Formats

• View PDF (PDF) 192 kB  


• Help with Viewers, Players, and Plug-ins

Table of Contents

• Background
• Objectives
• Methods
• Results
• Discussion
• Glossary
• References
• Full Report
• For More Copies

Background

Epilepsy is a clinical phenomenon in which a person has recurrent seizures due to a chronic underlying process.^1,2 Approximately 1 to 3 percent of people in the United States will develop epilepsy over the course of their lives.^2-4 Epilepsy begins most commonly during the first 9 years of life, plateaus over the next 30 years, dips in patients 40 to 59 years of age, and then rises again in the elderly.^1,4,5Seizures in epilepsy can result in status epilepticus, a life-threatening unrelenting seizure, or can result in car accidents or falls that can lead to morbidity or mortality. In addition, uncontrolled seizures can result in patients losing their jobs or driving privileges.^1,2,4 The main three types of seizures in patients with epilepsy include partial, generalized, and unclassified. There are several distinct subtypes of seizures.¹

Since 1993, the Food and Drug Administration (FDA) has approved several newer antiepileptic drugs (felbamate, gabapentin, lacosamide, lamotrigine, levetiracetam, oxcarbazepine, pregabalin, rufinamide, tiagabine, topiramate, vigabatrin, zonisamide) for the treatment of epilepsy.6 This offered clinicians and patients many new options over the older antiepileptic medications that were approved between 1953 and 1983 (phenytoin, 1953; primidone, 1954; ethosuximide, 1960; carbamazepine, 1968; clonazepam, 1975; divalproex, 1978; valproic acid, 1983). The comparative benefits and harms of older versus newer antiepileptic drugs have been assessed in numerous randomized controlled trials with varying results.^1,7,8

Another important issue in the management of epilepsy is generic substitution of innovator antiepileptic medications. The American Academy of Neurology has issued two position papers stating that there is concern with generic antiepileptic medication substitution and that physicians should specifically approve all generic substitutions.^9,10 The Italian League Against Epilepsy established a working group on generic products in epilepsy treatment. It concluded that generic medications offer a valuable and cost-effective choice in the management of epilepsy but that generic substitution is not recommended in patients who achieve seizure remission on an innovator product.¹¹ The FDA and the American Society of Health-System Pharmacists do not share the view that antiepileptic medications, or other narrow therapeutic index medications (medications where the difference between the minimum effective and minimum toxic concentrations are close together), should be treated differently as it pertains to generic substitution.^12-15

A comparative effectiveness review (CER) of the benefits and harms associated with newer versus older and innovator versus generic antiepileptic treatments is needed to clarify these issues.

Objectives

This CER utilized data on benefits and harms from direct comparative studies of newer versus older or innovator versus generic antiepileptic medications in patients with epilepsy. The analytic framework for the evaluation of effectiveness and safety of antiepileptic medication in patients with epilepsy is located in Figure A.

• Key Question 1: In patients with epilepsy, what is the comparative effectiveness/efficacy of antiepileptic medications on health outcomes: mortality, hospitalizations, office/emergency department visits, composite endpoint of medical service utilization, health-related quality of life, seizures, secondary seizure injury, status epilepticus, loss of driver’s license, and loss of employment?
• Key Question 2: In patients with epilepsy, what is the comparative effectiveness/efficacy of antiepileptic medications on intermediate outcomes: pharmacokinetics, the comparative dose of medication needed to control seizures, and switchback rates?
• Key Question 3: In patients with epilepsy, what is the comparative impact of antiepileptic medications on serious adverse events such as neurological adverse effects, hypotension, rash, suicidal ideation, mood and cognition, bone density, and cosmetic adverse effects?
• Key Question 4: In patients with epilepsy, what are the comparative benefits or harms for antiepileptic medications in subgroups of patients differentiated by seizure etiology, seizure type, gender, ethnicity, patient age, and patient pharmacogenetic profile; and by types of antiepileptic medication?

Figure A. Analytic framework for the evaluation of effectiveness and safety of antiepileptic medication in patients with epilepsy

KQ = Key Question

Methods

Input From Stakeholders

The Evidence-based Practice Center drafted a topic refinement document with proposed Key Questions after consulting with Key Informants. Our Key Informants did not have financial or other declared conflicts. The public was invited to comment on the topic refinement document and Key Questions. After reviewing the public commentary, responses to public commentary, and proposed revisions to the Key Questions, a preliminary protocol was generated and reviewed with the Technical Expert Panel. The Technical Expert Panel provided feedback on the feasibility and importance of our approach and provided their unique insight. Again, no conflict of interest was identified. The draft CER report underwent peer review and public commentary and revisions were made before the report was finalized.

Data Sources and Selection

Two independent investigators conducted systematic literature searches of MEDLINE (from 1950 to March 23, 2011), Web of Science from the earliest possible date through March 23, 2011, and the Cochrane Central Register of Controlled Trials and Cochrane Database of Systematic Reviews from the earliest possible date through March 23, 2011. Separate search strategies were employed for the older versus newer antiepileptic medication evaluation and for the innovator versus generic antiepileptic evaluation. No language restriction was imposed, and a manual search of references from reports of clinical trials or review articles was also conducted.

Studies were included in the evaluation of Key Questions if they: (1) compared older antiepileptic medications (pre-1993 FDA-approved medications: phenytoin, carbamazepine, carbamazepine sustained release (SR) or controlled release (CR), valproic acid, clonazepam, phenobarbital, ethosuximide, primidone) with newer antiepileptic medications (1993 or later FDA approvals) or compared innovator antiepileptic medications with generic antiepileptic medications; (2) conducted in patients with epilepsy; and (3) reported data on prespecified clinical or humanistic outcomes.

Data Extraction and Quality Assessment

Using a standardized data abstraction tool, two reviewers independently collected data, with disagreement resolved through discussion. The following information was obtained from each study, where applicable: author identification, year of publication, source of study funding, study design characteristics and methodological quality criteria, study population (including study inclusion and exclusion criteria, run-in period, study withdrawals, antiepileptic medication utilized, length of study, and duration of patient followup), patient baseline characteristics (gender, age, ethnicity), patient pharmacogenetic profile, seizure etiology (partial, generalized, specific epilepsy syndrome), seizure type (new onset, chronic disease), types of antiepileptic medication (individual drug names, drug class, Biopharmaceutics Classification System [BCS] class), comorbidities, and use of concurrent standard medical therapies. Intermediate and final health and harms outcomes were collected where applicable. Authors were contacted for clarification or to provide additional data, where applicable.

Validity assessment was performed using the recommendations in the Methods Guide for Effectiveness and Comparative Effectiveness Reviews.¹⁶ Each study was rated as good, fair, or poor. Applicability of individual studies was based on population, intervention, comparison, outcomes, and setting factors that limit overall applicability.

Data Synthesis and Analysis

In the evaluation of older versus newer antiepileptic medications, each newer antiepileptic medication was compared with an individual older epileptic medication as part of a direct comparative study. In the evaluation of innovator and generic medications, each innovator antiepileptic drug was compared with its corresponding generic medication separately as part of a direct comparative study. Comparative trials or studies could be qualitatively described or quantitatively synthesized. Controlled clinical trials could be pooled, as could controlled observational studies, but could not be pooled together.

When pooling continuous endpoints, a weighted mean difference along with 95 percent confidence intervals (CI) were calculated using a DerSimonian and Laird random effects model.¹⁷ In order to pool data of different antiepileptic medications together for continuous endpoints, we used an inverse variance weighting approach as standardized mean difference (mean difference between treatment and control groups divided by pooled standard deviation) and 95 percent CIs. In cases where mean change scores from baseline for each group were not reported, we calculated the difference between the mean baseline and mean followup scores for each group.¹⁸ When there was more than one treatment group versus control, each treatment group was treated as a separate trial for meta-analysis, dividing the control group sample size by the number of treatment arms.¹⁹

For dichotomous endpoints, weighted averages were reported as relative risks (RRs) with associated 95 percent CIs. As heterogeneity between included studies is expected, a DerSimonian and Laird random-effects model was used when pooling data and calculating RRs and 95 percent CIs.

Statistical heterogeneity was assessed using the I² statistic while Egger’s weighted regression statistics were used to assess for the presence of publication bias.

Statistics was performed using StatsDirect statistical software, version 2.7.8 (StatsDirect Ltd, Cheshire, England). A p-value of <0.05 was considered statistically significant for all analyses.¹⁹

For the section on medical service utilization, the data was available as incidence rate ratios in the individual observational studies, therefore the data was described, but not pooled. For the section on time to first seizure, data were expressed in the trials as hazard ratios and are reported and pooled.

In subgroup analyses for the older versus newer evaluation, we evaluated the results in those with new onset versus chronic (refractory) disease and by seizure type (partial, generalized, and specific epilepsy syndrome), gender, and age. In subgroup analyses for the innovator versus generic evaluation, innovator medications were specifically studied against known “A” rated generics, and innovator medications within a BCS class (I, II, or III) were compared with their corresponding generic medications within that same class.

Results

For the newer versus older antiepileptic medication literature search, 5,773 nonduplicative citations were identified. After title and abstract screening and full text review, 5,505 and 200 citations were excluded, respectively. Sixty-eight and 49 studies were available for qualitative and quantitative synthesis, respectively. Newer versus older comparisons were largely limited to studies using carbamazepine or valproic acid and to a lesser extent phenytoin and sustained/controlled-release carbamazepine. Comparisons versus clonazepam, phenobarbital, ethosuximide, or primidone were very limited or not conducted at all. Newer versus older comparisons were also largely limited to gabapentin, lamotrigine, levetiracetam, oxcarbazepine, topiramate and vigabatrin. Comparisons versus felbamate, lacosamide, pregabalin, tiagabine, and zonisamide were very limited or not conducted at all.

For the innovator versus generic antiepileptic medication literature search, 356 nonduplicative citations were identified. After title and abstract screening and full text review, 267 and 18 citations were excluded, respectively. Seventy-one and 18 studies were available for qualitative and quantitative synthesis, respectively. Innovator versus generic antiepileptic medication comparisons are limited predominantly to studies of carbamazepine and to a lesser extent phenytoin and valproic acid. The use of an “A” rated generic could only be verified in one controlled clinical trial and a minority of controlled observational studies.

A summary of the results with ratings of the strength of the body of evidence for all Key Questions can be found in Table A. Please refer to the appendix of the full report for ratings of the strength of evidence and applicability for individual studies. We conducted evaluations for each newer antiepileptic medication versus each older generic antiepileptic medication individually and then as all newer versus each individual older medication. Similarly, we conducted each innovator medication versus its generic comparator analysis separately and then evaluated all innovator versus all generic analyses for each endpoint. However, we are not able to provide all of the individual analyses in the limited space within the executive summary. Please see the full report for the detailed results of these individual agent analyses, which are less prone to clinical heterogeneity and vital to a full understanding of the topic area.

Key Question 1

Newer antiepileptic medications did not significantly impact the risk of mortality versus their older counterparts carbamazepine, phenytoin, or valproic acid.^20-29 However, many of these trials had followup times that might preclude observing an impact on a long-term outcome such as survival.

Switching from an innovator to a generic antiepileptic medication may increase the risk of hospitalization and hospital stay duration but may not increase outpatient service utilization.^12,30-32 Data supporting this is limited to four pharmaceutical industry-sponsored observational studies.^12,30-32 These studies compared the use of long tolerated innovator antiepileptic medication with short-term results yielded after switching. The controlled observational studies did not state that they were limited to “A” rated products. The switch was not blinded, so patients’ and clinicians’ emotional or anxiety-related triggers for medical service utilization could have occurred. Use of claims data increases the risk of missing or misclassified data. Three out of the four studies showed that rates of hospitalization were higher with generic use compared with innovator, and one study found no difference. For the endpoint of hospital stay duration, all four studies found that generic use was associated with longer hospital stay duration than innovator use. And for the endpoint of outpatient service utilization, two studies found generic use was associated with higher outpatient service utilization, and the other two studies found no difference between the generic and innovator groups.

Three separate, well-conducted controlled observational studies assessed a composite endpoint of medical service utilization.^33-35 They did not compare innovator with generic products but rather the switch between “A” rated versions of products (innovator to generic, generic to generic, or generic to innovator). Two of the studies were supported by the pharmaceutical industry, used similar methods, had a similar composite endpoint (emergency department visit, ambulance service utilization, or hospitalization) and derived similar results.^33,34 They matched for several important factors, limited the analyses to “A” rated products, and conducted subgroup analyses with similar results to the base case analysis. However, these studies did not control for comorbidities or changes in other medications and their associated dosages, which are known to impact seizure occurrence. As such, it is difficult to assure that the case population had the same baseline risk of an acute event requiring emergency services aside from their switch between antiepileptic medication versions. The third well-conducted case control study was sponsored by Express Scripts.³⁵ In this study, significant increases in hospitalization of emergency room visits were seen in unadjusted analyses (odds ratio [OR] 1.51 [1.29, 1.76]), but no significant difference was found after adjusting for confounders (OR 1.08 [0.91, 1.29]), although the direction of effect was the same as the unadjusted analyses. Unlike the other two trials, this study’s authors controlled for a person’s risk of epilepsy exacerbation, change in disease severity, drug interactions, poor adherence, and change in patient diagnosis. This suggests that the difference in magnitude between these three studies may be due to inadequate confounder adjustment and/or the inclusion of ambulance service utilization in the two previous studies. All three of these controlled observational trials were unblinded and used claims data. In total, two of the three observational studies suggest that switching from an antiepileptic medication to an “A” rated version of the product may increase the utilization of a composite of medical services (hospitalization, emergency department visit, with or without utilizing ambulance services for epilepsy).^33,35

Several markers of epilepsy control were used in randomized controlled trials to compare newer versus older antiepileptic medications. The risk of being seizure free for either 6–12 or 24 months was significantly lower for newer antiepileptic medications versus carbamazepine. The risk of withdrawing due to lack of efficacy was also significantly higher for newer antiepileptic medications versus carbamazepine. No differences in 6–12- or 24-month freedom from seizures were seen for newer antiepileptic medications versus valproic acid, although this was based on a single controlled clinical trial,²⁴ or for withdrawals due to lack of efficacy for newer antiepileptic medications versus phenytoin or valproic acid. The time to first seizure was increased for newer antiepileptic medications versus phenytoin, but not for newer antiepileptic medications versus carbamazepine or valproic acid. No significant difference in the risk of maintaining seizure freedom was seen when newer antiepileptic medications were compared versus carbamazepine, controlled/sustained-release carbamazepine, phenytoin, or valproic acid in controlled clinical trials, although data is limited for the comparison of newer antiepileptic medications versus controlled/sustained-release carbamazepine.

For the comparison of innovator antiepileptic medications with their respective generic versions, we found that seizure occurrence and frequency were not significantly different between groups in controlled clinical trials. In addition, there were no significant differences between innovator antiepileptic medications and their respective generic versions in terms of total withdrawals or withdrawals due to lack of efficacy in controlled clinical trials. In one controlled observational trial, there was a significant increase in withdrawals for any reason, but this trial had marked differences in several demographic variables (age, insurance type, and concomitant migraine headache and cerebral palsy) and the investigators did not conduct adjusted analyses.³⁶ This occurred even though many of the trials did not use FDA approved “A” rated generics. Many of these controlled clinical trials used a crossover design or randomized patients to either an innovator or generic product in a parallel fashion so they cannot be used to determine whether a switch from one antiepileptic medication to another “A” rated version would increase the risk of seizure occurrence or increase seizure frequency.

In 2010, a meta-analysis of seven trials on seizure occurrence following the use of generic versus innovator antiepileptic medications was published.³⁷ We did not include the trial by Wolf 1992 since it was comparing two established versions of a sustained-release carbamazepine product versus a new version that was not a generic of the original versions. The authors said they included data from Hartley 1991 but instead used the data from Hartley 1990. Even with these differences, our findings, using the six trials that were eligible for pooling within our analysis, are characteristically similar to that of their meta-analysis (OR 1.1 [0.9 to 1.2]).³⁷

Health-related quality of life, loss of driver’s license or employment, secondary seizure injury, and status epilepticus endpoints were unavailable or did not allow adequate data to determine comparative effectiveness.

Key Question 2

This section is specifically focused on innovator versus generic antiepileptic medications. The data were derived predominantly from carbamazepine trials and to a lesser extent phenytoin and lamotrigine trials. As such, there is limited ability to extrapolate to all antiepileptic medications with generic versions.

The average Cmax, Cmin, Css, Tmax, and AUC values from a population of patients receiving innovator antiepileptic medications are not significantly different from that of their generic versions. A population of patients should derive similar concentrations on an innovator to using generic antiepileptic medications. However, our data do not allow us to determine if an individual patient or subset of patients would have an over- or under-accentuated pharmacokinetic response if they were switched from one version of the medication to the other (innovator to generic, generic to generic, generic to innovator).

While 12 to 44 percent of patients in four observational studies switched back to innovator antiepileptics after taking a generic version of the medication, the main limitation of this type of data is that the patients and clinicians were not blinded. ^12,30-32 As such, the switchback from a generic to an innovator antiepileptic medication may or may not be due to real versus perceived differences in efficacy or adverse events.

Key Question 3

We could not adequately compare antiepileptic medications for hypotension, asthenia, ataxia, nystagmus, tremor, mood and cognition, or bone density.

Newer antiepileptic medications were not significantly different versus carbamazepine, carbamazepine SR/CR, phenytoin, valproic acid, or ethosuximide in risk of overall withdrawal and versus phenytoin, valproic acid, and ethosuximide in risk of withdrawal due to adverse events, although the phenytoin and ethosuximide evaluations for both outcomes are based on more limited data. Newer antiepileptic medications had a lower withdrawal rate due to adverse events but an offsetting higher withdrawal rate due to lack of efficacy versus carbamazepine and carbamazepine SR/CR.

Newer antiepileptic medications had a significantly lower risk of developing fatigue, somnolence, dizziness, and skin rash than carbamazepine; skin rash versus carbamazepine SR/CR; vomiting and gum hyperplasia versus phenytoin; fatigue, somnolence, nausea, and alopecia versus valproic acid; and somnolence versus ethosuximide. No significant differences in the risk of headache with newer versus older antiepileptic medications was seen. Data on adverse events was very limited for carbamazepine SR/CR and ethosuximide analyses. In no case did newer antiepileptic medications exhibit a higher risk of adverse events than older antiepileptic medications.

No significant differences were noted between innovator and generic antiepileptic medications for evaluated adverse events including headache, somnolence, diplopia, or skin rash. Given the similar blood concentrations between innovator versus generic antiepileptic medications, this would be anticipated, but it has to be noted that the crossover and parallel comparative trials establish the impact of starting patients on innovator or generic therapy and not the short-term impact of switching from one version of the medication to the other.

Key Question 4

The results of these a priori subgroup analyses are not very informative. Data were limited mostly to partial epilepsy, new onset epilepsy, and were generally in patients 18 years or younger. Gender, genetic profile, and polypharmacy’s impact on results could not be determined. Splitting our newer antiepileptic medication versus carbamazepine, phenytoin, valproic acid, or ethosuximide analyses by seizure etiology, seizure type, gender, and patient age, we had limited power to detect differences. The sample sizes of the trials in each subpopulation were lower than the overall population. Many trials were excluded from the subgroup analysis because they did not subdivide their populations. In many cases, one subpopulation was evaluated for an outcome but the other subpopulation was not. Therefore, we cannot identify a subpopulation for which differential effects on an outcome might have occurred based on subgroups. The results of the subgroup analysis were similar to the base case evaluations, although, in the subgroup analysis, the results were less likely to show significance.

Innovator versus generic controlled clinical trials and controlled observational studies did not provide data in prespecified subgroups based on seizure etiology or type, or on genetic profile. No controlled clinical trials and one controlled observational study reported data on gender, age, and polypharmacy impact on switchback rates from generic to innovator versions.¹² There was no statistically significant difference in women compared with men when switching back to innovator from generic versions of antiepileptic medications (HR 1.10 [0.97 to 1.24]; p=0.130). Younger patients were more likely to require a switchback to innovator medication compared with older patients (HR 0.993 [0.988 to 0.997]; p=0.002). Patients receiving polytherapy were no more or less likely to switch back to innovator (HR 1.23 [0.995 to 1.515]; p=0.056).

While data on BCS class for the innovator versus generic antiepileptic medication evaluation was presented directly in Key Questions 1, 2, and 3; the use of BCS class was not more instructive than individual agent evaluations.

Table A. Summary of results and strength of evidence

CR = controlled release; H = high; I = insufficient; L = low; M = moderate; MD = mean difference; SMD = standardized mean difference; NRCT = nonrandomized controlled trial; NNT = number needed to treat; OBS = observational study; RCT = randomized controlled trial; RR = relative risk; SOE = strength of evidence; SR = sustained release; WMD = weighted mean difference
Outcome	Type and Number of Studies	Pooled	Result/Conclusion	Strength of Evidence
KEY QUESTION 1 ENDPOINTS
MORTALITY: Newer vs. Carbamazepine Newer vs. Phenytoin Newer vs. Valproic Acid	6 RCTs 3 RCTs 3 RCTs	Yes Yes Yes	No effect, RR 0.75 (0.51, 1.12) No effect, RR 0.30 (0.05, 1.95) No effect, RR 0.94 (0.31, 2.80)	SOE: L SOE: L SOE: L
OUTPATIENT SERVICE UTILIZATION: Innovator vs. Generic	4 OBS	No	Similar utilization of outpatient services during generic medication periods.	SOE: L
HOSPITALIZATIONS: Newer vs. Carbamazepine Newer vs. Valproic Acid Newer vs. Ethosuximide Innovator vs. Generic	1 RCT 1 RCT 1 RCT 4 OBS	No No No No	Newer antiepileptic medications (lamotrigine) did not reduce the risk of hospitalization compared with carbamazepine. Newer antiepileptic medications (lamotrigine) did not reduce the risk of hospitalization compared with valproic acid. Newer antiepileptic medications (lamotrigine) did not reduce the risk of hospitalization compared with ethosuximide. Increased risk of hospitalizations during generic medication periods.	SOE: I SOE: I SOE: I SOE: L
HOSPITAL STAY DURATION: Innovator vs. Generic	4 OBS	No	Increased hospital stay during generic medication periods.	SOE: L
COMPOSITE OF MEDICAL SERVICE UTILIZATION (Ambulance service, hospitalization, or emergency department visit for epilepsy): Innovator vs. Generic	3 OBS	No	Increase in medical service utilization during periods when a patient’s antiepileptic medication is switched to an “A” rated version of the product (innovator to generic, generic to generic, generic to innovator).	SOE: I
HEALTH-RELATED QUALITY OF LIFE: Newer vs. Carbamazepine Newer vs. Carmabazepine CR/SR Newer vs. Phenytoin Newer vs. Valproic Acid	3 RCTs 1 RCT 2 RCTs 3 RCTs	No No No No	Different scales and subscales, data inconclusive Different scales and subscales, data inconclusive Different scales and subscales, data inconclusive Different scales and subscales, data inconclusive	SOE: I SOE: I SOE: I SOE: I
TIME TO FIRST SEIZURE: Newer vs. Carbamazepine Newer vs. Phenytoin Newer vs. Valproic Acid	4 RCTs 2 RCTs 1 RCT	Yes Yes Yes	No effect, (HR 1.14 [0.98, 1.33]) Time to seizure increased for newer vs. phenytoin. (HR 1.59 [1.04, 2.43]) No effect, (HR 0.8 [0.63, 1.02])	SOE: L SOE: L SOE: L
SEIZURE OCCURRENCE: Innovator vs. Generic	7 RCTs	Yes	No effect, [0.87 [0.64, 1.18])	SOE: L
SEIZURE FREEDOM FOR STUDY DURATION: Newer vs. Carbamazepine Newer vs. Carbamazepine CR/SR Newer vs. Phenytoin Newer vs. Valproic Acid	15 RCTs 2 RCTs 4 RCTs 12 RCTs	Yes Yes Yes Yes	No effect, (RR 0.94 [0.87, 1.03]) No effect, (RR, 0.90 [0.79, 1.02]) No effect, (RR 0.92 [0.85, 1.00]) No effect, (RR 0.97 [0.87, 1.08])	SOE: L SOE: M SOE: M SOE: M
SEIZURE FREQUENCY: Newer vs. Carbamazepine Newer vs. Phenytoin Newer vs. Valproic Acid Innovator vs. Generic	1 RCT 2 RCTs 1 RCT 3 RCTs	No No No Yes	No effect, [MD -3 [-6.32, 0.32]) Not enough data to evaluate effect Not enough data to evaluate effect No effect, [SMD 0.03 [-0.08, 0.14])	SOE: I SOE: I SOE: I SOE: L
SEIZURE REMISSION 6- to 12-Month: Newer vs. Carbamazepine   Newer vs. Valproic Acid 24-Month: Newer vs. Carbamazepine   Newer vs. Valproic Acid	2 RCTs   1 RCT   1 RCT   1 RCT	Yes   Yes   Yes   Yes	Patients on newer antiepileptic medications were less likely to have seizure remission vs. carbamazepine. (RR 0.81 [0.67, 0.99), NNT 9] No effect, (RR 0.97 [0.89,1.06])   Patients on newer antiepileptic medication were less likely to have seizure remission vs. carbamazepine. (RR 0.82 [0.72, 0.94), NNT 13] No effect, (RR 0.85 [0.73,1.00])	SOE:L   SOE: M   SOE: M   SOE: M
STATUS EPILEPTICUS, SECONDARY INJURY FROM SEIZURES, LOSS OF DRIVER’S LICENSE/EMPLOYMENT: Innovator vs. Generic	No data	No	No data	SOE: I
TOTAL WITHDRAWALS: Newer vs. Carbamazepine Newer vs. Carbamazepine CR/SR Newer vs. Phenytoin Newer vs. Valproic Acid Newer vs. Ethosuximide Innovator vs. Generic	14 RCTs 2 RCTs 3 RCTs 16 RCTs 1 RCT 9 RCTs + 1 NRCT	Yes Yes Yes Yes No Yes	No effect, (RR 0.90 [0.82, 1.00]) No effect, (RR 0.96 [0.78, 1.18]) No effect, (RR 0.91 [0.76, 1.09]) No effect, (RR 0.96 [0.85, 1.09]) No effect, (RR 0.95 [0.53, 1.71]) No effect, (RR 0.90 [0.39, 2.08])	SOE: L SOE: L SOE: L SOE: L SOE: I SOE: L
WITHDRAWALS DUE TO LACK OF EFFICACY: Newer vs. Carbamazepine Newer vs. Carbamazepine CR/SR Newer vs. Phenytoin Newer vs. Valproic Acid Innovator vs. Generic	10 RCTs 1 RCT   3 RCTs 11 RCTs 9 RCTs + 1 NRCT	Yes No   Yes Yes Yes	Withdrawals due to lack of efficacy increased with newer agents vs. carbamazepine. (RR 1.59 [1.25, 2.02), NNT 50] Withdrawals due to lack of efficacy increased with newer agents vs. carbamazepine CR/SR. (RR 2.43 [1.32, 4.52), NNT 16] No effect, (RR 1.03 [0.33, 3.23]) No effect, (RR 1.10 [0.77, 1.56]) No effect, (RR 1.02 [0.41, 2.54])	SOE: L SOE: I   SOE: L SOE: L SOE: L
KEY QUESTION 2 ENDPOINTS
Maximum Concentration: Innovator vs. Generic	7 RCTs + 1 NRCT	Yes	No effect, [SMD 0.10 [-0.13, 0.32])	SOE: L
Minimum Concentration: Innovator vs. Generic	5 RCTs + 1 NRCT	Yes	No effect, [SMD 0.05 [-0.21, 0.31])	SOE: L
Steady State Concentration: Innovator vs. Generic	7 RCTs	Yes	No effect, [SMD 0.18 [-0.09, 0.45])	SOE: L
Time to Maximal Concentration: Innovator vs. Generic	5 RCTs	Yes	No effect, [WMD 0.00 [-0.43, 0.43]) (Note: a WMD was calculated vs. an SMD for Tmax because only carbamazepine trials made up this evaluation).	SOE: I
AREA UNDER THE CURVE: Innovator vs. Generic	7 RCTs + 1 NRCT	Yes	No effect, [SMD 0.05 [-0.18, 0.28])	SOE: L
DOSE REQUIREMENTS FOR SEIZURE CONTROL: Innovator vs. Generic	No data	No	No data	SOE: I
SWITCHBACK RATES: Innovator vs. Generic	4 OBS	No	Switchback rates from a generic back to an innovator antiepileptic medication varied from 12.4% to 44.1%	SOE: L
KEY QUESTION 3 ENDPOINTS
WITHDRAWALS DUE TO ADVERSE EVENTS: Newer vs. Carbamazepine Newer vs. Carbamazepine CR/SR   Newer vs. Phenytoin Newer vs. Valproic Acid Newer vs. Ethosuximide Innovator vs. Generic	18 RCTs   2 RCTs   3 RCTs 16 RCTs 1 RCT 9 RCTs + 1 NRCT	Yes   Yes   Yes Yes No Yes	Withdrawals due to adverse events were reduced with newer antiepileptic medications vs. carbamazepine. (RR 0.62 [0.53, 0.73), NNT 13] Withdrawals due to adverse events were reduced with newer antiepileptic medications vs. carbamazepine CR/SR. (RR 0.69 [0.50, 0.95), NNT 16] No effect, [0.38 [0.14, 1.03]) No effect, (RR 0.90 [0.75, 1.08]) No effect, (RR 0.71 [0.45, 1.12]) No effect, (RR 0.79 [0.28, 2.20)	SOE: M   SOE: M   SOE: I SOE: L SOE: I SOE: L
HEADACHE: Newer vs. Carbamazepine Newer vs. Carbamazepine SR/CR Newer vs. Phenytoin Newer vs. Valproic Acid Newer vs. Ethosuximide Innovator vs. Generic	15 RCTs 2 RCTs 4 RCTs 15 RCTs 1 RCT 3 RCTs	Yes Yes Yes Yes No Yes	No effect, (RR 0.92 [0.78, 1.08]) No effect, (RR 0.83 [0.63, 1.10]) No effect, (RR 0.74[0.53, 1.02]) No effect, (RR 0.90 [0.70, 1.16]) No effect, (RR 0.66 [0.33, 1.29]) No effect, (RR 0.95 [0.55, 1.64])	SOE: L SOE: L SOE: L SOE: L SOE: I SOE: I
FATIGUE: Newer vs. Carbamazepine Newer vs. Carbamazepine SR/CR Newer vs. Phenytoin Newer vs. Valproic Acid Newer vs. Ethosuximide Innovator vs. Generic	7 RCTs 1 RCT 1 RCT 8 RCTs 1 RCT No data	Yes Yes No Yes No No	Risk of fatigue reduced with newer antiepileptic medications vs. carbamazepine. (RR 0.57 [0.41, 0.80), NNT 11] No effect, (RR 1.17 [0.80, 1.72]) No effect, (RR 1.05 [0.49, 2.25]) Risk of fatigue reduced with newer antiepileptic medications vs. valproic acid. (RR 0.61 [0.44, 0.85), NNT 23] No effect, (RR 0.90 [0.45, 1.80]) No data	SOE: L SOE: I SOE: I SOE: M SOE: I SOE: I
SOMNOLENCE: Newer vs. Carbamazepine   Newer vs. Carbamazepine SR/CR Newer vs. Phenytoin Newer vs. Valproic Acid Newer vs. Ethosuximide Innovator vs. Generic	8 RCTs   1 RCT 4 RCTs 9 RCTs 1 RCT   2 RCTs	Yes   No Yes Yes No   Yes	Risk of somnolence reduced with newer antiepileptic medications vs. carbamazepine. (RR 0.47 [0.36, 0.61), NNT 14] No effect, (RR 1.21 [0.75, 1.96]) No effect, (RR 0.72 [0.44, 1.18]) Risk of somnolence reduced with newer antiepileptic medications vs. valproic acid. (RR 0.65 [0.43, 0.98), NNT 25] Risk of somnolence reduced with newer antiepileptic medications vs. ethosuximide. (RR 0.22 [0.07, 0.70), NNT 15] No effect, (RR 0.90 [0.48, 1.70])	SOE: M   SOE: I SOE: I SOE: M SOE: I   SOE: L
DIZZINESS: Newer vs. Carbamazepine   Newer vs. Carbamazepine SR/CR Newer vs. Phenytoin Newer vs. Valproic Acid Newer vs. Ethosuximide Innovator vs. Generic	16 RCTs   2 RCTs 3 RCTs 12 RCTs 1 RCT No data	Yes   Yes Yes Yes No No	Risk of dizziness reduced with newer antiepileptic medications vs. carbamazepine. (RR 0.78 [0.67, 0.91), NNT 50] No effect, (RR 0.96 [0.56, 1.66]) No effect, (RR 0.67 [0.43, 1.05]) No effect, (RR 0.98 [0.71, 1.35]) No effect, (RR 0.46 [0.15, 1.38]) No data	SOE: M   SOE: L SOE: L SOE: L SOE: I SOE: I
COMBINED NEUROLOGICAL ADVERSE EVENTS: Innovator vs. Generic	1 RCT + 1 OBS	No	No effect, RCT: 4.3% vs 21.7%, p=0.189; OBS: 75.7 events per 1000 person years, 75.7 events per 1,000 person years, p=NS	SOE: L
DIPLOPIA: Innovator vs. Generic	2 RCT	Yes	No effect, [1.28 [0.38, 4.31])	SOE: L
HYPOTENSION, ASTHENIA, ATAXIA, NYSTAGMUS, TREMOR: Innovator vs. Generic	No data	No	No data	SOE: I
NAUSEA: Newer vs. Carbamazepine Newer vs. Carbamazepine SR/CR Newer vs. Phenytoin Newer vs. Valproic Acid Innovator vs. Generic	8 RCTs 1 RCT 4 RCTs 11 RCTs No data	Yes No Yes Yes No	No effect, (RR 0.69 [0.46, 1.02]) No effect, (RR 0.66 [0.39, 1.12]) No effect, (RR 0.88 [0.56, 1.37]) Risk of nausea reduced with newer antiepileptic medications vs. valproic acid. (RR 0.56 [0.41, 0.77), NNT 31] No data	SOE: L SOE: I SOE: L SOE: M SOE: I
VOMITING: Newer vs. Carbamazepine Newer vs. Phenytoin Newer vs. Valproic Acid Innovator vs. Generic	3 RCTs 1 RCT 5 RCTs No data	Yes No Yes No	No effect, (RR 1.25 [0.66, 2.35]) Risk of vomiting was reduced with newer antiepileptic medications vs. phenytoin (RR 0.09 [0.01, 0.89), NNT 19] No effect, (RR 0.69 [0.34, 1.42]) No data	SOE: L SOE: I SOE: L SOE: I
SKIN RASH: Newer vs. Carbamazepine   Newer vs. Carbamazepine SR/CR Newer vs. Phenytoin Newer vs. Valproic Acid Innovator vs. Generic	13 RCTs   2 RCTs   4 RCTs 10 RCTs 2 RCTs	Yes   Yes   Yes Yes Yes	Risk of skin rash was reduced with newer antiepileptic medications vs. carbamazepine (RR 0.52 [0.39, 0.69), NNT 24] Risk of skin rash was reduced with newer antiepileptic medications vs. carbamazepine SR/CR (RR 0.47 [0.25, 0.89), NNT 27] No effect, (RR 0.76 [0.34, 1.66]) No effect, (RR 1.17 [0.55, 2.48]) No effect, (RR 0.77 [0.17, 3.57])	SOE: M   SOE: L   SOE: I SOE: L SOE: I
SUICIDAL IDEATION: Newer vs. Carbamazepine Innovator vs. Generic	1 Obs No data	No No	Risk of attempted suicide was increased with gabapentin vs. carbamazepine (RR 13.92 [1.82, 106.38]). No data	SOE: I SOE: I
MOOD AND COGNITION: Newer vs. Carbamazepine Newer vs. Phenytoin Newer vs. Valproic Acid Innovator vs. Generic	4 RCTs 1 RCT 5 RCTs 1 RCT	No No No No	Different scales and subscales, data inconclusive No effect Different scales and subscales, data inconclusive Only cognition evaluated. No significant differences between innovator of generic but 4 of 5 cognitive test measures showed better scores for innovator than generic.	SOE: I SOE: I SOE: I SOE: I
BONE DENSITY: Newer vs. Older Innovator vs. Generic	No data No data	No No	No data No data	SOE: I SOE: I
ALOPECIA: Newer vs. Carbamazepine Newer vs. Valproic Acid	6 RCTs 8 RCTs	Yes Yes	No effect, (RR 0.60 [0.23, 1.58]) Risk of alopecia was reduced with newer antiepileptic medications vs. valproic acid. (RR 0.18 [0.10, 0.31), NNT 10]	SOE: L SOE: M
ACNE: Newer vs. Phenytoin	1 RCT	No	No effect, (RR 2.78 [0.82, 9.53])	SOE: I
GUM HYPERPLASIA: Newer vs. Phenytoin	2 RCTs	Yes	Risk of gum hyperplasia was reduced with newer antiepileptic medications vs. phenytoin. (RR 0.10 [0.04, 0.27), NNT 6]	SOE: H

Discussion

Overview of Key Findings

Our evaluation of newer versus older antiepileptic medications was predominantly limited to newer antiepileptic medication comparisons versus carbamazepine, valproic acid, and to a lesser extent phenytoin. Carbamazepine had advantages in epilepsy control over the newer antiepileptic medications it was compared with but had more adverse effects. The risk of withdrawing due to lack of efficacy was decreased, and withdrawing due to adverse events was increased, leading to a neutral effect on withdrawing for any reason. Valproic acid and phenytoin provided similar epilepsy control to newer antiepileptic medications, although newer antiepileptic medications had a longer time to first seizure versus phenytoin. While the risk of withdrawing due to adverse events was not significantly different with valproic acid or phenytoin versus newer antiepileptic medications, there were adverse events that occurred more commonly with these older antiepileptic medications. So, when qualitatively assessing the balance of benefits to harms, carbamazepine offers similar comparative effectiveness versus newer antiepileptic medications with greater benefit but more harms. Newer antiepileptic medications may have some advantages over valproic acid and phenytoin in comparative effectiveness with similar benefits but less harms.

In a patient who needs to initiate an antiepileptic medication, we could find no substantive differences in terms of benefits or harms associated with the use of a generic version versus an innovator product. Our data is limited predominantly to innovator versus generic versions of carbamazepine and to a lesser extent phenytoin and valproic acid. We could find no substantive differences in pharmacokinetic parameters between generic and innovator versions of the same antiepileptic medication either. While the source of the innovator and the generic (internationally versus domestically) may impact the variability in blood concentrations, the pharmacokinetic and final health outcomes results for initiating innovator versus generic medications seem congruent. In our literature set, patients were studied in a crossover or parallel design so when they were allocated to therapy or switched between therapies, the tendency for loss of efficacy or harm associated with switching might be similarly distributed across the groups. As such, this data cannot prove that intermediate or final health outcomes would be similar for the short-term period (several days to weeks) after an innovator or generic product is switched to another version of the medication versus maintaining the patient on their previous therapy. Switching from an innovator to a generic antiepileptic medication may increase the risk of hospitalization and hospital stay duration and may increase the risk of a composite of having an emergency department and hospitalization visit with or without ambulance service utilization. However, this is based on controlled observational study data, which has inherent limitations substantially reducing the strength of evidence. In addition, this data cannot be used to say that use of generic antiepileptic medications are less efficacious or safe than innovator versions for long-term therapy.12,30-35

Only one outcome, the risk of gum hyperplasia with phenytoin versus newer antiepileptic medications, had a high strength of evidence. For the outcomes reported in the executive summary, the strength of evidence was predominantly moderate to low for the newer versus older antiepileptic medication evaluation and low to insufficient for the innovator versus generic evaluation. In many cases, strength of evidence was reduced for issues of inconsistency and imprecision. Pooling multiple newer antiepileptic medication comparisons versus a single older antiepileptic medication enhanced power to detect differences but reduced consistency. Precision frequently was impacted negatively by having only a few small trials for an analysis. Analyses with only observational studies had a greater risk of bias which negatively impacted strength of evidence.

Applicability of evidence for both the newer versus older antiepileptic medication evaluation and the innovator versus generic evaluation was more evenly dispersed between insufficient, low, and moderate with no areas of high applicability. For the innovator versus generic evaluations, the lack of specification that the products were “A” rated generics and the multitude of studies conducted outside the United States limited applicability.  

Limitations

This CER is limited by heterogeneity. A heterogenous group of antiepileptic medications was placed into groups based on whether they were older, newer, innovator, or generic. A heterogenous group of epilepsy types was also lumped together, and the diagnosis and detection of these different types have changed over time, making clear subgroup evaluations using these trials difficult. Patients with different pharmacodynamic and pharmacokinetic genetic polymorphisms were not elucidated in virtually all the trials but these factors could make some medications more or less preferable. The use of studies from different countries, different time periods, utilizing differing study durations, mixing of patients with different baseline seizure frequency rates and different environmental triggers, and the use of “A” rated and non-“A” rated products in some analyses may have also introduced heterogeneity.

While there are some important differences between agents within the older and newer groups, we do not believe that the differences between groups are too marked to allow pooling. The drugs in our CER are all used to control or reduce seizure frequency, work in the central nervous system to cause their effect, are all given via the same route of adminstration, and many share aspects of their mechanism of action (for example, sodium channel or glutamate/glutyl-amino-butyric-acid effects) in a broad sense. We evaluate some of the major potential sources of heterogeneity in subgroup analyses.  Other sources of heterogeneity such as genomic differences, durations of therapy, mixing of patients with differing seizure frequencies should have been attenuated within a trial due to randomization. We are transparent in our presentation of the results since in the full report we provide individual agent comparisons and subgroup analyses for other potential sources of heterogeneity but could not report the results in the executive summary given wording limitations. 

We did not include every possible endpoint of interest. We had to make some choices as to which endpoints would be included and which would not and we wanted to make those decisions a priori. We included myriad endpoints that while not exhaustive, are very broad but may not contain a specific endpoint that a particular practitioner may wish to see. For instance, we included loss of job or driving privileges but did not include school performance. 

While we sought to evaluate the impact of newer versus older antiepileptic medications, only a few older antiepileptic medications were substantively evaluated and were compared to a greater or lesser extent with newer antiepileptic medications. In the full report, we provide the data for each individual newer antiepileptic medication versus each individual older antiepileptic medication. These data are more specific than the aggregate pooled data of all newer antiepileptic medication versus each older antiepileptic medication and decreases the clinical heterogeneity in the data. However, the power to detect differences in these individual analyses is substantially compromised.  With future direct comparative clinical trials, the ability to use individual newer versus individual older antiepileptic medication evaluations in agent selection could be enhanced. 

Our evaluations of newer versus older antiepileptic medications provide populationwide insight into comparative benefits and harms but cannot account for individual patient factors that may make the use of a certain antiepileptic medication more or less desirable. Factors such as pregnancy, the desire or possibility to become pregnant within a specified period of time, concomitant drugs and risk of serious drug interactions, and genetic polymorphisms or the ethnicities most likely to harbor polymorphisms that increase the risk of severe skin rashes can be used to select an optimal therapeutic choice for an individual patient. We need more information on the benefits and harms associated with older and newer antiepileptic medications in different seizure types, and it needs to be understood that the classification of epilepsy types is an evolving science.

Our innovator versus generic antiepileptic medications evaluation is limited by the small size, short-term nature, and the almost entire lack of clinical trials specifying that they were comparing “A” rated products. In the United States, generic substitution is done between products with an “A” rating by the FDA. The observational nature and lack of full accounting for confounders in other studies is also an important limitation. The observational study by Devine and colleagues demonstrates the potential impact of more fully accounting for confounders in observational studies. 

Future clinical trials should be conducted specifically evaluating the impact of switching patients from innovator to generic versions of medication. A proposed methodology would be to take a population of patients receiving either innovator or an “A” rated generic version of a medication and then randomize some patients to be switched and other patients to be maintained on initial therapy in a double blind manner.  This would eliminate the potential impact of clinician or patient apprehension about the switch on resource utilization or to increase the risk of experiencing a seizure or an adverse event either directly or indirectly through noncompliance or dose alteration.  Followup could be relatively brief (3 months) and should include a pharmacokinetic (using Bayesian population pharmacokinetics whereby only one or two samples from each patient would suffice) and final health outcome component (assessing for seizure occurrence, seizure frequency, health care utilization, and adverse events). Without randomization, blinding, and exclusive use of “A” rated products, future studies would share the substantial flaws of the current body of literature.

Our subgroup analyses could have been very important in helping identify which populations have an accentuated or attenuated effect versus the average, but due to a lack of power and methodological limitations, we were unable to generate data that could guide therapy in this manner. Future trials should report on their benefits and harms in these subpopulations even in the absence of power to judge significance because it allows systematic reviewers to pool the trials together. 

Endpoints such as bone fracture and concussion should be assessed in everyone, loss of job or driving privileges should be assessed in adults, and school performance should be evaluated in children. 

Glossary

“A” Rated Drug Products: Drug products that are considered to be therapeutically equivalent to other pharmaceutically equivalent products.  “A” products are those for which actual or potential bioequivalence problems have been resolved with adequate in vivo and/or in vitro evidence supporting bioequivalence.

Area Under the Curve (AUC): The area under the concentration versus time curve derived when an antiepileptic medication is dosed.  Also referred to as the total systemic exposure to the drug over time.

Bioequivalent Drug Products: Pharmaceutical equivalent or alternative products that display comparable bioavailability when studied under similar experimental conditions.

Biopharmaceutics Classification System (BCS): Classification of antiepileptic medications based on properties and relegated into four classes; high solubility/high permeability (Class I, optimal class with lowest risk of absorption variability), low solubility/high permeability (Class II), high solubility/low permeability (Class III), and low solubility/low permeability (Class IV).

Cmax: The maximal concentration of antiepileptic medication obtained after dosing.

Confidence Intervals (CIs): A range that is likely to include the given value. Usually presented as a percent (%). For example, a value with 95 percent confidence interval implies that when a measurement is made 100 times, it will fall within the given range 95 percent of the time.

Correlation Coefficient: A value (which usually ranges from zero to one) that indicates the degree of relationship between two variables. For example, a correlation coefficient of one would indicate a strong relationship.

Css: The concentration of antiepileptic medication obtained at steady state.

DerSimonian and Laird Random-Effects Model: A statistical method based on the assumption that the effects observed in different studies (in a meta-analysis) are truly different.

Egger’s Weighted Regression Statistics: A method of identifying and measuring publication bias. 

Epilepsy: A clinical phenomenon in which a person has recurrent seizures due to a chronic underlying process.  The main types of seizures include partial (simple partial, complex partial, partial with secondary generalization) and generalized (absence, tonic-clonic, tonic, atonic, myoclonic).

I2: Measure of degree of variation due to statistical heterogeneity. Usually reported as a percent ranging from 0 to100.

Meta-analysis: The process of extracting and pooling data from several studies investigating a similar topic to synthesize a final outcome.

Publication Bias: The possibility that published studies may not represent all the studies that have been conducted, and therefore, create bias by being left out of a meta-analysis. 

Q Statistic: A test to assess the presence of statistical heterogeneity among several studies.

Relative Risks (RRs):  The ratio of an event occurring in an exposed group to an event occurring in a non-exposed group in a given population. A ratio of one indicates no difference in the risk between the two groups.

Risk Difference: The absolute difference in the event rate between two comparison groups. A risk difference of zero indicates no difference between comparison groups.

Sensitivity Analysis: A “what if” analysis that helps determine the robustness of a study. Helps determine the degree of importance of each variable for a given outcome.

Standard Deviations (SDs): A measure of the variability of a dataset. For a simple dataset with numbers, standard deviation can be calculated using the following formula:
σ  = ((∑(x-xm))2/N)0.5
σ is standard deviation.
xm is the average.
∑(x-xm) is the sum of xm subtracted from each individual number x.
N is the total number of values.
Note: Other formulas also exist.

Statistical Heterogeneity: Variability in the observed effects among studies in a meta-analysis.

Therapeutic Equivalence: Drug products are considered to be therapeutic equivalents only if they are pharmaceutical equivalents and if they can be expected to have the same clinical effect and safety profile when administered to patients under the conditions specified in the labeling.

Tmax: The time from administration until the Cmax (see Cmax above) is obtained.

References

1. Lowenstein DH. Seizures and Epilepsy, Chapter 363. In: Fauci AS, Kasper DL, Longo DL. Harrison's Principles of Internal Medicine. 17th ed. New York, NY: McGraw-Hill; 2008.
2. Scheuer ML, Pedley TA. The evaluation and treatment of seizures. N Engl J Med 1990;323:1468-74. PMID: 2233919
3. Cascino GD, Sharbrough FW. Symposium on epilepsy. Foreword. Mayo Clinic Proc 1996;71:403-4. PMID: 8637267
4. Tatum IV WO, Benbadis S, Vale FL. The neurosurgical treatment of epilepsy. Arch Fam Med 2000;9:1142-6. PMID: 11115221
5. Middleton DB. Seizure Disorders, Chapter 64. In: Taylor RB. Family Medicine: Principles and Practice. 6th ed. New York, NY: Springer-Verlag; 2003.
6. French JA, Kanner AM, Baustista J, et al. Efficacy and tolerability of the new antiepileptic drugs I: treatment of new onset epilepsy: report of therapeutics and technology assessment subcommittee and quality standards subcommittee of the American Academy of Neurology and the American Epilepsy Society. Neurology 2004;62:1252-60. PMID: 15111659
7. Marson AG, Al-Kharusi AM, Alwaidh M, et al. The SANAD study of effectiveness of carbamazepine, gabapentin, lamotrigine, oxcarbazepine, or topiramate for treatment of partial epilepsy: an unblinded randomised controlled trial. Lancet 2007;369:1000-15. PMID: 17382827
8. Marson AG, Al-Kharusi AM, Alwaidh M, et al. The SANAD study of effectiveness of valproate, lamotrigine, or topiramate for generalised and unclassifiable epilepsy: an unblinded randomised controlled trial. Lancet 2007;369:1016-26. PMID: 17382828
9. Assessment: generic substitutionfor antiepileptic medication. Report of the therapeutics and technology assessment subcommittee of the American Academy of Neurology. Neurology 1990;40:1641-3. PMID: 2234417
10. Liow K, Barkley GL, Pollard JR, et al. Position statement on the coverage of anticonvulsant drugs for the treatment of epilepsy. Neurology 2007;68:1249-50. PMID: 17438213
11. Percucca E, Albani F, Capovilla G, et al. Recommendations of the Italian League against epilepsy working group on generic products of of antiepileptic drugs. Epilepsia 2006;47 (Suppl 5):16-20. PMID: 17239100
12. LeLorier J, Duh MS, Paradis PE, et al. Clinical consequences of generic substitution of lamotrigine for patients with epilepsy. Neurology 2008;70:2179-86. PMID: 18505997
13. American Society of Health-System Pharmacists. Comments from the FDA on narrow therapeutic index drugs and generic substitution. www.ashp.org/import/news/NewsCapsules/article.aspx?id=2 . Accessed October 22, 2010.
14. Williams RL. Therapeutic equivalence of generic drugs response to National Association of Board of Pharmacy.  www.fda.gov/Drugs/DevelopmentApprovalProcess/HowDrugsareDevelopedandApproved/
    ApprovalApplications/AbbreviatedNewDrugApplicationANDAGenerics/ucm073224.htm. Accessed October 22, 2010.
15. Nightingale SL. Therapeutic equivalence of generic drugs letter to health practitioners. www.fda.gov/Drugs/DevelopmentApprovalProcess/HowDrugsareDevelopedandApproved/
    ApprovalApplications/AbbreviatedNewDrugApplicationANDAGenerics/ucm073182.htm. Accessed October 22, 2010.
16. Methods guide for effectiveness and comparative effectiveness review. AHRQ Publication No. 10(11)-EHC063-EF. Rockville, MD: Agency for Healthcare Research and Quality. March 2011. Chapters available at: www.effectivehealthcare.ahrq.gov.
17. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177-88. PMID: 3802833
18. Follman D, Elliot P, Suh I, et al. Variance imputation for overviews of clinical trials with continuous response. J Clin Epidemiol 1992;45:769-73. PMID: 1619456
19. Higgins JPT, Green S, Editors. Cochrane handbook for systematic reviews of intervention. Posted February 2008. Updated February 2008. Version 5.0.0. Available at: www.cochrane-handbook.org .
20. Brodie MJ, Mumford JP. Double-blind substitution of vigabatrin and valproate in carbamazepine-resistant partial epilepsy. 012 Study group. Epilepsy Res 1999;34:199-205. PMID: 10210035
21. Brodie MJ, Overstall PW, Giorgi L. Multicentre, double-blind, randomised comparison between lamotrigine and carbamazepine in elderly patients with newly diagnosed epilepsy. The UK Lamotrigine Elderly Study Group. Epilepsy Res 1999;37:81-7. PMID: 10515178
22. Chadwick D. Safety and efficacy of vigabatrin and carbamazepine in newly diagnosed epilepsy: a multicentre randomised double-blind study. Vigabatrin European Monotherapy Study Group. Lancet 1999;354:13-9. PMID: 10406359
23. Fakhoury TA, Hammer AE, Vuong A, et al. Efficacy and tolerability of conversion to monotherapy with lamotrigine compared with valproate and carbamazepine in patients with epilepsy. Epilepsy Behav 2004;5:532-8. PMID: 15256191
24. Marson AG, Appleton R, Baker GA, et al. A randomised controlled trial examining the longer-term outcomes of standard versus new antiepileptic drugs. The SANAD trial. Health Technol Assess 2007;11:iii,iv, ix-x, 1-134. PMID: 17903391
25. Reunanen M, Dam M, Yuen AW. A randomised open multicentre comparative trial of lamotrigine and carbamazepine as monotherapy in patients with newly diagnosed or recurrent epilepsy. Epilepsy Res 1996;23:149-55. PMID: 8964276
26. Rowan AJ, Ramsay RE, Collins JF, et al. New onset geriatric epilepsy: a randomized study of gabapentin, lamotrigine, and carbamazepine. Neurology 2005;64:1868-73. PMID: 15955935
27. Bill PA, Vigonius U, Pohlmann H, et al. A double-blind controlled clinical trial of oxcarbazepine versus phenytoin in adults with previously untreated epilepsy. Epilepsy Res 1997;27:195-204. PMID: 9237054
28. Steiner TJ, Dellaportas CI, Findley LJ, et al. Lamotrigine monotherapy in newly diagnosed untreated epilepsy: a double-blind comparison with phenytoin. Epilepsia 1999;40:601-7. PMID: 10386529
29. Ramsay E, Faught E, Krumholz A, et al. Efficacy, tolerability, and safety of rapid initiation of topiramate versus phenytoin in patients with new-onset epilepsy: A randomized double-blind clinical trial. Epilepsia 2010;July:[Epub ahead of print]. PMID: 20633037
30. Labiner D, Paradis PE, Manjunath R, et al. Generic antiepileptic drugs and associated medical resource utilization in the United States. Neurology 2010;74:1566-74. PMID: 20393142
31. Duh MS, Paradis PE, Latremouille-Viau D, et al. The risks and costs of multiple-generic substitution of topiramate. Neurology 2009;72:2122-9. PMID: 19528520
32. Paradis PE, Latremouille-Viau D, Moore Y, et al. Projected economic impact of clinical findings of generic entry of topiramate on G4 European countries. Curr Med Res Opin 2009;25:1793-805. PMID: 19505202
33. Zachry WM,3rd, Doan QD, Clewell JD, et al. Case-control analysis of ambulance, emergency room, or inpatient hospital events for epilepsy and antiepileptic drug formulation changes. Epilepsia 2009;50:493-500. PMID: 18616554
34. Rascati KL, Richards KM, Johnsrud MT, et al. Effects of antiepileptic drug substitutions on epileptic events requiring acute care. Pharmacotherapy 2009;29:769-74. PMID: 19558250
35. Devine ST, Barron J, Behm A. Acute epilepsy exacerbations in patients switched between A-rated anti-epileptic drugs. Curr Med Res Opin 2010;26:455-63. PMID: 20014974
36. Garnett WR, Gilbert TD, O'Connor P. Patterns of care, outcomes, and direct health plan costs of antiepileptic therapy: a pharmacoeconomic analysis of the available carbamazepine formulations. Clin Ther 2005;27:1092-103. PMID: 16154489
37. Kesselheim AS, Stedman MR, Bubrick EJ, et al. Seizure outcomes following the use of generic versus brand-name antiepileptic drugs. Drugs 2010;70:605-21. PMID: 20329806

Full Report

This executive summary is part of the following document: Talati R, Scholle JM, Phung OJ, Baker WL, Baker EL, Ashaye A, Kluger J, Quercia R, Mather J, Giovanale S, Coleman CI, White CM. Effectiveness and Safety of Antiepileptic Medications in Patients with Epilepsy. Comparative Effectiveness Review No. 40. (Prepared by the University of Connecticut/Hartford Hospital Evidence-based Practice Center under Contract No. 20-2007-10067-I.) AHRQ Publication No. 11(12)-EHC082-EF. Rockville, MD: Agency for Healthcare Research and Quality. December 2011.

For More Copies

For more copies of Effectiveness and Safety of Antiepileptic Medications in Patients with Epilepsy: Comparative Effectiveness Review Executive Summary No. 40 (AHRQ Pub. No. 11(12)-EHC082-1), please call the AHRQ Clearinghouse at 1-800-358-9295.

Return to Top of Page