J Am Board Fam Pract: first published as 10.3122/15572625-11-3-233 on 1 May 1998. Downloaded from

BRIEF REPORTS -Associated Lactic

Darrell T. Hulisz, PhannD, Mark F. Bonfiglio, PhannD, and Richard D. Murray, MD

Metformin is a biguanide used to lower blood glu­ included NIDDM for 15 years, hypertension, cose levels in patients with non-insulin-dependent coronary artery disease, hypercholesterolemia, diabetes mellitus (NIDDM). Metformin notably congestive heart failure, and paroxysmal atrial fib­ has certain advantages over sulfonylureas because rillation. An echocardiogram 3 months before ad­ it does not cause hypoglycemia, nor does it pro­ mission showed left ventricular hypertrophy, mi­ mote weight gain or hyperinsulinemia.1 It also ap­ tral regurgitation, and an ejection fraction of 35 pears to have beneficial effects on blood lipid lev­ percent. The patient's home medications included els.2 Because of these advantages, metformin use is glipizide 10 mg twice a day, digoxin 0.125 mg daily, widely promoted, particularly to primary care prawsin 2 mg twice a day, furosemide 20 mg twice physicians, and it will likely be used extensively in a day, aspirin 325 mg daily, and metformin 2500 this country for treatment ofNIDDM. mg daily. She reported no history of ethanol abuse. Nevertheless, a potential complication of met­ The patient had started metformin therapy 2 formin use is its rare association with lactic acido­ months before admission, which was gradually sis, a condition that has a mortality rate of approx­ titrated to this dose. The most recent change in imately 50 percent.3,4 Accordingly, it is instructive dosage was made 1 week before admission from to describe a case of metformin-associated lactic 2000 to 2500 mg daily. acidosis. This condition developed in a patient Although findings from a physical examination who had no obvious risks for lactic acidosis (eg, re­ on admission were essentially unremarkable, the nal failure). Several teaching points emerge from patient had complaints of generalized fatigue. She this case report that have important implications had no evidence of retinopathy. Her vital signs in­ for metformin use, including altered physiology cluded a heart rate of 100 beats per minute, blood secondary to patient comorbidity, and potential pressure 176170 mmHg, respiratory rate IS/min, for metformin-drug interactions. This case under­ and temperature 37.9°C. Her lungs were clear to http://www.jabfm.org/ scores important patient considerations related to auscultation. An electrocardiogram on admission metformin prescribing. confirmed atrial fibrillation with a ventricular rate of 100 beats per minute. Her admission laboratory Case Report values included sodium 138 mEqlL, potassium 4.4 A 76-year-old, 65-kg woman was admitted to the mEqlL, chloride 95 mEqlL, carbon dioxide 20 hospital for progressive fatigue, shortness of mEqlL, blood urea nitrogen 30 mg/dL, and crea­ breath, and persistendy elevated blood glucose lev­ tinine 0.9 mg/dL. The patient's blood glucose was on 1 October 2021 by guest. Protected copyright. els as documented by home monitoring. Her fam­ 288 mg/dL. Her total white cell count was ily reported that the patient experienced episodes 9?00/pL, and glycosylated hemoglobin was 11.4 of lethargy, somnolence, and progressive percent. during the preceding week. Her medical history She had no indications of cardiac (jugular ve­ nous distention, S3 gallop, marked peripheral edema) or pulmonary decompensation. Her chest Submitted, revised, 22 October 1997. From the Department of Family Medicine, Case Western radiograph was clear and had no evidence of con­ Reserve University School of Medicine and University Hos­ gestive heart failure or infiltrate. An was pitals of Cleveland (DTH), Cleveland, and the College of Pharmacy, Ohio Northern University (DTH, MFB), Ada; the 23, and because of the known association of met­ Department of Internal Medicine, Akron General Medical formin with lactic acidosis, a serum lactate con­ Center (MFB), Akron; and the Department of Internal Medi­ cine, Northeastern Ohio Universities College of Medicine centration was measured, which was elevated to (RDM), Rootstown, Ohio. Address reprint requests to Dar­ 4.1 mEqlL (normal 0.7 - 2.1 mEqlL). Serum and rell T. Hulisz, PharmD, University Hospitals of Cleveland, University Family Medicine Foundation, Bolwell Suite 1200, urine ketones were not elevated, and there was no 11100 Euclid Ave, Cleveland, OH 44202. proteinuria. Arterial blood gas was not analyzed,

Metformin-Associated Lactic Acidosis 233 J Am Board Fam Pract: first published as 10.3122/15572625-11-3-233 on 1 May 1998. Downloaded from so could not be confinned. ity and mitochondrial transport of reducing Metfonnin was recognized as the potential agents, and thus enhance anaerobic . 1 cause oflactic acidosis and was discontinued. The This subsequent shift to anaerobic metabolism, in presumed acidosis resolved as the anion gap nor­ the presence of reduced insulin, increases produc­ malized, and a second lactate measurement ob­ tion of precursors for the Krebs cycle.6 The inhi­ tained 24 hours later was 1.9 mEqlL. Because the bition of pyruvate dehydrogenase results in a de­ patient remained in atrial fibrillation, she was creased ability to channel those precursors into transferred to a telemetry unit and was given aerobic metabolism, which, in turn, results in in­ amiodarone. Her subsequent hospitalization was creased metabolism of pyruvate to lactate and in­ uneventful, and she was discharged home. creases the net production. Addition­ ally, increased glucose utilization in the small Discussion intestine caused by biguanide drugs could theoret­ Clinical Features ofLactic Acidosis ically increase portal vein lactate levels. Physicians prescribing metformin should be fa­ Phenformin, another biguanide introduced miliar with the clinical manifestation of lactic aci­ into the market around 1950, has a mechanism of dosis, particularly when prescribing the drug for action similar to that of metformin, but a much elderly patients or those with predisposing fac­ greater propensity to cause lactic acidosis.6 Be­ tors. Signs and symptoms vary in actual case re­ cause of the high incidence of lactic acidosis asso­ ports but tend to be nonspecific and include nau­ ciated with phenfonnin. it was removed 20 years sea, , altered consciousness, fatigue, ago from clinical use in this country. Compared abdominal pain, and thirst.1,3,5 Treatment of lactic with metformin, patients taking phenformin have acidosis is supportive. Metformin is readily re­ a 10 to 20 times greater risk of developing lactic moved by hemodialysis, which would be appro­ acidosis.4 The chemical structure of phenformin priate for severe, life-threatening acidosis. In the includes a long side chain, which enhances its case we described, the patient's symptoms in­ lipophilicity and results in a greater affinity for cluded progressive fatigue, lethargy, and confu­ binding to mitochondrial membranes, which sion, which prompted her hospital admission. On could account for its greater ability to inhibit aer­ laboratory analysis, our patient had a decreased obic metabolism than metformin.1 serum bicarbonate level, an elevated anion gap, Additionally, after phenformin was withdrawn and elevated lactate concentration in the absence from the marketplace. it was found that certain of other factors known to be associated with an patients (eg, about 10 percent of white patients) http://www.jabfm.org/ anion-gap metabolic acidosis. The laboratory ab­ have an inherited defect in hydroxylation of the nonnalities as well as the associated symptoms re­ drug.8 This defect could have resulted in phen­ solved promptly after discontinuation of met­ fonnin accumulation. Metformin, in contrast, is formin. Fortunately, this episode was mild, and no not metabolized. The overall average estimated pennanent sequelae developed. The omission of incidence of metformin-induced lactic acidosis is an arterial blood gas was an oversight, as it is an rare: 0.03 cases per 1000 patient-years.1,3,4 This on 1 October 2021 by guest. Protected copyright. important part of the work-up for lactic acidosis. statistic is also cited by the manufacturer of the drug and is based upon worldwide surveillance Biguanide Mechanism ofAction and Lactic data. The frequency with which lactic acidosis de­ Acidosis iJevelopment velops in patients with risk factors is unknown and Metformin is thought to work by a combination might be much higher. of different mechanisms that include increasing glucose transport into glucose-utilizing cells and Predisposition to Metformin-Associated Lactic Acidosis decreasing hepatic .1,4,6 Met­ Renal insufficiency results in lowered clearance of fonlun might have an important effect of increas­ both lactate and metformin, increasing the risk of ing glucose transport across the cell membrane in lactic acidosis.6 Additionally, any state that results skeletal muscle.7 in tissue hypoperfusion can lead to tissue , The exact mechanism by which metformin and because pyruvate is converted to lactate (the alter­ phenfonnin cause lactic acidosis is uncertain. native pathway in the Krebs cycle) during tissue Biguanides reduce pyruvate dehydrogenase activ- hypoxia. Thus, in addition to renal failure, the fol-

234 JABFP May-June 1998 Vol. 11 No.3 J Am Board Fam Pract: first published as 10.3122/15572625-11-3-233 on 1 May 1998. Downloaded from lowing comorbid conditions can contribute to lac­ cretion. Metformin is a protonated cation at phys­ tate accumulation: pulmonary disease, liver fail­ iologic pH.! It has been hypothesized that certain ure, cardiac impairment, states, severe de­ drugs, such as cimetidine, might compete with re­ hydration, and microvascular disease. It is worth nal tubular secretion of metformin at the site of noting that some of the conditions (eg, renal in­ organic cation transport.12 In one report cimeti­ sufficiency, microvascular disease, and dehydra­ dine decreased the renal clearance of metformin tion) develop in poorly controlled or advanced by 27 percent.12 Examples of other drugs known stages of diabetes. Some clinicians have suggested to compete with organic cations for renal tubular that metformin would also be inappropriate for excretion include digoxin, procainamide, quini­ someone with chronic obstructive airway disease, dine, ranitidine, iodinated contrast media, and ischemic heart disease, or severe infection, in amiloride. 1 The effects of these drugs on met­ which lactate production might be increased or formin renal excretion are unknown, because tissue perfusion decreased.4,9 studies have yet to be published. A recommenda­ In the case described, our patient had a normal tion to temporarily discontinue metformin before serum creatinine level (0.9 mg/dL). Using the cor­ and after the administration of contrast media em­ rected Cockcroft-Gault method of calculating cre­ phasizes the need for caution when concurrent ad­ atinine c1earance,lO,ll this patient's estimated glo­ ministration of competing substances is required. merular filtration rate was approximately 60 In the case presented, the only medication that mL/min. The manufacturer of metformin states could theoretically compete for secretion was that metformin should not be used in patients with digoxin. serum creatinine values greater than 1.5 mg/dL for The patient described in this case was taking men, 1.4 mg/dL for women. These values, how­ furosemide, which reportedly can alter metformin ever, are less reliable as indicators of renal function pharmacokinetics.1 Metformin product informa­ in elderly or chronically debilitated patients. Met­ tion includes an unpublished report from a single­ formin should be prescribed with great caution in dose, normal volunteer study showing the ability any patient with moderate renal impairment (esti­ of furosemide to increase the maximum serum mated creatinine clearance 30- 60 mUmiu). concentration and area under the curve by 22 per­ Of particular importance was this patient's his­ cent and 15 percent, respectively. tory of severe obstructive pulmonary disease, In addition to concurrent medications, it might which was noted in her previous medical records. also be wise to consider a patient's use of alcoholic

Although her clinical condition was not consis­ beverages before prescribing metformin. Ethanol http://www.jabfm.org/ tent with an exacerbation of this disease process, might reduce both the conversion of lactate to her chronic pulmonary impairment might be glucose and hepatic extraction oflactate.3 Conse­ considered a risk factor for lactic acidosis. In ad­ quently, heavy consumption (eg, binge dition, although this patient did not have symp­ drinking) sufficient to cause hepatic impairment toms of acute decompensated heart failure, she could represent a relative contraindication to met­

did have a history of left ventricular impairment formin use. The patient described herein did not on 1 October 2021 by guest. Protected copyright. (ejection fraction 35 percent) and was in atrial fib­ have a history of ethanol use. rillation, both of which could have predisposed her to lactic acidosis by means of decreased pe­ Patient Selection Issues/or Metformin Use ripheral tissue perfusion. It is imperative to evaluate a patient's renal func­ We believe this case illustrates the potential for tion before prescribing metformin. Equally impor­ multiple risk factors to affect the overall risk of lac­ tant, however, is the consideration of any coinorbid tic acidosis associated with metformin use and em­ state associated with tissue hypoxia. Recent re­ phasizes the need for more than a cursory review of ports13,14 have confinned that metformin-induced associated conditions before implementing therapy. lactic acidosis is not necessarily due to metformin accumulation (as might be expected in renal fail­ Drug Interactions Contributing ure), and that the clinical severity and prognosis of to Metformin-Associated Lactic Acidosis acidosis is more dependent on associated patho­ Metformin does not undergo hepatic metabolism; logic conditions. In retrospect, one could conclude the main route of elimination is renal tubular se- that our patient was not an ideal candidate for

Metformin-Associated Lactic Acidosis 235 J Am Board Fam Pract: first published as 10.3122/15572625-11-3-233 on 1 May 1998. Downloaded from high-dose metfonnin therapy given her history of References congestive heart disease, pulmonary disease, coro­ 1. Lee AJ. Metformin in noninsulin-dependent dia­ nary artery disease, and atrial fibrillation. betes mellitus. Pharmacotherapy 1996;16:327-51. The ideal candidate for metfonnin monother...: 2. Schneider J, Erren T, Zofel P, Kaffarnik H. Met­ formin-induced changes in serum lipids, lipopro­ apy is an obese patient with NIDDM who has no teins, and apoproteins in non-insulin-dependent di­ serious comorbidity and who has failed diet and abetes mellitus. Atherosclerosis 1990;82:97-103. exercise therapy. 1 Concurrent hyperlipidemia 3. Gan SC, Barr J, Arieff AI, Pearl RG. Biguanide-as­ with poorly controlled NIDDM might also favor sociated lactic acidosis. Case report and review of the use of metformin. Further studies are needed the literature. Arch Intern Med 1992;152:2333-6. to establish the role of metforrnin for nonobese 4. Bailey C], Turner RC. Metformin. N EnglJ Med 1996;334:574-9. NIDDM patients and for those with serious co­ 5. GowardmanJR. Fatal metformin induced lactic aci­ morbidity before prescribing the drug as a first­ dosis: case report. N Z MedJ 1995;108:230-1. line therapy. 6. McGuinness ME, Talbert RL. Phenformin-induced Patients should be instructed to recognize po­ lactic acidosis: a forgotten adverse drug reaction. tential symptoms of lactic acidosis, and medica­ Ann Pharmacother 1993;27:1183-7. tions that can interfere with secretion should be 7. Klip A, Leiter LA. Cellular mechanism of action of avoided. Metforrnin should be temporarily with­ metformin. Diabetes Care 1990;13:696-704. held in patients receiving iodinated contrast 8. Oates NS, Shah RR, Idle JR, Smith RL. Influence of oxidation polymorphism on phenformin kinetics media or any other condition in which an acute and dynamics. Clin Pharmacol Ther 1983;34:827- decline in renal function might occur (eg, ag­ 34. gressive diuresis, excessive fluid loss from gas­ 9. Hutchison SM, Catterall JR. Metformin and lactic troenteritis, surgery, etc). It is possible to learn acidosis - a reminder. Br J Clin Pract 1987;41:673-4. which patients are at particular risk for develop­ 10. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16: ing metformin-associated lactic acidosis byap­ 31-41. plying general screening guidelines published in 11. D'Angio R, Platt DR, Gannon R. Creatinine clear­ the American Diabetes Association clinical prac­ ance: corrected versus uncorrected. Drug Intell Clin tice recornrnendations.1S•16 Pharm 1988;22:32-3. 12. Somogyi A, Stockley C, KealJ, Rolan P, Bochner F. Conclusion ~edu~ti?n ~f metformin renal tubular secretion by clmendme m man. Br J Clin PharmacoI1987;23: Metfonnin-associated lactic acidosis is a rare, pre­ 545-51. ventable, but life-threatening, adverse event. Pri­ http://www.jabfm.org/ 13. L~lau JD, La~roix C, Compagnon P, de Cagny B, mary care physicians should profile a patient's risk Rigaud JP, Blelchner G, et al. Role of metformin ac­ for developing this side effect by considering co­ cumulation in metformin-associated lactic acidosis. morbid diseases, concurrent diabetic complica­ Diabetes Care 1995;18:779-84. tions, concomitant medical therapy, and antici­ 14. LalauJD, Lacroix C, De Cagny B, Fournier A. Met­ formin-associated lactic acidosis in diabetic patients pated diagnostic and surgical procedures. Patients with acute renal failure. A critical analysis of its who are taking metfonnin should be instructed to pathogenesis and prognosis. Nephrol Dial Trans­ on 1 October 2021 by guest. Protected copyright. provide that information readily; the use of med­ plant 1994;9(SuppI4}:126-9. ical alert tags indicating metfonnin use might help 15. American Diabetes Association. Clinical practice reduce the incidence of lactic acidosis. Increased recommendations - 1995. Diabetes Care 1995;18 (Suppll}:1-96. drug surveillance is necessary to define further the 16. Kerr CPo Improving outcomes in diabetes: a review risk of developing lactic acidosis secondary to of the outpatient care of NIDDM patients. J Fam metfonnin use. Pract 1995;40:63-75.

236 JABFP May-June 1998 Vol. 11 No.3