Start here when vomiting and rapidly worsening confusion are being explained by infection or neurology alone. Severe non-cirrhotic hyperammonaemia can have unremarkable liver tests, so plasma ammonia, protein cessation, high-calorie glucose and early metabolic or critical care advice change the case.

Developmental regression, exaggerated startle, seizures or ataxia can be the start of a metabolic neurodegenerative story. The GM2 discussion is specialist, but it gives a clear reason to escalate progressive neurological symptoms despite nonspecific MRI changes.

Recurrent abdominal pain with headache, leg pain, palpitations or menstrual association should make the porphyria question hard to ignore. This is worth opening when repeated attendances have been split across specialties, because urinary porphobilinogen during symptoms is the concrete next step.

Severe motor delay with hypotonia, fluctuating dystonia, oculogyric crises, and a normal MRI should prompt a look at tyrosine hydroxylase deficiency. Niche, but worth keeping in mind because the disorder is treatable and early dyskinesia on L-dopa is not a reason to stop.

The clue is an isolated urinary glyceric acid rise that has not been split into D and L forms. Open this for clear teaching on enantiomer-specific testing, mitochondrial GLYCTK function, and why a genetic finding should not be forced to explain a broader developmental picture.

Delayed mitochondrial disease diagnosis is usually a recognition problem, not a testing problem. When seizures, stroke-like episodes, hypotonia, developmental delay and constipation accumulate across prior notes, revisit the timeline and raise neurology or genetics suspicion earlier.

RADs on EEG should widen, not close, the differential in refractory epilepsy. When mixed seizure types sit beside this pattern, add metabolic testing such as urinary organic acids and consider SDHA-related complex II deficiency rather than assuming POLG disease alone.

Rare, but not just academic. The episode explains why triheptanoin reduces metabolic crises yet leaves retinopathy and neuropathy behind, then makes the case for elamipretide through cardiolipin remodelling and biomarker thinking.