Epilepsy

By Malcolm P. Berger, M.D.

Tuesday, February 28, 2006
To review all that there is about the AEDs is beyond the scope of offlabel.us, but here is some overall information as of Feb, 2006.
 
The best outcome is seizure free and an acceptable tolerance of a medication.
Intolerance of a medication is better than failure, as the chance of future success is better with intolerance than failure.  In a recent study, Kwan and Brodie suggest that approximately 40% of those who do not tolerate their first AED will become seizure free, while only 20% of those who fail their first AED will become seizure free.  Applying this formula to the available data, here is what I calculate about the AEDs.
 
Medication                     Average dropout rate for adverse events          Eventual sz free score (theoretical)
Lamictal 150-200            8.8                                                                67.7            
Trileptal 2400                 11.25                                                             64.95          
Topamax 400                 31.8                                                               64.28            
Carbamazepine 600        20.65                                                             60.21          
Gabapentin 1500-1800   15.25                                                             62.27          
Valproate                        9.4                                                                 64.92          
Phenytoin                        16                                                                  69.6            
 
 
There is just not enough data for many medications for comparison, such as Keppra, Lyrica, Zonegran.
 
Ranked according to (sz free - dropout) (a measure of benefit and tolerability)
 
Lamictal (58.9)
Valproate (55.52)
Trileptal (53.7)
Phenytoin (53.6)
Gabapentin (47.02)
Carbamazepine (39.56)
Topamax (32.48)
 
Other factors (safety, cost, rapidity of need of control, pregnancy, insurance coverage, FDA approval) are not considered in this analysis.  Methodology of different studies makes above comparisons only approximate, but it is the best I can do with what is available to me at this time. 
 

Made with CityDesk