmus, TL hyperesthesia.	Acute	Multifocal CNS	6	NP	Mononuclear	Dexamethasone then prednisolone + cytarabin.	Survived to discharge, LTF.
2	8 y	MN	DLH	Lethargy, generalized ataxia, cervical hyperesthesia.	Unknown	Spinal (C6-T2)	6	0.37	Eosinophilic	Dexamethasone.	Resolved at 2 mo. FFU 5 mo.
3	10 mo	F	Siamese	Generalized tonic-clonic cluster seizures, decreased bilateral menace, bilateral horizontal nystagmus.	Acute	Multifocal brain (forebrain and central vestibular system)	10	0.21	Mixed	Dexamethasone then prednisolone, levetiracetam.	Survived to discharge, LTF.
4	6 y	F	Persian	Partial tonic seizures, ataxia, depressed mentation, decreased bilateral menace.	Acute	Forebrain	14	2.76	Mononuclear	Dexamethasone, phenobarbital.	Resolved at 2 mo. FFU 10 mo. Relapsed after vaccination.
5	11 y	FN	DLH	Obtunded mental status, non ambulatory right-sided hemiparesis, cervical hyperesthesia.	Acute	Multifocal CNS	15	0.52	Mixed mainly neutrophilic	Dexamethasone + lomustine.	Mild ataxia at 3 wk.
6	11 mo	F	DSH	Partial complex seizures, TL hyperesthesia.	Chronic	Multifocal CNS	18	0.33	Lymphocytic	Dexamethasone. Phenobarbitone added at recheck.	Mild partial seizures at 6 wk. Hyperesthesia at FFU 12 mo.
7	6 y	MN	DSH	Pelvic limb ataxia, decreased pelvic limbs postural reactions.	Acute	Spinal (T3-L3)	24	0.48	Mixed	Dexamethasone.	Survived to discharge, LTF.
8	4 y	MN	DSH	Lethargy, pyrexia, intentional tremors, ambulatory tetraparesis, vestibular ataxia, decreased bilateral menace.	Acute	Multifocal brain	53	0.33	Lymphocytic	Prednisolone + cytarabine, phenobarbitone.	Resolved at 2 wk. FFU 12 mo.
9	9 y	MN	DSH	Ambulatory paraparesis, decreased pelvic limbs postural reactions.	Acute	Spinal (T3-L3)	63	0.60	Mixed	Dexamethasone.	Mild pelvic limb ataxia. FFU 15 mo.
10	3 y	FN	DSH	Pyrexia, pelvic limbs ataxia and ambulatory paraparesis, TL hyperesthesia.	Acute	Spinal (L4-S3)	68	0.51	Lymphocytic	Dexamethasone.	Mild pelvic limb ataxia at 3 wk.
11	8 m	FN	DSH	Lethargy, pyrexia, obtunded mentation, circling to right, decreased left-sided postural reactions, decreased menace and PLR in left eye.	Acute	Forebrain	109	0.42	Neutrophilic	Prednisolone.	Resolved at 3 wk. FFU 9 mo.
12	4 y	MN	DSH	Obtunded mentation, ataxia, wide-based forelimbs stance, decreased postural reactions.	Unknown	Multifocal brain	110	1.67	Mixed	Dexamethasone then prednisolone.	Survived to discharge, LTF.
13	2 y	FN	DSH	Generalized tonic-clonic cluster seizures, ambulatory tetraparesis, absent bilateral PLR, horizontal nystagmus.	Unknown	Multifocal brain	130		Neutrophilic	Dexamethasone then prednisolone, diazepam, levetiracetam.	Mild ataxia and tetraparesis. FFU 9 m.
14	3 y	MN	DSH	Lethargy, pelvic limb ataxia, and ambulatory paraparesis.	Acute	Spinal (T3-L3)	344	22.25	Lymphocytic	Prednisolone + lomustine.	Resolved at 3 wk. FFU 2 wk.
15	7 y	MN	DSH	Status epilepticus, generalized tonic-clonic seizures, decreased menace on right eye.	Acute	Forebrain	6	0.52	Mononuclear	Dexamethasone + cytarabine. Phenobarbital, diazepam.	Good control of seizure and unremarkable neurologic examination 1 mo from discharge. FFU 4 mo.
16	4 y	MN	DSH	Generalized clonic cluster seizures, cervical hyperesthesia.	Acute	Forebrain	19	0.49	Lymphocytic	Dexamethasone. Phenobarbital, diazepam.	No seizures. FFU 5 wk.

M — male (intact); MN — male (neutered); F — female (intact); FN — female (neutered); TL — thoracolumbar; PLR — pupillary light reflex; LTF — lost to follow-up; FFU — final follow-up; TNCC — total nucleated cell count; DSH — domestic shorthair; CNS — central nervous system.

Signalment

Sixteen cats met the inclusion criteria. Median age was 9.4 y (range: 10 mo to 12 y). The cats were domestic shorthair or longhair (14/16), Persian (n = 1), and Siamese (n = 1).

Clinical signs

Onset of clinical signs was not specified by the owners in 3 cases. Clinical signs were reported as acute in the majority of cats (14/16). Systemic signs included obtunded mental status (5/16), anorexia (4/16), and pyrexia (3/16). The most common abnormalities on neurological examination were ataxia (8/16), paresis with decreased postural reactions (8/16), and spinal hyperesthesia (6/16) (see Table 1 for further details). Generalized (4/16) or partial (2/16) seizures were the presenting sign in 6 cats. Nystagmus was observed in 3 cats and decreased or absent pupillary light reflexes (PLR) and/or menace responses were noted in 3 cats. Localization based on neurological examination was multifocal brain in 4 cases, forebrain in 4 cases, multifocal CNS in 3 cases, and multifocal spinal cord in 5 cases.

Cerebrospinal fluid analysis

Cerebrospinal fluid nucleated cell count was mildly increased in 12/16 and moderately increased in 4/16 cats (median: 62.2 cells μL, range: 6 to 344) (Table 1). Pleocytosis was mixed in 5/16, lymphocytic in 5/16, mononuclear in 3/16, neutrophilic in 2/16, and eosinophilic in 1/16 cases. Cerebrospinal fluid protein was increased in 15/16 as the sample was of insufficient quantity for measurement in 1 case. Protein concentration was mildly increased in 12/16 (median: 0.50 g/L, range: 0.21 to 22.25 g/L). The PCR results in CSF for T. gondii and feline coronavirus were available in 3/16 cases and were all negative.

Magnetic resonance imaging findings

The MRI findings (Table 2) correlated with neuroanatomical localization in 14/16 cases. In 10/16 cats, MRI detected multifocal CNS lesions. Single lesions were observed in 4 cats and meningeal contrast enhancement only in the remaining 2 cases. Lesions tended to be T1W isointense or hypointense, T2W hyperintense, and were variable in their contrast uptake. Lesions were ill-defined in 9 cats and infiltrative in 3 cats. Mild subtentorial or cerebellar herniation was seen in 2/16 cases. Hydrocephalus was observed in 1 cat only.

Table 2.

Magnetic resonance imaging findings in 16 cases of feline meningoencephalomyelitis

Case	Region scanned	Lesion(s)	Margins	Hydrocephalus	Cerebellar herniation	T1WI	T2WI	Gad uptake
1	Brain	Diffuse intraparenchymal cerebral and cerebellar lesions.	Ill-defined	No	No	Iso	Iso	Yes
	T3-L3 spine	Unremarkable.
2	Cervical	Intramedullary lesion C6-T2.	Ill-defined	No	No	Hypo	Hyper	Yes
3	Brain	Multiple brainstem lesions.	Infiltrative	No	No	Iso	Hyper	No
4	Brain	Swollen cerebrum, decreased gray/white differentiation.	Ill-defined	Yes	Mild	Hypo	Hyper	Meninges
5	Brain, cervical	Intraparenchymal multifocal cerebral lesions and intramedullary lesion at C2 and C3 level.	Ill-defined	No	No	Iso	Hyper	Yes (cerebral) No (cervical)
6	Brain and T3-S3 spine	Meningeal contrast enhancement.	N/A	No	No	N/A	N/A	Meninges
7	T3-S3 spine	Intraparenchymal lesion at L4-L5.	Ill-defined	No	No	Iso	Hyper	Meninges
8	Brain	Intraparenchymal lesions in cerebrum and cerebellum (gray and white matter).	Infiltrative	No	Mild	Iso	Hyper	No
9	T3-S3	Intramedullary lesion extending at T13-L2 level.	Ill-defined	No	No	Iso	Hyper	Yes
10	T3-S3	Ventral meningeal contrast enhancement extending at L4-L6 level.	N/A	No	No	N/A	N/A	Meninges
11	Brain, cervical	Multiple lesions including right temporal lobe, left thalamus, and intramedullary lesion extending at C2-C3 level.	Ill-defined	No	No	Iso	Hyper	No
12	Brain	Multiple intraparenchymal lesions in cerebrum and patchy meningeal contrast enhancement.	Infiltrative	No	No	Iso	Hyper	Yes
13	Brain	Intraparenchymal lesion in cerebrum, brainstem, and cerebellum.	Ill-defined	No	No	Hypo	Iso	Yes
14	T3-L3	Intraparenchymal lesion extending at T13-L2 level.	Ill-defined	No	No	Iso	Hyper	No
15	Brain	Multifocal intraparenchymal cerebral lesions.	Ill-defined	No	No	Iso	Hyper	No
16	Brain	Multifocal meningeal contrast enhancement.	N/A	No	No	N/A	N/A	Meninges

Gad — gadoteric acid; N/A — not available.

Where the spinal cord was imaged, lesions were intraparenchymal in 3/5 and extraparenchymal in 2/5 cases. Meningeal contrast enhancement was seen in 2/5 cases.

Ancillary diagnostic investigations

Complete blood (cell) count (CBC) and serum biochemistry were performed in all cases; leucocytosis was documented in 5 cats. Toxoplasma gondii serology was performed in 9/16 cats; results were not supportive of recent or active infection (high IgM titer) in any case, but in 3 cats (cats 4, 9, 11) results indicated previous exposure (high IgG titer but IgM titer within normal limits). Other diagnostic investigations, including abdominal ultrasound (5/16 cats), inflated chest radiographs (2/16 cats), did not reveal significant concurrent disease in any cat.

Treatment

Dexamethasone alone was given to 9/16 cats, dexamethasone followed by prednisolone was administered to 4/16 cats, and prednisolone alone to 3/16 cats. Five cats received an additional immunomodulatory therapy, including cytarabine in 3 cats, and lomustine in 2 cats. Anticonvulsant medication was used in all cats presented with seizures, including phenobarbitone (4/6) and levetiracetam (2/6). Two cats, admitted for cluster seizure/status epilepticus, on admission received diazepam in combination with phenobarbitone (1 case) or levetiracetam (1 cat) to control seizures.

Outcome

Median duration of hospitalization was 6.8 d (range: 3 to 12 d). All cats survived to discharge (Table 1). Follow-up was unavailable in 4 cases. Median final follow-up was 5.2 mo (range: 0 to 16 mo) after presentation. At first re-examination, 7/12 cats were normal on neurological examination and remained asymptomatic at all subsequent re-examinations (2, 5, 9, 10, and 12 mo) (Table 1). The remaining cats that were presented for re-examination showed persistent mild neurological signs at first re-examination, including mild pelvic limb ataxia, mild tetraparesis, intermittent spinal hyperesthesia, and reduced frequency (> 50% reduction) of partial seizures. No further follow-up was available for 2 of these cats, while in the remaining 3 cats stable or improved clinical signs were recorded in subsequent re-examinations. One cat received an increased dose of dexamethasone, which lead to complete and long-term resolution of clinical signs (last follow-up 10 mo). This cat relapsed 1 y subsequent to remission, a few days after routine vaccination. Treatment was successfully stopped in the remaining 2 cats without deterioration (latest follow-up at 9 mo and 15 mo).

Discussion

This study represents the largest description of suspected FMUO. Findings indicate many similarities with canine MUO but also some important differences, particularly in apparent prognosis. Middle-aged and older cats appeared to be at highest risk, as the median age in this study was 9.4 y, although disease was seen in much older and younger cats. Age of dogs affected by MUO can also be highly variable (6 mo to 12 y) (7,12,13); however, young or middle-aged dogs have been more frequently reported (2,7,13). No breed or gender predisposition was observed, in contrast to canine studies describing a female predilection (2,7); however, the sample size of this study is unlikely to be sufficient to identify gender predispositions for feline MUO.

Clinical signs of FMUO typically corresponded to the distribution of CNS lesions. Onset of neurological signs was acute in most cases, similar to canine MUO (12,14), although the duration of signs may be underestimated in cats due to their temperament and often less time spent in close proximity to owners. Seizure activity, either partial or generalized, was the major reason for presentation in this study similar to data reported in multifocal granulomatous encephalitis (GME) in dogs (7), and NE in pugs (13) and Yorkshire terriers (2). Our findings highlight feline MUO as an important differential diagnosis for seizures, especially in young to middle-aged cats with no previous history of seizures.

Intracranial multifocal disease was the most common neuroanatomical localization, as reported in dogs (2) and previously described cases of presumed FMUO (8). Intracranial was the most common localization (50%), with equal distribution between multifocal and localized to the forebrain, while 31% of cats had a suspected lesion to the spine only, a similar percentage to the focal presentation of canine MUO (12).

Mild CSF pleocytosis was recorded in most cases in this study; previous reports of suspected FMUO show high cell count in suppurative disease but lower cell count in non-suppurative disease (8). Cerebrospinal fluid pleocytosis is highly variable in canine MUO and may even be normal (2). Severity of CSF pleocytosis did not appear to correlate with disease severity or prognosis in this study, as with dogs affected by MUO (9). Moreover, type of pleocytosis did not correlate with duration of clinical signs, being very variable leucocyte prevalence in the CSF with acute onset in most of the cases. As CSF pleocytosis is not specific for MUO, it is important to correlate it with characteristic magnetic resonance imaging (MRI) findings and negative infectious disease results, as were the inclusion criteria in this study.

Magnetic resonance imaging characteristics observed in most of the cases in this study are similar to those in canine MUO, in which the MRI findings appear to be variable, with prevalence of focal or multifocal, ill-defined hyperintense lesions on T2WI (4,6,15). T2WI hyperintensity has moderate sensitivity (68%) and a high predictive value (100%) for inflammatory CSF in dogs (4); however, gadolinium contrast has been reported to increase the sensitivity of MR for detecting inflammatory lesions (15,16). In this study, 2 cats (cases 1 and 13) with indistinct lesions on T2WI as isointense, showed marked contrast enhancement, which helped to detect and define the nature of brain pathology. Interestingly, most intracranial lesions affected white matter only, in keeping with previous data in dogs with GME (6) and necrotizing leukoencephalitis (NLE) (2). Unfortunately, FLAIR sequence, which has been previously reported to enhance sensitivity of MRI (15) was not available in all cases. In these 2 cats, however, contrast-enhancing lesions were also heterogeneously hyperintense in FLAIR. Further studies would be needed to better elucidate the MRI characteristics of presumed MUO in cats and the possible MRI differences with infectious encephalitis.

Feline limbic encephalitis (FLE) has been recently suspected to have a primary immune-mediated etiology, as 36% of cats presenting partial seizure with orofacial involvement showed increased concentrations of antibodies against voltage-gated potassium channel complexes (VGKC-complexes) (17). Diagnosis of suspected FLE is based on clinical signs, MRI findings, characterized by the typical bilateral hippocampal T1 hypo- and iso-intensity and T2 hyperintensity, and pleocytic CSF (17). None of the cats herein showed similar clinical and MR findings; however, an underlying immune-mediated inflammatory etiology may result in shared biochemical features with FMUO. Therefore, further studies are required to elucidate possible common underlying etiology and to test sera for VGKC-complexes in FMUO patients.

All cats in this study received dexamethasone only, dexamethasone followed by prednisolone, or prednisolone only; however, the sample size was too small for statistical comparisons. Corticosteroids were tapered over a range of weeks to months, depending on response to treatment and in 2 cats long-term (> 10 mo) therapy was required. Additional immunomodulatory drugs, including cytarabine and lomustine, were used in 5 cats alongside corticosteroids without adverse effects and complete remission of signs was recorded at the time of latest follow-up in all cases. Use of cytarabine and lomustine reflected clinician’s preference, mainly based on lack of response to corticosteroids while still hospitalized; however, complete information regarding this clinical decision is lacking in the records. Cytarabine and lomustine have been used extensively for chemotherapy against several cancers including lymphoma (14,18–20). Moreover, cytarabine (21,22) and lomustine (23) have also been used as immunomodulatory therapy in addition to steroids in canine MUO and were therefore used for their immunomodulatory effect in feline patients with suspected MUO in the present study. The dosages of these 2 drugs were based on previous descriptions of their use in feline patients (14,18–20). Due to the lack of definitive histopathological diagnosis, it may be reasonable to consider infiltrative neoplasia, including lymphoma, as the main differential diagnosis for these patients and the remission of clinical signs as secondary to use of chemotherapy in potentially misdiagnosed patients. The good evolution of the clinical signs in cats reported in the present study opens one of the major key discussions on MUO, both in canine and feline patients, which is the lack of definitive histopathological diagnosis. Brain biopsies do not represent one of the routine tests for diagnosis of MUO in animals, and in feline patients, the small brain volume, may be particularly challenging. Canine MUO is usually considered to be an ante-mortem diagnosis and further studies are required to increase confidence in diagnostic test findings, especially MRI and CSF analysis, in the diagnosis of feline MUO.

Despite the low number of cases with long-term follow-up, for the first time in veterinary literature, this study suggests prognostic data on feline MUO. All cats survived to discharge and more than half of the cats presented for re-examination were in clinical remission by 2 to 3 wk after diagnosis and remained so until final follow-up (2 to 12 mo). Median time to final follow-up was 5.2 mo, and it is possible that relapse could have occurred in reported cats at a later date than the final follow-up in this study. Overall prognosis in cats with inflammatory CSF has been reported as poor, with 77% of cats surviving less than 1 y (8,24). However inflammatory CSF has been reported in non-inflammatory conditions, including neoplasia, and in infectious CNS diseases, including feline infectious peritonitis (FIP), which may have an important impact on the poor prognosis for cats with inflammatory lesions in the CNS (4,5,24,25). Good prognosis in feline MUO is highlighted by the present study and inflammatory CSF in cats should not always be associated with a poor outcome. Moreover, in the 12/16 cats in which follow-up was available, feline MUO was associated with better medium-term prognosis than in dogs with MUO. Prognosis in dogs affected by MUO has been assessed in several studies, some of which had number of cases similar to the present study (26) and others with larger numbers (27). However, the prognosis remains guarded to poor (9,12,28) in most affected dogs, for which death was recorded in 50% to 56% (9,28) of the treated dogs.

This study has a number of limitations. A regimented treatment protocol was not used due to the retrospective nature of the study and the need for individual management of cases of MUO. The unpredictability of relapse, differences in disease severity, and variability in treatment response mean that prospective studies investigating treatment regimes is not always practical or ethical. The lack of follow-up in 25% of cases was disappointing and it is possible that some of these cats relapsed or died due to feline MUO; future studies with more complete follow-up are required to better understand the prognosis of this disease.

No histopathological confirmation of MUO was performed due to the survival of all cats. It is therefore not possible to define the precise nature of the inflammatory conditions affecting these cats. However canine MUO is usually considered to be an antemortem diagnosis and the aim of this study was to describe the clinical features of cats with this condition rather than determine its origin. Based on histological changes, several authors have suspected non-FIP viral encephalitides as the underlying cause of most cases of feline meningoencephalitis (8,13,14). While it is possible that false negative CSF PCR results occurred or agents not tested for by CSF PCR (including West Nile virus and Aujeszky’s disease virus) were the cause of disease in this study, it is highly unlikely given the improvement/resolution of clinical signs with immunosuppressive therapy. Ideally all cats would have had T. gondii serology performed to further exclude infection.

Further studies of feline MUO are needed to determine prognosis and treatment response, as well as highlight the corresponding histological changes and possible etiologies. Long-term prognosis of feline MUO remains unknown and large-scale, prospective randomized trials are necessary to compare different treatment protocols for feline MUO as described for canine MUO.

In conclusion, MUO is an important differential diagnosis for variable neurological signs in cats. There was no age, gender, or breed predisposition, although young to middle-aged cats were more commonly affected in this study. Disease onset was usually acute and compatible with multifocal intracranial lesions in most cases, although focal lesions of the brain and spinal cord were seen. Cerebrospinal fluid pleocytosis and MRI characteristics tended to be similar to most reports of canine MUO, with mixed or lymphocytic CSF and ill-defined hyperintensities on T2W and FLAIR sequences, which show patchy contrast uptake. Definitive conclusions regarding treatment cannot be made from this study, but immunosuppressive treatment appeared to be successful in achieving rapid complete or partial remission in all cases that were available for follow-up. Prognosis therefore appears to be better than for canine MUO, but further studies of feline MUO are indicated to better understand its origin, treatment and outcome. CVJ