A 10-month-old male castrated domestic shorthair cat presented with a four-week history of progressive gait incoordination and weakness. The physical examination upon arrival was normal. Neurological examination revealed ambulatory tetraparesis associated with proprioceptive ataxia in all four limbs. The mental state, menace response, and cranial and spinal nerve reflexes were normal. The neck palpation revealed no pain or muscle atrophy. Based on the neurological examination, the lesion was localized to the C1-C5 cervical spinal cord segments.
The complete blood count and serum chemistry tests were normal. Serological test results for feline leukemia virus, feline immunodeficiency virus, and Toxoplasma gondii were negative. Furthermore, no abnormal findings were detected on the chest radiograph and abdominal ultrasound.
Magnetic resonance imaging (MRI) of the cervical spine was performed under general anesthesia using a 0.35-Tesla magnet (Magnetom C, Siemens Healthineers, Berlin, Germany). The imaging protocol included T2-weighted imaging (T2WI) turbo spin echo (TSE) sequences in sagittal and transverse planes, a fat suppression sequence conducted as T1-weighted imaging TSE Dixon in the transverse plane, and a T1-weighted imaging (T1WI) TSE sequence in the transverse plane. Additionally, a T1WI sequence was acquired in both the transverse and dorsal planes following the IV administration of a gadolinium-based contrast agent (0.1 mmol/kg; Cyclolux Sanochemia Pharmazeutika GmbH, Neufeld an der Leitha, Austria).
The MRI revealed a focal dorsolateral spinal cord compression just caudally from the foramen magnum and extending to the mid-portion of the C2 vertebral level (Fig. 1). The lesion was hyperintense on T2WI and T1WI, hypointense on T1-Dixon, and showed no contrast uptake on T1WI after gadolinium administration. Based on these MRI characteristics, an extradural or intradural fat-containing lesion was initially suspected. Additionally, the left rectus capitis ventralis muscle (RCVm) exhibited an abnormal MRI signal. It demonstrated T2WI and T1WI Dixon hyperintensity, T1WI isointensity (relative to the epaxial muscles) and strong homogenous contrast enhancement on the T1WI post-contrast sequence.
Following the imaging investigations, cerebrospinal fluid (CSF) analysis was obtained from the lumbar cistern, revealing a normal nucleated cell count (3 cells/µL - reference range ≤ 8 cells/µL) and a normal protein concentration (0 g/L - reference range 0-0.3 g/L).
On the following day, surgery was performed. A dorsolateral hemilaminectomy was performed on the right side of C1 and the cranial aspect of C2, followed by a durectomy as no compressive material was found extradurally. The initial incision in the dura mater, executed with a #11 scalpel blade, was extended using fine scissors. Beneath and loosely attached to the dura, a pale-yellow, soft, and friable mass adhered to the pia mater became visible. En bloc - removal from the pial membrane achieved adequate spinal cord decompression. An absorbable gelatin sponge (Gelfoam; UpJohn, Kalamazoo, MI) was placed over the laminectomy site. No CSF leakage or hemorrhage was observed at the end of the procedure, and the site was closed routinely.
The mass was immediately immersed in 10% formalin and submitted for histopathological examination. In the laboratory, the mass was trimmed, embedded in paraffin, sectioned and routinely stained with hematoxylin-eosin (HE). Histological examination revealed the tissue to be homogenously composed of well-differentiated adipocytes with signet ring appearance supported by sparse capillaries and separated by delicate strands of fibrous tissue, hence indistinguishable from normal white fat tissue at untypical localisation. This led to the histological diagnosis of a benign intradural lipoma.
One day following surgery, the cat experienced a notable decline in its neurological function, resulting in tetraplegia. This rapid deterioration was believed to be associated with either iatrogenic damage resulting from spinal cord manipulation or linked to a reperfusion injury following decompression. While there was slight improvement in the next 3 days, marked by spontaneous voluntary movements in the left forelimb and hindlimb, a follow-up 15 days later revealed no further progress. Facing the possibility of a persistent debilitating condition, the cat was euthanized.
The necropsy was performed about 10 hours after euthanasia. There were no abnormalities seen outside the cervical spine. At the previous surgery site, a soft grey tissue proliferation was observed on the right dorso-lateral side of the upper cervical spinal cord (Fig. 2). This tissue exhibited a soft, jellyish consistency and was found to be strongly adherent to the dura and spinal cord. Moreover, upon its ventral exposure, the left RCVm was abnormally pale and, compared to the contralateral side, reduced in size.
Samples of both conspicuous tissues, contralateral RCVm and the spinal cord were taken, transferred into 10% neutral buffered formalin and routinely processed to HE stained sections. Microscopic evaluation of the soft tissue mass showed a severely cellular and highly vascularized proliferation of myofibroblasts arranged in disorganized sheets and nodules with a variable background of collagen fibers and/or myxoid stroma. The proliferation mingles with multiple small, invaded islands of adipose tissue and, over a large area, it infiltrated transpially and along blood vessels into the white matter of the associated spinal cord segments (Fig. 3). Moreover, the mass was multifocally invaded by large mononuclear infiltrates, mainly composed of plasma cells, histiocytes and macrophages. The affected spinal cord, on the other hand, showed focal fiber losses, eosinophilic spheroids and marked astrogliosis peppered with some microglial proliferates. Cytomorphology and growth pattern of the mass were consistent with a desmoid tumor.
On immunohistochemistry (IHC), the invading spindle cells were positive for vimentin and, in the peripheral aspect of the mass, also for smooth muscle actin, but were negative for glial fibrillary acidic protein (GFAP), and S100. Candidate markers β-catenin and calretinin yielded no immunopositivity in the invading cells as compared to the positive controls (Fig. 4).
RCVm indeed showed replacement of muscle fibers by a chronic severe coalescing interstitial fibrosis. No invasion of the neighboring tissue was seen. Similar, though less severe, the right RCVm also exhibited a focal proliferation of fibrocollagenous tissue within the endomysium and perimysium.