Nucpaedia
Nucpaedia
Neurology · Epilepsy

Epilepsy Imaging (SPECT & PET)

Snapshot

In pre-surgical epilepsy work-up, functional imaging localises the epileptogenic zone when MRI is normal or discordant. Interictal FDG PET shows focal hypometabolism (well established in temporal-lobe epilepsy), while ictal perfusion SPECT captures seizure-onset hyperperfusion; co-registering ictal–interictal SPECT with MRI (SISCOM) improves localisation. The two techniques are complementary.

The epileptogenic zone is typically hypometabolic between seizures and hyperperfused during them. Because HMPAO/ECD fix at injection, a tracer given at seizure onset “freezes” the ictal perfusion pattern for later imaging.

Interictal FDGFocal hypometabolism
Ictal SPECTOnset hyperperfusion
SISCOMCo-registration boosts yield

When to image

  • Pre-surgical localisation of drug-resistant focal epilepsy.
  • MRI-negative or MRI-discordant cases.
  • Confirming a suspected temporal-lobe focus.
  • Planning intracranial EEG placement.

Techniques

  • Interictal FDG PET: focal cortical hypometabolism at the epileptogenic zone (EEG monitoring during uptake is recommended to exclude subclinical seizures).
  • Ictal SPECT: inject HMPAO/ECD within seconds of seizure onset; hyperperfusion marks the focus.
  • Interictal SPECT: relatively insensitive on its own — used as the subtraction baseline.
  • SISCOM: subtract interictal from ictal SPECT and co-register to MRI.

How to read it

  • Concordant interictal hypometabolism and ictal hyperperfusion strongly localise the focus.
  • Temporal-lobe epilepsy localises more reliably than extratemporal.
  • Integrate with EEG, MRI and semiology in the multidisciplinary meeting.

Diagnostic performance

  • Interictal FDG PET is sensitive for temporal-lobe epilepsy localisation; extratemporal sensitivity is lower.
  • Ictal SPECT sensitivity is high with early injection and falls sharply with delay.
  • SISCOM improves localisation over visual SPECT reading.

Pitfalls

  • Late ictal injection captures seizure spread, mislocalising the onset.
  • Extratemporal and MRI-negative foci are harder to localise.
  • Hypometabolism can extend beyond the true epileptogenic zone.
Evidence & guidelines
  • FDG PET and ictal SPECT are established components of epilepsy-surgery evaluation.
  • SISCOM (ictal–interictal subtraction co-registered to MRI; O’Brien et al., Neurology 1998) improves localisation over visual SPECT reading.
  • Findings are interpreted within multidisciplinary surgical planning.
In depth
  • Ictal perfusion SPECT is the only practical routine technique that images the seizure-onset zone during a seizure, shown as hyperperfusion from neurovascular coupling to increased neuronal activity; ictal and interictal injections are given during video-EEG monitoring.
  • In temporal-lobe epilepsy, true ictal SPECT shows hyperperfusion of the whole temporal lobe; in the early postictal phase this evolves to medial temporal hyperperfusion with lateral temporal hypoperfusion.
  • Interictal SPECT alone has low sensitivity in temporal-lobe epilepsy (about 44% in meta-analysis, versus 75% postictal and 97% ictal), so it is used mainly as the baseline for ictal comparison.
  • SISCOM (subtraction ictal SPECT co-registered to MRI) is more sensitive and specific and gives objective localisation.
  • Interictal FDG-PET shows ipsilateral hypometabolism in most temporal-lobe epilepsy (about 85–90%) and is better for lateralisation than precise localisation; sensitivity is lower in extratemporal (for example frontal) epilepsy, at roughly 40–70%.
  • ¹¹C-flumazenil PET (GABA-A/benzodiazepine receptor) is highly sensitive in temporal-lobe epilepsy, up to about 100% for unilateral hippocampal sclerosis, and the abnormality is often smaller and better localised than the FDG hypometabolism.
  • Bilateral cerebellar hypometabolism is common in chronic epilepsy and has been linked to phenytoin; focal hypometabolism is found in only about a quarter of children with new-onset epilepsy, compared with 80–85% of adults with intractable seizures.

Sources: PMID 23970368 (Kumar & Chugani 2013) · PMID 23970368 · PMID 9476937 · PMID 9476937 (Devous 1998 meta-analysis) · EANM brain FDG PET guideline v3 (PMID 34882261)

Practise with a case
← BackNeurology