Abstracts

RATIONALE: Barbiturate anticonvulsants have been used in severe epileptic conditions including those associated with electroencephalographic findings termed periodic lateralizing epileptiform discharges (PLEDs). In the presence of PLEDs, barbiturates may be ineffective in stopping seizures or abnormal epileptiform discharges. We have described a model in which synaptic depression governs the transition between epileptiform activity and quiesence. Using this model, we are able to examine conditions in which seizure probability can be varied. When the seizure probability is high, our model predicts that barbiturate anticonvulsants will be ineffective. OBJECTIVE: After this presentation, the participants will be able to discuss a physiological explanation for the apparent ineffectiveness of barbiturates for the treatment of PLEDs.
METHODS: To address this paradox, we extended a model of CA3 bursting in the in-vitro slice preparation. Thresholds for burst initation and termination were calculated using experimentally obtained burst duration and interburst interval before and after application of barbiturate anticonvulsants. Baseline burst probability, a measure of CA3 network excitability, was also evaluated.
RESULTS: 1. The thresholds for burst initiation (burst start threshold) and burst termination (burst end threshold) were calculated from experimentally obtained burst duration and interburst intervals.
2. When burst probability was low, pentobarbital increased the burst start threshold by 17[plusminus]3% and burst end threshold by 66[plusminus]7% (n=10).
3. However, when burst probability was high, the burst start threshold was unchanged and the burst end threshold increased by 8[plusminus]3% (n=10).
CONCLUSIONS: 1. In high burst probability, the effects of barbiturate anticonvulsants on the bursting thresholds are attenuated compared to low burst probability.
2. This occurs because the threshold shift caused the synapses to oscillate in a region of the synaptic recovery curve where the rate of recovery was much slower. Synapses operating in high burst probability recover from synaptic depression partially and quickly, whereas synapses operating in lower burst probability recover more completely and slowly from synaptic depression.
3. This model of CA3 bursting, based on recovery from synaptic depression, provides one explanation for the apparent ineffectiveness of barbiturate anticonvulsants under conditions of high burst probability such as PLEDs.[figure1]
[Supported by: AES and NIH]