Intravenous Lipid Emulsion as Antidote Beyond Local Anesthetic Toxicity: A Systematic Review


  • Grant Cave MBChB, FACEM,

    1. From Hutt Hospital (GC), Lower Hutt, New Zealand; the University of Otago, Wellington School of Medicine and Health Sciences (GC), Wellington, New Zealand; Monash Medical Centre (GC), Clayton Road, Melbourne, Victoria, Australia; and the Department of Emergency Medicine, Waikato Hospital (MH), Hamilton, New Zealand.
    Search for more papers by this author
  • Martyn Harvey MBChB, FACEM

    1. From Hutt Hospital (GC), Lower Hutt, New Zealand; the University of Otago, Wellington School of Medicine and Health Sciences (GC), Wellington, New Zealand; Monash Medical Centre (GC), Clayton Road, Melbourne, Victoria, Australia; and the Department of Emergency Medicine, Waikato Hospital (MH), Hamilton, New Zealand.
    Search for more papers by this author

  • Presented at The Joint Faculty of Intensive Care Medicine ASM, Brisbane, Australia, June 12–14, 2009.

Address for correspondence and reprints: Martyn Harvey, MBChB, FACEM; e-mail:


Objectives:  The objective was to asses the efficacy of lipid emulsion as antidotal therapy outside the accepted setting of local anesthetic toxicity.

Methods:  Literature was accessed through PubMed, OVID (1966–February 2009), and EMBASE (1947–February 2009) using the search terms “intravenous” AND [“fat emulsion” OR “lipid emulsion” OR “Intralipid”] AND [“toxicity” OR “resuscitation” OR “rescue” OR “arrest” OR “antidote”]. Additional author and conference publication searches were undertaken. Publications describing the use of lipid emulsion as antidotal treatment in animals or humans were included.

Results:  Fourteen animal studies, one human study, and four case reports were identified. In animal models, intravenous lipid emulsion (ILE) has resulted in amelioration of toxicity associated with cyclic antidepressants, verapamil, propranolol, and thiopentone. Administration in human cases has resulted in successful resuscitation from combined bupropion/lamotrigine-induced cardiac arrest, reversal of sertraline/quetiapine-induced coma, and amelioration of verapamil- and beta blocker–induced shock.

Conclusions:  Management of overdose with highly lipophilic cardiotoxic medications should proceed in accord with established antidotal guidelines and early poisons center consultation. Data from animal experiments and human cases are limited, but suggestive that ILE may be helpful in potentially lethal cardiotoxicity or developed cardiac arrest attributable to such agents. Use of lipid emulsion as antidote remains a nascent field warranting further preclinical study and systematic reporting of human cases of use.

A convincing case has been made for the use of lipid emulsion in local anesthetic toxicity. In 1998, Weinberg et al.1 in a preliminary study demonstrated pretreatment or resuscitation with intravenous lipid emulsion (ILE) resulted in amelioration of bupivacaine-induced cardiotoxicity in rats. The same authors later confirmed these findings in intact dogs.2 Commensurate with the body of supportive animal work,1–4 human application of ILE has resulted in neurologically intact recovery in anecdotal case reports of cardiac arrest and neurologic toxicity attributable to local anesthetic agents.5–11 Use of lipid emulsion as antidote in this clinical scenario has subsequently gained endorsement in professional body guidelines for severe poisoning.12,13 The case for lipid as therapy in local anesthetic–induced cardiac arrest is made more compelling given that successful resuscitation in reported cases has invariably followed failure of conventional resuscitative measures, in effect, a positive outcome from an otherwise irretrievable situation.

Three postulated mechanisms have been forwarded to explain the beneficial effect of lipid.4 The first suggests establishment of a new pharmacokinetic equilibrium within an expanded plasma lipid phase, reducing free drug levels and thereby toxicity. Second, bupivacaine is known to inhibit carnitine acyltransferase, essential in transport of fatty acids across the inner mitochondrial membrane. ILE may overcome this inhibition and the metabolic effects of other similar metabolic poisons through “mass effect” alone or via some as yet unknown mechanism on preferred myocardial substrate. Third, fatty acids are known to increase cardiac myocyte calcium levels and may act via a direct inotropic action through increasing intramyocyte calcium concentration.14 This hypothesis may be of particular relevance to overdose with calcium channel blocking agents. All postulates permit the cogent hypothesis that the beneficial effect demonstrated in local anesthetic toxicity may generalize to agents of similar lipophilicity and/or inhibition to mitochondrial lipid metabolism.

Experimental evidence since the 1970s has demonstrated interaction between non–local anesthetic lipophilic drug moieties and lipid emulsions.15–17 Evaluation of lipid emulsions as therapy in resuscitation from lipophilic drug–induced cardiovascular collapse has nevertheless lagged behind that of local anesthetics. Recent editorial-level interest has focused attention on the potential benefit of ILE in cardiac arrest attributable to other lipophilic drugs.18,19 The objective of this review therefore was to determine the efficacy of ILE in animal models of poisoning and describe the outcomes associated with ILE therapy in poisoned humans, beyond the accepted setting of local anesthetic toxicity. A systematic review of the literature was the elected methodology.


Search Strategy

A comprehensive search of the literature was carried out to answer the research question, “Does the evidence support administration of ILE as antidote in lipophilic drug toxicity, beyond that of local anesthetics?” The research question was formulated as per QUOROM guidelines.20 The definition of ILE was taken to be that outlined in guidelines for use in local anesthetic toxicity.12,13

We defined the scope of this review to include articles describing the administration of lipid emulsion preparations as treatment in lipophilic drug toxicity in animal or human subjects. Specific inclusion criteria for animal studies were evidence of prior randomization in controlled evaluations of lipid emulsion, with either survival or physiologic parameters relevant to toxicity as endpoints. Studies were preferred in accordance with the usual hierarchy of evidence. An evaluation schema for prioritizing findings of animal studies is presented in Appendix 1. Any study or report describing human subjects treated with lipid emulsion in non–local anesthetic lipophilic drug poisoning was included.

Literature Search

MEDLINE-indexed literature was accessed utilizing the search engines PubMed, OVID (1966–February 2009), and EMBASE (1947–February 2009). The search involved both free text and Medical Subject Headings (MeSH) terms and included “intravenous” AND [“lipid emulsion” OR “fat emulsion” OR “Intralipid”] AND [“toxicity” OR “arrest” OR “rescue” OR “resuscitation” OR “antidote”]. Reports of animal and human studies were included. Articles pertaining to local anesthetic agents were excluded. Publications in any language were considered. Conference proceedings from the North American Congress of Clinical Toxicology (NACCT) and European Association of Poison Centers and Clinical Toxicology meetings were additionally hand searched for the years 2005 to 2008.

Two reviewers (GC, MH) independently examined results returned from the search to identify potentially relevant abstracts. Articles that clearly did not meet the review criteria according to title and abstract were not considered further. When the reviewers disagreed, a consensus was reached through discussion. The same two reviewers independently examined the full-text articles to determine which studies met inclusion guidelines. The bibliographies of identified papers were examined and cross-referenced to identify additional papers of relevance.

The following data were independently abstracted by the same reviewers: author, lipophilic toxin, and experimental model utilized in the case of animal study, relevant comparison treatment, outcome and method of measurement, type of analysis, and study findings. Human case reports were summarized to include causative agent(s), treatments administered, relevant investigations, time course, and patient outcome.


The literature search identified 149 potentially relevant citations. Electronic databases yielded 145 articles, and four were found with targeted hand searching. A total of 128 articles were excluded following review of abstracts. Following independent assessment of the full-text articles, an additional two articles were excluded (one animal and one human experiment) on the basis that phospholipid, rather than lipid emulsion, was studied. Therefore, 14 relevant animal studies, one human study, and four human case reports were included in this review. Studied agents included tricyclic antidepressants, verapamil, propranolol, atenolol, thiopentone, and organophosphates. A number of human case reports were additionally found via the educational Web site (G. Weinberg, MD) but due to the incomplete and non–peer-reviewed nature of these reports, they are not included in this review.

Animal Studies

Cyclic Antidepressants.  Lipid has been evaluated in four experimental models of tricyclic antidepressant toxicity. In a preliminary study, Yoav et al.21 demonstrated significantly lower mortality in rats administered clomipramine in an ILE vehicle when compared with saline control. Bania and Chu22 have further assessed the effect of ILE pretreatment on hemodynamic metrics and survival in amitriptyline toxic rats. A nonsignificant improvement in mean arterial pressure (MAP) and survival time for ILE-treated animals was observed; the authors concluded that an effect of the magnitude powered for (20 mm Hg MAP differential) had not been demonstrated.

Harvey and Cave23 have demonstrated accelerated recovery from clomipramine-induced hypotension when compared with both saline control and sodium bicarbonate in a whole rabbit model. Thirty sedated, instrumented rabbits were infused with clomipramine to target MAP 50% of baseline MAP. Animals were resuscitated with 12 mL/kg 20% Intralipid, 3 mL/kg 8.4% sodium bicarbonate, or 12 mL/kg 0.9% saline solution. Mean differences in MAP between Intralipid and bicarbonate groups were 19.4 mm Hg (95% confidence interval [CI] = 18.8 to 27.0 mm Hg) and 11.5 mm Hg (95% CI = 2.5 to 20.5 mm Hg) at 5 and 15 minutes, respectively. In the second phase of the experiment, eight rabbits underwent clomipramine infusion to target MAP of 25 mm Hg before receiving 8 mL/kg 20% Intralipid or 2 mL/kg 8.4% sodium bicarbonate as rescue therapy. External cardiac compressions and intravenous (IV) adrenaline were administered in the event of cardiovascular collapse. Spontaneous circulation was maintained in all ILE animals (n = 4). All saline-treated rabbits developed intractable pulseless electrical activity (n = 4; p = 0.023). Limitations of this study include lesser sodium channel blocking effect of clomipramine compared with more typical tricyclic antidepressants and the relatively low dose of sodium bicarbonate (NaHCO3) employed (3 and 2 mEq/kg in Phases 1 and 2, respectively).

In a subsequent experiment, the same investigators reproduced comparable hemodynamic improvement with application of intravenous fat emulsion (IFE) in a similar rabbit model of clomipramine poisoning.24 An elevated serum clomipramine concentration in the ILE group was observed contemporaneous with hemodynamic recovery. Pharmacokinetic analysis of clomipramine decay curves demonstrated a reduced initial volume of distribution of clomipramine in the IFE-treated group (15.9 L/kg vs. 5.7 L/kg; p = 0.0001). Peritoneal dialysis with 20% Intralipid commensurate with IV treatment furthermore resulted in a 16-fold increase in clomipramine extraction ratio for the lipid group (p = 0.009), albeit total clomipramine extraction was low in both study arms.

Verapamil.  In a preliminary study, Tebbutt et al.25 demonstrated co-infusion of Intralipid 20% in a rat model of verapamil toxicity to nearly double the LD50 and attenuate verapamil-induced bradycardia. Similarly, Bania and co-workers26 evaluated addition of ILE to standard resuscitative measures in a canine model of severe verapamil toxicity. Fourteen instrumented dogs were infused with verapamil until MAP was 50% baseline and were maintained at this level for 30 minutes. Intravenous verapamil infusion was subsequently continued at 2 mg/kg/hr. All dogs were given atropine and three doses of calcium chloride. Test animals additionally received 7 mL/kg 20% Intralipid, and control dogs a similar volume 0.9% saline solution. ILE animals demonstrated significantly greater MAP at 30, 45, and 60 minutes postrescue. One of seven saline-treated dogs survived to 120 minutes versus all seven ILE-treated animals (p = 0.01). Hemodynamic parameters (cardiac output, MAP) and partial pressure of oxygen were additionally superior in the ILE group. Limitations of this study included the ability to generalize from an IV protocol to the clinical scenario of enteral overdose and low survival in the control group, rendering comparison of hemodynamic and respiratory metrics problematic.

Subsequent to this, Perez et al.27 sought a dose–response relationship using a dose escalations study of ILE in a murine model of verapamil toxicity. Incremental ILE (6.2–37.6 mL/kg 20% ILE) was administered 5 minutes into continuous verapamil infusion in sedated rats. The maximal beneficial effect for ILE on survival was seen at 18.6 mL/kg (median survival time for controls = 37 minutes vs. 143 minutes for ILE; p = 0.0004), with greatest benefit to heart rate and MAP at 24.8 mL/kg. The authors posited some adverse effect for ILE at very high doses. Limitations of this study are the IV mode of verapamil delivery, which may not generalize to the enteral overdose situation, single bolus administration of ILE, and small numbers (n = 5 animals per group) studied.

The same authors performed an investigative study to elucidate the potential mechanism of action for lipid. Rats were infused with oxfenicine, an inhibitor of mitochondrial free fatty acid oxidation or placebo before IV Intralipid 20% at 15 mL/kg.28 Subsequent verapamil infusion was continued until death. No significant difference was observed in pulse rate, MAP, or median survival times between groups suggesting minimal contribution from any postulated augmentation of lipid metabolism. No oxfenicine-alone arm was included in this study, which to date has only been published in abstract form.

Beta Blockers.  In one preliminary study, Cave et al.29 have demonstrated amelioration of propranolol-induced ECG QRS prolongation and a trend toward attenuation of propranolol-induced bradycardia, in rats pretreated with ILE prior to propranolol infusion. A nonsignificant increase in survival time in the ILE-treated group was observed. Similarly, Bania et al.30 have demonstrated ILE pretreatment to increase MAP, but not affect heart rate, in rats undergoing continuous propranolol infusion compared with saline control.

Improved hemodynamic performance following lipid infusion was observed in a study of propranolol toxicity in whole rabbits by Harvey and Cave.31 Twenty sedated, mechanically ventilated rabbits underwent propranolol infusion at 4.2 mg/min to target MAP 60% baseline before resuscitation with 6 mL/kg IV 20% Intralipid or 6 mL/kg 0.9% saline solution over a 4-minute period. MAP proved greater in the Intralipid-treated group at termination of the monitoring interval (69 mm Hg Intralipid vs. 53 mm Hg saline; p = 0.029). No significant difference in pulse rate or rate of change of MAP was observed. Limitations of the study include rapid induction of hypotension, the absence of ongoing propranolol infusion as mimic for enteral overdose, and brief monitoring interval following rescue (15 minutes). In a similar model of atenolol-induced hypotension, no significant difference in MAP was observed following ILE infusion.32 The absence of observed signal in this study provides some inferential support for the “lipid sink” hypothesis, as significant lipid sequestration of the hydrophilic atenolol is improbable.

Exploratory work with additional lipophilic drug moieties has met with varying results. Thiopentone-induced bradypnea in rats showed greater improvement after ILE compared with saline as rescue therapy in one pilot study.33 Evaluation of intraperitoneal administration of lipid emulsion furthermore failed to increase the dose at which 50% were killed in mice undergoing lipid-soluble organophosphate poisoning with paraoxon.34 A summary of animal studies is presented in Table 1.

Table 1. 
Summary of Animal Studies
First Author; YearAgentSpeciesILE InterventionResults
  1. BP = blood pressure; ILE = intravenous lipid emulsion; MAP = mean arterial pressure; VD = volume of distribution.

Yoav; 200221ClomipramineRatILE vs. saline; co-infusion to deathIncreased survival in ILE group
Cave; 200533ThiopentoneRatILE vs. saline; rescueMore rapid reversal bradypnea in ILE
Bania; 200534ParoxonRatIntraperitoneal LE vs. salineNo difference survival
Bania; 200622AmitriptylineRatILE vs. salineNonsignificant increase in MAP in ILE
Cave; 200629PropranololRatILE vs. salineReduced QRS duration, trend to attenuation bradycardia
Bania; 200630PropranololRatILE vs. saline pretreatmentIncreased MAP in ILE
Tebbutt; 200625VerapamilRatILE vs. saline; co-infusion to deathIncreased LD50 in ILE group; reduction verapamil induced bradycardia in ILE
Harvey; 200723ClomipramineRabbitILE vs. bicarbonate vs. saline in hypotension; ILE vs. bicarbonate rescue before cardiac arrestGreater BP increment with ILE; greater survival in ILE group
Bania; 200726VerapamilDogILE vs. saline + atropine and calciumGreater survival in ILE; greater MAP in ILE
Bania; 200728VerapamilRatILE vs. saline in oxfenicine pretreatmentNo difference in pulse rate, MAP, or survival
Harvey; 200831PropranololRabbitILE vs. saline in hypotensionGreater MAP recovery in ILE
Harvey; 200924ClomipramineRabbitILE vs. saline in hypotension; pharmacokinetic evaluation clomipramine decayGreater MAP recovery in ILE; elevated serum clomipramine concentration and reduced Vd in ILE
Harvey; 200932AtenololRabbitILE vs. saline in hypotensionNo difference in MAP
Perez; 200927VerapamilRatILE dose rangingGreatest survival benefit @ 18.6 mL/kg; greatest benefit BP at 24.8 mL/kg

Limitations of Animal Studies.  A recurring problem throughout the above studies was a lack of enteral administration of offending lipophilic agent. Investigators invariably elected IV administration of toxin titrated either to dose or to physiologic endpoint. In addition, the majority of studies represent experiments designed to demonstrate efficacy alone, rather than provide head-to-head comparison with established antidotal therapies. As such, generalization of findings to commonly encountered clinical scenarios of overdose by the oral route must proceed with some caution.

Human Study

There is only one controlled human study in this field. A 5-hour infusion of lipid suspension resulted in a statistically nonsignificant 14% increase in plasma amitriptyline and nortriptyline levels compared with saline control, in four healthy volunteers taking amitriptyline at therapeutic dose for 8 days.35 These results from a small population in pharmacologic steady state are unlikely to generalize to the clinical scenario of acute tricyclic antidepressant poisoning.

Case Reports.  Use of ILE was associated with successful resuscitation from arrest secondary to combined bupropion/lamotrigine enteric overdose in a case report by Sirianni et al.36 A 17-year-old female with bipolar affective and attention deficit disorder developed seizure activity progressing to cardiac arrest with ventricular fibrillation, 10 hours following self-poisoning with up to 7.95 g of bupropion and 4 g of lamotrigine. Standard cardiopulmonary resuscitation in addition to repeated DC cardioversion; administration of amiodarone, magnesium, sodium bicarbonate, and calcium chloride, and high-dose inotrope infusion; in addition to unsuccessful attempted transcutaneous pacing, failed to restore a sustained perfusing cardiac rhythm. Seventy minutes following initial arrest, a single 100-mL bolus of 20% Intralipid was administered IV, and approximately 1 minute later a sustained palpable pulse was observed. Over the subsequent 15 minutes, ECG QRS duration decreased, sinus rhythm was restored, and pressor requirements reduced. An additional period of ventricular tachycardia developed at 90 minutes, but resolved with adrenaline and 1 minute of chest compressions only. Elevated serum triglyceride levels following successful resuscitation were associated with increased plasma bupropion concentration. A lengthy pediatric intensive care unit (PICU) admission complicated by significant acute lung injury followed prior to eventual transfer to a rehabilitation facility. At PICU discharge, the patient was conversant and talkative, with slight tremor, mild memory deficits, and fine motor incoordination.

Intravenous lipid emulsion has been used to reverse coma caused by sertraline and quetiapine overdose.37 The dosages taken were in an order associated with threat to life. A 61-year-old male presented to hospital 3.5 hours after ingestion with a patent airway, oxygenating and ventilating satisfactorily on supplemental oxygen, and a blood pressure of 86/64 mm Hg. His Glasgow Coma Scale (GCS) score was 3, and he proved unresponsive to flumazenil. Electrolytes, glucose, and ECG (sinus rhythm 80 beats/min with normal intervals) were within normal limits. ILE was administered as per the Association of Anaesthestists of Great Britain and Ireland (AAGBI) guidelines, with simultaneous improvement in level of consciousness to GCS score 9, averting intubation in this patient. While transfer to the nearest ICU (in this instance some distance away) was avoided, utilization of lipid infusion with the intention of negating intubation and airway protection cannot currently be recommended. An accompanying editorial discussed thresholds for use of ILE in this setting.19

Reversal of cardiovascular collapse has been reported in an abstract, with repeated-dose ILE given to a patient with combined verapamil and atenolol overdose.38 A 52-year-old comatose patient with verapamil/atenolol-induced shock refractory to conventional pharmacologic therapy received two doses of ILE as per AAGBI guidelines. Within minutes of initial application, the patient awoke and attempted to talk. Blood pressure and peripheral perfusion improved. Eight hours later the patient again became shocked, and evidenced further improvement in hemodynamic metrics following repeated ILE bolus. An intraarterial balloon pump was subsequently inserted, but the patient eventually died 12 hours after the second lipid infusion.

Another abstract reports ILE being used to treat a patient with atenolol poisoning.39 The ingestion was reportedly massive, although not quantified. Initial hemodynamics (pulse 32/min, systolic blood pressure 62 mm Hg) improved after initial conventional pharmacologic treatment, to pulse 43/min and blood pressure 96/53 mm Hg. One liter of ILE was administered over the course of 1 hour with further improvement to pulse 70/min and blood pressure 110/72. The patient survived. A summary of human case reports is presented in Table 2.

Table 2. 
Summary of Human Case Reports
First Author; YearAgentClinical ScenarioILE InterventionOutcome
  1. GCS = Glasgow Coma Scale score; ROSC = return of spontaneous circulation.

Sirianni; 200836Bupropion/lamotrigineIntractable cardiac arrest100 mL 20% Intralipid bolusROSC, patient survived.
Dolcourt; 200838Verapamil/atenololShock, coma1.5 mL/kg 20% Intralipid bolus repeatedInitial improvement in GCS and hemodynamic metrics before demise 12 hours later.
Harchelroad; 200839AtenololShock1000 mL ILE over 60 minutesAugmentation of blood pressure after conventional therapies, patient survived.
Finn; 200937Sertraline/quetiapineComa1.5 mL/kg 20% Intralipid bolus, 6 mL/kg infusionReversal of coma, intubation avoided, patient survived.


Literature identifies ILE as a potentially beneficial antidote in cardiovascular collapse or cardiac arrest secondary to overdose of highly lipophilic cardiotoxic medications, beyond the currently accepted indication of local anesthetic poisoning. Reported data provide evidence of efficacy, and plausible mechanism of action, in both intact animal models and human subjects.

Experimental work supports a causal relationship for the effects of ILE in targeted non–local anesthetic lipophilic drug poisoning. Of the 13 animal studies reviewed, eight have positive findings for the primary outcome variable,21,23–27,30,31 with two additional studies showing positive effect for secondary or post hoc endpoints.29,33 The remaining studies were small and powered to find large effects only.21,27,31,33 Additionally, evidence of consistency of effect is found in the hierarchical nature of studies for specific agents—the effect in earlier pilot murine studies was replicated in canine or rabbit models of verapamil, propranolol, and clomipramine cardiotoxicity. Strong effects were seen, both statistically and experimentally, in these more sophisticated models, showing a near “all or nothing” effect on survival.23,26 A strong temporal relationship between ILE administration and hemodynamic improvement was observed early in the experimental protocol in animal work on clompramine23,24 and propranolol.26 The only study dealing with dose response demonstrated positive findings for both survival and physiologic parameters.27

Findings in basic science work represent support for concept but do not establish efficacy in the clinical situation. Likewise, the dramatic recoveries observed in reported instances of human ILE application,36–39 although suggestive, fail to definitively identify ILE as causative in recovery. Attributing causality to lipid in the successful resuscitation from bupropion-induced cardiac arrest has been questioned given the confounding effect of numerous prior interventions.40 Furthermore, the submaximal sodium bicarbonate dosing prior to lipid application has raised doubt as to the perceived failure of standard therapy in this case of overwhelming sodium-channel blockade.40 The so-called SALT toxidrome (Shock, Altered mental status, Long QRS, and Terminal R in AVR) has been proposed by Lester41 to encapsulate manifestations of sodium channel blocker toxicity irrespective of offending agent. While there are differing additional mechanisms of toxicity between causative agents, local anesthetics are a unique cause of this toxidrome only inasmuch as recommendations for ILE presently pertain to them alone. With this to contextualize, and the presence of a strong temporal relationship (return of spontaneous circulation 1 minute after ILE 52 minutes after she last had a pulse), on balance of probabilities it appears that ILE was a lifesaving intervention for this patient.

Similarly, application of ILE in coma secondary to sertraline/quetiapine overdose was associated with rapid restoration or protective airway reflexes. The strong temporal relationship at a time (4 hours postingestion), when no other probable explanation for improvement avails, again suggests a causal role for ILE in the patient’s improvement.

Limited support for lipid may be drawn from the temporal association of hemodynamic improvement with application in the cases of verapamil/atenolol- and atenolol-induced cardiovascular collapse reported in NACCT abstracts. Attributing recovery to lipid in these cases is clouded by multiple simultaneous therapies, brevity of description of time course, and/or outcome in reporting, and in the instance of atenolol the hydrophilic nature of this beta blocker precluding significant expected benefit from lipid sequestration.

Mechanism.  Reestablishing equilibrium within an expanded plasma lipid phase (sink) with subsequent reduction in free drug levels has been proposed as one beneficial mechanism of action of lipid application. This hypothesis is supported by the commonality of high lipid solubility for agents demonstrated to respond to lipid infusion (lipid/aqueous partition coefficient: bupivacaine log P 3.64,4 bupropion log P 3.47,36 clomipramine log P 4.71,42 propranolol log P 3.65,43 verapamil octanol:ringer partition coefficient 67.844). In vitro studies have demonstrated high binding capacity of long-chain lipid emulsions for lipophilic local anesthetics in buffered solution45 and bupivacaine sequestration to an induced plasma lipid phase at a 12:1 ratio in rat plasma.1 Similarly, addition of phospholipid-containing liposomes to bovine serum reduced free amitriptyline concentration by a factor of 3.5.46 Accelerated myocardial bupivacaine washout has been demonstrated in isolated rat hearts commensurate with lipid perfusion.47 Perfusion of amitriptyline-intoxicated isolated rat hearts with phospholipid-based nanovesicles has likewise resulted in more rapid recovery in coronary perfusion pressure,48 suggesting augmented cardiac elimination. Long circulating lipid-containing nanocapsules have additionally been shown to sequester haloperidol, docetaxel, and paclitaxel in vitro and to increase blood levels of docetaxel due to in vivo drug sequestration.49

En vivo pharmacokinetic analysis of clomipramine decay curves following lipid application in whole rabbits has demonstrated reduced initial volume of distribution and elevated plasma clomipramine concentration consistent with intravascular lipid sequestration.24 Drug–lipid partitioning may be additionally inferred in the case report of combined bupiopion/lamotrigine overdose, wherein elevated serum bupropion levels were associated with elevation of serum triglyceride following ILE.36 Conversely, Litz et al.11 report serum levels of mepivacaine decreased more rapidly after lipid infusion than predicted by the published half-life, suggesting increased distribution or partitioning to alternate lipid stores.

Augmented myocardial free fatty acid uptake and utilization with increased myocyte high-energy phosphate production may additionally contribute to improved hemodynamic performance. A lipid bolus may, via mass action alone or by some as yet unknown mechanism, serve to overcome the impediments to mitochondrial oxidative phosphorylation known to be associated with bupivacaine,50 tricyclic antidepressants,51 and verapamil.52 Consistent with this hypothesis, Stehr et al.53 have reported a significant increase in cardiac performance in a study of bupivacaine toxicity in isolated rat hearts undergoing lipid perfusion, at concentrations too low to provide a lipid sink effect. Conversely, oxfenicine versus placebo pretreatment resulted in no difference in MAP or survival in an intact rat model of verapamil toxicity resuscitated with Intralipid,28 suggesting no beneficial metabolic effect for lipid infusion in this model.

Finally, ILE may restore myocardial function by increasing intracellular calcium concentration. Application of both saturated and unsaturated long-chain free fatty acids has been demonstrated to activate voltage-gated calcium channels in myocardial isolates.14,54 Augmentation of cardiac performance via this mechanism may be of particular importance in calcium channel blocker intoxication. Review of potential mechanisms suggests that the most fruitful policy for use of ILE as antidote is to limit consideration to highly lipophilic agents with potentially lethal cardiotoxicity.

Safety and Clinical Use.  Uptake of lipid emulsion by anesthesia departments worldwide has proven exponential.55,56 While certain parallels may be drawn between the initial experience of lipid in local anesthetic cardiotoxicity, and that of similarly lipophilic non–local anesthetic agents, important differences exist. Perhaps most significant is the antidotal landscape into which lipid is incorporated. In the case of local anesthetics, lipid emulsion provided a validated therapy for a feared complication of regional anesthesia, prior to which no antidote existed. Conversely, any recommendation for ILE in cardiotoxicity secondary to non–local anesthetic agents of high lipophilicity must incorporate both comparative analysis with established agent-specific antidotes and consideration of the potential deleterious effect of ILE on the efficacy of these therapies when administered concurrently. To date, many of the necessary comparative studies are yet to be performed.

Lipid emulsions have been associated with myriad adverse effects including allergic reactions, hyperthermia, thrombocytopenia, hypercoagulability, antineutrophil activity, pancreatitis, and elevated liver enzyme levels57 when infused as components of balanced parenteral nutrition regimens. Such adverse effects are likewise conceivable following use in resuscitation. Furthermore, the dosages described for lipid emulsion use in the acutely intoxicated patient (Table 3) are not the subject of clinical study, nor do they fall within manufacturer-recommended guidelines for the rate of infusion of lipid.58 There have been no demonstrated adverse events from the use of high-dose ILE as antidote in human cases to date, however. While this is reassuring, absence of evidence is not evidence of absence, and harm from use of high-dose ILE remains plausible.

Table 3. 
Suggested Lipid Emulsion Dosing
ReferenceBolus DosingInfusion Rate
  1. All infusions 20% Intralipid.

  2. *Association of Anaesthetists of Great Britain and Ireland: Guidelines for the Management of Severe Local-Anaesthetic Toxicity.12

  3. †Lipid Infusion Protocol. Guy Weinberg, MD:

AAGBI*1.5 mL/kg (repeat × 3)0.25 mL/kg/min rising to 0.5 mL/kg/min
Weinberg, MD†1.5 mL/kg (repeated for persistent asystole)0.25 mL/kg/min

Specific concerns have been raised about the potential for exacerbation or creation of an acute lung injury. The literature around the association between acute lung injury and ILE is not uniform in its findings, but does suggest that if an association exists it is transient.59 Of all reported cases of use of ILE as antidote in the peer-reviewed literature, only the case involving bupropion exhibited an acute lung injury, and this could be explained by prolonged shock alone. The only animal study measuring indices of gas transfer in the setting of large-dose ILE as antidote demonstrated improved oxygenation in the ILE group.26

The size of the lipid droplets in the blood has an impact of adverse effects for ILE. Droplets greater than 1 μm in size are phagocytosed by the reticular activating system and are more likely to cause microvascular blockage resulting in an inflammatory response. High-dose ILE will increase the tendency for coalescence of droplets with increased droplet size. Macrovascular embolization is also theoretically possible.

The pharmacokinetic effects of ILE may not prove universally favorable in toxicity or during attempted resuscitation. In oral overdose it is possible that an introduced intravascular lipid phase may actually increase enteric absorption of lipophilic toxin. Delayed toxicity is also theoretically possible if the toxin diffuses out or is released on metabolism of ILE.60 Consideration for use must furthermore take into account the potential deleterious interaction between ILE and established lipophilic drugs co-administered as therapy (e.g., amiodarone or insulin in therapeutic euglycemic hyperinsulinemia).

Two literature reports describe deleterious outcomes with lipid infusion in animal models of resuscitation. Mayr et al.61 report greater coronary perfusion pressure and increased survival rates in bupivacaine-induced cardiac arrest in a porcine model treated with vasopressin and adrenaline compared to ILE. This report is at variance with remaining animal reports in local anesthetic cardiotoxicity and may be explained by the concomitant utilization of asphyxia and bupivacaine in induction of asystole. Harvey et al.62 have additionally demonstrated lesser return of spontaneous circulation with ILE augmented resuscitation, compared with standard resuscitative measures alone, in a rabbit model of asphyxia-induced pulseless electrical activity. These authors implicate the adverse metabolic effects of augmenting lipid metabolism in the presence of significant hypoxia in the observed outcome.

The potential for adverse effects of lipid in asphyxia warrants attention in consideration of use. All available means to reverse asphyxia must be in place before administration. Limiting use to the setting of intractable arrest with significant whole body hypoxic/ischemic insults may preclude the maximal benefit from ILE. If such a situation could be anticipated on the basis of inexorable deterioration despite maximal therapy, earlier use of ILE in the clinical course may be warranted.

The expected benefit of ILE is difficult to quantify, but on the basis of limited animal and human data, this may range from minimal to life-preserving. Human cases of lipophilic drug poisoning presenting in extremis are, however, infrequent and do not lend themselves to prospective systemic evaluation. While awaiting human data, an approach to approval for antidotal therapy is the “animal efficacy, human safety” rule.63 We endorse the suggestion of Picard19 that the latter aspect be dealt with by individual clinicians and local ethics committees considering ILE as antidote (outside that of local anesthetic toxicity), only when patient deterioration is ongoing and life-threatening despite maximal available established therapies.

Given that case series are likely to provide a major route for refinement of dosing formulations and regimes, and documenting adverse events, it is probable that some cases particularly of unsuccessful or adverse use will not be widely disseminated. As knowledge in this nascent field progresses, guidance with a formal registry of all use, both successful and otherwise, could prove invaluable. A newly created registry of ILE use is accessible via the link


Management of overdose with highly lipophilic cardiotoxic medications should proceed in accord with established antidotal therapies and early consultation of a poisons center. Limited data from animal experiments and human cases suggest that intravenous lipid emulsion may be helpful in potentially lethal cardiotoxicity or developed cardiac arrest attributable to such agents. Use of lipid emulsion as an antidote remains a nascent field warranting further preclinical study and systematic reporting of human cases of use. A formal registry collating cases of lipid utilization could prove invaluable.

(Editor’s note added in proof: Very recently published reports include further examples of successful lipid-based resuscitation from toxic events related to verapamil,64 haloperidol,65 and moxidectin.66)


Appendix 1

System for evaluation of animal research. Criteria 1 and 2 are necessary for inclusion.

  • 1) Randomized, controlled trial?
    •  a. Evidence of similarity between groups
  • 2) Endpoints
    •  a. Survival
    •  b. Time to death
    •  c. Physiologic
  • 3) Powering
    •  a. Powered to mortality
  •  b. Powered to time to death
  •  c. Powered for physiologic effect
    •   i. Large effect, effect found
    •   ii. Small effect
    •   iii. Large effect, effect not found
    •   iv. Pilot
  • 4) Experimental model:
  •  Timing of ILE delivery

    •   i. Rescue
    •   ii. Pretreatment
    •   — Species, e.g., (dog > rabbit > rat)
    •   — Delivery of toxin
  • 5) Authors conclusions

Assessment of papers occurs along decreasing order of priority, e.g., a randomized controlled trial (RCT) with similar groups powered to large mortality difference using lipid as rescue in dogs carries more weight than an RCT with similar groups as a pilot measuring physiologic effects of ILE pretreatment in rats. ILE = intravenous lipid emulsion.