Multimodal Imaging Characteristics of a Migrating Oropharyngeal-Spinal Foreign Body in a Cat

Introduction

Although oropharyngeal foreign bodies are well documented in domestic animals, there are few reports of migration to the central nervous system (CNS).¹ Imaging modalities described include radiographs, computed tomography (CT), and MRI, with most diagnoses confirmed at postmortem examination.¹ Most cases describe hard objects, for example, a stick or sewing needle.^2,3 In one CT study evaluating nongastrointestinal foreign bodies, 10/13 foreign bodies were of plant origin.⁴ Few intracranial lesions attributable to migrating plant material are described in small animals, and the migration route is often not elucidated.⁵ There are cases of grass awns entering the spinal cord percutaneously or by unknown routes.^4,7 The prognosis for CNS foreign bodies is guarded depending on the extent of injury present and the timing of intervention, because there are a number of previously documented cases that responded favorably to surgical removal and/or appropriate medical management.^1,2,5–8 Here, we describe the MRI and CT characteristics of a migrating oropharyngeal-spinal foreign body in a 2.5 yr old female spayed domestic shorthair.

Case Report

A 2.5 yr old female spayed domestic shorthair presented for neurological evaluation of acute nonambulatory tetraparesis, dull mentation, and fever. The cat was initially presented to an emergency hospital for lethargy, trembling, and sneezing, 2 days after she was seen chewing on a dried Gypsophila spp. (“baby’s breath”). The owner recalled the cat retching and sneezing after this activity. The cat was indoor-only, and there were no external signs of trauma. At the emergency hospital, the rectal temperature was 40.6°C, but no neurologic signs were noted at that time. Complete blood count and chemistry were unremarkable. Systolic blood pressure was 122 mm Hg.

After 3 days of hospitalization at the emergency hospital on IV fluids, Cerenia^a 1 mg/kg q 24 hr, and Unasyn^b 24 mg/kg q 8 hr, the cat was discharged with a temperature of 39.7°C. The cat was found laterally recumbent the following day and referred for neurological evaluation.

Neurologic examination revealed plegia of the thoracic limbs, paresis of the pelvic limbs, and intermittent decerebellate posturing. The paw replacement test was absent in both thoracic limbs and the right pelvic limb but normal in the left pelvic limb. Mentally, the cat was obtunded but had normal cranial nerves and spinal reflexes. Neurolocalization was referable to the caudal brainstem.

MRI^c of the brain and cervical spine was performed under general anesthesia before and after IV contrast administration Gadoteridol^d 173 mg using a circular surface coil. Scan sequences and parameters included sagittal and transverse T1-weighted (T1-w) pre- and postcontrast (time of repetition [TR] 497 ms, time of echo [TE] 13 ms), sagittal and transverse T2- weighted (T2-w) (TR 2800 ms, TE 110 ms), sagittal short tau inversion recovery (TR 3000 ms, TE 25 ms, time of inversion [TI] 160) and transverse T2 fluid-attenuated inversion recovery (FLAIR; TR 6000 ms, TE 160 ms, TI 2000). Slice thickness was 3 mm with a 0.3 mm slice gap. Field of view was 100 × 100 mm.

On sagittal plane images (Figure 1), a 3.1 cm long, thin (0.16 cm) linear T1-w/T2-w hypointensity structure extended from the roof of the caudal nasopharynx/oropharynx at the longus capitis muscle caudodorsally through the atlanto-occipital joint, cervico-medullary junction, and the dorsal aspect of the atlanto-occipital joint into the caudal capital oblique muscle. The linear structure was partially visible in transverse plane images, but the plane of acquisition of the images resulted in slice thickness artifact on sequential slices (Figure 2). At the medial aspect of the left longus capitis muscle and just medial/central to this muscle at the level of the atlanto-occipital joint, the soft tissue was thickened and hyperintense compared with normal muscle bellies on T2-w images. Within the left longus capitis muscle and the caudal capital oblique muscle, there was a focal T2-w hyperintensity, which did not suppress on T2-w FLAIR images, and that contrast was enhanced on T1-w postcontrast images. There was increased fluid (T2-w hyperintense, T2-w FLAIR hypointense) at the ventral aspect of the atlanto-occipital joint. Within the medulla oblongata/cervicomedullary junction at the margins of the linear structure, an ill-defined increased T2-w hyperintensity was present that did not suppress on T2-w FLAIR images and had regional contrast enhancement. Within the dorsal aspect of the cervical spinal cord from this level and extending caudally to the level of C6, a more triangular (broad-based dorsally, tapering ventrally) T2-w hyperintensity was present, which did not suppress on T2-w FLAIR images, and decreased in intensity from cranial to caudal.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

CT of the head and cranial cervical spine was performed following MRI^e. No IV contrast was administered. Images were reconstructed in bone and soft-tissue algorithms with the following acquisition parameters: 120 kVp, 140 mA, 512 × 512 matrix, 1.25 mm slice thickness, 130 × 130 mm field of view. Best seen with a soft-tissue algorithm/soft-tissue window, at the same level of the linear structure described on MRI, a thin linear structure slightly hyperattenuating to adjacent parenchyma (HU 76) was identified. Similar thickening of the longus capitis muscle was also identified.

The combination of imaging findings was consistent with a penetrating foreign body from the oropharynx/nasopharynx through the longus capitis muscle, cervicomedullary junction and atlanto-occipital joint, and caudal capital oblique muscle. Based on the CT attenuation values, plant material such as an awn or splinter was suspected. Muscle changes were consistent with myositis, edema, and a small abscess in the left longus capitis muscle. Changes at the cervicomedullary junction were consistent with encephalomyelitis and/or edema. The dorsal cervical spinal cord changes were suspected to be due to inflammation/infection and concurrent edema given its shape and margination.

While under anesthesia, the foreign body was palpable intra-orally at the caudal aspect of the soft palate. Surgical removal was declined, and the cat recovered from anesthesia. Neurologic status upon recovery was similar to before anesthesia. Soon after, the owner elected euthanasia. On postmortem examination, a 3 cm plant stem (presumed Gypsophila spp.) was removed from the soft palate. Abscessation was present dorsal to and surrounding the site of entry in the soft palate. A full necropsy was not performed.

Discussion

This case report details a combined imaging approach (CT and MRI) to define location, migration, and regional damage to the brainstem, spinal cord, and adjacent musculature. The site of entry was con-firmed at the caudal nasopharynx/oropharynx, through the left longus capitis muscle. The retching and sneezing observed by the owner was likely consistent with the foreign body entry site.

The foreign body was suspected to be plant material based on CT attenuation values, similar to another study of a stick foreign body, which had a hyperattenuation of HU 120–180.¹ Imaging findings would have potentially aided in surgical planning, because the foreign body appeared more accessible from the nasopharynx than from a dorsal laminectomy approach. Epidural or subdural fluid collections and abscess can be present when there is a penetrating injury into the CNS.⁹ This was not identified in this patient, but a small abscess was suspected within the longus capitis muscle (Figure 2); however, no fluid pocketing was identified at postmortem examination. Ill-defined T2-w hyperintensity within the medulla oblongata/cervicomedullary junction that did not suppress on T2-w FLAIR images was indicative of edema and/or inflammation. The triangular T2-w/T2 FLAIR hyperintensity and T1-w hypointensity within the cervical spinal cord caudal to the foreign body in this patient were not likely to be consistent with infarction and edema, despite having similar characteristics.¹⁰ In a case report with aberrant microchip migration into the cervical spinal canal, there was a linear hyperintensity present within the longissimus cervicis muscle and the spinal cord, consistent with edema and/or hemorrhage.⁴ Although this patient was euthanized, there are case reports of successful surgical and/or medical management of CNS foreign bodies.^1,2,5–8

Without an obvious cerebellar lesion on MRI, the definitive cause of the decerebellate rigidity is unknown, although it may potentially be explained by changes in the brainstem because many natural disorders resulting in opisthotonus and extensor rigidity involve the cerebellum and brainstem.¹¹ Decerebellate posture is characterized by opisthotonus (marked head and neck extension dorsally), extensor rigidity of the thoracic limbs, and flexion of the hips, without mentation changes.¹² Mentation changes reported here were probably a reflection of brainstem involvement and interruption of the ascending reticular activating system, which is responsible for alertness. Classically, decerebellate posture is associated with rostral cerebellar lesions and is thought to be due to loss of the inhibitory output of the Purkinje cells on the brainstem upper motor neuron tracts (especially the lateral vestibulospinal tract), which is responsible for extensor tone to the limbs. Although no evidence of meningitis/cerebellitis or vascular compromise was identified on the MRI, the possibility of such causing the decerebellate posture cannot be completely excluded, because a full necropsy was not done.

Conclusion

We have described the characteristics of CT and MRI sequences in one case of an oropharyngeal-spinal migrating foreign body. The combination of CT and MRI in this patient allowed for identification of the presence of the foreign body, high suspicion of foreign body type, and regional parenchymal changes as well as the probable site and may have helped in surgical planning.