Differences in photosensibility in epilepsies of childhood and adolescence compared to adulthood.

The study by Y. Lu et al. (2008) reported here is a retrospective analysis of an original cohort of more than 1000 children who attended the Neuropediatric Department of Kiel University Hospital in Germany over a period of almost 30 years (1975-2002). To be included, patients had to have had at least two EEG studies with intermittent photic stimulation (IPS) performed between the ages of five and 15 years and to have had a syndromic diagnosis made according to the 1989 ILAE classification. Only a third of those with childhood absence epilepsy (CAE), rolandic and febrile seizures were included. The final cohort (566 in total; 322 boys and 244 girls) therefore consisted of 122 children with CAE, 103 with rolandic epilepsy, 117 with febrile seizures and 234 with other syndromes. One hundred and seventy seven (31%) of the children studied had, at some time in childhood or adolescence, a photoparoxysmal reaction (PPR) to IPS. This consisted of occipital spikes (grade 1 PPR) in six cases (1%), occipital spike and waves (grade 2 PPR) in 26 cases (5%), a more diffuse response (grade 3 PPR) in 62 cases (11%) and a generalised response (grade 4 PPR) in 83 cases (15%). These are remarkable findings. Previous studies have found prevalence rates of PPR in 2-10% of children and 5% of adults with epilepsy (Kasteleijn-Nolst Trenité 1989, Shiraishi et al. 2001). There are a number of possible explanations for this. The population studied by Y. Lu et al. was clearly highly selected. For example, it included children with both CAE and febrile seizures. However, there were more children with the former than the latter, which is not the case in the childhood epilepsy population as a whole (Hamdy et al. 2007), and children with definite epilepsy but no syndromic diagnosis were excluded. The authors do not make clear how the children in Kiel were selected for EEG, including repeat EEG studies. Why, for example, were children with febrile seizures studied, despite authorities viewing EEG as unhelpful? Even with the best intentions, selection bias is a potent mechanism of distortion. The study by Y. Lu et al. (2008), unlike most other studies, included all grades of PPR, including localised occipital spikes. The significance of spikes confined to the occipital regions during IPS is controversial. Not all authorities consider them as pathological. They may simply be evoked responses or else be a genetic trait unrelated to an increased propensity to epileptic seizures. However, even if all localised PPR are excluded, the photosensitivity rate was still 26%. The method Y. Lu et al. (2008) used to select subjects maximised detection of photosensitivity. To be included subjects had to have had at least two EEG studies performed between five and 15 years of age. It is well established that photosensitivity is highest in adolescence, as illustrated in the figure (figure 1). Following up patients over several years, including adolescence, with repeated EEGs is predestined to detect more positive cases than is a conventional incidence or point prevalence study. The study is, in effect, a mixture of point prevalence, incidence and cumulative incidence. A different methodology was used by Shiraishi et al. (2001). These authors studied 2,187 unselected patients from a Japanese epilepsy centre (age range 1-81; mean 24.2 yrs; 56% male), and found 37 patients (1.7%) with a generalized PPR. In a prospective comparative prevalence study performed in Caucasian patients with epilepsy attending a Dutch epilepsy centre, the prevalence of generalised PPR was 1.2 % in three to five year olds, 1.1 % in children aged five to ten and 7.6% in those of ten to 15 years (KasteleijnNolst Trenité 1989). Finally, the details of how photic stimulation was applied may be important. It is established that the longer the duration of stimulation the more likely it is that PPR will be obtained, especially in children. For this reason, adopting a standard methodology in studies of photosensitivity is crucial. In the study by Y. Lu et al. patients were divided into epilepsy syndromes. Moreover, this is the first study to look specifically at children with occasional seizures (neonatal and febrile). It, like previous studies from Japan (Shiraishi Epileptic Disord 2008; 10 (2): 144-5