Surgical removal of an intradural brainstem foreign body (microchip) with a ventral basioccipital approach in a kitten

Abstract

Case summary

A 2-month-old kitten was presented for ataxia and depressed mental status after implantation of a pet identification microchip. Radiographs were taken immediately by the referring veterinarian and showed a longitudinal metallic foreign body (electronic microchip) within the cervical vertebral canal at the craniocervical junction. A CT examination 2 days after the incident showed cranial migration of the microchip ventrally to the caudal brainstem. Ventral basioccipital craniectomy was immediately performed to retrieve the microchip by a ventral approach to the caudal brainstem. Postoperative recovery was uneventful and the cat was discharged 2 days later. At the 2-week follow-up, neurological examination of the cat was normal. No long-term complications were reported.

Relevance and novel information

This case report describes the intradural migration of a microchip and surgical removal via ventral basioccipital craniectomy.

Introduction

Microchip implantation is a common method of identification in domestic cats and dogs. In France, identification by ear tattoo or microchip is mandatory by 4 months of age in dogs and by 7 months in cats. Since 2011, a microchip has also been required for international travel with a domestic animal. The microchip should be implanted subcutaneously, either on the left side of the neck or between the scapulae.

There have been a few reports describing neurological complications after incorrect microchip placement in cats and dogs.^1–9 In the cases previously described, medical^1–3 or surgical^4,5 treatment was chosen, depending on the localisation, technical surgical considerations and severity of the neurological signs. Intraspinal extradural or intramedullary localisation was most common.

The ventral surgical approach to the caudal brainstem is a rarely used technique in veterinary medicine, previously described in cadaveric and in vivo studies in dogs ¹⁰ as well as a few case reports.^11–13 It has been considered an appropriate approach for accessing neoplastic lesions located ventral to the caudal brainstem. This case report presents the first description of an aberrant migration of an electronic microchip to an intradural extramedullary localisation and its surgical removal via a ventral basioccipital ostectomy in a kitten.

Case description

A 2-month-old kitten was referred to the neurology service after presenting with acute neurological signs after microchip implantation 5 days earlier.

Based on the clinical signs and dorsoventral radiographs obtained by the referring veterinarian, the microchip was suspected to be located within the vertebral canal (Figure 1). Orthogonal views were not performed by primary veterinarian.

Figure 1

Cervical radiographs of the kitten immediately after the microchip placement. Dorsoventral view. The microchip can be seen at the level of the craniocervical junction

A neurological examination revealed symmetrical general proprioceptive ataxia and mild tetraparesis. Postural reactions were mildly reduced in all four limbs, while spinal reflexes remained normal. The patient’s mental status was mildly decreased, with a reduced response to environmental stimuli. The only abnormality on cranial nerve examination was an absent menace response, consistent with age.

Radiographs performed after admission confirmed rostral migration of the microchip, which was located in the ventral part of the caudal fossa (Figure 2).

Figure 2

(a) Dorsoventral and (b) lateral radiographs of the cranium and cervical spine. The microchip is visualised as a longitudinal metallic foreign body

A CT examination was performed under general anaesthesia to further refine the location. The cat was premedicated with midazolam (0.2 mg/kg IV) and methadone (0.2 mg/kg IV), and anaesthesia was induced with propofol, dosed to effect. After tracheal intubation, the patient was maintained under anaesthesia with 100% oxygen and 1.5% isoflurane. Imaging was conducted using a 64-slice CT scanner (Aquilion 64; Toshiba). Images were obtained before and after intravenous (IV) contrast injection (iopamidol 300 mg/ml). CT confirmed that the microchip was located dorsally to the basioccipital bone and ventrally to the caudal brainstem (Figure 3), in an oblique position.

Figure 3

Bone window CT images of the head. (a) Sagittal and (b) transverse views showing the microchip at the level of the basal part of the occipital bone and laterally up to the petrosal part of the right temporal bone, under the caudal brainstem

A surgical intervention was performed immediately after the CT study. The cat was positioned in dorsal recumbency with the head and neck extended. A small round cushion was placed under the cat’s neck to enable adequate cervical extension. The surgical site was clipped and aseptically prepared, extending from the manubrium sterni to the rostral third of the mandible. This approach was similar to the previously described ventral atlantoaxial approach, ¹⁴ with adjustments made to access the caudal ventral brainstem. ¹⁰ A midline incision was made from the caudal third of the mandible to the mid-cervical region. Subcutaneous tissues were dissected to expose the sternohyoideus and sternocephalicus muscles. The sternohyoideus muscle was then dissected along the midline and retracted. The right carotid sheath and sternohyoideus muscles were separated by blunt dissection and retracted to the right, while the trachea and oesophagus were retracted to the left. The right sternothyroideus muscle was then transected at its rostral attachment, preserving the right cranial thyroid artery. The ventral process of C1 and the atlanto-occipital joint were palpated. The longus coli and longus capitis muscles were dissected in the midline using a periosteal elevator to expose the atlanto-occipital joint and ventral surface of the basioccipital bone. Dissection continued rostrally to expose the caudal two-thirds of basioccipital bone.

A ventral basioccipital craniectomy was performed using a 2 mm burr. To avoid damaging the regional vasculo nervous structures, several anatomical landmarks were respected. Laterally, the craniectomy was confined within the parasagittal planes of the bullae at the junction with basioccipital bone to avoid injury to structures running through the tympano-occipital fissure (including cranial nerves IX, X and XI, the vertebral and internal jugular veins, and the carotid artery) as well as those within the hypoglossal canal (such as the hypoglossal nerve and a branch of the vertebral vein). Rostrally, the craniectomy was limited by the caudal clinoid process of the dorsum sellae, avoiding damage to the cerebral arterial circle. ¹⁰

In this case, the caudal limit of the craniectomy was the caudal edge of the basioccipital bone, while the rostral margin was measured using CT images based on the position of the microchip. A mild right-sided asymmetry was maintained in the craniectomy approach because of the oblique position of the microchip.

Drilling continued until the microchip was visualised; it was found in a subdural location (Figure 4). The final craniectomy window measured 3.8 mm in width and 5.8 mm in length. A durotomy was performed using a number 11 surgical blade to access the subdural space. An abundant leak of clear cerebrospinal fluid (CSF) was encountered, which quickly subsided. The durotomy provided clear visualisation of the microchip, which was gently removed using small forceps. After removing the microchip, an attempt was made to suture the dura mater; however, because of the small window and the delicate structure in a patient this small, closure was not possible and the dura mater was left open.

Figure 4

Intraoperative images. (a) The grey microchip is visible through the dura mater within the craniectomy window (arrow). (b) The medulla oblongata is visualised after the durotomy and removal of the microchip

After saline flushing, the surgical site was closed. The longus capitis, longus colli and sternohyoideus muscles, as well as the subcutaneous tissues, were sutured using 3-0 synthetic resorbable suture material. The skin was closed with 3-0 synthetic non-absorbable sutures. A small amount of CSF leakage was still visible during closure of the sternohyoideus muscle but had stopped by the time the subcutaneous layer was closed. No significant haemorrhage occurred during the procedure.

An immediate postoperative CT examination was carried out and was unremarkable (Figure 5). The final size of the bone window was 6 mm in length and 4 mm in width.

Figure 5

Postoperative CT images of the head. (a) Sagittal, (b) transverse and (c) dorsal views, showing the area of the craniectomy

The cat was hospitalised for 48 h. Inpatient treatment included buprenorphine (0.02 mg/kg IV q8h) and meloxidyl (0.1 mg/kg IV q24h). An orthopaedic neck collar was fitted for the first 2 weeks postoperatively.

The microchip was sent for bacteriological analysis, and the cat received antibiotics (cefalexine 20 mg/kg PO q12h) until the results returned negative.

The day after surgery, the cat was ambulatory with mild ataxia and paresis, similar to its preoperative state. Its cranial nerve function and mental status were normal.

At the 2-week recheck, the cat was considered clinically normal. The neck collar and sutures were removed, showing good wound healing.

A telephone follow-up call with the owner, conducted at the time of writing (2 years postoperatively), confirmed complete recovery and the absence of any long-term complications.

Discussion

Microchip placement is a safe and routine procedure. Reports of traumatic complications are rare and mostly involve the cervical spinal cord and extradural or intradural/intramedullary placement.^1–9 Neurological deficits are typically acute and appear immediately after implantation; however, migration of the transponder and delayed complications may also occur. ⁴

Management of reported cases has included both medical^1,3 and surgical approaches.^4–8 Medical treatment is chosen if the microchip does not cause significant spinal cord compression or if surgery is considered too invasive. The surgical approach is dependent on the microchip’s location and typically involves a dorsal laminectomy or hemilaminectomy. In the present case, the decision between medical and surgical management was based on a risk-benefit assessment. The most frequent complication encountered during a ventral basioccipital approach is bleeding from the basilar artery. ¹⁰ Given the small size and young age of our patient, this type of complication could have been challenging to manage. Surgical removal of the chip was performed to prevent potential brain damage from further migration or inflammation. A long-term inflammatory response was also a risk associated with a medical approach.

The microchip’s intradural and extramedullary location in our patient is unusual. Within the vertebral canal, the dura mater is separated from the periosteal lining of the vertebrae, forming a significant epidural space, apart from the first two cervical vertebrae ventrally. At the level of foramen magnum and along the floor of the vertebral canal within vertebrae C1 and C2, the spinal dura mater fuses with the periosteum. As a result, the epidural space disappears at this region. ¹⁵

There are two possible explanations for the unusual intradural localisation of the microchip. The first is that the microchip was possibly injected directly into the intradural space at the C1–C2 level. The second is that the microchip may have initially been placed in an extradural location and subsequently migrated, lacerating the dura mater within the epidural space of the vertebral canal, to reach its final subdural position. This migration could have been facilitated by the cat’s head movements.

We suspect that the microchip needle was inadvertently inserted through the paravertebral muscles and implanted in close proximity to the cervical spine; however, this cannot be confirmed, as no soft tissue lesions were visible on the CT scan performed 5 days after implantation. As no bone lesions were identified, we suspect that the microchip either migrated through or was injected via the C2–C3 interarcuate space or the C1–C2/C2–C3 intervertebral foramen. An MRI scan could have provided additional information regarding the soft tissue lesions but it was deemed unnecessary, as it would not have affected the treatment plan. In addition, the microchip creates susceptibility artefacts, resulting in signal voids that limit the use of MRI in this case.

In the present case, the main intraoperative issue was the CSF leak and the inability to suture the dura mater after the durotomy. In human paediatric medicine, the incidence of CSF leakage after intradural cranial procedures has been reported at 4.4%. ¹⁶ This risk is heightened in surgeries involving the caudal fossa. Potential complications associated with CSF leakage include pseudomeningocele formation, delayed wound healing, surgical site infections, meningitis and pneumocephalus. ¹² None of these complications were observed in our patient. The impact of CSF leakage during veterinary intracranial surgery remains unclear.

Conclusions

This case report illustrates how a routine procedure such as microchip implantation can lead to dramatic consequences when performed incorrectly, including intracranial migration. This should encourage veterinarians to ensure proper subcutaneous needle placement and minimise patient movement during the procedure. Although the ventral surgical approach to the caudal brainstem has been rarely reported in dogs, this case illustrates its feasibility in cats, even in very small patients such as kittens. Limiting the dimensions of the basioccipital may help reduce intraoperative complications, such as haemorrhage from the basilar artery.