Acetone poisoning - a diagnostic dilemma

EDITOR: Acetone intoxication is a rarely reported substance involved in poisoning [1-3]. A literature search indicated that, in all cases reported, a clear history of acetone ingestion was obtained, thus facilitating management. We present a case where this information was not immediately available, and subsequent difficulties presented in the diagnosis and management. Case report A 47-yr-old female presented to A&E with respiratory distress. She had a past history of alcohol abuse and deliberate self-harm and was receiving citalopram for depression. She had no other medical problems of note. On physical examination, she had a slim build and was drowsy but rousable with a Glasgow coma scale of 12/15 and no focal neurological signs. She had a sinus tachycardia of 120 min−1 and a blood pressure of 180/120 mmHg. She had a respiratory rate of 40 min−1 with a fruity odour on her breath and SPO2 of 94% on air. Chest was clinically and radiologically clear with no signs of cardiac, hepatic or renal failure. She was hypothermic (35°C). Arterial blood gas analysis showed pH 6.9, PaCO2 1.4 kPa, PaO2 18.5 kPa and base excess −28.8 mmol L−1. There was negative toxicology screen involving ethanol, salicylate, paracetamol and barbiturates. The blood glucose was 15 mmol L−1 and a peripheral blood smear showed neutrophil leucocytosis and white cell count of 26 × 109 L−1. The calculated serum osmolarity was 288 but the measured osmolality was 310, giving an osmolar gap of 22. While the anion gap was 38, the blood lactate level was normal and urinanalysis was mildly positive for ketones. She did not require ventilation, but a bolus of intravenous (i.v.) sodium bicarbonate was administered, and she was transferred to intensive treatment unit (ITU) for further supportive care and management of her profound metabolic acidosis by haemofiltration. The National Toxicology Centre was contacted and suggested methanol or ethylene glycol, but not acetone as a potential aetiological agent. Treatment with i.v. ethanol was recommended. Gas chromatography of the patient's urine and blood confirmed the presence of acetone. The acidosis improved within hours of haemofiltration and i.v. ethanol was not administered. The osmolar gap, anion gap and pH were completely corrected by 18 h. She was transferred out of ITU after 30 h and subsequently sent home after a few days. Discussion Alcohol misuse is associated with concurrent intentional ingestion of acetone as well as methanol, ethylene glycol and acetonitrile (a common component of cosmetic nail varnish remover) [4]. In all of these, the presenting laboratory features of acidosis, with a high anion and osmolar gap, are similar and non-specific. Though acetone alone is a relatively non-toxic compound, methanol and ethylene glycol are converted to the highly toxic breakdown product formaldehyde, and acetonitrile is metabolized to cyanide. The toxicity of acetone is identical to ethanol, although the anaesthetic potency of acetone is much greater [5]. The main route of elimination of acetone is through the lungs. Both acetone and ethanol produce respiratory failure due to central nervous system depression in toxic doses [5]. The elimination of both at high doses follows zero order kinetics. The few reported signs of acetone poisoning, which may help to clinically differentiate ethanol from acetone ingestion, include red and swollen pharynx and soft palate erosions [5,6], but were not seen in this patient. It is likely that the acetone was diluted with another liquid and probably taken over a period of time. Acetone is metabolized to glucose [5,6], and the high blood sugar levels and ketonuria were in keeping with this. Diabetic keto-acidosis was considered as a possible diagnosis but was not in keeping with the overall clinical picture. Measurement of serum osmolality and osmolar gap have limited sensitivity and specificity. Gas chromatography, if available, offers the possibility of a definitive diagnosis. This case illustrates several interesting points: acetone containing nail polish removers are readily found at home, and acetone ingestion should be kept in mind when considering possible aetiological agents. Unless there is a clear history, it can be very difficult to distinguish clinically between all these agents. Finally, the recommended treatment of methanol, as well as ethylene glycol poisoning, is 100% ethanol. Ethanol administration will exacerbate acetone toxicity. It is important to consider acetone in the differential diagnosis of unexpected metabolic acidosis, especially in the presence of hyperglycaemia.