Medical case narrative
A 14-year-old whose thyroid blood tests were normal — the antibodies weren't
What happened
A 14-year-old girl in Korea, an above-average eighth-grade student (unnamed in the published report), was admitted with an acute illness nobody could explain: three generalised seizures, each lasting more than ten minutes, with drowsiness — preceded by three days of fever up to 38 °C, muscle weakness, joint pains, and a rash on both hands. On day 4 of the admission she developed acute psychosis: visual hallucinations, personality change, mood swings between depressive and manic states, poor concentration, insomnia, agitation. Printed honestly, as both reviewers required: her psychiatric phase lasted days, inside an acute brain-illness admission — there was no standing psychiatric label that labs later overturned. What was missed first was an autoimmune cause behind seizures no medication could stop.
What it looked like
Seizures that did not respond to three anti-seizure drugs (valproic acid, phenytoin, topiramate), an EEG showing generalised slowing, and then a psychosis arriving on top of an unexplained encephalopathy. Infection testing was negative, brain MRI and spinal fluid were normal, and the usual autoimmune markers were negative.
What the labs found
Because the cause was unclear, autoimmune status was checked — and the thyroid antibodies were high even though her thyroid function tests were entirely normal: TSH 4.78 µIU/mL, free T4 1.33 ng/dL, T3 89 ng/dL, all in range, alongside anti-thyroid peroxidase (anti-TPO) antibody of 280 IU/mL (normal under 100) and antithyroglobulin antibody of 138 IU/mL (normal under 100). A TSH alone would have missed this; the antibodies were the test. The diagnosis made was Hashimoto's encephalopathy — and its contested status travels with this entry, because the authors state it themselves: it is a diagnosis of exclusion, the literature shows no established causative link between thyroid autoimmunity and this encephalitis, and antibody levels do not track clinical status. The diagnosis here rests on exclusion plus response to steroids.
What changed
Intravenous methylprednisolone for three days, then oral prednisone. Her neuropsychiatric symptoms resolved and she was discharged on day 14; at one month she had been symptom-free and her antibody levels had normalised. But the headline of this entry includes the relapse, as both reviewers required: her seizures recurred while the steroid was being tapered. A partial, relapsing rescue — not a clean cure. No antipsychotics were used, which makes the steroid attribution cleaner; she was on three anti-seizure drugs throughout, which is also part of the record.
The test that told the story
Anti-thyroid antibody testing (anti-TPO and antithyroglobulin) — ordered when the standard workup came back empty, in a girl whose ordinary thyroid tests were normal.
Sources: Lee et al. 2012 (primary case report, all values; described by its authors as the first paediatric Hashimoto's encephalopathy report in Korea); Menon et al. 2017 systematic review of psychiatric presentations heralding Hashimoto's encephalopathy (46 cases), which independently extracted and coded this case.