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 indicato r 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 zon es and causes. Objective: To investigate whether this pattern may be r elated to the electroencephalographic and clinical features of the sei zure that precedes the positron emission tomographic (PET) scan with f ludeoxyglucose F 18. Patients and Methods: Fifty-three patients with C PSs 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 seiz ure 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. Re gions 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 rel ated to the time from the seizure to the PET scan with fludeoxyglucose F 18. Results: The hypometabolic area was limited to the epileptogeni c 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 differe d from both previous groups by having hypometabolism in the extralimbi c frontal lobe (P < .001), and subjects with secondarily generalized s eizures differed from all others because of cerebellar (P < .001) and parietal lobe (P < .05) reductions. The time between the seizure and t he performance of the PET scan did not correlate with the degree or ex tent of hypometabolism. Conclusions: Mechanisms involved in the genera tion of a seizure that precedes a PET scan with fludeoxyglucose F 18 s eem to influence the ''interictal'' hypometabolic pattern. Therefore, caution should be used when interpreting scans that are preceded by a nonhabitual seizure.