Share this article on:

Local Anesthetic Systemic Toxicity and Liposuction: Looking Back, Looking Forward

Weinberg, Guy MD

doi: 10.1213/ANE.0000000000001247
Editorials: Editorial

From the Department of Anesthesiology, University of Illinois at Chicago and Jesse Brown VA, Chicago, Illinois.

Accepted for publication January 8, 2016.

Funding: Departmental.

Conflict of Interest: See Disclosures at the end of the article.

Reprints will not be available from the author.

Address correspondence to Guy Weinberg, MD, Department of Anesthesiology, University of Illinois at Chicago, and Jesse Brown VA, 820 S Damen Ave., Chicago, IL 60612. Address e-mail to

The study by Klein and Jeske1 in this issue of Anesthesia & Analgesia uses recruited patients to address the question of safe dosing limits for tumescent lidocaine infiltration with and without subsequent liposuction. Novel features that distinguish this article from previous studies by the same author are a measure of the specific effect of liposuction on the pharmacokinetic profile of lidocaine used in the wetting solution and an estimate of the expected prevalence of attaining a toxic blood level for a given dose of tumescent lidocaine. The authors should be commended for undertaking such a challenging study and for adding useful information to the relevant literature. Importantly, the study also provides an opportunity to reconsider several topics relevant to patient safety: the long-standing issue of local anesthetic systemic toxicity (LAST) in liposuction, the ambiguities and limitations in recommended dosing of local anesthetic, generalizing about safety from a small study, and problems inherent in our individual and collective responses to rare events.

The experimental design was a crossover pharmacokinetic study of lidocaine plasma concentrations after tumescent infiltration of a dilute (0.09%) lidocaine solution comparing each patient’s results with and without liposuction. Lidocaine plasma concentrations were followed over 24 hours, and key metrics included peak concentration and area under the concentration curve. They specifically tested the hypothesis that lidocaine at doses much higher than the Food and Drug Administration’s (FDA) limit of 7 mg/kg provided in the package insert is reliably safe for tumescent infiltration and, in their words, would “represent a nonsignificant risk of harm.” Comparing treatment with and without liposuction allowed testing the corollary that liposuction removed sufficient lidocaine that the data from tumescent infiltration alone could not inform dosing limits for infiltration followed by liposuction. They also tested for adverse effects, a relationship between total dose and peak serum lidocaine concentration and sought to estimate a frequency with which a given dose would result in a peak concentration that exceeded 6 μg/mL, a presumed threshold for clinical toxicity.

The data are represented very accessibly in Figure 1 (taken from their Table 3), which shows the dot plots for all serum lidocaine concentrations for all patients and treatments. Nine patients received 45 mg/kg lidocaine, 3 received approximately 30 mg/kg, 1 received 19 mg/kg, and in all cases, volunteers received treatment both with and without subsequent liposuction. A subset of volunteers received another infiltration of approximately 22 mg/kg to obtain limited dose-response data for infiltration without liposuction. One patient received the larger lidocaine dose with liposuction but no crossover. Their key findings were that liposuction reliably removed a substantial amount of lidocaine, reducing the area under the concentration-time curves (and, presumably the fraction of total lidocaine) by approximately 30% of a given dose. The maximal serum concentration was also reduced by liposuction, although this comparison was made less robust by combining a range of doses to obtain means. A strong, linear relationship between dose and peak concentration was maintained for tumescent infiltration without liposuction (r2 = 0.85) but did not hold after liposuction. Neither did liposuction affect the time to achieving the maximal concentration that occurred on average at about 12 hours after treatment (range 8–24 hours). Notably, the highest lidocaine concentration measured was 4.4 μg/mL and all but 1 volunteer had measurable lidocaine levels at 24 hours, the highest at that time point being approximately 3 μg/kg. Notably, 1 volunteer with a body mass index of 20 reported nausea at 12 hours when his (peak) lidocaine level was 4.3 μg/mL.

Figure 1

Figure 1

It is useful to put this study in context. The premise dates to the earliest use of tumescent solution infiltration for liposuction, a method first described by Dr. Klein almost 30 years ago. Since then, he has been a leading voice for patient safety in plastic surgery. In a field that can sometimes resemble the wild West where some surgeons might use whatever dose and drug suits them that day, Dr. Klein has been steadfast, conveying in the literature the need for a reliable process that minimizes risk of LAST by formulating tumescent solution with lidocaine (not the more lipid-soluble local anesthetics), adding dilute epinephrine to minimize absorption, limiting total anesthetic dose, and preparing for resuscitation. Nevertheless, in a seeming paradox, he has also insisted for decades that the FDA package insert is wrong to apply a 7 mg/kg limit to lidocaine in tumescent infiltration, arguing logically and consistently that, at least in this context, much higher doses are acceptable. His call in 19962 for a rational approach to dosing for tumescent infiltration prefigures almost verbatim the Introduction and Discussion of the current article. Moreover, the conclusions in this article support Dr. Klein’s previous stance from as early as 1990 that large doses of lidocaine (e.g., 35 mg/kg) are safe and prudent for tumescent infiltration.3 Nor is the delayed (approximately 12 hours) peak in lidocaine blood concentration a new finding; this provided a rationale as early as 1990 for using lidocaine in place of longer-acting, more cardiotoxic local anesthetics in formulating the wetting solution.4 Thus, the current article is the most recent effort in a long push to have the FDA reconsider their published limits for lidocaine dosing. One wonders what additional studies the FDA would require to modify their stance.

Historically, this argument mirrors a similar long-standing dispute among anesthesiologists about the rationale and value of publishing guidelines proposing maximal local anesthetic dosing limits. Rosenberg et al.5 pointed out in 2004 that these recommended maximal doses are not evidence based and fail to take into account the site of injection and various patient factors such as age and organ failure. This editorial led for calls on the one hand to “abandon blanket maximum recommended doses of local anesthetics”6 and provoked understandable alarm among others who were concerned that this stance left the door open to unlimited dosing.7 In this sense, I think Dr. Klein has it right. Rational dosing limits are plausible and helpful but must be evidence based, taking context into account and still recognizing that setting limits does not completely protect one from harm. The story does not end there.

The authors touch all the right bases in describing their study’s limitations (e.g., small sample size, sampling error, an unrepresentative study population, and unique surgical system). Nevertheless, I am concerned that their overall tone in this regard is perfunctory, that their message of “…a nonsignificant risk of harm to patients” is dismissive and that this message is easily misinterpreted to mean that clinicians can perform tumescent infiltration with a carefree attitude, blind to residual dangers. Even the authors exhibit a bit of denial in this regard. They describe the patient experiencing nausea 12 hours after the operation, and 2 sentences later state that, “…adverse events with the large doses of lidocaine with epinephrine are infrequent.” This patient received the higher lidocaine dose, attained the second highest blood lidocaine concentration in the study, had a known risk factor for LAST (low body mass index), and experienced a complaint consistent with LAST prodrome, all making it more likely in my view that the symptoms were connected to tumescent lidocaine infiltration. Therefore, I draw the opposite conclusion: a rate of 2.4% to 9.1% suggests that LAST in this setting is not so rare an event (the range reflects the choice of denominators: 1/41 treatments, 1/21 treatments with 45 mg/kg, or 1/11 treatments with 45 mg/kg but no liposuction). Moreover, their small sample size inherently limits conclusions about the prevalence of a complication that was not observed and certainly does not excuse ignoring or minimizing one that was; note that even a mild complaint like nausea can be a sentinel of worsening LAST.

Minimizing the impact of rare events plays into our natural tendency to practice experiential or anecdotal medicine. That is, we tend to ignore risks until we experience them.8 This common behavior abounds in and outside medicine. Every time a driver tailgates or does not wear a seatbelt but gets home without suffering harm, an essentially unsafe behavior is encouraged, even if the driver knows better. The authors assure us repeatedly that using their guidance for dosing minimizes risk, but their benchmark for an acceptable level of safety is 1 event per 2000 procedures. How does that compare with the risk of driving unsafely? Does their conclusion effectively suggest we should drive without a seatbelt? Recognizing rare events requires our improving systems to reduce those risks and Dr. Klein’s work makes a very useful contribution in providing guidance for 1 aspect of that. Nevertheless, if one accepts the 1/2000 risk rate (or, as the one symptomatic volunteer suggests, possibly a much higher risk rate), then there is an obvious need for us to acknowledge and prepare for such events, they do occur. The question of assigning an “acceptable” rate of risk is tied closely to the controversy around recommending a maximal allowable dose of local anesthetic. Both topics, like taste, can be reduced to a matter of individual preference or (risk) tolerance and cannot be precisely defined. However, a graph (Figure 1, taken from their Table 3) of the calculated risk (Cmax > 6 μg/mL) versus dose (without liposuction) has an elbow indicating that doses >40 mg/kg carry an “increased” risk. Unfortunately, such a plot does not include outliers with extreme sensitivity to LAST, and those patients provide that adverse events can likely never be completely eliminated.

Many reviews attest to the overall safety of liposuction. Nevertheless, others confirm that there are risks involved. A sentinel article by Rao et al.9 pointed to the potential for fatal LAST in these procedures. General anesthesia may have contributed to these events, because other studies indicate that fatalities during tumescent liposuction occur almost exclusively when done under general anesthesia. However, based on the doses alone (10–40 mg/kg), Klein and Jeske1 maintain that their data make it unlikely the 4 fatalities reported by Rao et al. were related to lidocaine toxicity. This blanket statement ignores the fact that some patients are extremely sensitive to LAST or the possibility that other surgeons do not practice with the same level of technical expertise as Dr. Klein. My personal interest in LAST dates to a near miss in just such an outlier: an index patient who developed life-threatening arrhythmias after a tiny dose of local anesthetic (albeit bupivacaine) delivered by tumescent infiltration for liposuction.10 The patient was found to have severe carnitine deficiency, and this has subsequently been confirmed to substantially lower the threshold for LAST-induced cardiac toxicity.11 Such patients are not always identifiable before surgery, and along with sporadic system failures and physician errors, they make it impossible to eliminate completely the risk of LAST in what is generally a very safe operation. Moreover, it is likely that cardiac arrest related to LAST during liposuction is underreported or underdiagnosed. For instance, a questionnaire study by Lehnhardt et al.12 described 72 severe (including 23 fatal) complications of liposuction in Germany between 1998 and 2002. Six patients experienced cardiac arrest and 4 of these had only tumescent anesthesia or tumescent anesthesia with sedation. In none of those cases was documentation sufficient to identify a cause of the cardiac arrest. Are we safe to assume they were not related to LAST?

Klein and Jeske have done good work to establish a safe dosing strategy for tumescent local anesthetic infiltration. However, it is time to change the focus of the discussion from defining acceptable practice to acknowledging the continued risk for rare events like LAST. Then, we can take necessary steps to identify other system improvements to further minimize those risks and prepare to manage such problems when they occur.

Back to Top | Article Outline


Name: Guy Weinberg, MD.

Contribution: This author helped write the manuscript.

Conflict of Interest: Guy Weinberg is a co-founder of ResQ Pharma, Inc. Dr. Weinberg was awarded a US patent related to lipid resuscitation and established the educational Web site:

Attestation: Guy Weinberg approved the final manuscript.

This manuscript was handled by: Ken B. Johnson, MD.

Back to Top | Article Outline


1. Klein J, Jeske D. Maximum safe dosages of tumescent lidocaine. Anesthesia Analgesia. 2016;122:1350–9
2. Klein JA. Tumescent technique for local anesthesia. West J Med. 1996;164:517
3. Klein JA. Tumescent technique for regional anesthesia permits lidocaine doses of 35 mg/kg for liposuction. J Dermatol Surg Oncol. 1990;16:248–63
4. Klein JA. The tumescent technique. Anesthesia and modified liposuction technique. Dermatol Clin. 1990;8:425–37
5. Rosenberg PH, Veering BT, Urmey WF. Maximum recommended doses of local anesthetics: a multifactorial concept. Reg Anesth Pain Med. 2004;29:564–75
6. Heavner JE. Let’s abandon blanket maximum recommended doses of local anesthetics. Reg Anesth Pain Med. 2004;29:524
7. Reynolds F. Maximum recommended doses of local anesthetics: a constant cause of confusion. Reg Anesth Pain Med. 2005;30:314–6
8. Barron G, Leider S, Stack J. The effect of safe experience on a warnings’ impact: Sex, drugs, and rock-n-roll. Organ Behav Hum Decis Process. 2008;106:125–42
9. Rao RB, Ely SF, Hoffman RS. Deaths related to liposuction. N Engl J Med. 1999;340:1471–5
10. Weinberg GL, Laurito CE, Geldner P, Pygon BH, Burton BK. Malignant ventricular dysrhythmias in a patient with isovaleric acidemia receiving general and local anesthesia for suction lipectomy. J Clin Anesth. 1997;9:668–70
11. Wong GK, Crawford MW. Carnitine deficiency increases susceptibility to bupivacaine-induced cardiotoxicity in rats. Anesthesiology. 2011;114:1417–24
12. Lehnhardt M, Homann HH, Daigeler A, Hauser J, Palka P, Steinau HU. Major and lethal complications of liposuction: a review of 72 cases in Germany between 1998 and 2002. Plast Reconstr Surg. 2008;121:396e–403e
© 2016 International Anesthesia Research Society