Accidental Intravenous Levobupivacaine

Abstract

The accidental IV injection of bupivacaine has been responsible for the death of a number of patients, including pregnant women during attempted epidural anesthesia (1) and with sudden cuff deflation during IV regional anesthesia (Bier block) (2). Levobupivacaine, the isolated S(-)-isomer of bupivacaine, is less toxic than the racemic mixture in animal studies and in small-dose injections in humans (3–5). It is being developed for clinical use as a less toxic alternative to racemic bupivacaine. We report the first case of an accidental IV injection of a large dose of levobupivacaine (142.5 mg) in a patient during attempted epidural anesthesia. No arrythmias or other severe cardiovascular effects were observed. Transient agitation was the only symptom of systemic toxicity, and the patient recovered fully. Case Report A 77-yr-old, 74-kg female, with history of osteoarthritis, asthma, and stable, treated hypertension (enalapril 10 mg PO QD), presented for elective total hip arthroplasty under epidural anesthesia. Written, informed consent was obtained to include her in a randomized research study of postoperative epidural analgesia. After infusion of 500 mL of lactated Ringer’s solution and administration of oxygen (2 L/min via nasal cannula), midazolam 2.5 mg was given 15 min before epidural placement. With the patient in the left lateral decubitus position, the epidural space was located, without difficulty, at a depth of 5.0 cm using an 18-gauge Tuohy needle and a loss of resistance to saline technique at the L2-3 interspace. A 20-gauge end-hole epidural catheter was advanced 4 cm into the epidural space without apparent paresthesia, blood, or cerebrospinal fluid. The patient was turned to the supine position, and after negative aspiration, 3 mL of 0.75% levobupivacaine with 1:200,000 epinephrine (15 μg) was injected as an epidural “test dose” (baseline HR [heart rate] = 90 bpm, BP [blood pressure] = 153/68 mm Hg, Figure 1). Five minutes later (HR = 86 bpm, BP = 146/49 mm Hg), incremental dosing of a planned additional 17 mL of 0.75% levobupivacaine without epinephrine was started. The patient remained communicative, and hemodynamic measurements remained unchanged (HR = 86 bpm, BP = 147/53 mm Hg). During the final 5 mL of injection, the patient became disoriented and drowsy, and her speech became slurred. Immediately thereafter, she became excited and began to shout and writhe about. The epidural injection was immediately halted, 19 mL of 0.75% levobupivacaine having been injected. Suspecting an accidental IV injection, two 40-mg doses of IV thiopental were empirically administered for seizure prophylaxis, and the supplemental oxygen was changed to 6 L via face mask. Signs of excitation abated, the patient continued to breathe spontaneously, and pulse oximetry remained 99%–100% throughout. Arterial blood gases were drawn (arterial pH = 7.37, PaO2 = 340 mm Hg, PaCO2 = 47 mm Hg, HCO3− = 26 mmol/L), as her hemodynamic values remained stable. No other therapeutic interventions were required. Figure 1: Real-time recording of hemodynamic variables (heart rate [HR], systolic blood pressure [BP], and diastolic BP) during accidental IV injection of levopupivacaine.Within 10 min, the patient was awake, oriented, and without complaint or recall of the preceding events. She was turned to the left lateral position, and, after removing the epidural catheter about 1 cm, frank blood could be freely aspirated. A thorough examination revealed no sign of even partial sensory or motor blockade. Because the patient remained clinically stable and her mental status had returned to baseline, we decided to proceed. The planned surgical procedure was completed using spinal anesthesia (hypobaric tetracaine 6 mg with 0.2 mg epinephrine) without further difficulties. Venous blood samples drawn 14 and 120 min after cessation of the intended epidural injection, resulted in serum levobupivacaine levels of 2.7 μg/mL and 1.1 μg/mL, respectively. The patient continued through an otherwise uneventful convalescence and was discharged from the hospital on the fourth postoperative day. Discussion Levobupivacaine, the newest amide local anesthetic, is thought to be less cardio- and neurotoxic than its racemic parent in laboratory and preclinical studies (3–6). This first case of accidental IV injection of levobupivacaine into a patient during attempted epidural anesthesia, without subsequent adverse consequences, further supports this conclusion. The incidence of local anesthetic-induced seizures has been demonstrated in two reports to be more frequent during initiation of brachial plexus anesthesia (0.199%, 0.075%) than during epidural anesthesia (0.012%, 0.013%) (7,8). However, this case shows that, even when the most reliable precautionary measures are taken (i.e., negative aspiration, “test dosing” with epinephrine, the slow administration of fractionated doses of drug while in continuous communication with the patient), accidental intravascular injection of large volumes of local anesthetic and potential systemic toxicity can still occur during attempted epidural anesthesia. We can only presume that the tip of our uniorifice catheter was juxtaposed to the vein wall and that the negative pressure generated by the syringe caused it to become occluded or that a small clot was causing a ball-valve effect. It is unclear why this patient did not demonstrate a “positive” response to 15 μg of epinephrine. β-adrenergic blockers may make an epinephrine-containing “test dose” unreliable, but angiotensin-converting enzyme inhibitors should not. The shortest time to onset of any pinprick anesthesia in 59 patients given either 0.5% or 0.75% levobupivacaine (15 mL during a period of 7 minutes and 45 seconds) was 6 ± 4 (mean ± SD) minutes after the end of the injection (9), so the absence of initial signs of blockade is of no use in indicating impending problems, as has been suggested (10). Several factors may have contributed to the absence of significant central nervous system or cardiovascular side effects in our patient. First, she had received a small amount of midazolam (2.5 mg IV) for sedation shortly before the attempted epidural, and benzodiazepines increase the seizure threshold for IV local anesthetics (11). Second, it is not clear what peak serum levobupivacaine was attained in this patient. Unfortunately, sufficient blood was not saved from the arterial blood gas drawn at 8 minutes to measure a levobupivacaine level. When venous blood was specifically drawn 14 minutes after the end of injection, the levobupivacaine level detected (2.7 μg/mL) could still be above the toxicity threshold (>2.0–4.0 μg/mL) (12) if bupivacaine had been injected. The peak concentration 10–15 minutes earlier would have definitely been much higher. Severe impairment of intraventricular conduction does not ordinarily occur unless bupivacaine levels exceed 8 μg/mL or other conditions, which decrease this threshold, exist (e.g., electrolyte imbalances, concomitant cardioactive drugs). Our patient was receiving enalapril for essential hypertension. Although a variety of other cardiovascular drugs, which slow intraventricular conduction (glycosides, class IC antiarrhythmics, tricyclic antidepressants, and β-adrenergic blockers) (13,14), are felt to potentiate the cardiotoxic effects of bupivacaine, neither Ca++-channel blockers nor angiotensin converting enzyme inhibitors have shown this tendency. Currently, there is no known level for levobupivacaine that can be considered toxic. The levobupivacaine blood concentrations measured (2.7 μg/mL at 14 minutes, and 1.1 μg/mL at 120 minutes) are evidently less than the clinical toxic threshold. In a cross-over study of slow (10 mg/min) IV infusion of levobupivacaine in 12 healthy volunteers, central nervous system symptoms first appeared at a larger mean total dose (54.0 mg versus 45.6 mg), producing a higher resultant plasma level (2.38 μg/mL versus 1.87 μg/mL) than for racemic bupivacaine (5). Third, despite the period of agitation, our patient could ventilate normally, maintaining an arterial pH of 7.37 and PaCO2 of 47 mm Hg. Acidosis increases the cardiotoxicity of IV local anesthetics (15). Finally, levobupivacaine itself may be inherently safer than racemic bupivacaine. Although the levobupivacaine level was not drawn until well past the expected peak, the level that was detected (2.7 μg/mL after 14 minutes) is equal to or greater than those obtained in many previous cases of accidental IV racemic bupivacaine injection in which seizures or severe cardiac arrhythmias occurred. Serum levels as low as 1.8 (16), 2.03, 2.74 (17), and 2.3 μg/mL (18) of racemic bupivacaine have been detected within 5 minutes of injecting patients having seizures. Our intent in reporting this case is to warn anesthesia providers that all local anesthetics are potentially harmful, that levobupivacaine may be less toxic relative to other potential local anesthetics, and that early recognition and the rapid initiation of measures to prevent symptom progression are the mainstays of therapy.