Case Presentation: A 44-year-old male hospitalized for an extended period with severe diarrhea and polyarthritis developed worsening anion gap metabolic acidosis (AGMA) over the course of 10 days. The patient was being treated for septic arthritis in the background of polyarticular gout with vancomycin, ceftriaxone, colchicine, indomethacin, and acetaminophen. Acetaminophen was dosed as needed on hospital days 0-10 (mean 1530 mg/day), then scheduled at 1000 mg every 6 hours (4000 mg/day) starting on hospital day 11. The anion gap (AG) became elevated to 15 meq/L on hospital day 8; bicarbonate (HCO3-) level at this time was 14 meq/L (Figure). Glucose remained normal throughout the stay, but the AG and HCO3- continued to worsen (Figure). As the AG worsened, daily monitoring revealed decreased oral intake during the hospital stay. On hospital day 15, Beta-hydroxybutyrate (BHB) level was 0.2 mmol/L (normal); repeat level taken on hospital day 16 was 2.2 mmol/L. Acetaminophen was discontinued on hospital day 16. On hospital day 17, the patient developed altered mental status; labs showed AG of 28 meq/L, HCO3- of 5 meq/L (Figure), BHB of 8.7 mmol/L, and venous blood gas (VBG) pH of 7.18. A urine organic acid screen showed markedly elevated pyroglutamic acid (5-oxoproline), 3-OH-butyric acid, and acetoacetic acid, and moderately elevated lactic acid. The patient was treated with a bolus of lactated ringers followed by 5% dextrose in water, sodium bicarbonate, acetylcysteine, and thiamine. Arterial blood gas after treatment had begun showed a pH of 7.38. The patient recovered and laboratory values normalized within 24 hours. Upon stabilization on hospital day 18, VBG showed a pH of 7.48, HCO3- was 22 meq/L, and AG was 11 meq/L (Figure).
Discussion: We present this case to highlight the importance of considering a broad differential in cases of AGMA. The GOLD MARK mnemonic is helpful for a review of common and uncommon causes of AGMA. This patient had an AGMA due to chronic acetaminophen use causing 5-oxoproline acidosis, compounded by a starvation ketoacidosis and mild lactic acidosis. In this case, we would first like to highlight starvation ketoacidosis as an important cause of AGMA in the hospital; and the importance of continuing to consider this as a cause of AGMA as the clinical course changes. Secondly, we would like to highlight 5-oxoproline acidosis as a rare but important cause of AGMA. Importantly, 5-oxoproline acidosis occurs not secondary to markedly elevated acetaminophen levels, but in a patient with normal acetaminophen levels in whom appropriate dose chronic use causes high 5-oxoproline levels.
Conclusions: A broad differential should be considered for cases of AGMA in the hospital, including less common causes such as starvation ketoacidosis and 5-oxoproline acidosis.
