Prognosis in meningoencephalitis of unknown origin in dogs: Risk factors associated with survival, clinical relapse, and long‐term disability

Abstract Background Meningoencephalitis of unknown origin (MUO) comprises a group of noninfectious inflammatory diseases affecting the central nervous system of dogs. Previous studies have reported individual risk factors for survival but prognostication for MUO remains challenging. Objectives Identify clinical prognostic variables in dogs with MUO. Animals A retrospective study of 447 dogs presented to 2 UK referral hospitals and diagnosed with MUO. Methods Medical records of dogs diagnosed with MUO were retrospectively reviewed. Multivariable logistic regression was used for the identification of risk factors for survival and Cox proportional hazards analysis for the identification of risk factors for clinical relapse. Results Eighty‐two percent (366/447) of dogs with presumptive MUO survived to discharge and 63.5% (284/447) were alive at 6 months; 36% of the latter (103/284) had persistent neurological deficits. Breed (pugs; P = .03), epileptic seizures (P < .001), paresis (P < .001), and higher neurodisability scale (NDS) score (P < .001) at presentation were negatively associated with survival to 6 months. Dogs with persistent deficits had higher NDS scores on presentation (P = .001). Median follow‐up time was 11 months (interquartile range [IQR], 1‐24) and 50.6% (160/316) relapsed during treatment (median time to relapse, 7 months; IQR, 2‐15). Incomplete resolution of the clinical signs during the 6 months after diagnosis (P < .001), higher NDS score (P < .001), and longer duration of the clinical signs (P < .001) were associated with relapse. Conclusions and Clinical Importance Knowledge of risk factors associated with survival, incomplete recovery and clinical relapse in MUO can help guide monitoring and treatment and improve owner communications regarding prognosis for this debilitating disease.

treatment and improve owner communications regarding prognosis for this debilitating disease.

K E Y W O R D S
canine, mortality, MUO, outcome, prognosis, relapse

| INTRODUCTION
Meningoencephalitis of unknown origin (MUO) in dogs is a heterogeneous disorder with a broad spectrum of clinical signs and presentations.It comprises a group of idiopathic inflammatory diseases that can only be definitively diagnosed by histopathologic evaluation of the brain, including granulomatous meningoencephalomyelitis, necrotizing meningoencephalomyelitis, and necrotizing leukoencephalitis. 1,2The true incidence of these conditions remains uncertain but, in referral populations, MUO seems to be the most common inflammatory condition affecting the central nervous system in dogs. 3The wide range of clinical signs, imaging findings, and pathologic lesions in affected dogs makes studying MUO in a clinical population challenging, and reported outcomes therefore remain variable.
Despite initiation of appropriate treatment, 25% to 33% of dogs with MUO are reported to die within 1 week of diagnosis, 4,5 and 31% within 100 days of diagnosis. 6Previous studies have identified that younger age at diagnosis 7 and early diagnosis (within 7 days of development of clinical signs) 8 are associated with longer survival times, whereas seizures 5,9,10 and obtundation 5,6,10 are associated with shorter survival times.Recently, a neurodisability scale (NDS) has been shown to be a reliable clinical assessment tool for dogs with MUO, but initial data from a small population of dogs did not show an association between their scores and outcome. 11though some information regarding the prognosis for survival in MUO is available, most studies include relatively small numbers of dogs and findings have not been consistent across studies.Treatment efficacy may be improved by identifying specific MUO subpopulations at higher risk of treatment failure or relapse despite having received disease-modifying treatments.We, therefore, aimed to evaluate a large population of dogs with MUO to identify clinical variables associated with prognosis: survival at 6 months after diagnosis, persistent neurodisability, and clinical relapse.The following data were extracted from the medical records: age, sex, neuter status, breed, year of presentation, season of the year, onset, duration and progression of the clinical signs before diagnosis, neurological deficits identified on presentation, neurolocalization (categorized as forebrain, brainstem, cerebellum, central vestibular [used when clinical signs were insufficient to localize as brainstem or cerebellum], optic nerves/chiasm, multifocal, or no neurological deficits when only pain was identified), and survival to discharge.The findings of the neurological examination recorded in the patient files were used to retrospectively assign the NDS 11 for each dog (calculated by the same assessor in all dogs).Onset of clinical signs was categorized as acute (<24 hours), subacute (between 24 and 72 hours), and chronic (>72 hours) and duration of the clinical signs was ≤7 days or >7 days before presentation.The results of the CSF analysis (TNCC and protein concentration) also were recorded, as well as the treatment received.The initial treatment received also was recorded and further categorized as corticosteroids alone, corticosteroids and IV cytosine arabinoside (200 mg/m 2 ) or corticosteroids and SC cytosine arabinoside (200 mg/m 2 ).Long-term treatment was categorized as corticosteroids alone, corticosteroids and repeated administrations of SC cytosine arabinoside for a minimum of 4 treatment cycles, and corticosteroids and another PO immunomodulatory drug.For statistical analysis, breeds with >15 individuals represented were compared with the remainder of the study population.

| MATERIALS AND METHODS
The outcome was assessed at the last follow-up appointment or telephone conversation with the owners or referring veterinary surgeons.Episodes of suspected relapse (recurrence of similar or new neurological deficits and, when possible, repeat MRI of the brain and CSF analysis) were recorded as well as response to rescue treatment when available.
Statistical analysis was performed using the software SPSS 27.0 (SPSS Inc., Chicago, Illinois).Continuous data were assessed for normality using the Shapiro-Wilk test.Descriptive statistics are reported for continuous variables using mean (SD) for approximately normally distributed variables and median (interquartile range [IQR]) for variables with skewed distributions, and frequencies (with 95% confidence intervals [CI] where appropriate) are reported for categorical variables.The Mann-Whitney test was used to assess the association between the NDS score and resolution of clinical signs.
Univariable binary logistic regression was performed to identify clinical variables associated with survival at 6 months after diagnosis (using the information from last examination or telephone conversation).Before multivariable analysis, all variables were assessed for correlation using Spearman's rank correlation coefficients.If Spearman's rank correlation coefficient was >.8, the most statistically significant or biologically plausible variable was selected.Any independent variable demonstrating some association on preliminary univariable analysis (P < .2) was considered for inclusion in a multivariable model.Multivariable models were constructed using a manual backwards stepwise removal approach; variables with P < .05were retained, with model fit assessed using the Hosmer-Lemeshow statistic.Cox proportional hazards analysis was used to identify clinical variables associated with long-term relapse using the same approach, with the assumption of proportional hazards assessed by evaluating the significance of incorporating time-dependent covariate interaction terms and evidence for improved model fit, and by inspection of the Schoenfeld residuals of the final model.The receiver operating characteristic (ROC) curve was used to determine the usefulness of the NDS score for predicting survival at 6 months after diagnosis of MUO; the corresponding area under the curve (AUC) was calculated to evaluate the predictive capability and the Youden index was used to calculate the optimal cut-off value.Survival times were assessed using Kaplan-Meier plots and were compared between groups by log-rank analysis.

| RESULTS
A total of 633 client-owned dogs diagnosed with suspected MUO were initially identified from the hospital records of the 2 contributing institutions over a 12-year period (2010-2022).After a review of the patient files, 447 dogs met the inclusion criteria (Figure 1).

The median age at presentation was 48 months (IQR, 26-73).
There were 246 (55%) females (145 neutered) and 201 (45%) males   4).Incorporated time-dependent interaction terms for the final variables were not significant and the Schoenfeld residual plots indicated that the proportional hazards assumption was not violated.

| DISCUSSION
We identified clinically relevant prognostic variables associated with survival at 6 months after diagnosis, clinical relapse, and long-term neurological impairment in these dogs with MUO.A total of 63.5% of dogs were alive at 6 months and 50.6% of dogs that survived to discharge subsequently relapsed.The presence of epileptic seizures and paresis at the time of presentation and a higher NDS score were associated with increased risk of death by 6 months after diagnosis.Pugs were also less likely to survive.Incomplete resolution of the clinical signs, higher NDS score, and longer duration of clinical signs were associated with higher risk of relapse.Furthermore, our recently validated NDS was shown to be an important clinical assessment tool because the score was associated with all 3 outcome measures.
The 6-month survival rate identified in our study is similar to the 69% previously described at 100 days after diagnosis, 6 and only slightly lower than the 67% to 74% reported 1 week after diagnosis. 4,5is finding suggests that patients that survive the initial stage of the disease tend to have a more stable course and respond at least partially to treatment.[13] Relapses occurred at many different time points and, in some cases, over 12 months after diagnosis.It is therefore likely that many studies with shorter follow-up times, including ours, fail to identify these patients and that the relapse rate in MUO is in fact higher than has been reported.
The occurrence of epileptic seizures as 1 of the initial clinical signs was associated with decreased survival at 6 months, similar to previous reports. 5,9,10Seizures might be associated with a more severe clinical phenotype, but owners also could be more likely to euthanize these patients because of limited tolerance for this clinical sign or the additional treatment entailed and adverse effects.The majority of dogs presenting with seizures had cluster seizures (50.7%) and a substantial proportion (24.3%) had at least 1 episode of status epilepticus.
An association between the frequency of cluster seizures and euthanasia has been previously identified in patients with idiopathic epilepsy, 14 and epilepsy has been shown to affect the quality of life of the dogs and their owners. 15,16The importance of adequate anticonvulsive treatment has been emphasized in pugs diagnosed with necrotizing meningoencephalomyelitis (NME) because anticonvulsive drugs were the only treatment associated with longer survival in 1 study. 17 view of our results and the previous literature, it is possible that, in MUO patients presenting with seizures, aggressive anticonvulsant treatment to decrease seizure frequency (and therefore improve owner perception of quality of life) ultimately could improve survival.
Obtundation also has been previously reported to be associated with shorter survival times, 5,6,10 but this association was not found in our population and may reflect the subjectivity involved in evaluation of this clinical sign.We nonetheless identified paresis as another clinical sign associated with decreased survival that has not been reported previously.4][25] It has been suggested that motor symptoms reflect either more extensive or more persistent functional effects or lesions. 26Interestingly, no association was found between specific neurolocalizations (such as brainstem) and survival.It is likely that increased risk of death is related to the location of the lesions within the central nervous system, but this association was not identified in our population and may be difficult to establish when the majority of patients present with multifocal signs.
We identified that pug dogs were less likely to survive to 6 months after diagnosis compared with other breeds.This breed is commonly affected by NME, 17,24,[27][28][29] a subtype of MUO that causes extensive necrosis, which could at least partially explain the increase in mortality.
Nonetheless, other breeds affected with necrotizing forms of MUO such as the Chihuahua, Maltese, and Yorkshire terrier also were included in the analysis and did not show an increased risk of mortality.
Previous studies have reported short survival times for pugs diagnosed with NME but many of these studies included only necropsy-confirmed cases, which likely biased the results toward more severely affected patients. 17,28Serum glial fibrillary acidic protein (GFAP) concentrations were found to be significantly higher in pug dogs with NME compared with healthy controls and other breeds affected with NME, suggesting a breed-specific fragility of astrocytes. 30,31In humans with MS, higher serum concentrations of GFAP were associated with disease severity and higher MRI lesion count, 32,33 suggesting an important role of astrogliosis in advanced stages of disease.The NDS has been recently proposed as a reliable clinical scale for initial assessment of patients with MUO. 11Our study found an association between the NDS score and increased risk of death within the 6 months after diagnosis, relapse, and incomplete resolution of clinical signs within that time.5][36] Identification of a higher EDSS score at baseline increases the level of concern for disease progression in patients with MS and more frequent assessments are recommended in such patients. 37 could have more severe or extensive lesions, which would be reflected in the higher NDS scores.21]38 Additionally, dogs that presented within 7 days of the onset of clinical signs were less likely to relapse.A previous study had identified that dogs presented within the initial 7 days of clinical signs had longer survival times, but relapse was not evaluated in that population. 8It is likely that early treatment may influence outcome as in human patients with MS, starting treatment soon after initial symptom onset minimizes long-term disability. 39,40r study had some limitations.As previously discussed, the most important limitation was the lack of standardized treatment, which likely had an effect on individual survival times and likelihood of relapse.The NDS score was calculated based on retrospective data available from the patient records and not collected prospectively. 11e lack of histopathologic examination in most cases makes misdiagnosis a possibility.The diagnosis of relapse was, for the majority of cases, based on the recurrence of clinical signs and response to changes in the treatment protocol, with only a minority of cases having repeat diagnostic investigations.It is therefore possible that in some cases the clinical signs could have been caused by a second condition.Cases were included over several years and treatment and standards of care may have improved over time, thus influencing outcome.Nonetheless, despite the long study period, there is still no clear evidence of superiority of any specific treatment regimen with multicenter randomized controlled trials required to determine the superiority of a specific treatment. 23 showed that pugs, the presence of seizures and paresis at the time of presentation, and a higher NDS score were associated with an increased risk of death at 6 months after diagnosis in dogs with MUO.

CONFLICT OF INTEREST DECLARATION
Authors declare no conflict of interest.

OFF-LABEL ANTIMICROBIAL DECLARATION
Authors declare no off-label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION
Approved by the IACUC of the University of Liverpool (VREC840).
The medical records of dogs diagnosed with MUO at the Small Animal Teaching Hospital (SATH) of the University of Liverpool and the Royal Veterinary College (RVC) were retrospectively reviewed.Ethical approval for use of data was granted by the Ethics Committee of the University of Liverpool (VREC840).Dogs meeting the following criteria were included in the study: (1) clinical examination consistent with focal or multifocal intracranial neuroanatomical localization;(2) >6 months of age; (3a) multiple, single or diffuse intra-axial hyperintensities on T2-weighted magnetic resonance imaging (MRI) and mononuclear pleocytosis (total nucleated cell count [TNCC] >5/μL) on cerebrospinal fluid (CSF) analysis, or (3b) normal brain MRI but mononuclear pleocytosis.In dogs with MRI findings consistent with MUO (3a) but with signs of increased intracranial pressure (ICP) on imaging (namely transtentorial or foramen magnum herniation), CSF collection was considered contraindicated and not performed, but these cases were included.Dogs were excluded if the TNCC was normal on CSF analysis (independent of their MRI findings), if follow-up time was <6 months (except for those dogs dying during this period), or if the owners elected to euthanize at the time of diagnosis without treatment.The MRI examinations were performed using a 1.5 T (Philips Ingenia CX, Philips Healthcare) or a 1 T (Siemens Magnetom, Siemens Healthcare) scanner.

A
higher NDS score, incomplete resolution of the clinical signs, and longer duration of clinical signs before presentation were associated with higher risk of relapse.Awareness of risk factors associated with survival and relapse in MUO can help guide treatment and decisionmaking in the management of this debilitating disease.Dogs with recognized risk factors for early death or relapse may benefit from closer monitoring (more frequent re-examinations by veterinary neurologists) or different treatment protocols.Additional studies are required to determine if this approach would improve the outcome in these high-risk cases.ACKNOWLEDGMENT No funding was received for this study.This work was presented at the 35th ESVN-ECVN Symposium, September 2023, Venice, Italy.

Table 1
and the findings of the neurological examination in Table2.Findings of the neurological examination and presenting neurological signs in 447 dogs diagnosed with MUO.Final multivariable logistic regression model for risk factors associated with survival at 6 months after diagnosis in dogs with MUO.
Necropsy confirmation was available in 24 cases (17 granulomatous meningoencephalitis and 7 necrotizing meningoencephalitis). F I G U R E 1 Study protocol for data analysis.n,number of dogs.T A B L E 1 Descriptive statistics for clinical and demographic information for the 447 dogs diagnosed with MUO.n = 447 Median age (IQR) 48 months (26-73) All dogs received immunosuppressive doses of dexamethasone (0.3-0.5 mg/kg) at the time of diagnosis (30 dogs did not recover from anesthesia and so did not receive further treatment) alongside cytosine arabinoside in 84.4% (352/417) of dogs (SC injection in 38.8% and IV constant rate infusion in 45.6%).In 22/352 dogs, cytarabine administration was delayed by a few days while waiting for infectious titer results, but it was administered within 1 week of diagnosis.died in the 3 weeks after discharge and thus were not included in any of the treatment groups.The treatment groups included 74/352 (21%) dogs treated with corticosteroids alone, 244/352 (69.3%) dogs treated with corticosteroids and repeated SC administrations of cytosine arabinoside, and 34/352 (9.7%) dogs treated with corticosteroids and another PO immunomodulatory drug.Several other immunomodulatory medications were used across the latter 2 groups and included cyclosporine (44), leflunomide (10), azathioprine (7), mycophenolate mofetil (3), procarbazine (2), and lomustine (2).Where this information was clearly identifiable in the clinical records, it was noted that in 36% (103/284) of dogs, despite clinical improvement, resolution of the clinical signs had not occurred by 6 months after diagnosis.A significant difference (P = .001)wasfound between the NDS score on presentation of the dogs that had permanent residual neurological deficits (median, 6; range, 2-12) compared with those that did not (median, 5; range, 1-13).On univariable logistic regression analysis, breed, center, progression of clinical signs, neurolocalization, clinical signs of increased ICP, epileptic seizures, obtundation, paresis, tremors, decreased gag reflex, protocol of cytarabine administration at time of diagnosis, and NDS score showed some evidence of association with survival to 6 months.However, on multivariable analysis, only breed (pugs; P = .03),epilepchargeand had detailed follow-up data (median time to relapse, 7 months; IQR, 2-15).On univariable Cox regression analysis, age, weight, breed, year of diagnosis, onset and progression of clinical signs, neurolocalization, CSF protein concentration, incomplete resolution of clinical signs during the 6 months after diagnosis, long-term treatment and NDS score showed some evidence of association with T A B L E 2 a Including tetraparesis, hemiparesis, and paraparesis.bAny combination of the following clinical signs along with mentation changes (obtundation, stupor, or coma): decreased physiological nystagmus, decerebellate, or decerebrate rigidity, pupil size changes.cDogs presenting with spinal or unlocalized pain and lethargy.T A B L E 3