When patients are inadequately controlled on AED monotherapy, there is not always a “right” answer to the question of whether it is better to switch to a different monotherapy or to try adjunctive therapy, as there are several important factors to consider. For instance, it is important to know whether the patient has failed a drug because of tolerability (eg, the patient is dizzy, nauseated, or too sleepy) or because of poor efficacy (eg, the dose is optimized but the patient is still having seizures). If the drug is both poorly tolerated and has not appreciably reduced the seizure frequency, trying something new makes sense. The failure of 2 or more AEDs, whether alone or in combination, is generally considered to be refractory epilepsy.

When the decision is made to use adjunctive therapy, the adjunct should preferably have a different mechanism of action (MOA) than the initial agent. Consequently, if the patient is on carbamazepine, we are not going to add another sodium channel blocker. The adjunctive therapy should be a different class of medication (eg, levetiracetam, perampanel). In the past, we did not have clearly different mechanisms to choose from, but that is no longer the case. We currently have not only sodium channel blockers, but also agents that enhance GABAergic inhibition, agents that bind synaptic vesicle protein 2A, and 1 agent that antagonizes the AMPA receptor through noncompetitive inhibition. Other AEDs with novel MOAs are in the pipeline. So, the goal should be to practice “rational polytherapy.” (Incidentally, using medications with different MOAs is practiced routinely in other treatments, such as antihypertensives and antibiotics.) Assuming that the type of epilepsy has been correctly diagnosed and treated accordingly, no single AED has been proven to have substantially better efficacy than any other AED; where they differ is in their MOAs, frequency of administration, tolerability, adverse effects, drug-drug interactions, and impact on comorbidities. These are the differentiators that should guide the selection of a particular AED over another as an adjunct in any given case of epilepsy. If sleep is an issue for the patient, some AEDs are more sedative than others. Further, some AEDs, such as valproic acid and topiramate, are effective at migraine prevention, while others, such as valproic acid and lamotrigine, are mood stabilizers. Levetiracetam and perampanel can worsen depression and anxiety in some patients. Weight is also an important consideration, as certain AEDs are linked to weight gain and others to weight loss. Ultimately, if the patient has been seizure free for approximately 1 year using rational polytherapy, the goal in many cases would be a timed, careful withdrawal of 1 of the medications; however, the risks of seizure recurrence vary among individuals and types of epilepsy.