Persistent seizures following failed surgery - Ictal SPECT and correlation with video EEG and follow-up

Introduction: A significant proportion (30-70%) of seizure patients (pts) have recurrent seizures post surgically. Up to 40-60% of these pts may benefit from re-evaluation and repeat surgical resections. Follow-up MRI is generally not useful in these patients due to post-surgical changes. Recently, ictal SPECT (IC) has been shown to be useful for the localization of seizure focus. We retrospectively analyzed the utility of IC in patients with unsuccessful surgical resections. Methods: Review of medical charts of all pts with IC from 1995 to 2001 was performed to identify pts with recurrent seizures who had IC studies. A total of 15 pts (male = 6, female 9; mean age = 17.9 yrs., range 6-31 yr.) were identified. All pts received 74-140 MBq of Tc-99m ECD each within 30 sec (range 14-81 sec) of seizure onset and were imaged within 6 hr post injection. For the inter-ictal (INT) scan, they received a similar dose and were imaged within 15 min post-injection on a triple-head gamma camera (Triad Trionix, Twinsburg, Ohio). The reconstructed transverse images were co-registered and normalized to the total counts of the INT. Greater than 10% increase in counts on the IC compared to the INT was considered positive for ictally-enhanced perfusion. The ICs were classified as localizable, lateralizable, or discordant with respect to vEEG. Results: Ictal SPECT and vEEG were concordant and lateralized the seizure focus in 14/15 pts, while both modalities were concordant and localized the foci in 7/15 pts, they were discordant for laterialization in 1 pt, and for localization in 8 pts. The MRI in 13/15 pts prior to the IC, revealed post surgical changes at the resection site, while in two pts MRI showed residual/recurrent tumor. Repeat resections and post-surgical follow-up (30 days to > 2 yrs) were available in 6/15 pts, 5 of these pts with repeat surgical resections of concordantly localized seizure foci by IC and vEEG, had good post-surgical outcome (Engel's I/II), while in the remaining one pt, both modalities were discordant for localization and had poor outcome (Engel's III/IV). Conclusion: Our study demonstrates that ictal SPECT may be useful in the re-evaluation of surgical failures. Repeat resections of seizure on-set foci identified by ictal SPECT and concordant with vEEG appear to have good prognosis following repeat surgical resections