Prolonged Coma and Quadriplegia After Accidental Subarachnoid Injection of a Local Anesthetic with an Opiate

Accidental subdural or subarachnoid injection of a local anesthetic is an uncommon complication of epidural anesthesia (1–4). The present case documents an unusual feature of accidental subarachnoid injection of a local anesthetic together with an opiate. It was characterized by a delayed onset and a prolonged duration of coma and quadriplegia, with slow but complete recovery. Case Report On April 11, 1998, at 11:00 AM, a healthy, 80-kg, 160-cm, 22-yr-old, gravida 1 patient, without any family history of anesthetic intolerance or acute intermittent porphyria, presented in early labor at the delivery room in her 42nd week of gestation. At 9:00 PM, after augmentation of labor by amniotomy and oxytocin when the cervix was dilated to 4 cm, epidural analgesia was requested. After prehydration and with the patient in a sitting position, epidural puncture was attempted using a loss of resistance to air technique with an 18-gauge Tuohy needle by a midline approach at the L2–3 interspace. The L2–3 interspace could not be easily identified on the first try because of the patient’s obesity and subcutaneous edema, and she complained of a sudden onset of sharp pain in the cervico-occipital region, reminiscent of subarachnoid air injection. Aspiration for cerebrospinal fluid (CSF) was negative, and the needle was withdrawn. A second attempt, at L1–2, using the same position and technique, was successful, and a Portex (SIMS Portex Ltd., Hythe, Kent, UK) epidural catheter, outside diameter of 0.9 mm, was easily threaded 3 cm cephalad beyond the needle point. Aspiration was negative through both the needle and the catheter. The total amount of air used to identify the epidural space was approximately 10 mL. The patient was placed in a supine position with a left lateral tilt, and a 3-mL test dose of 2% lidocaine without adrenaline, was injected. It is our policy not to include epinephrine in the test dose in epidural anesthesia in parturients in labor. At 9:30 PM, after another negative aspiration from the catheter, and a negative test dose, 10 mL of 0.25% bupivacaine with 100 μg of fentanyl, was injected in divided doses of 4 mL, 3 mL, and 3 mL over 30 min, achieving an adequate sensory block for labor pain (T10–L2). At 11:00 PM, the patient vomited, and her blood pressure decreased from 120/80 to 90/40 mm Hg without any electrocardiogram changes. Pulse rate remained stable at 100 bpm. Oxygen was given via a face mask. Ten milligrams of ephedrine and 10 mg of metoclopramide HCl were injected IV. Naloxone 0.4 mg IV was subsequently added, because her pupils were constricted. A few minutes later, the patient became somnolent with complete symmetrical lower limb weakness and sensory loss to the T6 level. The hypotension episode lasted 5–7 min, and over the following 2 h, the blood pressure was maintained at a level of 80–90/45–50 mm Hg by intermittent 5-mg boluses of IV ephedrine (total dose, 30 mg) and the infusion of normal saline 1.5 liters. At 11:40 PM, the cervix was completely dilated, and vacuum extraction was applied to expedite delivery. A normal 3.2-kg female infant was delivered, with Apgar scores 9 and 10, at 1 min and 5 min, respectively, and with a cord blood pH of 7.1. The epidural catheter was removed with a negative aspiration test for CSF, after delivery of the infant and placenta. At 12:30 AM on April 12, the patient showed progressive motor weakness of the upper limbs and evidence of respiratory failure. Arterial blood gases indicated a combined metabolic and respiratory acidosis pH of 7.175, PaCO2 of 45.4, PaO2 of 137.6, B.E of 12.4, HCO3 of 16.2, and SaO2 of 99%. Sodium thiopental and succinylcholine were administered, followed by endotracheal intubation and mechanical ventilation. The patient was transferred to the intensive surgical care unit. Therapy was supportive and included IV fluids, intermittent ephedrine, and maintenance of assisted pulmonary ventilation. Neurologic evaluation at 2:30 AM revealed the patient to be in coma, with spontaneous eye movement, slightly constricted pupils (2 mm), reaction to light, normal doll’s eye movement, and no papilledema. Supra-orbital pressure evoked symmetrical facial grimacing, but no limb movement. Corneal reflex remained intact bilaterally. Jaw jerk could not be elicited. Palatal movement could not be assessed, as she had an oral endotracheal tube in place. There was a good tracheal reflex to carinal stimulation, with spontaneous diaphragmatic ventilation. No spontaneous limb movement or limb movement in response to pain stimuli could be obtained. A flaccid tetraplegia with complete loss of tendon reflexes was present as was atony of the anal sphincter on rectal examination. Since the time of endotracheal intubation, we had refrained from administering any sedative or analgesic drugs. As this neurologic state persisted, cerebral and spinal axial computed tomography was performed at 9:00 AM. This showed air bubbles in the subarachnoid space of the spinal cord and the brain (Fig. 1). No signs of hematoma or edema were demonstrated. Another IV injection of 0.4 mg of naloxone was ineffective. Laboratory examination, which included serum potassium, calcium and magnesium ions, creatine phosphokinase, acetylcholine receptor antibodies, and pseudocholinesterase levels, eliminated metabolic and electrolytic disturbances or myasthenia.Figure 1: Axial computed tomography scan of the brain at the midconvexity level, showing pneumocephalus of the frontal lobe and an air bubble in the supra cerebellar cistern.In order to exclude a drug error or the possibility the injection of some unusual substance via the epidural catheter, the lidocaine ampule, the bupivacaine vial, and the fentanyl ampule were checked both before their administration and afterward by two people—the senior anesthesiologist and the midwife, who remained with the patient throughout the entire incident until her transfer from the labor ward to the intensive surgical care unit. At 10:00 AM, the patient was stuporous, but she began moving her fingers and toes on verbal command and could breathe using her chest muscles. On April 13 at 3:00 AM, she was still somnolent, but moved her limbs. Extubation was performed at 10:00 PM. The following morning, the patient became aggressive and bit the attending nurse. By 3:00 PM, she was fully awake and coherent. She had complete amnesia for all the events from the onset of labor onward. On discharge from the hospital on the sixth postpartum day and 1 mo later, her neurologic status and psychiatric status were normal. Discussion Inadvertent dural puncture and extensive block is one of the distressing complications associated with lumbar extradural analgesia. Depending on the technique, its incidence ranges from 0.4%–7.5%. According to Morgan (5), unexpected extensive conduction block can occur as a result of 1) high or total spinal anesthesia from the first anesthetic dose injected into the subarachnoid space, despite negative aspiration for CSF and a negative test dose; 2) high block on a subsequent top-up after a known prior dural tap; or 3) high or total spinal on subsequent top-ups caused by migration of the catheter. The case presented here is the first report of a delayed onset and prolonged duration of total spinal anesthesia after the accidental subarachnoid injection of a local anesthetic combined with an opiate. At first, the clinical picture was suggestive of a subdural injection of the local anesthetic, a well recognized complication (1,6). The block was extensive, the onset delayed, and the aspiration test for CSF was negative. However, the block subsequently became more extensive and profound, ascended bilaterally, and was accompanied by prolonged coma—a picture more reminiscent of total spinal anesthesia of prolonged duration rather than a subdural injection. Furthermore, this diagnosis is supported by the subsequent finding of air bubbles in the subarachnoid space of the brain. Our presumptive explanation is that a large quantity of local anesthetic with fentanyl, passed intrathecally through the puncture made during the first epidural puncture attempt. The ineffectiveness of naloxone administration emphasizes the greater role of the local anesthetic in such a complication. Previous observations of contrast material injected through a catheter placed accidentally in the subdural space (7) revealed that the material ascended unilaterally and freely over a considerable distance. Fractional epidural administration of the local anesthetic dose failed to prevent this serious mishap. Despite repeated checking of all the substances injected via the epidural catheter in this patient, there was no evidence that some unrecognized agent or that a dose, larger than that stipulated of the local anesthetic and opiate combination had been administered in error. The prolonged coma demanded that organic or metabolic causes be excluded (8). Whereas the clinical scenario of a “wet tap,” followed by reinsertion of a catheter at another site, is not uncommon, this procedure demands special precautions, and one must anticipate the possible entry of drugs injected extradurally into the subarachnoid space. This could be accomplished by using a slow infusion pump for the extradural injection or by using ultra-low concentrations of local anesthesia in such a situation. Awareness of this complication would enable earlier diagnosis and optimal management to be instituted. The authors wish to thank Dr. M. L. Cohen from the Department of Anesthesiology, Assaf Harofeh Medical Center, for his advice and editing assistance.