[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure 1.
Participant Disposition Diagram
Participant Disposition Diagram

aThree participants completed only the first 3 pharmacokinetic assessments.

AED indicates antiepileptic drug; PKA, pharmacokinetic assessment.

Figure 2.
Mean Lamotrigine Plasma Concentrations After Single-Dose Administration of 1 Branded Lamotrigine and 2 Disparate Generic Lamotrigine Products
Mean Lamotrigine Plasma Concentrations After Single-Dose Administration of 1 Branded Lamotrigine and 2 Disparate Generic Lamotrigine Products

Main graph, lamotrigine plasma concentration from 0 to 96 hours. Inset, lamotrigine plasma concentration from 0 to 360 minutes (0-6 hours). gLTG indicates generic lamotrigine; error bars, SD

Figure 3.
Mean Ratio of Drug Formulation Estimates for 3 Lamotrigine Drug Products
Mean Ratio of Drug Formulation Estimates for 3 Lamotrigine Drug Products

The area under the concentration–time curve (AUC0-96) is designated as an orange circle (percentage point estimate [PE]) and orange line (90% CI); Cmax is designated as a blue circle (percentage PE) and blue line (90% CI). Brand indicates branded lamotrigine; FDA, US Food and Drug Administration; and gLTG, generic lamotrigine.

Table 1.  
Participant Demographic and Baseline Characteristics
Participant Demographic and Baseline Characteristics
Table 2.  
Lamotrigine Within-Subject Variability
Lamotrigine Within-Subject Variability
1.
Shaw  SJ, Hartman  AL.  The controversy over generic antiepileptic drugs.  J Pediatr Pharmacol Ther. 2010;15(2):81-93.PubMedGoogle Scholar
2.
Center for Drug Evaluation and Research; US Food and Drug Administration. Guidance for Industry: Bioavailability and Bioequivalence Studies Submitted in NDAs or INDs—General Considerations. https://www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm389370.pdf. Published March 2014. Accessed August 15, 2016.
3.
Center for Drug Evaluation and Research; US Food and Drug Administration. Guidance for Industry: Statistical Approaches to Establishing Bioequivalence. https://www.fda.gov/downloads/drugs/guidances/ucm070244.pdf. Published January 2001. Accessed August 15, 2016.
4.
Center for Drug Evaluation and Research; US Food and Drug Administration. Guidance: Statistical Procedures for Bioequivalence Studies Using a Standard Two-Treatment Crossover Design. http://www.reocities.com/HotSprings/Spa/6896/stat-two.pdf. Published July 1992. Accessed August 15, 2016.
5.
Galgatte  UC, Jamdade  VR, Aute  PP, Chaudhari  PD.  Study on requirements of bioequivalence for registration of pharmaceutical products in USA, Europe and Canada.  Saudi Pharm J. 2014;22(5):391-402.PubMedGoogle ScholarCrossref
6.
Westlake  WJ.  Use of confidence intervals in analysis of comparative bioavailability trials.  J Pharm Sci. 1972;61(8):1340-1341.PubMedGoogle ScholarCrossref
7.
Cabana  BE.  Assessment of 75/75 rule: FDA viewpoint.  J Pharm Sci. 1983;72(1):98-100.PubMedGoogle ScholarCrossref
8.
Metzler  CM, Huang  DC.  Statistical methods for bioavailability and bioequivalence.  Clin Res Pr Drug Regul Aff. 1983;1(2):109-132.Google Scholar
9.
US Food and Drug Administration.  Report by the Bioequivalence Task Force on Recommendations From the Bioequivalence Hearing Conducted by the Food and Drug Administration, 9/29-10/1/1986. Silver Spring, MD: US Food & Drug Administration; 1988:29-33.
10.
Sander  JW, Ryvlin  P, Stefan  H, Booth  DR, Bauer  J.  Generic substitution of antiepileptic drugs.  Expert Rev Neurother. 2010;10(12):1887-1898.PubMedGoogle ScholarCrossref
11.
Henderson  JD, Esham  RH.  Generic substitution: issues for problematic drugs.  South Med J. 2001;94(1):16-21.PubMedGoogle ScholarCrossref
12.
Davit  BM, Nwakama  PE, Buehler  GJ,  et al.  Comparing generic and innovator drugs: a review of 12 years of bioequivalence data from the United States Food and Drug Administration.  Ann Pharmacother. 2009;43(10):1583-1597.PubMedGoogle ScholarCrossref
13.
Berg  MJ, Gross  RA, Haskins  LS, Zingaro  WM, Tomaszewski  KJ.  Generic substitution in the treatment of epilepsy: patient and physician perceptions.  Epilepsy Behav. 2008;13(4):693-699.PubMedGoogle ScholarCrossref
14.
Gidal  BE, Tomson  T.  Debate: substitution of generic drugs in epilepsy: is there cause for concern?  Epilepsia. 2008;49(suppl 9):56-62.PubMedGoogle ScholarCrossref
15.
Ting  TY, Jiang  W, Lionberger  R,  et al.  Generic lamotrigine versus brand-name Lamictal bioequivalence in patients with epilepsy: a field test of the FDA bioequivalence standard.  Epilepsia. 2015;56(9):1415-1424.PubMedGoogle ScholarCrossref
16.
Privitera  MD, Welty  TE, Gidal  BE,  et al.  Generic-to-generic lamotrigine switches in people with epilepsy: the randomised controlled EQUIGEN trial.  Lancet Neurol. 2016;15(4):365-372.PubMedGoogle ScholarCrossref
17.
Jackson  AJ.  Prediction of steady-state bioequivalence relationships using single dose data, I—linear kinetics.  Biopharm Drug Dispos. 1987;8(5):483-496.PubMedGoogle ScholarCrossref
18.
Duh  MS, Andermann  F, Paradis  PE, Weiner  J, Manjunath  R, Crémieux  PY.  The economic consequences of generic substitution for antiepileptic drugs in a public payer setting: the case of lamotrigine.  Dis Manag. 2007;10(4):216-225.PubMedGoogle ScholarCrossref
19.
LeLorier  J, Duh  MS, Paradis  PE,  et al.  Clinical consequences of generic substitution of lamotrigine for patients with epilepsy.  Neurology. 2008;70(22, pt 2):2179-2186.PubMedGoogle ScholarCrossref
20.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194.PubMedGoogle ScholarCrossref
21.
Kwan  P, Arzimanoglou  A, Berg  AT,  et al.  Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies  [published correction appears in Epilepsia. 2010;51(9):1922].  Epilepsia. 2010;51(6):1069-1077.PubMedGoogle ScholarCrossref
22.
Naranjo  CA, Busto  U, Sellers  EM,  et al.  A method for estimating the probability of adverse drug reactions.  Clin Pharmacol Ther. 1981;30(2):239-245.PubMedGoogle ScholarCrossref
23.
Diaz  FJ, Berg  MJ, Krebill  R,  et al.  Random-effects linear modeling and sample size tables for two special crossover designs of average bioequivalence studies: the four-period, two-sequence, two-formulation and six-period, three-sequence, three-formulation designs.  Clin Pharmacokinet. 2013;52(12):1033-1043.PubMedGoogle ScholarCrossref
24.
Schall  R, Endrenyi  L, Ring  A.  Residuals and outliers in replicate design crossover studies.  J Biopharm Stat. 2010;20(4):835-849.PubMedGoogle ScholarCrossref
25.
Tothfalusi  L, Endrenyi  L, Arieta  AG.  Evaluation of bioequivalence for highly variable drugs with scaled average bioequivalence.  Clin Pharmacokinet. 2009;48(11):725-743.PubMedGoogle ScholarCrossref
26.
Center for Drug Evaluation and Research. Briefing information for the FDA meeting of the Advisory Committee for Pharmaceutical Science and Clinical Pharmacology. Topic 1: bioequivalence (BE) and quality standards for narrow therapeutics index (NTI) drug products. Meeting of the Advisory Committee for Pharmaceutical Science and Clinical Pharmacology; July 26, 2011. https://www.fda.gov/ucm/groups/fdagov-public/@fdagov-afda-adcom/documents/document/ucm263465.pdf. Accessed August 15, 2016.
27.
Zhang  X, Zheng  N, Lionberger  RA, Yu  LX.  Innovative approaches for demonstration of bioequivalence: the US FDA perspective.  Ther Deliv. 2013;4(6):725-740.PubMedGoogle ScholarCrossref
28.
Tamargo  J, Le Heuzey  JY, Mabo  P.  Narrow therapeutic index drugs: a clinical pharmacological consideration to flecainide.  Eur J Clin Pharmacol. 2015;71(5):549-567.PubMedGoogle ScholarCrossref
Original Investigation
August 2017

Bioequivalence Between Generic and Branded Lamotrigine in People With Epilepsy: The EQUIGEN Randomized Clinical Trial

Author Affiliations
  • 1Department of Neurology, University of Rochester School of Medicine and Dentistry, Rochester, New York
  • 2College of Pharmacy and Health Sciences, Drake University, Des Moines, Iowa
  • 3School of Pharmacy and Department of Neurology, University of Wisconsin–Madison
  • 4Department of Biostatistics, The University of Kansas Medical Center, Kansas City
  • 5Department of Neurology, University of Alabama at Birmingham
  • 6Department of Neurology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts
  • 7Department of Neurology, University of Pennsylvania, Philadelphia
  • 8Zeeh Pharmaceutical Experiment Station, University of Wisconsin–Madison
  • 9Office of Research and Standards, Office of Generic Drugs, Center for Drug Evaluation and Research, US Food and Drug Administration (FDA), Silver Spring, MD
  • 10SuccinctChoice Medical Communications, Chicago, Illinois
  • 11Department of Neurology, University of Cincinnati Medical Center, Cincinnati, Ohio
JAMA Neurol. 2017;74(8):919-926. doi:10.1001/jamaneurol.2017.0497
Key Points

Question  Are branded-to-generic and generic-to-generic lamotrigine switches bioequivalent in people with epilepsy?

Findings  In this randomized clinical trial involving 50 adults with epilepsy, bioequivalence between branded and generic lamotrigine products was established for 2 key pharmacokinetic measures (area under the concentration–time curve and maximal concentration).

Meaning  Branded and generic lamotrigine products can be substituted with an expectation of bioequivalence.

Abstract

Importance  Switching between generic antiepileptic drugs is a highly debated issue that affects both clinical care and overall health care costs.

Objective  To evaluate the single-dose pharmacokinetic bioequivalence of 3 (1 branded and 2 generic drugs) on-market, immediate-release lamotrigine drug products.

Design, Setting, and Participants  The Equivalence Among Antiepileptic Drug Generic and Brand Products in People With Epilepsy (EQUIGEN) single-dose study is a crossover, prospective, sequence-randomized, replicate pharmacokinetic study conducted at 5 US academic epilepsy centers. Fifty adults (≥18 years) with epilepsy who were taking concomitant antiepileptic drugs and not currently receiving lamotrigine were enrolled between July 18, 2013, and January 19, 2015. Every participant was randomly assigned to 1 of 3 equivalent sequences, each comprising 6 study periods, during which they had blood draws before and after medication administration. Forty-nine participants were included in intention-to-treat analyses.

Interventions  Participants received a single 25-mg dose of immediate-release lamotrigine at the start of each period, with the branded and the 2 most disparate generic products each studied twice. Lamotrigine was selected as the antiepileptic drug of interest because of its wide use, publications indicating problems with generic switches, and complaints to the US Food and Drug Administration regarding generic products. Both participants and study personnel were blinded to the specific generic products selected.

Main Outcomes and Measures  The primary outcome was bioequivalence between products. Maximum plasma concentration (Cmax) and area under the concentration–time curve (AUC) were compared, and average bioequivalence (ABE) was established if the 90% CIs of the ratios of the 2 products were within equivalence limits (80%-125%).

Results  Of the 50 randomized participants, 49 (98%) received all 3 lamotrigine products and completed at least 3 pharmacokinetic assessments and 46 (92%) completed all 6 pharmacokinetic assessments. Among the 49 participants, 28 (57%) were men and 21 (43%) were women, 42 (86%) self-identified as white, and 46 (16) years was the mean (SD) age. The 3 drug products were considered bioequivalent because the 90% CIs were within equivalence limits (lowest and highest CI limits for Cmax, 92.6% and 110.4%; for AUC0-96, 96.9% and 101.9%). Replicate testing demonstrated no significant differences in within-subject variability across the 3 products (likelihood ratios, χ22 for log-transformed variables: AUC0-96, 2.58; Cmax, 0.64; and AUC0-∞, 4.05; P ≥ .13) and that the 3 products were also bioequivalent according to scaled ABE and individual bioequivalence criteria with no subject × formulation interaction (Cmax, 0.00; AUC0-96, 0.54; and AUC0-∞, 0.36; P ≥ .76).

Conclusions and Relevance  This study provides evidence that the disparate lamotrigine products studied are bioequivalent when tested in people with epilepsy taking concomitant antiepileptic drugs.

Trial Registration  clinicaltrials.gov Identifier: NCT01733394

Introduction

Equivalence between generic and branded antiepileptic drugs (AEDs) is an issue that affects both clinical care and health care costs.1 Regulatory agencies, such as the US Food and Drug Administration (FDA), determine the equivalence of generic and branded products on the basis of 2 key pharmacokinetic (PK) measures: area under the concentration–time curve (AUC) and maximal concentration (Cmax).2 As differences in systemic drug exposure up to 20% were concluded not to be clinically significant, bioequivalence between 2 products is established if the 90% CI of the ratios of the generic to the reference compound for the AUC and Cmax fall within 80% to 125%.2-5 This range, established from the FDA Bioequivalence Hearing, was based on both statistical analysis and expert opinion6-9; although some consider this range too broad, the actual difference between products that satisfy these requirements is typically less than 10%.10-12

Bioequivalence data suggest that generic AEDs can be interchanged for branded or other generic drug products, but patients and clinicians share reservations about indiscriminate substitution.13,14 Many concerns stem from conflicting medical literature and uncertainties regarding the design of bioequivalence studies. Within the past year, 2 randomized, prospective, well-controlled, chronic-dosing bioequivalence studies of people with epilepsy independently demonstrated bioequivalence between a branded product and the most widely used generic lamotrigine product15 as well as between the 2 most disparate generic lamotrigine products available in the commercial market.16 As secondary end points, these studies detected no loss of seizure control or no emergence of any new adverse effect when switching from branded to generic lamotrigine15 or from generic to generic lamotrigine.16 As such, problems associated with switching between AEDs were not likely to be related to PK differences with generic products.15,16

Single-dose bioequivalence studies are perceived to be more sensitive than multiple-dose studies in detecting formulation differences because each participant starts and ends the PK assessment with essentially nondetectable drug plasma concentrations.2 This notion is supported by simulation data suggesting that the probability of failing the test of bioequivalence decreases considerably with multiple-dose administration.17 Consequently, multiple-dose studies are generally not recommended by the FDA for either immediate-release or modified-release products.2 Therefore, it is possible that differences in AUC and Cmax between branded and generic drug products were not identified in the previous studies. In addition, the previous studies did not fully evaluate and compare the variability (eg, within-subject variability and subject × formulation interaction variability) of branded and generic lamotrigine.

The Equivalence Among Antiepileptic Drug Generic and Brand Products in People With Epilepsy (EQUIGEN) study group evaluated the single-dose PK profiles of branded lamotrigine (Lamictal; GlaxoSmithKline PLC) and the 2 most disparate generic lamotrigine (generic lamotrigine–high and generic lamotrigine–low) drug products available on the market using a prospective, replicate, single-dose study design in people with epilepsy taking concomitant AEDs. Lamotrigine was selected as the AED to be assessed because of its wide clinical use, publications indicating problems with generic switches, and complaints to the FDA regarding generic products.18,19

Methods
Participants

Between July 18, 2013, and January 19, 2015, 59 adults (aged ≥18 years) with epilepsy from 5 US academic epilepsy centers (University of Rochester Medical Center, University of Alabama at Birmingham, Brigham and Women’s Hospital Epilepsy Center, University of Pennsylvania, and University of Cincinnati Medical Center) who met inclusion and exclusion criteria were approached to enroll as participants. The study was conducted in accordance with Declaration of Helsinki20 ethical principles, the FDA Good Clinical Practice standards, principles of informed consent, and requirements of public registration of clinical trials. Prior to study initiation, each investigator obtained approval from a local institutional review board and from the FDA Research Involving Human Subject Committee (the complete trial protocol is available in Supplement 1). Written informed consent was obtained from each participant at enrollment. An independent medical monitor assessed the study data with particular consideration of subject safety. Data analyses were conducted from July 1, 2015, to July 31, 2016.

The single-dose PK profiles of 3 lamotrigine drug products (2 on-market, disparate, generic lamotrigine drug products [generic lamotrigine–high and generic lamotrigine–low] and 1 branded lamotrigine product) were compared in adults with epilepsy taking a stable, lamotrigine-free AED treatment regimen, including adults with refractory epilepsy (as defined by the International League Against Epilepsy criteria21). Potential participants were excluded if they had received lamotrigine, valproate (all forms), estrogens, rifampin, orlistat, felbamate, or sertraline hydrochloride within 28 days of enrollment. Also excluded were adults with a progressive central nervous system disorder, a history of alcohol or substance use disorder, psychogenic seizures within the past 2 years, any clinically significant psychiatric illness, psychological or behavioral problems, clinically significant illness or laboratory abnormality, or known medication nonadherence. Those with any history of allergic reaction to lamotrigine, 2 or more allergic reactions to an AED, 1 serious hypersensitivity reaction to an AED, or a history of adverse effects associated with the prior use of lamotrigine were excluded as well. Complaints of previous sensitivity to switching between lamotrigine products were assessed using a modified Naranjo criteria22 by the same method as in the EQUIGEN chronic study.16

Procedures

Disparate generic lamotrigine products were selected using a 3-step process. In step 1, products with the most disparate Cmax and AUC were identified from the Abbreviated New Drug Applications submitted to the FDA by all generic drug manufacturers. In step 2, in vitro dissolution testing, plus content and impurities analysis, was conducted by an independent laboratory (Zeeh Pharmaceutical Experiment Station) to evaluate product performance. The most disparate products emerging after steps 1 and 2 with different excipient composition and availability on the market were chosen for the study. Both participants and study personnel were blinded to the specific generic products selected and their predicted exposure (ie, high or low).

After completion of a 2-day to 30-day randomization period that allowed for flexibility in scheduling at the inpatient PK facility, every participant was randomly assigned to one of 3 sequences (Figure 1), each comprising 6 study periods. All participants received a single, subtherapeutic (25-mg) dose of each lamotrigine product (1 branded, 1 generic lamotrigine–high, and 1 generic lamotrigine–low; eFigure in Supplement 2); the 25-mg dose was selected to minimize the risk of adverse events, but the dose was adequate for PK assessment with the highly sensitive assay used. The product sequence order was based on FDA guidance criteria and was modified in accordance with discussions between the EQUIGEN committee and the FDA.23 Single doses of study medication were administered to each participant after an overnight fast during an in-facility, 12-hour PK session; 4 additional outpatient samples were drawn at daily intervals after dose administration. Each in-facility PK testing was separated by a 12-day to 23-day washout period; consistent washout periods of 14 days were preferred. During the study, the participants continued their usual concomitant medications, including AEDs, without change.

Commercial lamotrigine products were administered in this study. The site personnel opened the study medication bottle, confirmed the medication was correct, placed the medication in an opaque container (small envelope), and gave the medication to the participant in a manner that prevented the participant from seeing the tablet and was therefore masked to it. Although the site personnel verified the brand or generic status of the medication administered, they were blinded to which generic lamotrigine was predicted with high exposure (generic lamotrigine–high) and which was predicted with low exposure (generic lamotrigine–low). The site investigators were blinded to both drug and participant sequence, did not see the tablets, and were not informed of the manufacturers of the tested generic products.

Each 96-hour assessment began with admission to a testing facility. Blood samples were drawn just before dose administration (pre-dose trough [0 hours]); the dose was administered; and then blood samples were obtained at 20, 40, 60, 80, 100, 120, 140, 160, 180, 200, and 220 minutes as well as at 4, 4.5, 5, 6, 8, 10, and 12 hours within 5 minutes of the specified time. After discharge from each PK facility, the participant had 4 additional blood samples drawn at 24, 48, 72, and 96 hours (±4 hours) following dose administration. Plasma samples were analyzed for lamotrigine level at a central laboratory (University of Rochester Medical Center Laboratories) using a validated high-performance liquid chromatography with tandem mass spectrometry method with a quantification limit of 10 ng/mL.

The primary objective was to determine if there were significant bioequivalence deviations in Cmax or AUC between the branded, generic lamotrigine–high, and generic lamotrigine–low products. Within-subject variances of the branded and generic products, as well as subject × interaction variances, were evaluated as primary outcomes. These variances were used to assess individual bioequivalence and detect subject × formulation interaction outliers.24 The secondary objectives were to assess bioequivalence in various study population subgroups (eg, by sex, age, and concomitant enzyme-inducing AED status) and scaled average bioequivalence (ABE). In addition, adverse effects, including changes in seizure frequency, were reported in the participant paper diary and captured at study interviews.

Statistical Analysis

The methods and software used in this study for sample size computations have previously been published.23 Briefly, for testing the null hypothesis of non-mean bioequivalence for Cmax between the 2 lamotrigine generic products, a sample size of 45 participants would provide a power of 90%. Equivalence of lamotrigine products was determined by comparing their Cmax, AUC to time 96 hours (AUC0-96), and AUC extrapolated to infinity (AUC0-∞) using ABE criteria. Pharmacokinetic equivalence between the 2 lamotrigine generic products was established if the 90% CIs of the ratio of the 2 products for Cmax and AUC were within the range of 80% to 125%. The AUC0-96 and AUC0-∞ were calculated using the linear trapezoidal rule with PK modeling software (PK Solutions, version 2.0; Summit Research Services). All PK analyses were conducted blinded to the drug product. Maximum concentration (Cmax) and time to maximum concentration were calculated by visual inspection.16

Random-effects linear models of log-transformed Cmax and AUCs were fitted using data from all participants and all periods. Specific products, treatment sequences, and study periods were included in the model as independent variables. The intercept and the regression coefficients corresponding to the variables representing the products were treated as random coefficients; all other regression coefficients were treated as fixed effects. Because lamotrigine is inducible, a variable representing participants who received hepatic enzyme-inducing AEDs vs those who did not receive concomitant enzyme-inducing AEDs was also included in the model as an independent variable. An interaction term between variables representing induction and the variables representing the generic products was also examined. The statistical model provided all within-subject variances needed to perform ABE and scaled ABE analyses to compare the 3 drug products. Scaled ABE was examined for Cmax and both AUCs through 95% upper confidence bounds of linearized criteria using the Howe approximation.25 There was no evidence supporting models containing subject × formulation interaction variances; therefore, individual bioequivalence analyses reduced to scaled ABE analyses in these lamotrigine data.15 Statistical models included 3 residual variances—1 per product; within-subject coefficients of variation of lamotrigine were computed as follows:

Image description not available.

where σ2 is a residual variance under an approximate log-normal distribution that was verified with residual analyses. Outlier analyses were conducted using studentized residuals based on contrasts.24 One randomized participant who did not receive all 3 lamotrigine products was excluded per protocol. The other 49 participants were included in intention-to-treat analyses.

Results
Participant Disposition, Demographics, and Baseline Characteristics

Of the 59 adults screened, 50 (85%) were enrolled and randomized into 3 separate sequences (sequence 1, 17 participants [34%]; sequence 2, 16 [32%]; and sequence 3, 17 [34%]; Figure 1). Study retention was high; 49 of 50 participants received all 3 lamotrigine products, completed at least 3 PK assessments, and were included in the ABE analysis population. Forty-six participants (92%) completed all 6 PK periods and were included in the replicate analysis population.

Among the 49 participants, 28 (57%) were men and 21 (43%) were women, 42 (86%) self-identified as white, and 46 (16) years was the mean (SD) age. Demographics and baseline disease characteristics for the participants in the ABE analysis population are presented in Table 1.

Pharmacokinetic Evaluation

Plasma concentrations pooled from all sequences following a single 25-mg dose of branded lamotrigine and 2 disparate generic lamotrigine drug products over 96 hours are presented in Figure 2. Across all 6 periods, 7 of the 285 samples taken at time 0 had trace amounts of lamotrigine (all <18 ng/mL), which were considered too small to affect the results.

Across the time course, no significant differences in PK profiles of the branded lamotrigine and the 2 generic lamotrigine products were observed (Figure 3). No participant demonstrated an aberrant PK profile, to suggest an absence of bioequivalence in an individual. The 3 drug products were considered bioequivalent because the 90% CIs of relative bioavailabilities were within the 80%-125% equivalence limits for both Cmax and AUC, for all comparisons across the 3 drug products. The relative bioavailabilities (and the 90% CIs) comparing generic lamotrigine-high vs brand products were: 102.2% (98.7%-105.8%) for Cmax; 99.0% (96.9%-101.2%) for AUC0-96; and 97.8% (94.9%-100.8%) for AUC0-∞ (Figure 3). The relative bioavailabilities comparing generic lamotrigine-low vs brand were: 96.0% (92.6%-99.6%) for Cmax; 99.4% (97.6%-101.2%) for AUC0-96; and 98.5% (95.9%-101.2%) for AUC0-∞. Finally, the relative bioavailabilities comparing generic lamotrigine-high vs generic lamotrigine-low were: 106.4% (102.6%-110.4%) for Cmax; 99.6% (97.3%-101.9%) for AUC0-96; and 99.3% (96.6%-102.0%) for AUC0-∞.

Replicate exposure to the same product demonstrated that there was no evidence for differences in within-subject variabilities across the 3 products. For log-transformed AUC0-96, after excluding an influential observation, likelihood ratio χ22 = 2.58; P = .27; for log-transformed Cmax, including all observations, χ22 = 0.64; P = .73; and for log-transformed AUC0-∞, including all observations, χ22 = 4.05; P = .13. Confidence intervals for within-subject coefficients of variation for AUC0-96, AUC0-∞, and Cmax overlapped across the 3 products (Table 2).

After controlling for sequence and period, there were no significant subject × formulation interaction variabilities: for log-transformed Cmax, likelihood ratio χ22 =0.00, P=.99; for log-transformed AUC0-96 , χ22=0.54, P=.76 (after excluding one influential observation); and for log-transformed AUC0-∞, χ22=0.36, P=.83. The drug products were also found to be bioequivalent according to the scaled ABE criteria for log-transformed Cmax, AUC0-96, and AUC0-∞, with the 95% upper confidence bounds for linearized criteria less than zero. The lack of significant subject × formulation interaction variability or of differential within‐subject variability between products imply that individual bioequivalence analyses reduce to scaled ABE analyses in this lamotrigine data.

Observed bioequivalence of AUCs and Cmax between the different lamotrigine drug products was not affected by potential confounding factors (eg, treatment sequence, study period, and age). According to a random-effects regression model of the log of Cmax, after controlling for product, sequence, period, and site and taking inducers, biological sex did not have a significant effect on Cmax (F1,39.8 = 1.10; P = .75). An additional analysis showed that there was a nonsignificant interaction between biological sex and products (F2,145 = 2.8; P = .07). As expected, when controlling for product, treatment sequence, study period, and study site, concomitant hepatic enzyme–inducing AEDs significantly reduced the lamotrigine AUC0-96 by a mean of 33% across all products (95% CI, −43.7% to −20.4%; P < .001) but produced only a small (5.5%), nonsignificant decrease in Cmax (95% CI, −16.6% to 7.0%; P = .36). However, after controlling for this interaction, the conclusions on bioequivalence remained the same for both AUC0-96 and Cmax.

Adverse Events

Overall, single doses of lamotrigine were well tolerated with no obvious differences in the adverse events reported between any of the drug products. Among the 49 participants, 132 adverse events were reported. Most of these adverse events were mild and considered not related to study medication. Study-related, treatment-emergent adverse events with an incidence of 2 or more were “rash” (coinciding with the location of adhesive tape for the intravenous catheter), pain at the phlebotomy site, headaches, fatigue, nightmares, and nausea. The low frequency of adverse events precluded any systematic observation of an association with a single product. Three serious adverse events were reported during this study (death resulting from a motor vehicle crash, acute gallbladder obstruction, and vaginal bleeding), but none of them was considered associated with treatment.

Discussion

To inform the debate about generic AED equivalence, we performed a single-dose replicate study (among people with epilepsy taking concomitant AEDs) comparing bioequivalence between disparate generic lamotrigine products and bioequivalence of each generic product to branded lamotrigine. Confidence intervals comparing the extent (AUC) and rate (Cmax) of absorption of these tested products were all well within the FDA ABE range, thus establishing bioequivalence. This study provides further support to the bioequivalence found in 2 recently reported long-term–dose studies,15,16 especially because single-dose study designs are considered more sensitive to PK differences than are long-term–dose studies.2

Most of the published literature concerning the therapeutic implications of branded-to-generic or generic-to-generic substitutions is uncontrolled or comprises retrospective studies that did not account for confounding factors (eg, adherence and participant bias). This study’s prospective replicate design enabled us to determine the within-subject variations over time in the PK profile for the 3 products and to explore subject × formulation interaction. Replicate testing demonstrated variability of approximately 8% (AUC) and 15% (Cmax) within each of the individual products. The finding that biological variation was similar for the branded and both generic products suggests that this variation may account for some of the differences reported when switching among branded and generic products.

Biological factors are presumed to have similar effects regardless of the product (branded or generic) taken by the participant. Establishing the amount of variation from these factors for the branded product is important for understanding how much variation among generic products is acceptable. That is, if there is a wide variation in the PK measures from a replicate study of the branded product, then similar wide variations among generic products should be acceptable. By contrast, if the differences in a replicate study of the branded product are small, then wider variations among generic products may be unacceptable.

Within-subject variability is one of the factors proposed to characterize narrow therapeutic index drugs, which are required to meet reference-scaled bioequivalence limits using the scaled ABE method.26-28 The PK results for the products tested in our study fell well within scaled ABE bounds. Individual bioequivalence was also established with this set of lamotrigine data given that individual bioequivalence analyses reduced to scaled ABE analyses because there was no subject × formulation interaction.

Limitations

This study evaluated 2 different generic lamotrigine products from different manufacturers. We sought to study the most disparate products, but the original Abbreviated New Drug Applications data on these products had relatively narrow low-to-high CI ranges for both Cmax and AUCs. Our determination of the disparate designation (high vs low) was primarily based on the findings from an in vitro dissolution study conducted at an independent laboratory. Ideally, all available generic products should be studied and all AEDs that have caused concerns should be studied, but doing so is highly impractical. This study was designed in an attempt to provide data that can be extrapolated to other drugs. Of note, we studied an immediate-release AED formulation; our findings may not be applicable to modified-release formulations that have a variety of formulation and PK complexity.

An additional limitation of this study was possible selection bias. Participants who may have previously experienced changes in seizure frequency or increased adverse events with generic switching could be less likely to volunteer for a study involving generic substitution. No participant enrolled in this study met the criteria for a previous problem with generic switching. By contrast, in the EQUIGEN long-term–dose study, 4 participants (12%) met the criterion for switching sensitivity.16 We expected people with a history of switching sensitivity to be less willing to participate in the long-term–dose study than in the single-dose study because of the low dose of lamotrigine in the single-dose study. We doubt that history of switching sensitivity had an effect on recruitment in this single-dose study; however, it may be desirable to replicate the findings in enriched participants.

The primary aim of this study was not to test the validity of the FDA bioequivalence requirements for standard or narrow therapeutic index drugs, but questions have been raised regarding the clinical validity of bioequivalence. Analysis by the FDA of more than 2000 single-dose bioequivalence studies of generic products approved according to the current standard revealed that there was less than a 5% mean difference in bioavailability between generic and branded products.12 As such, it seems reasonable to assume that similar results would be seen if approved generic products were tested with our study design.

Conclusions

Despite the inclusion of participants with multiple clinically relevant variables, the 3 lamotrigine products tested demonstrated essentially identical PK profiles and were all well within the FDA bioequivalence standards. In fact, our results fell within the standards for narrow therapeutic index drugs and showed no evidence of differences in within-subject variability between the 3 products.

Back to top
Article Information

Corresponding Author: Michel Berg, MD, Department of Neurology, University of Rochester School of Medicine and Dentistry, 919 Westfall Rd, Clinton Crossings Bldg C, Ste 220, Rochester, NY 14618 (michel_berg@urmc.rochester.edu).

Accepted for Publication: March 14, 2017.

Published Online: June 26, 2017. doi:10.1001/jamaneurol.2017.0497

Author Contributions: Dr Berg had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Berg, Welty, Gidal, Diaz, Szaflarski, Elder, W. Jiang, X. Jiang, Privitera.

Acquisition, analysis, or interpretation of data: Berg, Welty, Gidal, Diaz, Krebill, Szaflarski, Dworetzky, Pollard, Elder, W. Jiang, Switzer, Privitera.

Drafting of the manuscript: Berg, Gidal, Krebill, Switzer, Privitera.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Berg, Diaz, Krebill.

Obtained funding: Berg, Privitera.

Administrative, technical, or material support: Berg, Welty, Szaflarski, Dworetzky, Pollard, Elder, W. Jiang.

Study supervision: Berg, Welty, Gidal, Szaflarski, Elder, W. Jiang, X. Jiang, Privitera.

Conflict of Interest Disclosures: Dr Berg reported being a site investigator for industry-sponsored research by Upsher-Smith Laboratories, Sunovion, Neuropace, Lundbeck, Pfizer, King Pharmaceuticals, Sage Therapeutics, and Acorda Therapeutics. Dr Welty reported serving as chair of the treatments committee of the American Epilepsy Society and receiving personal fees from Upsher-Smith Laboratories and Eisai. Dr Gidal reported receiving personal fees from UCB Pharma, Upsher-Smith Laboratories, Eisai, and Sunovion. Dr Szaflarski reported being a paid consultant for GW Pharmaceuticals, Upsher-Smith Laboratories, Sage Pharmaceuticals, and Biomedical Systems; serving on the editorial boards of Epilepsy & Behavior, Epilepsy Currents (contributing editor), Journal of Epileptology (associate editor), Journal of Medical Science, Folia Medica Copernicana, Restorative Neurology and Neuroscience (associate editor), and Conference Papers in Medicine; receiving research funding from the US Department of Defense, US FDA, American Epilepsy Society, Sage Pharmaceuticals, Eisai, UCB Pharma, the National Institutes of Health (NIH)/National Institute of Neurological Disorders and Stroke, the State of Alabama (for Carly’s Law), and the University of Alabama at Birmingham; and being an expert witness in legal proceedings. Dr Dworetzky reported receiving personal fees from Sleep Med and Best Doctors. Dr Pollard reported receiving grants from GlaxoSmithKline, Lundbeck, SK Pharmaceuticals, Upsher-Smith Laboratories, Eisai, and Cognizance Biomarkers; receiving personal fees from Lundbeck; and being part owner of a company executing an NIH Small Business Innovation Research program for biomarker research, which at this time has no financial value. Dr Elder reported owning GlaxoSmithKline stock and receiving personal fees from Mylan and Teva. Drs W. Jiang and X. Jiang reported being current employees of the FDA. Dr Switzer reported being a current employee of SuccinctChoice Medical Communications. Dr Privitera reported receiving grants from UCB Pharma, GW Pharmaceuticals, and Neuren; receiving personal fees from Astellas Pharma, Sage Pharmaceuticals, and Upsher-Smith Laboratories; and serving as an expert witness in legal proceedings. No other disclosures were reported.

Funding/Support: This study was funded by contract HHSF223201110112A from the FDA, a grant from the Epilepsy Foundation, and a generous gift from the American Epilepsy Society.

Role of the Funder/Sponsor: The funding sources had a role in the design of the study. Drs W. Jiang and X. Jiang, as employees of the FDA (a sponsor of the study), were involved in the analysis of the data as well as the preparation, review, and approval of the manuscript. They were included in the decision to submit the manuscript for publication.

Additional Contributions: The following were the study site coordinators: Nichol McBee, MPH, CCRP, and Noeleen Ostapkovich, MS, Division of Brain Injury Outcomes, Johns Hopkins University; Donna Schwieterman, RA, University of Cincinnati; Diane Smith, CCRP, University of Rochester; Nancy Cohen, RN, University of Alabama; Nichelle Llewellyn, BS, Brigham and Women’s Hospital; and Meryl Lozano, MBE, and Emily Acton, PBAC, University of Pennsylvania. Statistical analyses were performed by Francisco J. Diaz, PhD, and Ron Krebill, MPH, University of Kansas Medical Center. All of these contributors received compensation with funding for salary support.

References
1.
Shaw  SJ, Hartman  AL.  The controversy over generic antiepileptic drugs.  J Pediatr Pharmacol Ther. 2010;15(2):81-93.PubMedGoogle Scholar
2.
Center for Drug Evaluation and Research; US Food and Drug Administration. Guidance for Industry: Bioavailability and Bioequivalence Studies Submitted in NDAs or INDs—General Considerations. https://www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm389370.pdf. Published March 2014. Accessed August 15, 2016.
3.
Center for Drug Evaluation and Research; US Food and Drug Administration. Guidance for Industry: Statistical Approaches to Establishing Bioequivalence. https://www.fda.gov/downloads/drugs/guidances/ucm070244.pdf. Published January 2001. Accessed August 15, 2016.
4.
Center for Drug Evaluation and Research; US Food and Drug Administration. Guidance: Statistical Procedures for Bioequivalence Studies Using a Standard Two-Treatment Crossover Design. http://www.reocities.com/HotSprings/Spa/6896/stat-two.pdf. Published July 1992. Accessed August 15, 2016.
5.
Galgatte  UC, Jamdade  VR, Aute  PP, Chaudhari  PD.  Study on requirements of bioequivalence for registration of pharmaceutical products in USA, Europe and Canada.  Saudi Pharm J. 2014;22(5):391-402.PubMedGoogle ScholarCrossref
6.
Westlake  WJ.  Use of confidence intervals in analysis of comparative bioavailability trials.  J Pharm Sci. 1972;61(8):1340-1341.PubMedGoogle ScholarCrossref
7.
Cabana  BE.  Assessment of 75/75 rule: FDA viewpoint.  J Pharm Sci. 1983;72(1):98-100.PubMedGoogle ScholarCrossref
8.
Metzler  CM, Huang  DC.  Statistical methods for bioavailability and bioequivalence.  Clin Res Pr Drug Regul Aff. 1983;1(2):109-132.Google Scholar
9.
US Food and Drug Administration.  Report by the Bioequivalence Task Force on Recommendations From the Bioequivalence Hearing Conducted by the Food and Drug Administration, 9/29-10/1/1986. Silver Spring, MD: US Food & Drug Administration; 1988:29-33.
10.
Sander  JW, Ryvlin  P, Stefan  H, Booth  DR, Bauer  J.  Generic substitution of antiepileptic drugs.  Expert Rev Neurother. 2010;10(12):1887-1898.PubMedGoogle ScholarCrossref
11.
Henderson  JD, Esham  RH.  Generic substitution: issues for problematic drugs.  South Med J. 2001;94(1):16-21.PubMedGoogle ScholarCrossref
12.
Davit  BM, Nwakama  PE, Buehler  GJ,  et al.  Comparing generic and innovator drugs: a review of 12 years of bioequivalence data from the United States Food and Drug Administration.  Ann Pharmacother. 2009;43(10):1583-1597.PubMedGoogle ScholarCrossref
13.
Berg  MJ, Gross  RA, Haskins  LS, Zingaro  WM, Tomaszewski  KJ.  Generic substitution in the treatment of epilepsy: patient and physician perceptions.  Epilepsy Behav. 2008;13(4):693-699.PubMedGoogle ScholarCrossref
14.
Gidal  BE, Tomson  T.  Debate: substitution of generic drugs in epilepsy: is there cause for concern?  Epilepsia. 2008;49(suppl 9):56-62.PubMedGoogle ScholarCrossref
15.
Ting  TY, Jiang  W, Lionberger  R,  et al.  Generic lamotrigine versus brand-name Lamictal bioequivalence in patients with epilepsy: a field test of the FDA bioequivalence standard.  Epilepsia. 2015;56(9):1415-1424.PubMedGoogle ScholarCrossref
16.
Privitera  MD, Welty  TE, Gidal  BE,  et al.  Generic-to-generic lamotrigine switches in people with epilepsy: the randomised controlled EQUIGEN trial.  Lancet Neurol. 2016;15(4):365-372.PubMedGoogle ScholarCrossref
17.
Jackson  AJ.  Prediction of steady-state bioequivalence relationships using single dose data, I—linear kinetics.  Biopharm Drug Dispos. 1987;8(5):483-496.PubMedGoogle ScholarCrossref
18.
Duh  MS, Andermann  F, Paradis  PE, Weiner  J, Manjunath  R, Crémieux  PY.  The economic consequences of generic substitution for antiepileptic drugs in a public payer setting: the case of lamotrigine.  Dis Manag. 2007;10(4):216-225.PubMedGoogle ScholarCrossref
19.
LeLorier  J, Duh  MS, Paradis  PE,  et al.  Clinical consequences of generic substitution of lamotrigine for patients with epilepsy.  Neurology. 2008;70(22, pt 2):2179-2186.PubMedGoogle ScholarCrossref
20.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194.PubMedGoogle ScholarCrossref
21.
Kwan  P, Arzimanoglou  A, Berg  AT,  et al.  Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies  [published correction appears in Epilepsia. 2010;51(9):1922].  Epilepsia. 2010;51(6):1069-1077.PubMedGoogle ScholarCrossref
22.
Naranjo  CA, Busto  U, Sellers  EM,  et al.  A method for estimating the probability of adverse drug reactions.  Clin Pharmacol Ther. 1981;30(2):239-245.PubMedGoogle ScholarCrossref
23.
Diaz  FJ, Berg  MJ, Krebill  R,  et al.  Random-effects linear modeling and sample size tables for two special crossover designs of average bioequivalence studies: the four-period, two-sequence, two-formulation and six-period, three-sequence, three-formulation designs.  Clin Pharmacokinet. 2013;52(12):1033-1043.PubMedGoogle ScholarCrossref
24.
Schall  R, Endrenyi  L, Ring  A.  Residuals and outliers in replicate design crossover studies.  J Biopharm Stat. 2010;20(4):835-849.PubMedGoogle ScholarCrossref
25.
Tothfalusi  L, Endrenyi  L, Arieta  AG.  Evaluation of bioequivalence for highly variable drugs with scaled average bioequivalence.  Clin Pharmacokinet. 2009;48(11):725-743.PubMedGoogle ScholarCrossref
26.
Center for Drug Evaluation and Research. Briefing information for the FDA meeting of the Advisory Committee for Pharmaceutical Science and Clinical Pharmacology. Topic 1: bioequivalence (BE) and quality standards for narrow therapeutics index (NTI) drug products. Meeting of the Advisory Committee for Pharmaceutical Science and Clinical Pharmacology; July 26, 2011. https://www.fda.gov/ucm/groups/fdagov-public/@fdagov-afda-adcom/documents/document/ucm263465.pdf. Accessed August 15, 2016.
27.
Zhang  X, Zheng  N, Lionberger  RA, Yu  LX.  Innovative approaches for demonstration of bioequivalence: the US FDA perspective.  Ther Deliv. 2013;4(6):725-740.PubMedGoogle ScholarCrossref
28.
Tamargo  J, Le Heuzey  JY, Mabo  P.  Narrow therapeutic index drugs: a clinical pharmacological consideration to flecainide.  Eur J Clin Pharmacol. 2015;71(5):549-567.PubMedGoogle ScholarCrossref
×