Case description

A 7-month-old female spayed Siberian cat was presented with a 4-day history of acute progressive dullness, left hemiparesis and ataxia. There was no known history of trauma. The cat was up to date with core vaccinations, had no travel history outside of the UK and was fed a veterinary practice own-brand grain-free chicken-based diet designed for kittens. The cat had intermittent mixed-pattern diarrhoea since ownership. Prior faecal analysis detected both Giardia species and feline coronavirus (FCoV) by PCR (IDEXX Feline Diarrhoea Panel; IDEXX Laboratories). The Giardia species infection had been managed with fenbendazole (50 mg/kg PO q24h) initially and, owing to compliance issues, subsequently switched to metronidazole (10 mg/kg PO q12h for 5 days), which improved faecal consistency. No medications had been administered in the 3 weeks prior to presentation and during hospitalisation other than those associated with performance of diagnostic procedures (eg, anaesthetic agents and contrast agent).

On physical examination the cat was in lean body condition (body condition score 4/9) and weighed 2.3 kg. Ophthalmic examination revealed bilateral iris pigment changes suggestive of uveitis (Figure 1). Neurological examination (Table 1) documented dull mental status and absent proprioception (hopping and paw placement) in the left forelimb and hindlimb, resulting in left-sided hemiparesis. The rest of the examination was unremarkable.

Table 1

Neurological examination findings for the cat at time of admission

Figure 1

Ophthalmic examination revealed bilateral iris pigment changes (arrow) suggestive of uveitis

Haematology (Table 2) revealed a moderate, microcytic, non-regenerative anaemia, lymphopenia and neutrophil toxicity. Blood biochemistry (Table 2) revealed marked hyperproteinaemia secondary to an extreme hyperglobulinaemia, with an albumin:globulin ratio of 0.2; a mild hypoalbuminaemia; mildly increased alanine aminotransferase, creatine kinase and aspartate aminotransferase activities; a low creatinine; and mild hyperglycaemia. Toxoplasma gondii, feline leukaemia virus and feline immunodeficiency virus serology was negative, while feline coronavirus serology was positive with a high titre (Table 2).

Table 2

Haematology, blood biochemistry and infectious disease screening for the cat at time of admission

An MRI study of the brain and cervical spine, including postcontrast images, was performed under general anaesthesia (see document 1 in the  supplementary material) and revealed dorsal and ventral dilation of the subarachnoid space that extended from the obex to the first cervical segments, and showed abrupt narrowing at the level of the C2–C3 intervertebral space (Figure 2a). This resulted in dorsoventral flattening of the cord at this level on transverse images, which was more marked on the right (Figure 2b). Immediately caudally to the end of this lesion and throughout most of the cervical spine, the cord assumed a swollen appearance and showed irregular distributed T2-weighted hyperintensity within the left dorsal compartment at the level of C1, and diffusely surrounding the central canal from C2 to C6 (Figure 2a). There was marked leptomeningeal contrast enhancement surrounding the cranial cervical cord (Figure 2c,d) and along the brainstem, suggestive of an inflammatory process.

Figure 2

(a) Midline, sagittal, T2-weighted (T2W) MRI of the brain and cranial cervical spinal cord and (b) transverse T2W, (c) T1-weighted (T1W) and (d) T1W post-gadolinium MRI of the vertebral column at the level of mid-C1 (indicated by the solid yellow line on image [a]). Marked compression of the caudal medulla oblongata and cranial cervical spinal cord compression by a partially right-lateralised subarachnoid diverticulum is visible, particularly on T2W images (a,b; arrows). Leptomeningeal contrast enhancement is seen surrounding the cranial cervical cord (d; arrows)

Analysis of cerebrospinal fluid (CSF) collected by cisternal puncture documented a marked neutrophilic pleocytosis (Table 3; Figure 3). The presence of FCoV RNA within the CSF at a moderate-to-high level was revealed using reverse transcriptase quantitative PCR (RT-qPCR).³

Table 3

Cerebrospinal fluid analysis results

Figure 3

(a) Cisternal cerebrospinal fluid cytology revealed a predominantly neutrophilic pleocytosis, with (b) non-degenerative neutrophils (arrowheads) and large mononuclear cells (arrows). (c) Less frequently reactive lymphocytes (arrowhead) and plasma cells (arrow) were noted. Modified Wright’s stain (cytospin preparation; (a) × 20) and (b,c) × 50 objective magnification). Courtesy of Elpida Sarvani, Diagnostic Laboratories, Langford Vets

The cat was subsequently euthanased owing to further decline in its condition, in conjunction with the strong suspicion of FIP and poor associated prognosis. The owner consented to a post-mortem examination, which was performed (by ENB) immediately after euthanasia. There was no evidence of effusion within any body cavity. Both kidneys had multiple, pale, mass lesions present within the cortices (Figure 4a). Upon removal of the overlying bone and meninges the spinal cord appeared bruised at the level of the previously noted compression (Figure 4b).

Figure 4

At post mortem examination (a) multiple, pale, mass lesions (arrow; left kidney shown) were visible within the renal cortices. Following removal of the overlying bone and meninges, (b) the dorsal surface of the spinal cord appeared bruised at the level C2 (arrow)

Histopathology of the brain and spinal cord documented marked inflammation expanding the meninges, particularly around the medulla, mid-brain (most marked around the thalamus and hippocampus), ventral forebrain and spinal cord (at all levels examined). Inflammation was mixed, with many plasma cells, histiocytic cells and moderate numbers of neutrophils. There was some free blood and eosinophilic fibrinous fluid within the meninges, and the occasional vessel walls were degenerate or contained a proliferation of fibroblastic cells within their lumen. The choroid plexuses at these levels were markedly inflamed and hyperplastic. There was also mild inflammation within the ventricles. Elsewhere in the body there was mild-to-moderate pyogranulomatous uveitis, extensive pyogranulomatous nephritis and mild hepatic portal inflammation (predominantly plasma cells and histiocytic cells). Within the mesenteric lymph nodes and spleen there was evidence of immune stimulation with active lymphoid follicles, mild increases in plasma cells and histiocytes. The small and large intestine showed evidence of mild inflammation, alongside the presence of protozoa within the small intestine (likely Giardia species based on prior faecal analysis).

Discussion

This is the first report, to our knowledge, to describe a subarachnoid diverticulum associated with FIP.

Spinal arachnoid diverticula (SAD) are characterised by a focal dilation of the meninges permitting the accumulation of CSF, which, in turn, compresses the spinal cord.⁴ SAD are considered rare in cats and often thoracolumbar in localisation, with only one case report of acquired cervical SAD found in the literature.⁵ Intervertebral disc disease, spinal cord trauma and inflammatory spinal cord disease have been suggested as possible causes of acquired SAD in cats, dogs and humans;^4,5 however, as most cases are not imaged prior to the development of SAD, the presence of a congenital lesion cannot be ruled out. Clinical signs attributed to the presence of SAD are consistent with a chronic, progressive myelopathy with neurolocalisation dependent upon the site of compression.⁶ Differentials initially considered for the subarachnoid diverticulum in this case included a primary/congenital structural lesion exacerbated by acquired inflammatory infiltrate or, more likely, an acquired structural lesion secondary to inflammatory infiltrate causing adhesions. Ante-mortem, we were strongly suspicious of an FIP-associated arachnoiditis as the origin of the SAD, supported by the MRI findings and presence of FCoV within the CSF. Ultimately, post-mortem collected samples revealed the presence of granulomatous to pyogranulomatous inflammation affecting the meninges, ventricles, choroid plexus and spinal cord, consistent with a diagnosis of neurological FIP, with evidence of additional non-effusive FIP changes involving the eyes, liver and kidneys.⁷

Treatment options for SAD are limited, with most feline case reports describing surgical decompression.⁵ In a large case series of dogs with SAD, the outcome of conservatively (ie, medically, often with the glucocorticoid prednisolone) managed cases was relatively poor (n = 50; neurological signs were stable in 30% and improved in 30%), when compared with the outcome of surgically managed cases (n = 46; neurological signs were stable in 3% and improved in 82%);⁴ however, this was a retrospective case series and it is possible that more severely affected cases were elected to be managed medically. Recurrence of clinical signs after surgery for SAD has been reported infrequently for both cats and dogs, with the median time from surgery to recurrence of SAD being 20.5 months.⁶

In the absence of the experimental antiviral drug GS-441524, prognosis associated with neurological FIP is generally grave,⁸ although there are occasional reports of cats surviving clinical FIP.⁷ The presence of neurological signs has also been associated with poor response to, or relapse following, initial trial doses of recently described antiviral drugs GC376 and GS-441524, which have been successful in the management of effusive FIP.^9,10 A median survival onset of clinical signs of 14 days (range 2–115 days) in 24 cats with neurological FIP has been reported in one study,¹¹ comparable to that of the effusive form of the disease.¹² Neurological FIP describes a subset of cats that present with predominantly neurological signs.¹³ Around 40% of cats with the non-effusive (ie, ‘dry’) form of FIP have neurological signs, although neurological signs also occur in around 5–10% of cats with effusive (ie, ‘wet’) FIP.^11,13,14 No evidence of effusions were found in this case. Management is aimed at supportive palliative care, with corticosteroids or other immunosuppressive medications administered in an attempt to slow disease progression. In the absence of readily available antivirals shown to be effective in the management of FIP,^9,10 and given the dire prognosis, surgical decompression of the SAD was not considered to be ethical in this case.

The clinicopathological results were also suggestive of FIP. The microcytic, non-regenerative anaemia was considered most consistent with anaemia of chronic inflammatory disease with iron sequestration, with hepatic dysfunction or absolute iron deficiency considered less likely. Microcytosis, with or without anaemia, is frequently reported in FIP.¹⁴ The lymphopenia and neutrophil changes were consistent with a systemic inflammatory process. Although the differential list for hyperglobulinaemia is wide,¹⁵ the degree of hyperglobulinaemia was considered most consistent with a chronic inflammatory or infectious disease such as FIP, or a lymphoproliferative disease (eg, lymphoma or malignant plasma cell disorders). The mild hypoalbuminaemia was considered most likely compensatory to the severe hyperglobulinaemia and resultant elevated blood osmolarity. The elevations in liver and muscle enzyme activities were considered most likely secondary to systemic inflammation or tissue hypoxia resulting from increased blood viscosity. The low creatinine was attributed to reduced muscle mass.

FIP is the most common infectious cause of neurological disease in cats.¹⁶ A marked inflammatory reaction is triggered by release of inflammatory cytokines from FCoV-laden monocytes/macrophages resulting in multisystemic pyogranulomatous vasculitis.⁷ MRI abnormalities detected in FIP include meningeal contrast enhancement, ependymal contrast enhancement, ventriculomegaly (inflammation may cause obstruction of the ventricular system, which may cause a secondary mass effect and compromise surrounding parenchyma), syrinx formation and foramen magnum herniation.¹¹ Commonly recognised spinal signs include hindlimb or generalised ataxia and paraspinal hyperaesthesia, with associated spinal pathology, which may contain a syrinx or myelitis.

Ideally, prompt diagnosis of FIP is required to minimise further distress to patient and client through additional diagnostic procedures; however, ante-mortem diagnosis of FIP can be challenging, particularly for neurological FIP where minimally invasive sampling of the affected tissue is limited by the skull and vertebral column, and advanced imaging studies are not 100% sensitive. In a study that described the MRI features of feline meningoencephalitis, 2/8 cats with neurological FIP had unremarkable MRI studies.¹⁷ Further, cats with non-effusive FIP are less likely to show fever or classical haematological and serum biochemical changes (eg, lymphopenia, hyperbilirubinaemia, hyperproteinaemia and hypoalbuminaemia).¹⁴ Confirmation of FIP using immunocytochemistry to co-localise the presence of FCoV within macrophages or by the demonstration of viral RNA within CSF are limited by sensitivity and specificity. One study that used immunocytochemical staining of CSF preparations for FCoV antigen to diagnose FIP had a specificity of 83% and a sensitivity of 85%.¹⁸ Interestingly, more of the cats with FIP but without neurological deficits were positive using this assay than those with neurological deficits. One study that examined the sensitivity of RT-qPCR of CSF in the diagnosis of FIP to be 42.1% overall, rising to 85.7% for seven cats with neurological or ocular involvement.¹⁹ Another study looking at the use of RT-qPCR found viral RNA in the CSF of 50% of 10 cats with FIP that had samples collected post mortem, although whether neurological signs were present in these cats was not reported.³ Specificity in both RT-qPCR studies was 100% with a total of 34 samples from cats without FIP.³ Unfortunately, owing to limitations of RT-qPCR and immunocytochemistry, many consider that definitive diagnosis of FIP requires immunohistology to co-localise pyogranulomatous vasculitis with presence of FCoV in macrophages.⁷

Conclusions

To our knowledge, this unusual presentation is the first reported case of a subarachnoid diverticulum associated with FIP. Although not definitive, MRI can be a sensitive method for detecting evidence of meningoencephalitis, further supporting a diagnosis of neurological FIP alongside compatible signalment, clinical history, physical examination findings, haematology, serum biochemistry and CSF analysis.