Pattern of interictal hypometabolism in PET scans with fludeoxyglucose F 18 reflects prior seizure types in patients with mesial temporal lobe seizures

Background: Interictal hypometabolism is routinely used as an indicator of the epileptogenic zone in patients with complex partial seizures (CPSs). However, the regional pattern of hypometabolism varies without obvious reasons, even among patients with identical epileptogenic zones and causes. Objective: To investigate whether this pattern may be related to the electroencephalographic and clinical features of the seizure that precedes the positron emission tomographic (PET) scan with fludeoxyglucose F 18. Patients and Methods: Fifty-three patients with CPSs were first classified into 4 groups, depending on the features of the seizure that preceded the PET scan (determined from findings from electroencephalography and video telemetry). In 14 patients, this seizure was focal limbic (characterized by aurae or staring spells); in 18, the CPS was widespread limbic (including automatisms). Ten patients had a CPS with posturing, and 11 had a secondarily generalized CPS. Regions with a hemisphere-normalized concentration of fludeoxyglucose F 18 below the 95% confidence interval of values from 8 control subjects were defined as hypometabolic. The location of these regions was then compared among the 4 groups, and the degree of hypometabolism was related to the time from the seizure to the PET scan with fludeoxyglucose F 18. Results: The hypometabolic area was limited to the epileptogenic zone if the preceding seizure was focal limbic, whereas in patients with widespread limbic seizures, it included one or several additional areas of the limbic cortex (P = .03). Patients with posturing differed from both previous groups by having hypometabolism in the extralimbic frontal lobe (P < .001), and subjects with secondarily generalized seizures differed from all others because of cerebellar (P < .001) and parietal lobe (P < .05) reductions. The time between the seizure and the performance of the PET scan did not correlate with the degree or extent of hypometabolism. Conclusions: Mechanisms involved in the generation of a seizure that precedes a PET scan with fludeoxyglucose F 18 seem to influence the ''interictal'' hypometabolic pattern. Therefore, caution should be used when interpreting scans that are preceded by a nonhabitual seizure.