An unusual cause of interference in a salicylate assay caused by mitochondrial acetoacetyl-CoA thiolase deficiency

Introduction

Mitochondrial acetoacetyl-CoA thiolase deficiency (OMIM# 203750), also known as alpha-methylacetoacetic aciduria; 2-methyl-3-hydroxybutyric acidaemia; mitochondrial acetoacetyl-CoA thiolase deficiency; T2 deficiency; 3-oxothiolase deficiency and 3-ketothiolase deficiency is a rare autosomal recessive condition associated with mutations on the acetyl-CoA acetyltransferase-1 gene on chromosome 11q22.3-q23.1 and was first described in 1972. ¹ There is no single common mutation and a variety of different mutations have been described in association with the condition. ² To date, 50–60 cases have been reported worldwide. ³ Presentation tends to occur in childhood (before the age of 4 years). Most cases present with their first ketoacidotic event between five months and two years of age. ² The disorder is characterized by intermittent episodes of severe vomiting and diarrhoea, accompanied by a severe metabolic acidosis and ketosis. In addition, there is often haematemesis. In severe cases, there may be coma and other neurological sequelae or even death. Other symptoms and complications may include cardiomyopathy, prolonged QT interval, poor weight gain, renal failure and short stature. ^4,5 If the patient escapes neurological damage, they are asymptomatic between acute episodes. The clinical expression of beta-ketothiolase deficiency is variable, but these acute ‘decompensation’ episodes tend to be most severe in infancy and their frequency falls with the age – they are uncommon after the age of 10. ⁶

A case is reported of a young child who presented with diarrhoea and vomiting, who was subsequently found to have mitochondrial acetoacetyl-CoA thiolase deficiency. Initial investigations revealed detectable salicylate concentrations of 36 mg/L despite the parents denying its administration. This caused a degree of diagnostic uncertainty until it was confirmed that the high concentrations of acetoacetate which are found in this condition can cause a false-positive reaction in the Trinder method for salicylate. ⁷

Case report

An 18-month-old boy presented to the Accident and Emergency unit (A&E) via his General Practitioner with a short two-day history of acute diarrhoea and vomiting. On examination, he was dehydrated with sunken eyes and decreased skin turgor. He was conscious and responsive to stimuli although crying and clearly distressed. He was noted to have rapid Kussmaul breathing and this had been highlighted previously by the patient's mother as a particular concern.

The patient was the second child born to non-consanguineous parents. There was no previous family history of note. The elder daughter had had some diarrhoea and vomiting in the preceding days and so the provisional diagnosis was of viral gastroenteritis.

Initial investigations revealed a marked metabolic acidosis (Table 1) with a pH of 6.94. He smelt strongly of ketones, and the suspicion of ketoacidosis was confirmed by the finding of Multistix ++++ for ketones in his urine. Diabetes was excluded by the finding of a normal plasma glucose (6.1 mmol/L) and several point of care glucose results within the reference range.

Treatment was commenced with an intravenous saline bolus and continuous saline infusion was started. He was also given half-bicarbonate correction and broad-spectrum antibiotics were commenced as a precaution.

Repeat arterial blood gases 30 min after fluid treatment started showed that the acidosis had improved slightly (arterial blood gases showed pH 6.94) and the patient was admitted into the paediatric ward. Over the course of the next 20 h his vomiting and diarrhoea settled, but he remained acidotic with persistent Kussmaul respiration. Despite large volumes of bicarbonate correction, he remained acidotic with blood pH 7.2 and urine samples remained strongly positive for ketones.

Although salicylate was withdrawn from all paediatric medications over 20 years ago due to concerns over its association with Reye's syndrome, the clinical features were such that the medical team felt that salicylate toxicity should be ruled out as a possible cause of the patient's symptoms. Salicylate was indeed detected in the plasma, albeit at a low concentration of 36 mg/L (adult therapeutic range for analgesia is quoted as 50–150 mg/L [Olympus kit insert]). The parents were adamant that no salicylate-containing medications had been given and an alternative explanation was sought from the laboratory. Discussion with the Consultant Clinical Scientist subsequently alerted the medical team to the possibility of high acetoacetate concentrations causing assay interference.

A urine specimen for organic acids had already been collected and following discussions with the paediatric tertiary referral centre, the patient was transferred by ambulance for further investigation and treatment. Several days later the results of urine organic acid analysis became available and this confirmed a diagnosis of mitochondrial acetoacetyl- CoA thiolase deficiency. This rare autosomal recessive disorder is characterized by an intermittent episodes of severe metabolic acidosis and ketosis accompanied by vomiting, diarrhoea and coma. In severe cases it can be fatal.

Discussion

This patient presented with marked metabolic acidosis after a short, two-day illness with diarrhoea and vomiting. Such ketoacidosis is commonly found in young children due to intercurrent illness and prolonged fasting, as falling glucose and insulin concentrations result in activation of fatty acid oxidation. The patient, however, failed to respond to fluids and remained acidotic despite large volumes of bicarbonate being given. Hypoglycaemia was a possible cause for the ketoacidosis, however both the laboratory plasma glucose and point-of-care results were consistently within the reference range. The patient had been brought to A&E directly by the parents and so out-of-hospital administration of glucose was felt to be unlikely.

Salicylate overdose is normally associated with respiratory alkalosis and hyperventilation due to stimulation of the central nervous system. However, in children the period of respiratory alkalosis can be very brief and in such circumstances the initial presentation is often with metabolic acidosis. Symptoms of salicylate toxicity include tinnitus, hyperthermia and nausea and vomiting and therefore it is understandable that a request for a standard salicylate and paracetamol screen was made. There was an initial period of diagnostic uncertainty when a routine screen indicated the possible presence of salicylate; however, the persistent ketosis strongly suggested a metabolic cause for the patient's symptoms and an alternative cause for the detectable salicylate was sought.

Paediatric metabolic acidosis is an important presenting clinical feature and even if identified as a keto- or lactic acidosis it is important that a sample for urinary organic acids be collected as soon as possible after admission in order to rule out a possible inborn error of metabolism.

The high concentrations of acetoacetate associated with this patient's condition were responsible for the interference in the salicylate assay (Trinder method) and a false-positive result was generated. Mitochondrial acetoacetyl-CoA thiolase deficiency has been reported previously as clinically mimicking salicylate toxicity. ^8,9 High concentrations of acetoacetate may be associated with a number of conditions including diabetic ketoacidosis and ketoacidosis, resulting from intercurrent illness and low food intake in young children. In addition, there are a wide range of individually rare organic acidaemias and other inborn errors of metabolism in which acetoacetate concentrations may be increased. While mitochondrial acetoacetyl-CoA thiolase deficiency is very rare, this case serves as a reminder to be aware of possible sources of false-positives with this salicylate assay.