To the Editor:

We read with interest the report of Sanders et al. on the results of their ongoing prospective study. In their study, the authors found isolated midline spikes in six (2.6%) of 228 patients, compared with a frequency of 0.2% in our study (1). They postulate that the higher frequency in their study is due to an improved detection with digital EEG. Although the latter might indeed help to detect cases otherwise missed with paper EEG, especially when midline electrodes are not used, a number of differences between the two studies also could explain this discrepancy. For instance, in our study, we included in the denominator all EEGs performed in our laboratories, including those of neonatal, pediatric, and adult patients. In the study by Sanders et al., only children ranging in age from 1 to 14 years were included. In addition, the majority (74%) of patients included in their study had seizures, whereas the percentage of patients referred to our laboratories for a seizure disorder is substantially lower. Finally, in our study, the denominator included all EEGs performed over a 10-year period, whereas the denominator in the study by Sanders et al. consisted of the number of patients, many of whom had sequential EEGs.

Based on our findings, we have to disagree with the conclusion of the Sanders et al. stating that midline spikes are significantly associated with benign childhood focal seizures and particularly with the Panayiotopoulos syndrome. None of our patients with isolated midline spikes had seizure semiology that was consistent with either of those two electroclinical syndromes, despite a careful characterization of the semiologies based on a review of the medical records and telephone interviews with patients or their parents when needed.

We strongly believe that to start defining the electroclinical syndrome of midline spikes, it is imperative to exclude patients with any other additional epileptiform discharges. Midline spikes can be associated with focal epileptiform discharges at other locations in various epilepsy syndromes, including patients with benign epilepsy of childhood with centrotemporal sharp waves (2). In our study, all patients had isolated midline spikes, and all patients displayed the epileptiform activity during wakefulness. Only with the application of strict criteria will we be able to characterize better the clinical features, semiology, and neuroimaging abnormalities associated with this syndrome.