Iatrogenic Traumatic Brain Injury During Tooth Extraction

Introduction

Traumatic brain injury (TBI) is a common disorder in veterinary medicine, most often the result of the patient being hit by a motor vehicle. Other forms of TBI resulting from penetrating gunshot wounds or bite wounds are also somewhat common. Iatrogenic orbital penetration is a relatively common complication of tooth extraction.¹ However, iatrogenic TBI during dental procedures rarely has been reported. In one case, a patient developed a retrobulbar and intracranial abscess following tooth extraction and subsequently died 48 hr after the procedure.¹ This case report describes the MRI findings and successful treatment of a dog that suffered iatrogenic TBI during extraction of tooth 210 when a dental elevator slipped.

Case Report

An 8 yr old spayed female Yorkshire terrier was referred to Massachusetts Veterinary Referral Hospital (MVRH) for evaluation of progressive neurological signs following a dental cleaning and extraction of several teeth. The referring veterinarian performed a routine dental cleaning and extracted several teeth 2 days prior to admission to MVRH. While extracting tooth 210, the dental elevator slipped and the veterinarian thought it entered only the left retrobulbar space. The dog had a slow recovery from anesthesia and scleral hemorrhage was noted. The patient was discharged that evening to the owner with instructions to give meloxicam (0.1 mg/kg per os [PO] q 24 hr) and clindamycin (6.9 mg/kg PO q 12 hr). The dog's clinical signs progressed the evening of discharge to agitation, vocalizing, and difficulty standing and walking. Several hours later, the dog was unable to stand or walk, was falling to the right, and appeared to be blind in the right eye.

The dog was taken to a local emergency clinic for evaluation. Physical examination revealed mild tachycardia (150 beats/min), mild tachypnea (50 breaths/min), mild erythema in the vicinity of teeth 209 and 210, a 1 cm × 3–4 mm bruise on the left upper eyelid, and severe scleral hemorrhage with exophthalmos oculus sinister (OS). Neurological examination revealed an absent menace response oculus dexter (OD) and slow to absent pupillary light reflex oculus uterque (OU). The remainder of the neurological exam was reportedly normal. Intraocular pressures were normal (17–26 mm Hg OU) and fluorescein staining was negative. The dog was prescribed buprenorphine (0.018 mg/kg IV q 8 hr) for pain management. Mannitol (1 g/kg IV) was administered at the time of admission to reduce intracranial pressure and was repeated twice more at 4 hr intervals. Ten hours after admission, the dog was still ataxic, but would trot for a short period of time. By 13 hr after admission, the dog's ability to walk had improved significantly and she was falling only when changing directions suddenly. At that point in time, the emergency veterinarian started IV fluids (0.9% Na chloride at 7 mL/hr) for dehydration. The dog was discharged a few hours later. At discharge, the dog was able to walk without assistance, but was circling to the left.

The patient was re-evaluated the following morning by its regular veterinarian who then referred the patient to MVRH. Physical examination revealed a normal heart rate (124 beats/min) and respiratory rate (40 breaths/min). The dog had a bruised upper eyelid OD, severe episcleral hemorrhage OS, a small ventral episcleral hemorrhage OD, and pain opening the mouth. Neurological examination revealed that the dog was ambulatory with circling to the left and occasional falling to the right, right hemiparesis, and mild to moderate proprioceptive ataxia in all four limbs. There was no menace response OD, and the patient did not track cotton balls thrown in the right visual field. The remainder of the cranial nerve examination was within normal limits. Absent conscious proprioception and hopping were noted in the right pelvic limb with delayed conscious proprioception and hopping in the right thoracic limb. Postural reactions on the left side and the remainder of the neurological examination were within normal limits. Neuroanatomical localization was to the left forebrain (cerebral hemisphere/thalamus). Screening blood tests were normal other than elevated lactate (2.5 mmol/L; reference range, 0.3–1.8 mmol/L). In-house prothrombin time and activated partial thromboplastin time were within normal limits.

On day 2, MRI of the head was performed using a 1.5 tesla magnet^a. The dog was premedicated with morphine^b (0.3 mg/kg intramuscular [IM]) and dexmedetomidine^c (0.01 mg/kg IM). General anesthesia was induced with propofol^d (5 mg/kg IV), administered to effect to allow endotracheal intubation, and maintained with isoflurane^e. A routine MRI scan of the head was performed, including sagittal (repetition time [TR], 2900 msec; echo time [TE], 120 msec) and transverse T2-weighted images (T2WI; TR, 3000 msec; TE, 100 msec), transverse fluid-attenuated inversion recovery (TR, 7000 msec; TE, 90 msec), transverse T2*HEMO images (TR, 605 msec; TE, 23 msec), and precontrast transverse T1-weighted images (T1WI; TR, 602 msec; TE, 18 msec). IV contrast agent^f (0.1 mmol/kg IV) was administered and transverse (TR, 602 msec; TE, 18 msec), sagittal (TR, 412 msec; TE, 18 msec), and dorsal (TR, 400 msec; TE, 18 msec) postcontrast T1WI images were obtained.

The MRI (Figure 1) showed a linear tract on T2WI (parasagittal) extending caudodorsally from the left orbit, through a skull fracture into the brain, ending in the left parietal lobe that was consistent with intracranial penetration of the dental elevator. On transverse images, the tract was hyperintense and relatively well defined on T2WI and poorly defined on T1WI. There was a hyperintense region at the dorsal extent of the tract on T2WI that was hypointense on T1WI, suggestive of fluid accumulation. On transverse T2*HEMO images, there was hypointensity in the ventral portion of the tract that was consistent with hemorrhage. There also was an area of hypointensity in each lateral ventricle, suggestive of intraventricular hemorrhage. The linear tract was observed to pass through the left lateral ventricle on transverse images at the level of the thalamus, explaining the likely source of intraventricular hemorrhage. Following administration of the IV contrast agent, there was mild blushing contrast enhancement along the tract. There was also a moderate amount of T2-weighted hyperintense signal in the surrounding brain parenchyma that was isointense on T1WI but that did not contrast enhance, suggestive of cerebral edema.

View Figure

• Download as Powerpoint
• Download Figure

Cerebrospinal fluid (CSF) analysis was performed and a mild mixed-cell pleocytosis was identified. The nucleated cell count was 6 cells/μL (reference range, <5 cells/μL; differential cytology: 58% nondegenerate neutrophils, 35% small to medium lymphocytes, and 7% large mononuclear cells). Cytology also demonstrated the presence of red blood cells within large mononuclear cells (erythrophagocytosis), suggestive of hemorrhage prior to CSF sampling.

The patient recovered normally following MRI and CSF sampling and was discharged the same day with clindamycin^g (10 mg/kg PO q 12 hr). CSF aerobic and anaerobic bacterial cultures were negative, but the clindamycin was continued because CSF bacterial culture can be negative even in the face of histologically confirmed bacterial meningoencephalitis.

One month after the initial MRI, the patient was re-examined at MVRH for a follow-up MRI. The owner reported that the dog was less ataxic and had started to use stairs again but still appeared blind in the right eye. The neurological exam was improved but still not normal. The patient was able to walk a straight line without falling but still had absent postural reactions in the right limbs and absent vision OD. Neurolocalization of the lesion was still to the left forebrain. Repeat MRI showed continued presence of the linear tract through the brain as described above, but the cerebral edema was significantly reduced. The clindamycin was discontinued and the patient was discharged.

Three months later, the patient was brought in for another MRI. Neurologic examination revealed a normal gait and intact vision in both eyes. The dog still had delayed postural reactions in the right thoracic and pelvic limbs. Neurolocalization was still to the left forebrain. MRI at that time showed that the tract was smaller than in the previous studies.

The patient was evaluated once again 1 yr after the original injury. The neurological examination was within normal limits other than continued mild postural reaction deficits in the right thoracic and pelvic limbs. MRI at that time showed continued presence of traumatic brain injury, but the lesion was smaller than at the time of original injury (Figure 2).

View Figure

• Download as Powerpoint
• Download Figure

Discussion

Ophthalmic complications of dental disease are common in dogs and cats because of the proximity of the caudal maxillary teeth and the ventral aspect of the orbit.² Several cases of iatrogenic injury to periorbital structures have been reported previously, but to the author's knowledge, only one case of intracranial disease following tooth extraction has been reported in the veterinary literature.¹ In that case, computed tomography showed a retrobulbar abscess with extension through the frontal bone into the calvaria with an associated intracranial abscess. The patient died 48 hr following the dental procedures despite draining the retrobulbar abscess, IV fluids, and IV antibioitics.¹

Penetration of the orbital floor during tooth extraction is relatively uncommon but can occur due to multifactorial causes, including regionally thin bony structures, periodontal pathology, and improper extraction techniques.¹ In dogs and cats, the bony orbit is incomplete and the ventral aspect of the orbit is comprised of soft tissues, including the zygomatic salivary gland, orbital fat, and medial pterygoid muscles.^1,2 The roots of the maxillary fourth premolar and both molar teeth are located within the maxillary bone in close proximity to the orbit.^1–3 Periodontal pathology can weaken the bone, leading to orbital penetration during tooth extraction.^3,4 Additionally, the orbits in brachycephalic breeds are positioned more rostrally than mesaticephalic and dolichocephalic breeds, increasing the risk of iatrogenic injury to the orbit and globe during tooth extraction.²

Several recommendations have been reported to reduce the risk of iatrogenic trauma to surrounding soft-tissue structures. Dental radiographs ideally should be obtained prior to tooth extraction to help gauge the depth of the roots and to assess periodontal pathology.¹ Teeth with multiple roots should be sectioned prior to extraction.^1,3,4 The elevator should be applied with gentle and steady rotational pressure for 10–30 sec at a time and advanced apically using a slow, twisting action.^1,3,4 Finally, to help reduce the risk of accidental slippage, it is recommended that the elevator be held down the shaft of the instrument and a finger placed near the tip of the elevator to act as a “stop” should the elevator slip.¹

Conclusion

Fortunately, the patient in this case report survived the iatrogenic traumatic brain injury with minimal residual neurological deficits. With attention to detail and proper handling of dental instruments, iatrogenic orbital and brain injury should remain an uncommon occurrence in veterinary medicine.