Paroxysmal exercise-induced dystonia associated with hypoglycaemia induced by an insulinoma.

Sirs: Paroxysmal exercise-induced dystonia (PED) is a rare movement disorder that belongs to the family of paroxysmal dyskinesias [1, 2]. We report a patient who had PED associated with an insulinoma. The patient was a 28 year old male Chinese, previously well, who sustained a mild closed head injury and a diagnosis of concussion. CT of the head was normal. Two months later, he complained of painless dystonia of both feet and toes while jogging, which consistently occurred after 20–30 minutes, always starting with the right side. The episodes were stereotyped, lasted 15–30 minutes, and worsened with continued exertion, forcing him to stop jogging. The upper limbs, head and trunk were spared. These episodes resolved spontaneously after rest. His wife also noticed episodes of decreased responsiveness lasting 1–2 minutes. These often occurred in the early morning before breakfast, and sometimes before dinner. There were no generalised seizures, lip smacking or automatisms. These episodes often resolved after food or sweet drinks. He also had increased appetite, and gained 5 kilograms in 6 months. Neurological and general examination were normal. Complete blood counts, electrolytes, calcium, magnesium, caeruloplasmin and 2 electroencephalograms were normal. Brain MRI done 3 months and 7 months after the head injury showed a focus of diffuse axonal injury in the left globus pallidus (Fig. 1). He was fasted overnight and exercised on a treadmill. Pre-exercise tests showed serum glucose of 1.8 mmol/L. Concurrent plasma insulin (7.6 mU/L, normal 0–22.7 mU/L) and C-peptide levels (1.7 μg/L, normal 0.1–3.2 μg/L) were inappropriately raised for the degree of hypoglycaemia. He developed dystonia of his feet after 20 minutes which worsened with continued exertion. After stopping the exercise, the dystonia improved over 30 minutes, and resolved after a chocolate drink. Serum glucose after the drink was 2.1 mmol/L. He was found on abdominal imaging to have a 1 centimetre lesion in the pancreatic tail which was excised; histology confirmed an insulinoma. Two months post-operatively, he was able to jog for 45 minutes without dystonia with a fasting glucose of 5.3 mmol/L, and has remained completely symptom-free since. Paroxysmal exercise-induced dystonia is a form of paroxysmal dyskinesia that may be familial or sporadic. Attacks are precipitated by prolonged exercise, often running or walking, and feet are involved in 75 % of cases [3]. A channelopathy is suspected [4]. Optimal treatment is uncertain. Hypoglycaemia usually manifests neurologically as confusion, coma or seizures [5]; paroxysmal dyskinesias are uncommon [6–8]. Insulinomas have rarely been described to cause paroxysmal dyskinesias [8]. There has only been one previous report of PED associated with an insulinoma which resolved after removal of the insulinoma [9]. Head injury has been associated with the development of focal dystonia or hemidystonia, with lesions seen in the contralateral basal ganglia or thalamus [10, 11]. Mild head injury can cause subcortical lesions seen only on MRI, due to diffuse axonal injury [12]. Paroxysmal hemidystonia resulting from head injury has previously been reported [13, 14] due to abnormalities in the contralateral basal ganglia. Postulated mechanisms include aberrant neuronal sprouting and ephaptic transmission [11]. Insulinomas are slow-growing tumours; it is likely that the insulinoma was present before head injury, yet his PED only manifested after the injury. Vincent described a patient with a lentiform nucleus vascular malformation who developed contralateral hemichoreoathetosis during hyperglycaemia, which resolved with euglycaemia [15]. In our patient, the dual insults of hypoglycaemia (from his insulinoma), and his globus pallidus lesion (from his head injury) possibly combined to cause his PED. The structural lesion may have increased vulnerability to his pre-existing chronic hypoglycaemia, resulting in PED. In conclusion, hypoglycaemia should be excluded in all patients LETTER TO THE EDITORS