To Wean or Not to Wean, That Is the Question; the Downside of Maintaining Antiepileptic Drugs After Pediatric Epilepsy Surgery

Commentary

Intelligence Quotient Improves After Antiepileptic Drug Withdrawal Following Pediatric Epilepsy Surgery.

Boshuisen K, van Schooneveld MM, Uiterwaal CS, Cross JH, Harrison S, Polster T, Daehn M, Djimjadi S, Yalnizoglu D, Turanli G, Sassen R, Hoppe C, Kuczaty S, Barba C, Kahane P, Schubert-Bast S, Reuner G, Bast T, Strobl K, Mayer H, de Saint-Martin A, Seegmuller C, Laurent A, Arzimanoglou A, Braun KP; TimeToStop cognitive outcome study group. Ann Neurol 2015;1:104–114.

OBJECTIVE: Antiepileptic drugs (AEDs) have cognitive side effects that, particularly in children, may affect intellectual functioning. With the TimeToStop (TTS) study, we showed that timing of AED withdrawal does not majorly influence long-term seizure outcomes. We now aimed to evaluate the effect of AED withdrawal on postoperative intelligence quotient (IQ), and change in IQ (delta IQ) following pediatric epilepsy surgery. METHODS: We collected IQ scores of children from the TTS cohort with both pre- and postoperative neuropsychological assessments (NPAs; n = 301) and analyzed whether reduction of AEDs prior to the latest NPA was related to postoperative IQ and delta IQ, using linear regression analyses. Factors previously identified as independently relating to (delta) IQ, and currently identified predictors of (delta) IQ, were considered possible confounders and used for adjustment. Additionally, we adjusted for a compound propensity score that contained previously identified determinants of timing of AED withdrawal. RESULTS: Mean interval to the latest NPA was 19.8 ± 18.9 months. Reduction of AEDs at the latest NPA significantly improved postoperative IQ and delta IQ (adjusted regression coefficient [RC] = 3.4, 95% confidence interval [CI] = 0.6–6.2, p = 0.018 and RC = 4.5, 95% CI = 1.7–7.4, p = 0.002), as did complete withdrawal (RC = 4.8, 95% CI = 1.4–8.3, p = 0.006 and RC = 5.1, 95% CI = 1.5–8.7, p = 0.006). AED reduction also predicted ≥10-point IQ increase (p = 0.019). The higher the number of AEDs reduced, the higher was the IQ (gain) after surgery (RC = 2.2, 95% CI = 0.6–3.7, p = 0.007 and RC = 2.6, 95% CI = 1.0–4.2, p = 0.001, IQ points per AED reduced). INTERPRETATION: Start of AED withdrawal, number of AEDs reduced, and complete AED withdrawal were associated with improved postoperative IQ scores and gain in IQ, independent of other determinants of cognitive outcome.

Surgical intervention in pediatric epilepsy is increasingly sought after as a potentially curative treatment for seizures. The opportunity to benefit from brain development without the interference of seizures is an important consideration in the surgical candidate selection process and has undoubtedly contributed to a trend toward earlier surgical intervention in pediatric epilepsy. Continued interference of antiepileptic drugs (AEDs) for an extended time after successful surgery may be counterproductive to the goal of enhancing the development of neurocognitive networks, as the research from the TimeToStop (TTS) study suggests.

Earlier age at surgery has demonstrated promise in enhancing neuropsychological development progress and may prevent catastrophic cognitive outcomes associated with early onset epilepsies in some cases (1).

Research on timing of discontinuation of AEDs postsurgery put forth in the TTS study is important in highlighting the potential differential benefits of seizure freedom and freedom from the effects of AEDs, which may inhibit cognitive development as well as seizures. The primary findings highlight 301 patients with pre- and postsurgical neuropsychological evaluations (performed on average approximately 20 months postsurgery, give or take 19 months) analyzed to better appreciate the contribution of AED discontinuation following epilepsy surgery on cognitive outcome. The greatest predictors of postsurgical intelligence quotient (IQ) were presurgical IQ, age at time of surgery, etiology, the number of AEDs at time of surgery, and type of surgery. After these variables, along with sex and duration of epilepsy, were controlled for statistically, patients who had begun or completed AED withdrawal at the time of latest neuropsychological evaluation had the highest IQ increase. An important difference between the AED stable and AED wean groups was time between surgery and neuropsychological evaluation; specifically, the patients who started to wean had neuropsychological evaluation an average of 10 months later than those who were maintained at presurgical AED levels. The authors state that this difference was controlled for statistically.

The decision to wean antiepileptic drugs is complex, involving patient etiology, completeness of surgical resection, postsurgical electrical activity, epilepsy center protocol, epileptologist experience, and parent/patient preference.

Long-term seizure freedom following epilepsy surgery is less likely for some etiologies, such as focal cortical dysplasia, and certainly for nonlesional focal epilepsy (2); in these situations, the epileptologist may be uncertain that all cortex involved in the epileptogenic zone was resected and therefore may wish to prolong reliance on AEDs to inhibit reestablishment of a seizure network. When the surgery planned is considered palliative, reduction in AEDs may be a goal rather than elimination. Standard practice in many epilepsy centers involves a set period of time (often at least 1 year) to monitor postsurgical patients prior to drug weaning, provided that they remain seizure free and the EEG appears normal or near normal. It is also understandable that after a period of living with intractable epilepsy, some parents may be hesitant to wean their child from antiepileptic drugs—despite recommendations—for fear that doing so may result in seizure reoccurrence following a hard-won reprieve. By contrast, many parents are quite eager to discontinue medications and have their children free from side effects as soon as possible.

In reading this research report from the TTS study, the large cohort of patients gathered and the coordinated effort across so many centers are clearly impressive. In appreciating the generalizability of these results, it is important to understand the observational nature of this research. The patients who were weaning or weaned from medications at the time of postsurgical neuropsychological evaluation arrived at that status via a complex clinical decision-making process set in motion prior to surgery. Unlike random assignment in experimental trials, differences between groups cannot be accounted for by independent variables alone. Patients who were discontinued from AEDs earlier than others were likely self-selected due to variables that, more than likely, also affect postoperative recovery of function. The authors of the TTS study acknowledge this and made several statistical adjustments. Yet, it is important to appreciate that mathematical adjustments cannot fully equalize unequal groups. The longer recovery time for the weaning group allowed for neurocognitive networks to develop without the interference of seizures for longer than their counterparts who had not yet begun to wean AEDs; the importance of this difference cannot be overstated. As such, it is not clear that the difference between these groups is due to medications alone.

Nonetheless, the affect of AEDs on cognition in the developing brain is critically important to understand. Appreciating how AEDs affect long-term potentiation and may inhibit the establishment of optimal neurocognitive networks is within the grasp of modern cognitive neuroscience (3). Combining this knowledge with clinical observation may well lead to changes in standards of care that address seizure freedom as well as cognitive developmental preservation in choosing an appropriate time to begin weaning AEDs following successful pediatric epilepsy surgery.

The confounding of recovery time following successful epilepsy surgery and AED withdrawal is necessary to consider in reviewing all studies of neuropsychological outcome following pediatric epilepsy surgery. Beyond IQ, the patients' processing speed, memory, attention, mood, and executive functions are all subject to potential improvement with medication discontinuation or reduction. Cognitive improvements cannot be attributed to surgical intervention and seizure control alone, although untangling the unique effects of AED weaning on cognitive recovery may prove to be quite challenging.