Non-barbiturate, drug-induced reversible loss of brainstem reflexes

Criteria for establishing brain death require the clinician to rule out reversible causes for loss of brainstem function. Aside from barbiturates, few medications are known to abolish brainstem reflexes entirely.

Case report. A 60-year-old woman underwent a five-vessel coronary artery bypass graft after receiving streptokinase and heparin for an acute myocardial infarction.

Postoperatively, she was alert, following commands, and moving all extremities. Early on postoperative day 2, she developed pulseless ventricular fibrillation requiring cardiopulmonary resuscitation and defibrillation. She received a lidocaine bolus (100 mg) followed by lidocaine drip (2 mg/min). She was alert following this episode. Nine hours later, she experienced ventricular tachycardia requiring cardioversion. She was bolused with lidocaine (75 mg), her lidocaine drip was increased (4 mg/min), and amiodarone and bretylium drips were started. She was alert and following commands after this episode.

Later on postoperative day 2, she lost her distal pulses. She was believed to have embolized from her aorta and was placed on heparin. Heparin was then discontinued due to presumptive heparin-induced thrombocytopenia and thrombosis syndrome (HITTS).

During the afternoon of postoperative day 2, she became less responsive. Because of grimacing thought to reflect pain, fentanyl (25 mcg/min) was begun at 9:00 PM. By 10:00 PM, she had no eye opening or extremity movement(Glasgow Coma Scale [GCS] = 3). Pupils remained reactive at 5 mm. Between 11:00 PM and midnight, lidocaine was decreased (3 mg/min) (GCS = 7 at midnight and 4:00 AM on postoperative day 3). She opened her eyes to painful stimuli; pupils remained 5 mm and reactive. Between 5:00 and 6:00 AM on postoperative day 3, the investigational thrombin inhibitor, argatroban, was begun for treatment of HITTS. At 7:00 AM, she was unresponsive with unreactive pupils. The fentanyl was discontinued.

At 9:00 AM, neurologic evaluation for possible brain death revealed blood pressure 120/90, respiratory rate = 14 (not breathing above the ventilator), heart rate = 100, and temperature = 39.2°C. Pupils were 6 mm and unreactive. There were no corneal, gag, or oculocephalic reflexes. Limbs were flaccid without spontaneous movement, withdrawal, or posturing to noxious stimuli. She was areflexic except for grade 2+ reflexes at the right biceps and brachioradialis. Plantar responses were mute. Apnea test was precluded by hemodynamic instability.

Abnormal laboratory test results included glucose = 130 mg/dL, creatinine= 1.3 mg/dL, total bilirubin = 3.3 mg/dL, aspartate transaminase = 319 IU/L, alanine transaminase = 160 IU/L, lactic dehydrogenase = 3,918 IU/mL, partial thromboplastin time = 39 seconds, prothrombin time = 16.7 seconds, total protein = 2.5 g/dL, albumin = 1.2 g/dL, calcium = 6.0 mg/dL, cholesterol < 50 mg/dL, hematocrit = 33%, and platelet count = 36,000/µL.

Medications included IV famotidine, levothyroxine, argatroban, amiodarone, dopamine, bretylium, phenylephrine hydrochloride, and lidocaine. Phenylephrine hydrochloride was discontinued at 5:00 PM on postoperative day 2 and restarted between 6:00 and 7:00 AM on postoperative day 3. Famotidine and bretylium were stopped between 9:00 and 10:00 AM. Lidocaine was discontinued at 11:00 AM. The lidocaine level was 11.2 mcg/mL (normal range = 1.5 to 5).

A brainstem stroke was considered, but she was not considered stable enough to undergo a CT scan.

By noon, she was opening her eyes to voice. Pupils were 5 mm and reactive; GCS = 10. At 2:00 PM, eyebrows raised and eyes opened to voice. Oculocephalic reflexes were intact. She grimaced to pain in both upper extremities but appeared to have no sensation in her necrotic lower extremities. She flexed her right arm to pain but not her left; reflexes were somewhat brisker on the right. She gradually awoke, and was intermittently following commands and moving both upper extremities. Fentanyl was restarted. While sedated, her brainstem function continued to be intact.

After undergoing bilateral above-the-knee amputations and dialysis for transient renal failure, she was transferred to the rehabilitation unit on postoperative day 31 and discharged to home on postoperative day 47. Neuropsychologic evaluation revealed mild cognitive impairment, but no neurologic deficits were noted otherwise.

Discussion. Our patient never had an apnea test, EEG, or radionucleotide brain scan. She never met criteria for brain death. She did, however, have a reversible loss of all apparent brainstem function. The relative contribution of central versus peripheral nervous system involvement is difficult to discern. Because of the known CNS effects of lidocaine and the temporal correlation between its discontinuation and her recovery, we propose that her neurologic condition was, at least in part, due to this agent. CNS toxicities have been noted with levels >5 mcg/mL¹ and coma has been reported.^2,3 Lidocaine has been associated with abolition of brainstem auditory evoked potentials.⁴ A study in dogs revealed that lidocaine had a "barbiturate-like" effect, abolishing electrocortical activity and reducing oxygen and glucose consumption.⁵

An interaction between lidocaine and one or more of the other drugs is possible. Fentanyl was discontinued 2 hours earlier but may have been exerting sedative effects. Although the loss of brainstem reflexes correlated with the introduction of argatroban, these reflexes recovered despite its continuation. There have been anecdotal reports of headache with argatroban but it has not been associated with any other neurologic side effects(personal communication, Texas Biotechnology Corporation, Houston, TX, 1998).

This case highlights the importance of seeking reversible causes of absent brainstem function when evaluating patients for brain death.

Acknowledgment

We thank Dr. Frederick Marshall for his helpful input and Ms. Donna LaDonna for preparing the manuscript.