What causes an elevated anion gap outside of DKA?

What literature is there on other causes outside of DKA, such as seizures, elevating anion gap?

Comment by InpharmD Researcher

High anion gap metabolic acidosis (HAGMA) occurs when unmeasured anions accumulate due to metabolic disturbances, toxic ingestions, or impaired renal excretion. Common causes are summarized by the GOLDMARK mnemonic: Glycols, Oxoproline (from chronic acetaminophen use), L-lactic acid, D-lactic acid, Methanol, Aspirin (salicylate), Renal failure, and Ketoacidosis. Less frequent causes include D-lactic acidosis in patients with short bowel syndrome and 5-oxoproline accumulation in malnourished patients using acetaminophen. Elevated anion gap has also been reported in association with generalized seizures, although the evidence supporting its use for predicting seizure is limited. Accurate interpretation requires consideration of mixed acid-base disorders, potential laboratory interferences, and correlation with clinical context and blood gas measurements. Of note, an additional search did not identify any new evidence that would change the overall understanding of causes or clinical interpretation of HAGMA.

A targeted literature search was conducted in PubMed and Google Scholar for narrative reviews and case series evaluating causes of elevated anion gap metabolic acidosis outside of diabetic ketoacidosis, including seizure-related acid–base disturbances. The search identified reviews describing established diagnostic mnemonics and reports associating seizures with transient anion gap elevations, while noting limited supporting data.

Background

Several review articles highlight the evaluation and causes of high anion gap metabolic acidosis (HAGMA), which arises when unmeasured anions accumulate due to metabolic processes, toxic ingestions, or impaired renal excretion. The most common causes are summarized by the mnemonic GOLDMARK: Glycols (ethylene, propylene, and diethylene glycol), 5-oxoproline (associated with chronic acetaminophen use), L-lactic acid, D-lactic acid, Methanol, Aspirin (salicylate), Renal failure, and Ketoacidosis. Less common but clinically important scenarios include chronic acetaminophen ingestion, particularly in malnourished patients, which can cause accumulation of 5-oxoproline (pyroglutamic acid) even at therapeutic or subtherapeutic doses, producing HAGMA without acute hepatotoxicity. [1], [2], [3], [4]

D-lactic acidosis may occur in patients with short bowel syndrome or other malabsorptive disorders, where bacterial fermentation of unabsorbed carbohydrates generates D-lactate that is metabolized slowly, leading to HAGMA and neurological symptoms such as confusion, ataxia, and slurred speech. Of note, elevated anion gap metabolic acidosis has been associated with generalized seizures, with case studies showing higher anion gaps and lower serum bicarbonate in seizure patients; the Denver Seizure Score uses these values to predict seizure likelihood in cases of unwitnessed loss of consciousness. However, data supporting this approach are limited, and anion gap changes alone should not be used to definitively diagnose seizure without clinical correlation. [1], [2], [3], [4]

HAGMA can also be part of mixed acid-base disorders. Salicylate toxicity, for example, produces both HAGMA and respiratory alkalosis through stimulation of the respiratory center and metabolic effects on oxidative phosphorylation, glycolysis, lipolysis, and hepatic ketogenesis. Stepwise evaluation of the anion gap, including calculating the change in anion gap relative to bicarbonate (Δ[AG]/Δ[HCO₃⁻]), helps identify additional disturbances, while rare contributors such as elevated phosphate, monoclonal paraproteins, or laboratory artifacts should also be considered. Careful interpretation requires attention to patient history, clinical context, and limitations of laboratory measurements. [1], [2], [3], [4]

A 2023 narrative review describes a structured approach to evaluating high anion gap metabolic acidosis and summarizes a broad range of etiologies beyond diabetic ketoacidosis using established and expanded diagnostic mnemonics. The authors reference the traditional MUDPILES mnemonic, which encompasses Methanol, Uremia, Diabetic ketoacidosis, Propylene glycol/Paraldehyde, Iron or Isoniazid, Lactic acidosis, Ethylene glycol, and Salicylates, as well as the more contemporary GOLDMARK mnemonic, which includes Glycols (ethylene and propylene), Oxoproline, L-lactate, D-lactate, Methanol, Aspirin, Renal failure, and Ketoacidosis, to summarize common etiologies of an elevated anion gap. The review emphasizes that these mnemonics, while useful, do not capture all potential causes and therefore introduces an expanded differential (CUTE DIMPLES) that incorporates additional toxic, metabolic, and medication-related etiologies, including citrate excess from anticoagulation, sodium thiosulfate therapy, cyanide exposure, toluene, iron, isoniazid, paraldehyde, medication-associated lactic acidosis (e.g., metformin, linezolid, propofol), starvation or ethanol-related ketoacidosis, and euglycemic diabetic ketoacidosis associated with sodium-glucose cotransporter-2 inhibitors. The authors also describe pseudohypobicarbonatemia due to laboratory interference from substances such as severe hypertriglyceridemia or paraproteins as a condition that can falsely elevate the calculated anion gap without representing true metabolic acidosis, indicating the importance of correlating serum chemistry results with blood gas analysis when evaluating an elevated anion gap. [5]

Literature Review

A search of the published medical literature revealed 1 study investigating the researchable question:

What literature is there on other causes outside of DKA, such as seizures, elevating anion gap?

Level of evidence

D - Case reports or unreliable data Read more→

Please see Table 1 for your response.

How I Evaluate a High Anion Gap Metabolic Acidosis
Design	Case report
Case presentation	A 36-year-old woman with classical Hodgkin lymphoma was admitted with a 3-week history of nausea, vomiting, diarrhea, and abdominal pain, occurring after initiation of chemotherapy 5 weeks prior with brentuximab vedotin, doxorubicin, vinblastine, and dacarbazine; home medications before lymphoma diagnosis were estradiol and spironolactone. On admission, vital signs included blood pressure 131/84 mm Hg, heart rate 131 beats/min, and temperature 36.1°C, with physical examination notable only for general discomfort. Initial serum chemistry demonstrated sodium 132 mEq/L, potassium 4.6 mEq/L, chloride 99 mEq/L, total CO₂ 15 mEq/L, BUN 14 mg/dl, creatinine 0.6 mg/dl, glucose 169 mg/dl, albumin 3.5 g/dl, and an anion gap of 18 mEq/L (reported as 10 mEq/L three months prior), with serum lactate 4.7 mmol/L and urinalysis showing 4+ ketonuria; computed tomography of the abdomen and pelvis was normal. The following day, total CO₂ decreased to 13 mEq/L with anion gap 17 mEq/L, and after administration of 1 L isotonic bicarbonate and 2600 mg oral sodium bicarbonate, total CO₂ decreased further to 6 mEq/L with anion gap increasing to 23 mEq/L, prompting additional evaluation. Arterial blood gas showed pH 7.39, pCO₂ 32 mm Hg, and calculated bicarbonate 19 mEq/L, with lactate on blood gas 2.9 mmol/L; salicylate, methanol, ethanol, and acetaminophen testing was negative, and calculated osmolal gap was 2 mOsm/L. Nephrology consultation noted the discordance between measured total CO₂ and calculated bicarbonate with normal arterial pH, and subsequent testing revealed triglycerides 4546 mg/dl; repeat abdominal/pelvic computed tomography demonstrated fat stranding around the pancreatic head consistent with pancreatitis attributed to hypertriglyceridemia. The case conclusion was pseudohypobicarbonatemia due to hypertriglyceridemia with resolving lactic acidosis and ketoacidosis, treated with intravenous fluids, opioids, fenofibrate, and fish oil with discontinuation of estradiol, after which triglycerides decreased, total CO₂ increased to 23 mEq/L, and the anion gap normalized without further bicarbonate administration.
Study Author Conclusions	We concluded that, at the time of consultation, the patient had pseudohypobicarbonatemia from hypertriglyceridemia and resolving lactic acidosis and ketoacidosis. The etiology of the hypertriglyceridemia was believed to be multifactorial. We suspect that ketoacidosis superimposed on a background of exposure to estradiol and brentuximab, which is associated with hypertriglyceridemia, triggered severe hypertriglyceridemia. The patient was treated with intravenous fluids, opioids, fenofibrate, and fish oil, and the estradiol was discontinued. The serum triglyceride level decreased over the next several days, the total CO₂ increased to 23 mEq/L, and the anion gap returned to normal without administration of bicarbonate.

References:
[1] Azim A, Hu B, Gilligan S, et al. How I Evaluate a High Anion Gap Metabolic Acidosis. Clin J Am Soc Nephrol. 2024;19(4):525-527. doi:10.2215/CJN.0000000000000381