Acepromazine seems to lower the seizure threshold in some animals and can apparently allow a seizure to occur that might not have without the lowering of the seizure threshold. Paradoxically, it can raise the seizure threshold associated with some anesthetic agents (most notably ketamine).
Diazepam (Valium Rx) is helpful in reducing seizure activity in most dogs but it does have a paradoxical excitatory effect in a few animals. I don’t know if this stimulates seizure activity but it does occur.
When to medicate to control seizure activity is a really debatable question. The pro treatment side of the argument for early treatment is that “mirroring” and “kindling” of seizures are recognized in dogs. Mirroring is when an area of the brain causing seizures on one side induces the development of an area causing seizures in the same place in the other half of the brain. Kindling is the process in which seizures make it easier for other seizures to occur in effect lowering the seizure threshold a little bit every time one happens. The con side of the argument mostly revolves around the side effects of the most consistently successful seizure control medication in dogs, phenobarbital. It can cause incoordination and a general lethargy for several weeks on first administration. Most dogs overcome these effects in a few weeks, though. It also causes increased hunger, often increased water consumption and therefore urination and it causes severe liver damage in some patients. Not many, but enough to be very worrisome.
Primidone (Rx) is commonly used in dogs to avoid keeping controlled substances on hand but it is more likely to be toxic to the liver and is not a good first choice for seizure control. Seizures themselves are very unlikely to kill a dog, but it does sometimes happen as well. So the question is, when are the seizures severe enough or frequent enough to absolutely warrant treatment? I think we work out a different answer in almost every case. We try to make our best guess as to what is best for each individual patient. The traditional guidelines in veterinary medicine have been seizures that last for longer than 5 minutes (actual seizure activity) or seizures that are occurring more than once a month. We probably stick reasonably close to these guidelines but are a little quicker to consider seizure medications now that there is pretty good evidence for the kindling theory. Once it seems pretty apparent that the seizures are going to continue to get closer and closer we sometimes treat now even if they are till more than a month apart.