Bupivacaine Cardiotoxicity: From Observation to Evidence

  • Youssefzadeh, Kevin
| Mar 28, 2022

Yousefzadeh, KevinThe CSA Committee on the History of Anesthesia congratulates Dr. Kevin Youssefzadeh, the 2nd place winner of the 2022 CSA History of Anesthesia Essay Contest!

Dr. Youssefzadeh, a CA-2 resident at UCLA, is a native Angeleno. When not at the hospital, he enjoys taking long walks with friends and family.

Synthesized in 1956, bupivacaine was introduced to the market in 1965 and quickly became an anesthetic of choice for regional anesthesia.1 However, peculiar cases of cardiovascular collapse after bupivacaine administration began to be reported in the 1970s. Many of these cases occurred in the apparent absence of significant hypoxia or metabolic derangement. While the adverse respiratory and central nervous system (CNS) effects of local anesthetics were well known, the cardiac toxicity of highly lipophilic local anesthetics was not yet understood. Contributions from several Californian anesthesiologists led to the recognition of bupivacaine’s cardiac risk. This resulted in improved patient safety and sparked debate about some general principles of anesthetic practice.

In 1979, Stanford Assistant Professor George A. Albright, MD (1931-2020), submitted a manuscript to Anesthesiology entitled “Cardiac Arrest Following Regional Anesthesia with Etidocaine or Bupivacaine.”2 In it, he presented five cases during which administration of highly lipid soluble local anesthetics resulted in almost immediate cardiovascular collapse. The rapidity with which the patients deteriorated suggested an absence of preceding hypoxia or hypotension. Dr. Albright noted that resuscitation was difficult, often requiring prolonged cardiac massage.2 After examining these cases, he astutely hypothesized that the cardiovascular toxicity threshold of bupivacaine and etidocaine was relatively low compared to other agents at similar clinical doses. He recommended further study of the potential cardiovascular toxicity of these agents, along with an investigation of possible treatment modalities.

When C. Philip Larson, Jr., MDCM, Dr. Albright’s Chair of Anesthesia at Stanford and Editor-in-Chief of Anesthesiology, received the paper for publication, he was alarmed. “At the time I had no suspicion of the high rates of toxicity with bupivacaine and thought it was serious enough to publish,” Dr. Larson recalled during an interview on November 6, 2021. He shared Dr. Albright’s manuscript with a council of senior scientists and clinicians for review. The reviewers disagreed with its findings and did not recommend publication based on what they viewed as merely anecdotal findings and a lack of scientific evidence.

However, Dr. Larson still found the case reports compelling. “I had major concern over the findings of the paper and made the decision to publish in light of the review board’s critique…. I did not believe that the editorial would be found controversial,” he stated. Due to the reviewers’ recommendation, he chose to publish the manuscript not as an original article, but as an editorial instead. Its publication sparked a heated debate, with many practicing anesthesiologists agreeing with the reviewers’ conclusion that the fatalities were most likely due to hypoxia, metabolic derangement, or improper resuscitation, rather than any intrinsic cardiotoxicity of the drugs. However, after publication of the editorial, several more case reports and animal studies demonstrating bupivacaine toxicity were published.

On August 23, 1983, the Food and Drug Administration (FDA), concerned by the case reports in Dr. Albright’s editorial and other publications, sent a letter to all doctors in the United States recommending discontinuation of the use of 0.75% bupivacaine in epidural obstetric anesthesia. In October of that year, the FDA held a special meeting of the Anesthetic and Life Support Drugs Advisory Committee. During the meeting, Dr. Albright presented 49 cases of cardiac arrest after bupivacaine administration, with 16 of the 21 associated deaths having occurred in obstetric patients who had received the drug epidurally.1

Ultimately, the committee upheld its recommendation to discontinue use of 0.75% bupivacaine in obstetric epidural anesthesia and also contraindicated its use in intravenous regional anesthesia. Additionally, committee members recommended the use of a test dose containing epinephrine, slow incremental administration, and availability of resuscitation supplies whenever performing regional anesthesia—practices that persist today.1 Critics of the decision argued that even if bupivacaine did have cardiotoxic properties—a theory that had yet to be rigorously proven—the cumulative dose of bupivacaine was more likely to blame than the specific concentration itself.3,4 

International governing agencies soon followed the FDA and discontinued the use of 0.75% bupivacaine for obstetric anesthesia in the United Kingdom and Canada. This sparked further critique internationally given that no cases of cardiovascular toxicity related to bupivacaine use had been reported in those countries. Several international journals published editorials dismissing the safety concerns over bupivacaine and insisting that cases of cardiovascular collapse could be attributed to suboptimal resuscitation practices or poor procedural technique leading to intravascular injection.3,4

With definitive proof of bupivacaine’s cardiac toxicity still lacking, the January 1984 issue of Anesthesiology finally provided conclusive evidence. A paper titled “Bupivacaine-induced Cardiac Arrhythmias in Sheep,” written by University of California, San Francisco Professor Sol Shnider, MD (1929-1994), and colleagues, revolutionized the debate. The authors had performed intravenous injections of lidocaine and bupivacaine at epidural analgesic doses to gather data on any potential CNS and cardiac toxicity in sheep. The animal subjects’ pH, pCO2, pO2, base excess, and serum potassium were measured at frequent intervals after injection. No significant metabolic disturbances were seen even during seizure activity, which was quickly terminated with thiopental. However, after administration of bupivacaine at therapeutic and supratherapeutic doses, all subjects demonstrated serious arrhythmias and/or EKG changes. Approximately 50% of subjects in the therapeutic-dose bupivacaine group developed either atrioventricular block, supraventricular tachycardia, ST changes, or premature ventricular contractions after administration. 11% of subjects in the therapeutic-dose and 80% of subjects in the supratherapeutic-dose groups developed ventricular tachycardia. By contrast, no subjects in the lidocaine group developed a serious arrhythmia. Overall, the results demonstrated that “in the absence of marked hypoxia, respiratory or metabolic acidosis, hyperkalemia, or hypotension, serious cardiac arrhythmia occurs after bupivacaine but not lidocaine [administration].”5

The work of several Californian anesthesiologists made possible the discovery of bupivacaine’s direct cardiotoxic effects. Their efforts led to significant changes in practice and spurred the development of safer alternative agents, including levobupivacaine and ropivacaine. More than 40 years later, the discussions that surrounded this story are still relevant today. Any medication we use can be potentially toxic with improper administration or at supratherapeutic doses. So how do we as a professional community agree on guidelines to promote an optimal margin of safety?


1. Feldman HS. Toxicity of Local Anesthetic Agents. In: Rice S, Fish K, eds. Anesthetic Toxicity. Boca Raton, FL: CRC Press; 1994:107-133.

2. Albright GA. Cardiac arrest following regional anesthesia with etidocaine or bupivacaine. Anesthesiology. 1979;51(4):285-7.

3. Writer WD, Davies JM, Strunin L. Trial by media: the bupivacaine story. Can Anaesth Soc J. 1984;31(1):1-4.

4. Shannon J, Douglas J. Thirty years after the bupivacaine controversy: what have we learned? Can J Anaesth. 2010;57(4):289-92.

5. Kotelko DM, Shnider SM, Dailey PA, et al. Bupivacaine-induced cardiac arrhythmias in sheep. Anesthesiology. 1984;60(1):10-8. 

Leave a comment