Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=ictx20

Download by: [American Academy of Clinical Toxicology] Date: 01 May 2017, At: 04:54

Clinical Toxicology

ISSN: 1556-3650 (Print) 1556-9519 (Online) Journal homepage: http://www.tandfonline.com/loi/ictx20

Drug-induced hyperlactatemia

Eike Blohm, Jeffrey Lai & Mark Neavyn

To cite this article: Eike Blohm, Jeffrey Lai & Mark Neavyn (2017): Drug-induced hyperlactatemia,Clinical Toxicology, DOI: 10.1080/15563650.2017.1317348

To link to this article: http://dx.doi.org/10.1080/15563650.2017.1317348

Published online: 27 Apr 2017.

Submit your article to this journal

Article views: 28

View related articles

View Crossmark data

REVIEW

Drug-induced hyperlactatemia

Eike Blohm, Jeffrey Lai and Mark Neavyn

Department of Emergency Medicine, Division of Toxicology, University of Massachusetts Medical Center, Worcester, MA, USA

ABSTRACTBackground: Hyperlactatemia is common in critically ill patients and has a variety of etiologies.Medication toxicity remains an uncommon cause that providers often fail to recognize. In this article,we review several medications that cause hyperlactatemia in either therapeutic or supratherapeuticdosing. When known, the incidence, mortality, pathophysiology, and treatment options are discussed.Methods: We performed a literature search using PUBMED and Google Scholar for English languagearticles published after 1980 regarding medication induced hyperlactatemia and its management. Oursearch string resulted in 798 articles of which 138 articles met inclusion criteria and were relevant tothe topic of our review.Conclusions: Hyperlactatemia is a relatively rare but life-threatening toxicity of various medicationclasses. Discontinuation of the drug is always advised, and some toxicities are subject to specific anti-dotal treatment. If there is no apparent medical cause for hyperlactatemia (sepsis, hypotension, hyp-oxia), clinicians should consider a toxicological etiology.

ARTICLE HISTORYReceived 9 January 2017Revised 3 April 2017Accepted 4 April 2017Published online 24 April2017

KEYWORDSLactic acidosis;hyperlactatemia; metforminassociated lactic acidosis;mitochondrial toxicity

Introduction

Hyperlactatemia commonly occurs in critically ill patients andserves as a clinical marker for illness severity and risk of mor-tality. While sepsis, cardiogenic shock, or hemorrhage are themost common causes of hyperlactatemia [1], a subset ofpatients have no apparent cause. Clinicians should considerdrug-induced hyperlactatemia in this patient group. This art-icle provides an overview of medications associated with ele-vated lactate concentrations, grouped by their underlyingpathophysiology of increased pyruvate production, metabol-ism to lactate, interference with the lactate metabolism (Coricycle), inhibition of mitochondrial protein synthesis, inhibitionof the electron transport chain (ETC), and uncoupling of oxi-dative phosphorylation. For reasons of clinical applicability,we include substances that interfere with the lactate assay toproduce a false elevation of serum lactate concentrations.Where available, we present incidence, treatment options,and data on morbidity and mortality.

Multiple medications cause hyperlactatemia by interfer-ence with cellular energy production. In order to meet theenergy demands of cells, glucose undergoes cytosolic gly-colysis, an oxygen-independent process that generates pyru-vate, adenosine triphosphate (ATP), and reducednicotinamide adenine dinucleotide (NADH). If oxygen is avail-able to the cell, pyruvate enters the Krebs cycle. Underanaerobic conditions, lactate dehydrogenase (LDH) reducespyruvate to lactate in order to regenerate oxidized nicotina-mide adenine dinucleotide (NADþ) so that glycolysis cancontinue.

Pyruvate dehydrogenase converts pyruvate to acetyl-CoAthat enters the mitochondrial Krebs cycle. The Krebs cycle isa series of enzymatic reactions that produces ATP, guano-sine-50-triphosphate (GTP), NADH, and succinate. NADHdonates two electrons to complex I of the ETC. As electronspass from complex I to complex III and complex IV, eachcomplex transfers hydrogen ions into the mitochondrial intra-membrane space. Succinate donates electrons to complex IIand contributes to the hydrogen gradient between the inter-membrane space and the mitochondrial matrix. The hydro-gen ion gradient provides the proton motive force for theproduction of ATP by complex V.

At physiologic equilibrium, endogenous lactate productionoccurs at a rate of 20mmol/kg/day. Metabolism of lactate iscentralized in the liver, which oxidizes approximately 70% oflactate via pyruvate to glucose (Figure 1); the remainderundergoes renal excretion. Glucose either re-enters the circu-lation if energy demand persists, or is otherwise stored asglycogen [2].

Hyperlactatemia results from an imbalance of lactate pro-duction and metabolism. The most common medical causeof hyperlactatemia is tissue hypoxia (type A hyperlactatemia)as is the case in severe sepsis, seizure, or trauma. In contrast,most toxicological causes are independent of tissue oxygenavailability (type B hyperlactatemia). This distinction isimportant as both type A and type B hyperlactatemia canpresent with hemodynamic compromise. A tenuous hemo-dynamic state is often the cause of type A hyperlactatemiabut typically a consequence of type B hyperlactatemia.

CONTACT Jeffrey Lai eike.blohm@gmail.com Department of Emergency Medicine, Division of Toxicology, University of Massachusetts Medical Center,Worcester, MA, USA� 2017 Informa UK Limited, trading as Taylor & Francis Group

CLINICAL TOXICOLOGY, 2017http://dx.doi.org/10.1080/15563650.2017.1317348

Therefore, hemodynamic stabilization does not suffice toresolve drug-induced hyperlactatemia.

Methods

We performed a target search of available English literatureon pubmed.gov and scholar.google.com from 1980 toAugust 2016. Search terms used included medication OR iat-rogenic OR albuterol OR epinephrine OR beta-agonist ORpropylene glycol OR lorazepam OR barbiturate OR phenobar-bital OR lactated ringers OR biguanide OR phenformin ORmetformin OR oxazolidinone OR linezolid OR eperezolid ORNRTIs OR stavudine OR sodium nitroprusside OR salicylateOR aspirin OR bismuth subsalicylate OR valproic acid ORethylene glycol OR propofol. We combined each search termwith overdose OR hyperlactatemia OR lactic acidosis OR tox-icity. Based on our clinical experience, we combined specificmedication names with the terms dialysis OR hemofiltrationOR extracorporeal removal OR antidote OR management. Wescreened 798 unique titles for relevancy. Abstracts that per-tained to the development, diagnosis, treatment, or progno-sis of hyperlactatemia from causes other than medicationeffect (e.g., sepsis) were excluded. We obtained full-lengtharticles where appropriate and accessible, for a total of 138articles. We followed citations and incorporated relevantinformation.

Results

Increased pyruvate production

Stimulation of beta-2 adrenergic receptors leads to mobiliza-tion of energy stores. In the adrenergic state, glycogenolysis,lipolysis, and gluconeogenesis significantly increase abovebaseline [3]. In the medical setting, administration of beta-2agonists occurs at supraphysiologic doses and with intent toease bronchoconstriction rather than meeting metabolicneeds. This results in energy supply surpassing energydemand and subsequently high tissue pyruvate concentra-tion [4]. Cells shunt pyruvate to lactate to regenerate thecell’s nicotinamide adenine dinucleotide (NADþ) pool.

Of particular clinical importance is hyperlactatemia frombeta-agonists in patients with severe acute asthma.

A volunteer study (n¼ 9) showed a rise in serum lactate con-centration from 1.1 to 2.3mmol/L in the absence of hypox-emia during the intravenous administration of albuterol at20 lg/min for 30minutes [5]. Another study of four healthyadults showed lactate increased to 4.3mmol/L (from0.8mmol/L) after intravenous administration of albuterol at0.7 lg/kg over 90minutes [6]. This complication occurs in upto 40% of patients [3]. The associated acidemia stimulatesthe respiratory center and worsens the patient’s hyperpnea[7,8]. Clinicians may interpret this as worsening of thepatient’s respiratory status, resulting in more aggressive treat-ment with beta-agonists. This hypothesized positive feedbackloop leads to dynamic hyperinflation, intrinsic positive end-expiratory pressure (auto-PEEP), and ultimately respiratoryfailure [9].

Treatment of beta-agonist induced hyperlactatemia con-sists of discontinuing the offending agent if the underlyingclinical condition allows. Several case reports describe thisphenomenon in patients who received intravenous albuterol.None of these patients received epinephrine, which may bepreferred in severe status asthmaticus. While a degree ofhyperlactatemia is tolerable with use of beta-agonists forrespiratory support, clinicians must be vigilant once hyperlac-tatemia develops in patients with acute severe asthma [10].

Metabolism of medication to lactate

Propylene glycol (PG) is a dihydroxy alcohol used as a vehiclefor hydrophobic medication. While 45% is renally excreted,the remainder is hepatically metabolized by alcohol dehydro-genase and aldehyde dehydrogenase. The final metabolite ofthis pathway is racemic mixture of D-lactate and L-lactate[11,12]. In patients with both intact renal and hepatic func-tion, serum half-life of PG is approximately five hours [13].

The concentration of PG varies by medication, with up to80% by volume in the case of intravenous lorazepam (seeTable 1). Some pharmaceutical formulations also containpolyethylene glycol, but the majority of toxicity is likely dueto PG [11]. There are three distinct risk factors for the devel-opment of PG-induced hyperlactatemia: amount of PGinfused, renal impairment [13], or small distributive volume(e.g., children <4 years) [14]. For example, a 54-year-old mandeveloped hyperlactatemia (6.1mmol/L) three days after hewas started on continuous lorazepam infusion (range5–20mg/h) for severe alcohol withdrawal. Serum PG concen-tration measured 810mg/dL [15]. A 50-year old man mis-takenly received lorazepam at 2mg/min instead of mg/h and

Table 1. Propylene glycol content of common parenteral medications.

Medication PG volume %

Lorazepam 80Phenobarbital 70Diazepam 40Digoxin 40Hydralazine 40Phenytoin 40Trimethoprim/sulfamethoxazole 40Etomidate 35Nitroglycerin 30Esmolol 25Chlordiazepoxide 20

Figure 1. Peripheral tissues produce lactate during glycolysis. An energydependent hepatic process (Cori cycle) converts lactate back to usable energysubstrates.

2 E. BLOHM ET AL.

within 10 hours developed severe acidosis (pH 6.9), hyperlac-tatemia (18.6mmol/L) and had a PG concentration of659mg/dL [16]. A retrospective study of critically ill patients(n¼ 33) found the PG accumulation occurs at lorazepam infu-sion rates exceeding 4mg/h [17]. As such, patients sufferingsevere alcohol withdrawal constitute the majority of reportedcases due to the reliance on large doses of benzodiazepinesover the course of their illness [12,17].

An elevation in the serum osmolal gap of more than10 mOsm/L strongly predicts accumulation of PG [18,19].Other laboratory findings include presence of an anion gapmetabolic acidosis and acute kidney injury [17].

Initial treatment consists of cessation of the offendingdrug. For patients requiring benzodiazepines, midazolam pro-vides a therapeutic alternative that does not contain PG. Thealcohol-dehydrogenase inhibitor 4-methylpyrazole (fomepi-zole) stops the formation of lactate in the setting of PG tox-icity [15,16]. However, PG will remain in circulation withfomepizole treatment, especially in the setting of renal insuf-ficiency. The compound itself has significant toxicity. Animalstudies demonstrate an increase in cardiomotor vagus activ-ity combined with decrease of sympathetic output [20]. Thisleads to a decrease of mean arterial pressure and heart rate,to the point of asystole [21,22]. Hence, its role in the treat-ment of PG toxicity remains unclear [11]. Hemodialysis hasbeen successfully employed to remove PG and correct serumpH, and should be considered in patients with significanttoxicity in the setting of hepatic and renal failure [14]. Othertoxicity associated with PG includes nephrotoxicity in 21% ofpatients receiving lorazepam infusions exceeding 4mg/h [17]and hyperosmolality [14].

A common medication that directly adds lactate is lac-tated Ringer’s solution (LR). This crystalloid contains28mmol/L of lactate. After fluid administration, 1 L of LRleads to a theoretical increase of 4.6mmol/L of lactate serumconcentration in an average adult if no lactate clearanceoccurs. No clinical effect was demonstrated in healthy humanvolunteers (n¼ 24) who received 1 L of intravenous LR overone hour. No volunteer exceeded a lactate concentration of2mmol/L. Lactate concentrations did not rise and remainedsimilar to the control group (1 L of 0.9% sodium chloridesolution) [23]. Patients with malignancy may have a higherbasal lactate production due to the Warburg effect the pref-erential use of glycolysis for ATP production by cancer cells.This process alone can exceed hepatic lactate clearance [24].While there are no human data, a transient elevation ofserum lactate concentration occurs in dogs with lymphomathat receive LR. The animals received 4.125mL/kg/h for sixhours, with the highest lactate recorded of 4.2mmol/L [25].The effect on critically ill patients without malignancyremains unknown.

Interference with the Cori cycle

The Cori cycle is a metabolic pathway that shifts the burdenof lactate metabolism from peripheral tissues to the liver.Hepatocytes use lactate for gluconeogenesis, and shuttle glu-cose back to peripheral tissues. Biguanides such as

metformin and phenformin reduce hepatic gluconeogenesis.This aspect of the Cori cycle converts lactate to pyruvate andsubsequently glucose in an energy-dependent pathway.The incidence of phenformin-associated hyperlactatemia isapproximately 60 per 100,000 patient-years and carries amortality rate of 50% and can occur at therapeutic dosing(e.g., 30mg PO BID) [26]. In contrast, metformin-associatedhyperlactatemia has an incidence of 3 per 100,000 patient-years [27]. Phenformin was removed from the US market in1978, but is still available in Italy, Brazil, and China [26]. It hassince been identified in multiple Chinese “herbal” prepara-tions in the US [28].

A subtle variation in mechanism of action accounts forthis difference in incidence. Phenformin increases lactate pro-duction in myocytes and interferes with oxidation of lactateto pyruvate whereas metformin does not affect myocytesand increases oxidation of lactate to pyruvate by 25% butdecreases conversion of pyruvate to glucose by 37% [29].Furthermore, phenformin inhibits complex I of the mitochon-drial ETC, thereby hindering hepatocellular ATP productionand subsequently gluconeogenesis. Metformin binds poorlyto mitochondrial membranes and only exerts this effect atsupratherapeutic concentrations [27]. Elevated serum lactateconcentrations are common in therapeutic dosing of phen-formin [30] but not metformin [31].

Metformin undergoes renal tubular excretion withoutmetabolism. Reduced renal function (e.g., after receiving iod-ine-based contrast) or overdose increase serum metforminconcentrations and place the patient at risk for hyperlactate-mia. Other risk factors include hypoxemia, alcohol abuse, liverfailure, myocardial infarction, sepsis, and shock [32].

Lactate concentration and acidemia correlate with serummetformin concentration and peak at around six hours in theevent of an acute overdose. In 22 cases identified in a litera-ture review, a pH of less than 6.9 or a serum lactate concen-trations of more than 25mmol/L correlated with a mortality of83%, whereas a peak serum metformin concentration ofgreater than 50lg/ml predicted a mortality of 38% [33].However, a patient survived without sequelae despite a pH of6.59, lactate concentration of 40mmol/L, and metformin con-centration of 160lg/ml [34]. A systematic review of a volun-tary pharmacovigilance database maintained by Merck Serono(the maker of GlucophageVR ) failed to demonstrate an associ-ation between lactate concentration or serum pH and surviv-ability among a cohort of 56 patients. Inclusion criteria for thisstudy included a blood lactate concentration of >10mmol/L,an arterial blood pH <7.0, and a measurable serum metforminconcentration. These criteria represent more severe metabolicderangements, yet the overall mortality for this group was53%. In addition, survival rates are likely underestimated sinceapproximately 1/3 of the patients in this analysis were septicor had multiple other drug ingestions [35]. The overall mortal-ity of metformin-associated hyperlactatemia is likely less than50%. In fact, a case series of 36 patients with metformin over-dose of greater than 3g had no deaths reported, althoughthis cohort had less severe metabolic derangements thanthose reported in the previous study [36].

Patients may present with clinical signs of acidosis such asKussmaul respirations, or feature altered mental status,

CLINICAL TOXICOLOGY 3

abdominal pain, gastrointestinal distress, or hemodynamicinstability due to reduced catecholamine-receptor binding atpH measurements less than 6.9 [34]. The clinical appearancein combination with laboratory findings may mimic periton-itis. Several patients underwent diagnostic laparotomies priorto identification of metformin toxicity as the underlyingcause [37–39].

Treatment options include crystalloid administration [40],bicarbonate therapy [41], and extracorporeal removal [38].Continuous venovenous high flow hemofiltration dialysis isthe recommended modality for enhanced elimination.Metformin is a small molecule and not protein bound, buthas a high volume of distribution (>3 L/kg) [42] which neces-sitates prolonged treatment sessions (average 15 h, up to24 h) [38]. In addition to drug removal, enhanced eliminationtechniques correct the pH imbalance. No data exist regardingthe effect of extracorporeal removal on mortality. Suggestedindications for hemodialysis include a pH <6.9, lactate>25mmol/L, metformin concentration >50mcg/ml, or rap-idly deteriorating clinical status [33].

Inhibition of mitochondrial protein synthesis

The mitochondrion is a cell organelle found in eukaryotesthat facilitates energy production through oxidative phos-phorylation. When mitochondria become dysfunctional, cyto-solic glycolysis is the only remaining means of ATP synthesis.This organelle is the descendent of an alphaproteobacterialendosymbiont, and thus is phylogenetically related to bac-teria [43]. Consequently, some mitochondrial structuresclosely resemble bacterial structures, as is the case with mito-chondrial ribosomes.

Oxazolidinone antibiotics such as linezolid and eperezolidare treatment options for drug-resistant Gram-positive bac-teria including methicillin-resistant Staphylococcus aureus(MRSA) and vancomycin-resistant enterococci (VRE) [44]. Thisdrug class binds to the bacterial 23S ribosome subunit andprevents assembly of the 30S-50S ribosome complex [45].

Patients with oxazolidinone-associated hyperlactatemiashow reduced oxygen consumption and extraction. Theunderlying cause is a polymorphism in mitochondrial DNA(mtDNA) (m.2706A>G) that encodes the 16S subunit of themitochondrial ribosome. This polymorphism increases dru-g–ribosome binding and subsequently inhibits mitochondrialprotein synthesis similar to the drug’s antimicrobial mechan-ism of action [46]. A post-mortem muscle biopsy of a patientwith linezolid-associated hyperlactatemia showed reducedactivity of complexes I, III, and IV of the ETC, all of which areencoded in part by mtDNA. Complex II, entirely encoded byDNA of the cell’s nucleus, was not diminished [47].

In a retrospective cohort study (n¼ 72), about 3% ofpatients developed oxazolidinone-associated hyperlactatemiadefined as a pH< 7.25 and serum lactate concentrationexceeding 4mmol/L. Onset of toxicity occurred with delayamong this group, on average between week 1 and 7 oftherapy. This delay likely represents the gradual depletion ofmitochondrial proteins in the setting of impaired protein syn-thesis [48]. While in this study 80% of those who developed

hyperlactatemia had either acute or chronic renal failure, thistrend did not continue in an analysis by the same authors of35 cases reported in the literature [48]. In addition, linezolidundergoes hepatic metabolism and does not require adjust-ment for renal insufficiency [49]. Hyperlactatemia arises asearly as four hours after intravenous infusion of linezolid, butthis phenomenon has only occurred in a single case report,and the patient rapidly died hours later, so causation isunclear since no other case with similar time course exists[50]. Major risk factors remain unknown, but patients withmtDNA m.2706A>G polymorphism [51] or MELAS syndrome(mitochondrial encephalomyopathy, lactic acidosis, andstroke-like symptoms) appear susceptible [52]. Presentingsymptoms include gastrointestinal distress, tachypnea/hyperpnea, and hypotension. Some patients remain asymp-tomatic [48].

Treatment of oxazolidinone-associated hyperlactatemiaconsists of withdrawal of the offending agent, hemodynamicsupport, and pH correction in severe cases. Renal replace-ment therapy (RRT) increases lactate clearance by only 3%,but addresses the underlying cause and lowers concentrationof the offending drug (approximately 30% reduction in line-zolid concentration over 10 hours of hemodialysis) [53].Linezolid is 31% protein bound and has a volume of distribu-tion of 50 L [54]. The majority of patients improve over1–15 days after drug discontinuation.

Nucleoside reverse transcriptase inhibitors (NRTIs) are amainstay of HIV therapy. These drugs resemble deoxynucleo-tides, which the viral reverse transcriptase enzyme uses toproduce viral DNA from RNA. Unlike deoxynucleotides, NRTIslack the 30-hydroxyl group and cannot form the 50–30

phosphodiester bond to elongate the DNA chain.Similar to reverse transcriptase, mtDNA polymerase c lacks

proofreading ability and will incorporate NRTIs into mtDNA[55]. Once mtDNA has incurred significant damage, theorganelle can no longer produce components of complexes I,III, and IV of the ETC. As mitochondrial respiration slows, cellsswitch to anaerobic metabolism leading to significant lactateproduction.

NRTIs also inhibit b-oxidation of fatty acids. Subsequently,the rate of glycolysis increases to provide pyruvate to anyremaining mitochondrial respiration. In turn, upregulated gly-colysis increases lactate production. Furthermore, the fattyacids accumulate in the cytosol. Subsequent hepatic steatosisreduces lactate clearance rates [56].

With the exception of abacavir, all NRTIs are associatedwith hyperlactatemia [57]. Estimates of incidence of hyperlac-tatemia with acidosis vary widely from 1 to 73.9 cases per1000 patient-years. Therapy with stavudine and didanosinecarries the highest risk for hyperlactatemia, which correspondto their increased affinity for mtDNA polymerase c observedin vitro [57,58].

Risk factors for NRTI-associated hyperlactatemia includereduced creatinine clearance (<70mL/min), low CD4þ count(<200 cells/mm3), and female sex [57,58]. Clinicians shouldsuspect this condition in patients on NRTIs who present withabdominal pain, emesis, or pancreatitis [59,60]. NRTI-associ-ated hyperlactatemia with acidosis has a mortality rate of

4 E. BLOHM ET AL.

40–70% despite maximal supportive care [57]. Death occurswithin 1–17 days after diagnosis [61].

L-Carnitine reduces mortality in NRTI-associated hyperlac-tatemia with acidosis. It serves as a co-factor in the ETC andtransports long-chain fatty acids across the mitochondrialmembrane, which enhances b-oxidation. Patients shouldreceive 50mg/kg daily (100mg/kg daily if on continuous dia-lysis). This recommendation is based on a small study (n¼ 6)[61]. The efficacy of serum alkalinization, prostaglandins, andthiamine administration has not been systemically studied.Withdrawal of the offending agent is paramount.

Non-nucleoside reverse transcription inhibitors (NNRTIs),protease inhibitors (PIs), fusion Inhibitors, entry inhibitors, andintegrase inhibitors are not associated with hyperlactatemia.

Inhibition of the electron transport chain

The ETC is the final biochemical pathway of mitochondrialrespiration. It consists of a series of five complexes, of whichcomplex I, III, and IV create a chemical gradient by pumpinghydrogen ion into the intermembrane space. Complex Vuses the potential energy of the chemical gradient to pro-duce ATP.

Propofol is a commonly used sedative-hypnotic for bothprocedural sedation and for the sedation of ventilatedpatients. Infusion rates above 4mg/h and prolonged therapycan cause propofol-related infusion syndrome (PRIS). Onsetof PRIS occurs between 20 and 60 hours of continuous pro-pofol infusion in 23% of patients, but is delayed by over60 hours in 57% of patients [62]. The syndrome is a constella-tion of hyperlactatemia, metabolic acidosis, rhabdomyolysis,hepatomegaly, hyperlipidaemia, and renal failure.

Some patients with PRIS develop dysrhythmias. In an ana-lysis of seven cases, the initial abnormality on electrocardio-gram (EKG) was a Brugada-pattern with coved-type elevationsof the ST-segment in the precordial leads. The QT interval sub-sequently lengthened and patients developed ventriculartachycardia followed by recalcitrant ventricular fibrillation[62,63]. Despite withdrawal of the offending drug and max-imum supportive therapy, mortality persists at 35% [62].

PRIS develops by several different mechanisms. Highdoses (3mg/kg/h in one case report) directly inhibit complexI and IV of the ETC [64]. High dose propofol also uncouplesthe ETC by altering the structure of F1F0ATPase (complex V).Dissipation of the proton gradient generates heat. The result-ing hyperthermia is a very poor prognostic factor, andpatients are at risk of heart failure and sudden death [62].

In vitro studies show that propofol interferes with electrontransfer among complex I, II, and III in a dose dependedmanner. Propofol mimics coenzyme-Q and accepts electronsfrom the ETC but does not donate electrons to the nextdownstream ETC complex [64,65]. Furthermore, propofolinhibits carnitine palmityl transferase I, rendering mitochon-dria unable to import long-chain acylcarnitine esters [66].Impaired fatty acid metabolism leads to hepatic steatosis andreduced lactate clearance. The degree of hepatic dysfunctionas measured by liver function tests and serum ammonia con-centration is directly proportional to the duration of drug

exposure [62,67]. Prolonged infusion increases the risk fordysrhythmias and can produce Brugada-type EKG. This maybe due to proarrhythmic properties of free fatty acids con-tained in the propofol infusion (10% by weight). A propofolinfusion of 1mg/kg/h will meet a patient’s daily dietary fattyacid requirements, but clinical practice often demands higherdoses [66,67]. Lastly, the inability of muscle tissues to metab-olize fatty acids leads to rhabdomyolysis.

Critical care providers should diagnose PRIS based ondrug exposure history, and laboratory tests indicating hyper-lactatemia, rhabdomyolysis, hypertriglyceridemia, and acide-mia. Immediate withdrawal of the drug is required, and theclinician must limit additional triglyceride administration (e.g.,total parenteral nutrition). In addition to supportive care,coenzyme-Q administration is of theoretical benefit, but effi-cacy is unproven [64].

Sodium nitroprusside (SNP) is a simple molecule of a nitricoxide and five cyanate groups bound to a central iron atom.The desired effect is the release of nitric oxide for the pur-pose of blood pressure control. The release of the molecule’scyanate groups produces clinically significant toxicity.

During normal oxidative phosphorylation in the ETC, cyto-chrome c shuttles electrons from complex III to complex IV(also called cytochrome oxidase or cytochrome a3). ComplexIV temporarily accepts the electrons by reducing its trivalentiron moieties, and subsequently donates four electrons tomolecular oxygen. The reduced oxygen ions bind hydrogenions to form water. Cyanide binds the ferric ion (Fe3þ) incomplex IV and prevents any further electron transfer [68].This profoundly inhibits the ETC as oxygen cannot bereduced without the enzymatic activity of complex IV.Subsequently, cells switch to lactate-producing anaerobicmetabolism [69]. In addition to hyperlactatemia and acidosis,cellular ATP depletion causes altered mental status, seizures,myocardial infarction, and other signs of organ damage [70].

Thiosulfate sulfurtransferase (“rhodanase”) is the endogen-ous detoxification system that converts cyanide to thiocyan-ate. SNP infusion rates generally should not exceed2mcg/kg/min as this represents the approximate endogen-ous metabolic capacity for cyanide [70]. Approximately 50mgSNP administration suffices to deplete thiosulfate stores, anddetoxification fails [70,71]. If the patient needs emergentblood pressure control, rates as high as 10mcg/kg/min canbe administered for several minutes [70,71]. Malnourishedpatients are at increased risk of toxicity due to lower thiosul-fate stores. Furthermore, 98.5% of cyanide is sequestered inred blood cells [72]. Hemolysis due to any reason (e.g., pro-longed cardiopulmonary bypass time) increases free cyanideand accelerates toxicity.

Several treatments are available for SNP-associated cyan-ide toxicity. The concomitant infusion of sodium thiosulfatewith SNP delays and blunts toxicity [73–75]. Amyl nitrite,sodium nitrite, and hydroxocobalamin are alternative anti-dotes should toxicity develop.

In a series of patients with aortic dissection treated withSNP, 11.5% of patients (18/157) who received nitroprussideinfusion for blood pressure control developed clinically sig-nificant cyanide toxicity. None received prophylactic thiosul-fate [76].

CLINICAL TOXICOLOGY 5

Barbiturates inhibit complex I of the ETC in tissue culture[77]. The clinical significance of this is not fully understood.Both a 73-year-old man and 21-year-old woman developedhyperlactatemia after administration of thiopental for refrac-tory status epilepticus (303mg/kg over 48 h and 840mg/kgover 150 h, respectively) [78]. However, both patients alsodeveloped non-occlusive mesenteric ischemia, which in itselfproduces hyperlactatemia. It is unclear if tissue necrosisoccurred due to impaired mitochondrial function. In healthyindividuals, barbiturate administration does not produce clin-ically measurable effects.

Valproic acid is a widely used antiepileptic and mood sta-bilizer. It is associated with hepatotoxicity and hyperammo-nemia at therapeutic dosing. In overdose, valproic acidproduces hyperlactatemia and acidosis [79,80]. Mitochondrialcomplex I and IV were the suspected targets of inhibition[81]. However, the effect on ETC complexes was not consist-ently reproduced in vitro, depending on the energy substrateused for oxidative phosphorylation [82]. Current data suggestthat the metabolite valproyl-CoA inhibits dihydrolipoamidedehydrogenase, an enzymatic subunit of pyruvate dehydro-genase, and thus prevents entry of some energy substratesinto the citric acid cycle [83].

Pharmacokinetic data show protein binding of 91.6% attherapeutic dosing, which suggests extracorporeal removal tobe of low efficacy [84]. In overdose, drug molecules exceedavailable protein binding sites and up to 85% of drugremains unbound and therefore dialyzable [85]. Continuousvenovenous hemodialysis has been used successfully in thetreatment of valproic acid induced hyperlactatemia with acid-emia [79,80]. The Extracorporeal Treatments in Poisoning(EXTRIP) workgroup recommends dialysis for a pH of lessthan 7.1 in the setting of acute valproate toxicity [86].

Multiple antidepressants and mood stabilizers inhibit theETC in vitro. Porcine mitochondria had reduced oxygen useduring exposure to supratherapeutic concentrations ofdesipramine (inhibition of complex I, IV), amitriptyline (inhib-ition of complex I, II, IV), imipramine (inhibition of complex I,II, IV), citalopram (inhibition of complex I, II), mirtazapine(inhibition of complex I, IV), olanzapine (inhibition of complexI IV), and venlafaxine (inhibition of complex II, IV) [81].

Clinically, this effect appears to be rare. A 55-year-old manwho ingested 6000mg venlafaxine developed hyperlactate-mia (peak 8.6mmol/L) which resolved with hydration andactivated charcoal [87]. A 81-year-old man had been on cita-lopram 20mg BID for three weeks was started on linezolid600mg BID for osteomyelitis. After 3 weeks of combinedtherapy, he developed hyperlactatemia (peak 29.1mmol/L)and died shortly thereafter. It is unclear whether he devel-oped serotonin syndrome (tremor/hyperthermia reported,but no clonus or seizure), experienced linezolid-inducedhyperlactatemia, citalopram-induced hyperlactatemia, or anycombination thereof [88].

Uncoupling of oxidative phosphorylation

The final step of mitochondrial respiration involves complexV (ATP synthase). This enzyme phosphorylates adenosine

diphosphate (ADP) to ATP, using the potential energy storedin the mitochondrial chemical hydrogen gradient created bythe ETC.

Poisoning with salicylates such as aspirin and bismuth sub-salicylate (“Pepto-Bismol”) accounts for approximately 13,000annual poisonings and 47 deaths in the United States [89].Salicylic acid is a weak acid (pK¼ 3.0) which deprotonates atalkaline pH but remains unionized at acidic pH [90].

The ETC produces a hydrogen ion gradient between themitochondrial intermembrane space (pH 6.88) and mitochon-drial matrix (pH 7.78) that serves as the proton motive forcefor ATP synthesis by complex V [91]. The salicylate ion dif-fuses down its concentration gradient from the cytosol (pH7.59) into the mitochondrial intermembrane space. The lowerpH in the intermembrane space shifts the equilibrium towardthe unionized state, and the salicylate ions bind protons.Unlike salicylate, salicylic acid does not carry a charge andcan diffuse rapidly (permeability coefficient 0.7 cm/s) intoboth the cytosol and the mitochondrial matrix [90]. As thesecompartments have relatively alkaline pH, salicylic aciddeprotonates to salicylate and the cycle repeats.

Salicylate therefore acts as a “proton shuttle” from themitochondrial intermembrane space to the surrounding com-partments. As protons bypass complex V and no longer con-tribute to ATP synthesis, salicylate effectively uncouples theETC [90,92]. This process is exothermic and patients maydevelop hyperthermia [89]. Hyperlactatemia results fromincreased rates of glycolysis in an effort to meet metabolicdemands [93]. In addition, salicylic acid accumulation in themitochondrial matrix exerts osmotic pressure. The subse-quent swelling contributes to mitochondrial dysfunction [94].

Treatment of salicylism takes advantage of the three com-partment model (tissue, serum, and urine). Alkalinization of acompartment traps salicylate in its ionized form in that parti-tion. Administration of sodium bicarbonate with a goal serumpH of 7.55 and goal urine pH of 8.0 creates a gradient favor-ing salicylate excretion [95]. Severely poisoned patients (ASAconcentration >90mg/dL, cerebral edema/seizure, acute lunginjury, profound acidosis, worsening clinical status) requirehemodialysis [95,96].

Laboratory errors in lactate measurement

Intoxication with ethylene glycol (EG) results in anion-gapmetabolic acidosis due to metabolism to glycolic acid andglyoxylic acid. Several cases reported describe massive hyper-lactatemia (greater 15mmol/L, as high as 42mmol/L) meas-ured on point-of-care testing, but near-normal serum lactateconcentration in laboratory-based assays [97–101]. In combin-ation with severe acidosis and depressed mental status, aclinician may suspect mesenteric ischemia or other condi-tions as a cause, delaying diagnosis and treatment forEG toxicity.

Two enzymatic laboratory assays are available to measureserum lactate concentrations. Lactate analyzers either use lac-tate oxidase (LOD) or LDH. LOD converts lactate to H2O2; theresultant colorimetric or amperometric changes are quanti-fied by the analyzer. Glycolate and glyoxylate structurally

6 E. BLOHM ET AL.

resemble lactate and serve as a substrate for LOD, but notLDH. For unclear reasons, amperometric quantification ismore susceptible to produce factitious lactate elevation thancolorimetry. Most point-of-care lactate assays and blood gasanalyzers use LOD and amperometry. Interestingly, the bac-terial origin of the LOD affects the accuracy of the assay.Amperometric assays using Pediococcus sp. LOD are suscep-tible to interference, whereas amperometric assays withAerococcus viridans derived LOD are not [102].

The interference with point-of-care assays but not labora-tory-based assays can also serve as a diagnostic tool. Patientsfound to have a “lactate gap” between two assays should besuspected to have EG poisoning [103]. While there is a linearcorrelation between serum glycolate concentrations andspurious serum lactate concentrations [104], the degree ofinterference depends on the particular point-of-care assay[102]. Therefore, we cannot recommend the use point-of-carelactate concentrations as a surrogate quantitative marker forthese toxic metabolites.

Discussion

Hyperlactatemia may result from therapeutic use or overdoseof a variety of medications and is easy to overlook. Truehyperlactatemia generally portends a high risk of mortality.Clinicians often suspect and investigate common causes ofhyperlactatemia such as sepsis or mesenteric ischemia first,resulting in a delay of diagnosis and even unnecessary inva-sive workup. Therefore, emergency medicine physicians,intensive care providers, and doctors prescribing the

medications discussed above should remain cognizant andvigilant of this complication.

The pathophysiologic mechanism of action ranges frompyruvate overproduction, to mitochondrial impairment(Figure 2), or interference with lactate metabolism. It isunclear why some patients develop hyperlactatemia whereasothers do not despite comparable medication administration.In addition, patients may develop hyperlactatemia weeksafter initiation of drug therapy, as is the case in NRTIs.

Treatment generally includes withdrawal of the offendingdrug and often includes hemodialysis to remove the toxinand to correct serum pH. Some causative agents are treat-able with targeted therapies. Delay or failure of recognition isassociated with very poor outcome.

This literature review has certain limitations. The availabledata are often limited to case reports for a specific drug as acause for hyperlactatemia rather than large scale populationstudies or randomized controlled trials. In addition, the truecause of medication-induced hyperlactatemia may go unrec-ognized and thus underreported. At the same time, cliniciansmay erroneously attribute hyperlactatemia to medicationexposure. Most patients who receive the medications associ-ated with hyperlactatemia are medically complex and poten-tially critically ill. Therefore, clinicians should consider theconditions outlined above only in parallel to other causes ofhyperlactatemia.

Disclosure statement

The authors report no conflicts of interest. The authors alone are respon-sible for the content and writing of this article.

Figure 2. A graphical representation of the site of medication-induced mitochondrial inhibition.

CLINICAL TOXICOLOGY 7

References

[1] Haas SA, Lange T, Saugel B, et al. Severe hyperlactatemia, lactateclearance and mortality in unselected critically ill patients.Intensive Care Med. 2016;42:202–210.

[2] Ingelfinger JR, Kraut JA, Madias NE. Lactic acidosis. N Engl JMed. 2014;371:2309–2319.

[3] Rodrigo GJ, Rodrigo C. Elevated plasma lactate level associatedwith high dose inhaled albuterol therapy in acute severeasthma. Emerg Med J. 2005;22:404–408.

[4] Suzuki K, Tanaka S, Uchida T, et al. Catecholamine release indu-ces elevation in plasma lactate levels in patients undergoingadrenalectomy for pheochromocytoma. J Clin Anesth. 2014;26:616–622.

[5] Tobin AE, Pellizzer A-M, Santamaria JD. Mechanisms by whichsystemic salbutamol increases ventilation. Respirology. 2006;11:182–187.

[6] Phillips PJ, Vedig AE, Jones PL, et al. Metabolic and cardiovascu-lar side effects of the beta 2-adrenoceptor agonists salbutamoland rimiterol. Br J Clin Pharmacol. 1980;9:483–491.

[7] Dodda VR, Spiro P. Can albuterol be blamed for lactic acidosis?Respir Care. 2012;57:2115–2118.

[8] Saxena R, Marais G. Salbutamol: beware of the paradox! BMJCase Rep. 2010;2010:bcr0120102665.

[9] Creagh-Brown BC, Ball J. An under-recognized complication oftreatment of acute severe asthma. Am J Emerg Med.2008;26:514.e1–514.e3.

[10] Tobin A, Santamaria J. Respiratory failure precipitated by salbu-tamol. Intern Med J. 2005;35:199–200.

[11] Mullins ME, Barnes BJ. Hyperosmolar metabolic acidosis andintravenous Lorazepam. N Engl J Med. 2002;347:857–858.

[12] Miller MA, Forni A, Yogaratnam D. Propylene glycol-induced lac-tic acidosis in a patient receiving continuous infusion pentobar-bital. Ann Pharmacother. 2008;42:1502–1506.

[13] Cawley MJ. Short-term lorazepam infusion and concern for pro-pylene glycol toxicity: case report and review. Pharmacotherapy.2001;21:1140–1144.

[14] Parker MG, Fraser GL, Watson DM, et al. Removal of propyleneglycol and correction of increased osmolar gap by hemodialysisin a patient on high dose lorazepam infusion therapy. IntensiveCare Med. 2002;28:81–84.

[15] Zar T, Yusufzai I, Sullivan A, et al. Acute kidney injury, hyperos-molality and metabolic acidosis associated with lorazepam. NatClin Pract Nephrol. 2007;3:515–520.

[16] Zosel A, Egelhoff E, Heard K. Severe lactic acidosis after an iatro-genic propylene glycol overdose. Pharmacotherapy. 2010;30:219–219.

[17] Horinek EL, Kiser TH, Fish DN, et al. Propylene glycol accumula-tion in critically ill patients receiving continuous intravenous lor-azepam infusions. Ann Pharmacother. 2009;43:1964–1971.

[18] Arroliga AC, Shehab N, McCarthy K, et al. Relationship of con-tinuous infusion lorazepam to serum propylene glycol concen-tration in critically ill adults. Crit Care Med. 2004;32:1709–1714.

[19] Barnes BJ, Gerst C, Smith JR, et al. Osmol gap as a surrogatemarker for serum propylene glycol concentrations in patientsreceiving lorazepam for sedation. Pharmacotherapy. 2006;26:23–33.

[20] al-khudhairi D. Autonomic reflexes and the cardiovascular effectsof propylene glycol. Br J Anaesth. 1986;58:897–902.

[21] Doenicke A, Nebauer AE, Hoernecke R, et al. Osmolalities of pro-pylene glycol-containing drug formulations for parenteral use.Should propylene glycol be used as a solvent? Anesth Analg.1992;75:431–435.

[22] Wheless JW. Pediatric use of intravenous and intramuscularphenytoin: lessons learned. J Child Neurol. 1998;13(Suppl 1):S11–S14. discussion S30–S32.

[23] Didwania A, Miller J, Kassel D, et al. Effect of intravenous lac-tated Ringer's solution infusion on the circulating lactate con-centration: Part 3. Results of a prospective, randomized, double-blind, placebo-controlled trial. Crit Care Med. 1997;25:1851–1854.

[24] Gardner A, Griffiths J. A case of type B lactic acidosis as a com-plication of chronic myelomonocytic leukaemia: a case reportand review of the literature. J Med Case Rep. 2015;9:16.

[25] Vail DM, Ogilvie GK, Fettman MJ, et al. Exacerbation of hyperlac-tatemia by infusion of lactated Ringer's solution in dogs withlymphoma. J Vet Intern Med. 1990;4:228–232.

[26] Fimognari FL, Pastorelli R, Incalzi RA. Phenformin-induced lacticacidosis in an older diabetic patient: a recurrent drama (phenfor-min and lactic acidosis). Diabetes Care. 2006;29:950–951.

[27] Crofford OB. Metformin. N Engl J Med. 1995;333:588–589.[28] Aschwanden C. Herbs for health, but how safe are they? Bull

World Health Organ. 2001;79:691–692.[29] Stumvoll M, Nurjhan N, Perriello G, et al. Metabolic effects of

metformin in non-insulin-dependent diabetes mellitus. N Engl JMed. 1995;333:550–554.

[30] Kreisberg RA, Pennington LF, Boshell BR. Lactate turnover andgluconeogenesis in obesity. Effect of phenformin. Diabetes.1970;19:64–69.

[31] Salpeter SR, Greyber E, Pasternak GA, et al. Risk of fatal and non-fatal lactic acidosis with metformin use in type 2 diabetes melli-tus. Cochrane Database Syst Rev. 2010;CD002967.

[32] Spiller HA, Sawyer TS. Toxicology of oral antidiabetic medica-tions. Am J Health Syst Pharm. 2006;63:929–938.

[33] Dell'Aglio DM, Perino LJ, Kazzi Z, et al. Acute metformin over-dose: examining serum pH, lactate level, and metformin concen-trations in survivors versus nonsurvivors: a systematic review ofthe literature. Ann Emerg Med. 2009;54:818–823.

[34] Dell'Aglio DM, Perino LJ, Todino JD, et al. Metformin overdosewith a resultant serum pH of 6.59: survival without sequalae.J Emerg Med. 2010;39:e77–e80.

[35] Kajbaf F, Lalau J-D. The prognostic value of blood pH and lac-tate and metformin concentrations in severe metformin-associ-ated lactic acidosis. BMC Pharmacol Toxicol. 2013;14:22.

[36] McNamara K, Isbister GK. Hyperlactataemia and clinical severityof acute metformin overdose. Intern Med J. 2015;45:402–408.

[37] Ashall V, Dawes T. Metformin and lactic acidosis. Br J Anaesth.2008;101:876–877.

[38] Dichtwald S, Weinbroum AA, Sorkine P, et al. Metformin-associ-ated lactic acidosis following acute kidney injury. Efficacioustreatment with continuous renal replacement therapy. DiabetMed. 2012;29:245–250.

[39] Biradar V, Moran JL, Peake SL, et al. Metformin-associatedlactic acidosis (MALA): clinical profile and outcomes in patientsadmitted to the intensive care unit. Crit Care Resusc.2010;12:191–195.

[40] Bebarta VS, Pead J, Varney SM. Lacticemia after acute overdoseof metformin in an adolescent managed without intravenoussodium bicarbonate or extracorporeal therapy. Pediatr EmergCare. 2015;31:589–590.

[41] Ansari RSH, Mady AF, Qutub HO, et al. Metformin-related acid-osis in a woman while performing Haj: a conservative approach.Saudi J Kidney Dis Transpl. 2015;26:125–127.

[42] Peters N, Jay N, Barraud D, et al. Metformin-associated lacticacidosis in an intensive care unit. Crit Care. 2008;12:R149.

[43] Wang Z, Wu M. An integrated phylogenomic approach towardpinpointing the origin of mitochondria. Sci Rep. 2015;5:7949.

[44] Djibr�e M, Pham T, Denis M, et al. Fatal lactic acidosis associatedwith linezolid therapy. Infection. 2015;43:125–126.

[45] Barnhill AE, Brewer MT, Carlson SA. Adverse effects of antimicro-bials via predictable or idiosyncratic inhibition of host mitochon-drial components. Antimicrob Agents Chemother. 2012;56:4046–4051.

[46] Palenzuela L, Hirano M. Does linezolid cause lactic acidosis byinhibiting mitochondrial protein synthesis? Clin Infect Dis.2015;9:16.

[47] Protti A, Ronchi D, Bassi G, et al. Changes in whole-body oxygenconsumption and skeletal muscle mitochondria during linezolid-induced lactic acidosis. Crit Care Med. 2016;7:e579–e582.

[48] Im JH, Baek JH, Kwon HY, et al. Incidence and risk factors of line-zolid-induced lactic acidosis. Int J Infect Dis. 2015;31:47–52.

8 E. BLOHM ET AL.

[49] Brier ME, Stalker DJ, Aronoff GR, et al. Pharmacokinetics of line-zolid in subjects with renal dysfunction. Antimicrob AgentsChemother. 2003;47:2775–2780.

[50] Contou D, Fichet J, Grimaldi D, et al. Early life-threatening lacticacidosis following a single infusion of linezolid. Int J AntimicrobAgents. 2011;38:84–85.

[51] Del Pozo JL, Fern�andez-Ros N, S�aez E, et al. Linezolid-inducedlactic acidosis in two liver transplant patients with the mitochon-drial DNA A2706G polymorphism. Antimicrob AgentsChemother. 2014;58:4227–4229.

[52] Cope TE, McFarland R, Schaefer A. Rapid-onset, linezolid-inducedlactic acidosis in MELAS. Mitochondrion. 2011;11:992–993.

[53] Fiaccadori E. Removal of linezolid by conventional intermittenthemodialysis, sustained low-efficiency dialysis, or continuousvenovenous hemofiltration in patients with acute renal failure.Crit Care Med. 2004;32:2437–2442.

[54] Sawyer AJ, Haley HL, Baty SR, et al. Linezolid-induced lactic acid-osis corrected with sustained low-efficiency dialysis: a casereport. Am J Kidney Dis. 2014;64:457–459.

[55] Tolomeo M, Mancuso S, Todaro M, et al. Mitochondrial disrup-tion and apoptosis in lymphocytes of an HIV infected patientaffected by lactic acidosis after treatment with highly active anti-retroviral therapy. J Clin Pathol. 2003;56:147–151.

[56] Claessens Y-E, Chiche J-D, Mira J-P, et al. Bench-to-bedsidereview: severe lactic acidosis in HIV patients treated with nucleo-side analogue reverse transcriptase inhibitors. Crit Care.2003;7:226–232.

[57] Bonnet F, Bonarek M, Morlat P, et al. Risk factors for lactic acid-osis in HIV-infected patients treated with nucleoside reverse-transcriptase inhibitors: a case-control study. Clin Infect Dis.2003;36:1324–1328.

[58] Dragovic G, Jevtovic D. The role of nucleoside reverse transcript-ase inhibitors usage in the incidence of hyperlactatemia and lac-tic acidosis in HIV/AIDS patients. Biomed Pharmacother.2012;66:308–311.

[59] Coghlan M. Symptomatic lactic acidosis in hospitalized antiretro-viral-treated patients with human immunodeficiency virus infec-tion: a report of 12 cases. Clin Infect Dis. 2001;33:1914–1921.

[60] Miller KD. Lactic acidosis and hepatic steatosis associated withuse of stavudine: report of four cases. Ann Intern Med.2000;133:192.

[61] Claessens Y-E, Cariou A, Monchi M, et al. Detecting life-threaten-ing lactic acidosis related to nucleoside-analog treatment ofhuman immunodeficiency virus-infected patients, and treatmentwith l-carnitine. Crit Care Med. 2003;31:1042–1047.

[62] Kraj�cov�a A, Waldauf P, And�el M, et al. Propofol infusion syn-drome: a structured review of experimental studies and 153published case reports. Crit Care. 2015;19:1383.

[63] Vernooy K, Delhaas T, Cremer OL, et al. Electrocardiographicchanges predicting sudden death in propofol-related infusionsyndrome. Heart Rhythm. 2006;3:131–137.

[64] Vanlander AV, Okun JG, de Jaeger A, et al. Possible pathogenicmechanism of propofol infusion syndrome involves coenzyme q.Anesthesiology. 2015;122:343–352.

[65] Schenkman KA, Yan S. Propofol impairment of mitochondrialrespiration in isolated perfused guinea pig hearts determined byreflectance spectroscopy. Crit Care Med. 2000;28:172–177.

[66] Wolf A, Weir P, Segar P, et al. Impaired fatty acid oxidation inpropofol infusion syndrome. Lancet. 2001;357:606–607.

[67] Rison RA, Ko DY. Isolated fatty liver from prolonged propofoluse in a pediatric patient with refractory status epilepticus. ClinNeurol Neurosurg. 2009;111:558–561.

[68] Beasley DMG, Glass WI. Cyanide poisoning: pathophysiology andtreatment recommendations. Occup Med (Lond). 1998;48:427–431.

[69] Huzar TF, George T, Cross JM. Carbon monoxide and cyanidetoxicity: etiology, pathophysiology and treatment in inhalationinjury. Expert Rev Resp Med. 2013;7:159–170.

[70] Udeh CI, Ting M, Arango M, et al. Delayed presentation of nitro-prusside-induced cyanide toxicity. Ann Thorac Surg. 2015;99:1432–1434.

[71] Belani K, Hottinger D, Kozhimannil T, et al. Sodium nitroprussidein 2014: a clinical concepts review. J Anaesthesiol ClinPharmacol. 2014;30:462–471.

[72] Pasch T, Schulz V, Hoppelsh€auser G. Nitroprusside-induced for-mation of cyanide and its detoxication with thiosulfate duringdeliberate hypotension. J Cardiovasc Pharmacol. 1983;5:77–85.

[73] Hall V, Guest J. Sodium nitroprusside-induced cyanide intoxica-tion and prevention with sodium thiosulfate prophylaxis. Am JCrit Care. 1992;1:19–25.

[74] Curry SC. Prevention of fetal and maternal cyanide toxicity fromnitroprusside with coinfusion of sodium thiosulfate in gravidewes. Anesth Analg. 1997;8:1121–1126.

[75] Meyer S, Baghai A, Sailer NL, et al. Lactic acidosis caused bysodium nitroprusside in a newborn with congenital heart dis-ease. Eur J Pediatr. 2005;164:253–254.

[76] Huang X, Hou L, Tang J, et al. Central nervous system toxicity ofsodium nitroprusside in treatment of patients with aortic dissec-tion. J Huazhong Univ Sci Technol Med Sci. 2012;32:927–930.

[77] Anderson CM, Norquist BA, Vesce S, et al. Barbiturates inducemitochondrial depolarization and potentiate excitotoxic neur-onal death. J Neurosci. 2002;22:9203–9209.

[78] Cereda C, Berger MM, Rossetti AO. Bowel ischemia: a rare com-plication of thiopental treatment for status epilepticus. NeurocritCare. 2008;10:355–358.

[79] Alluin A, Jezequel J, Gauthier N, et al. Cas d’une intoxications�ev�ere �a l’acide valproïque: int�eret d’un traitement par �epurationextrar�enale combin�e �a la l-carnitine. Ann Fr Anesth Reanim.2011;30:752–754.

[80] van den Broek MPH, Sikma MA, Ververs TF, et al. Severe valproicacid intoxication: case study on the unbound fraction and theapplicability of extracorporeal elimination. Eur J Emerg Med.2009;16:330–332.

[81] Hroudova J, Fisar Z. Activities of respiratory chain complexesand citrate synthase influenced by pharmacologically differentantidepressants and mood stabilizers. Neuro Endocrinol Lett.2010;31:336–342.

[82] Komulainen T, Lodge T, Hinttala RB, et al. Sodium valproateinduces mitochondrial respiration dysfunction in HepG2 in vitrocell model. Toxicology. 2015;331:47–56.

[83] Lu�ıs PBM, Ruiter JPN, Aires CCP, et al. Valproic acid metabolitesinhibit dihydrolipoyl dehydrogenase activity leading to impaired2-oxoglutarate-driven oxidative phosphorylation. BiochimBiophys Acta (BBA) – Bioenerg. 2007;1767:1126–1133.

[84] Gugler R, Mueller G. Plasma protein binding of valproic acid inhealthy subjects and in patients with renal disease. Br J ClinPharmacol. 1978;5:441–446.

[85] Thanacoody RHK. Extracorporeal elimination in acute valproicacid poisoning. Clin Toxicol. 2009;47:609–616.

[86] Ghannoum M, Lalibert�e M, Nolin TD, et al. Extracorporeal treat-ment for valproic acid poisoning: systematic review and recom-mendations from the EXTRIP workgroup. Clin Toxicol. 2015;53:454–465.

[87] Iragavarapu C, Gupta T, Chugh SS, et al. Type B lactic acidosisassociated with venlafaxine overdose. Am J Ther. 2016;23:e1082–e1084.

[88] Bernard L. Serotonin syndrome after concomitant treatmentwith linezolid and citalopram. Clin Infect Dis. 2003;36:1197.

[89] O'Malley GF. Emergency department management of the salicyl-ate-poisoned patient. Emerg Med Clin N Am. 2007;25:333–346.

[90] Gutknecht J. Salicylates and proton transport through lipidbilayer membranes: a model for salicylate-induced uncouplingand swelling in mitochondria. J Membr Biol. 1990;115:253–260.

[91] Porcelli AM, Ghelli A, Zanna C, et al. pH difference across theouter mitochondrial membrane measured with a green fluores-cent protein mutant. Biochem Biophys Res Commun. 2005;326:799–804.

[92] Saeedi S, Rocher F, Bonmort J, et al. Early membrane eventsinduced by salicylic acid in motor cells of the Mimosa pudicapulvinus. J Exp Bot. 2013;64:1829–1836.

CLINICAL TOXICOLOGY 9

[93] Petrescu I, Tarba C. Uncoupling effects of diclofenac and aspirinin the perfused liver and isolated hepatic mitochondria of rat.Biochim Biophys Acta. 1997;1318:385–394.

[94] You K. Salicylate and mitochondrial injury in Reye’s syndrome.Science. 1983;221:163–165.

[95] American College of Medical Toxicology. Guidance document:management priorities in salicylate toxicity. J Med Toxicol.2015;11:149–152.

[96] Juurlink DN, Gosselin S, Kielstein JT, et al. Extracorporeal treat-ment for salicylate poisoning: systematic review and recommen-dations from the EXTRIP Workgroup. Ann Emerg Med.2015;66:165–181.

[97] Sandberg Y, Rood PPM, Russcher H, et al. Falsely elevated lac-tate in severe ethylene glycol intoxication. Neth J Med. 2010;68:320–323.

[98] Pernet P, B�en�eteau-Burnat B, Vaubourdolle M, et al. False eleva-tion of blood lactate reveals ethylene glycol poisoning. Am JEmerg Med. 2009;27:132.e1–132.e2.

[99] Brindley PG, Butler MS, Cembrowski G, et al. Falsely elevatedpoint-of-care lactate measurement after ingestion of ethyleneglycol. CMAJ. 2007;176:1097–1099.

[100] Verelst S, Vermeersch P, Desmet K. Ethylene glycol poisoningpresenting with a falsely elevated lactate level. Clin Toxicol.2009;47:236–238.

[101] Woo MY, Greenway DC, Nadler SP, et al. Artifactual elevation oflactate in ethylene glycol poisoning. J Emerg Med. 2003;25:289–293.

[102] Tintu A, Rouwet E, Russcher H. Interference of ethylene glycolwith (l)-lactate measurement is assay-dependent. Ann ClinBiochem. 2013;50:70–72.

[103] Chaudhry SD, Pandurangan M, Pinnell AE. Lactate gap andethylene glycol poisoning. Eur J Anaesthesiol. 2008;25:511–513.

[104] Manini AF, Hoffman RS, McMartin KE, et al. Relationship betweenserum glycolate and falsely elevated lactate in severe ethyleneglycol poisoning. J Anal Toxicol. 2009;33:174–176.

10 E. BLOHM ET AL.