Generic Modified-Release Antiepileptic Drugs: No Difference Within Formulations, but Important Differences Across Formulations

Commentary

Assessing Bioequivalence of Generic Modified-Release Antiepileptic Drugs.

Johnson EL, Chang YT, Davit B, Gidal BE, Krauss GL. Neurology 2016;86(17):1597–1604. doi:10.1212/WNL.0000000000002607.

OBJECTIVES: The purpose of this study was to determine how closely generic modified-release antiepileptic drugs (MR-AEDs) resemble reference (brand) formulations by comparing peak concentrations (Cmax), total absorption (area under the curve [AUC]), time to Cmax (Tmax), intersubject variability, and food effects between generic and reference products. METHODS: We tabulated Cmax and AUC data from the bioequivalence (BE) studies used to support the approvals of generic Food and Drug Administration-approved MR-AEDs. We compared differences in 90% confidence intervals of the generic/reference AUC and Cmax geometric mean ratios, and intersubject variability, Tmax and delivery profiles and food effects. RESULTS:

Forty-two MR-AED formulations were studied in 3,175 healthy participants without epilepsy in 97 BE studies. BE ratios for AUC and Cmax were similar between most generic and reference products: AUC ratios varied by >15% in 11.4% of BE studies; Cmax varied by >15% in 25.8% of studies. Tmax was more variable, with >30% difference in 13 studies (usually delayed in the fed compared to fasting BE studies). Generic and reference MR products had similar intersubject variability. Immediate-release AEDs showed less intersubject variability in AUC than did MR-AEDs. CONCLUSIONS: Most generic and reference MR-AEDs have similar AUC and Cmax values. Ratios for some products, however, are near acceptance limits and Tmax values may vary. Food effects are common with MR-AED products. High variability in pharmacokinetic values for once-a-day MR-AEDs suggests their major advantage compared to immediate-release AED formulations may be the convenience of less frequent dosing to improve adherence.

The recent assessment of the bioequivalence of generic modified-release antiepileptic drugs compared with brand (reference) formulations was an ambitious undertaking, which required obtaining and analyzing data from the FDA Center for Drug Evaluation and Research, Office of Generic Drugs. While the challenge of getting this information is not emphasized in this report, it should be recognized as a near-miracle to be reading an analysis of FDA clinical trial data performed by independent, nonindustry, nongovernment investigators. This is especially remarkable considering that the modified-release medicine technology is considered “CCI” (confidential commercial information) and is a legally protected trade secret.

With the advent of the publicly accessible clinicaltrial.gov Web site in 2007, registration with the FDA was required for “applicable clinical trials,” which includes all controlled trials, and excludes phase 1 trials. A proposal to expand the completeness and transparency of the clinical trial data available to the public was put forth in November 21, 2014 in the Federal Register (1). The genesis of the proposal was fallout from a 2009 Cochrane review of oseltamivir phosphate (Tamiflu, Genentech Inc, a Member of the Roche Group, South San Francisco, CA), in which it was reported that only two out of ten efficacy studies were published, and therefore only these two papers, a fraction of the available data, could be used for evidence (2). In response to critiques about the validity of the conclusions resulting from publication bias, the Cochrane review authors sought and obtained the submerged portion of the iceberg of data from Roche and published an update in 2014 (3). In the second report, efficacy was still reported as modest, but the adverse event profile was more fully understood, specifically regarding renal and neuropsychiatric risks. Yet this unfolding of events raised an unhappy question around public health cost-effectiveness that has not yet been resolved: Did this fragmented availability of data enable the stockpiling of Tamiflu, with this stockpiling endorsed by the Centers for Disease Control and Prevention (4) and the World Health Organization (5), and enriching Roche by billions of dollars while costing taxpayers in kind?

Professor Tom Jefferson and epidemiologist Peter Doshi are the main activists in lobbying for greater FDA and EMA transparency (6), as well as authors of the Tamiflu Cochrane reviews. They were hopefully pleased when “Clinical Trials Registration and Results Information Submission” was published on September 21, 2016 in the Federal Register (7). This document provides the ‘final rules’ on the topic, after incorporation of public comments on the initial publication of November 2014, and will take effect January 18, 2017. Key elements of this lengthy document are firstly, the completeness of capture of clinical trials, with all acceptable clinic trials regardless of the approval status of the studied products to be registered. Completeness is a problem, as stated in a NEJM commentary (8), with 224,000 trials registered with ClinicalTrials.gov, but of these only 23,000 posting results. Secondly, the final rule provides more guidance on definitions of scientifically valid trials and on mandatory registration data elements. For unapproved products, trial results in general must be posted within 1 year of study completion. Expanded transparency is also part of the rule, with requirements to report baseline demographic information as well as specific details on outcome measures and adverse events (7). While much more rigorous summary data is called for, it is not clear that the course of individual subjects will be discernible even with the revisions.

To wrap up the intriguing Tamiflu story, in June 2016, authors affiliated with the WHO, AC Hurt and H Kelly, published a paper in which the benefits of stockpiling Tamiflu were re-examined. One of their concluding assertions was “data on outpatients with relatively mild disease should not form the basis for policies on the management of more severe disease” (9). In August 2016, a generic version of Tamiflu was approved.

Johnson et al. of the present study cited lack of access to the modified-release (MR) technology for each formulation as a limitation and acknowledged that this is confidential commercial information. The authors also stated that they did not have access to the individual case reports. These factors limited their ability to associate the MR technology with the bioequivalence variations, as well as their ability to analyze detailed, participant-level information. They also cited the limited generalizability of their analysis; the studies were performed in subjects without epilepsy as single-dose crossover studies. Therefore, important information such as pharmacokinetic and pharmacodynamic drug interactions and adverse effects in an at-risk population could not be addressed. However, they have strength in numbers, as data on 3,175 subjects were available, which highlights the potential of being able to mine the FDA clinical trial data at a “granular” level. It is clearly a challenge to obtain bioequivalence data at steady state: the recent EQUIGEN study comparing several generic formulations of lamotrigine enrolled 35 patients and had 33 completers (6). The Field test study of generic versus brand lamotrigine in subjects who were historically “generic-brittle” enrolled 34 patients (7). The valuable information obtained in these two painstaking studies was, simply stated, that no difference was found between generics or when comparing generic with brand regarding pharmacokinetics, adverse effects, or seizure occurrence.

The authors of the present report focused on extended-release and delayed-release formulations, which actually encompass only part of “modified-release” formulations. MR formulations include extended-release products, which strictly speaking means a dosage form that allows at least a twofold reduction in dosage frequency compared with its immediate-release (IR) form. Terms for this formulation include ER (extended release), SR (sustained release), XR (extended release), XL (extended length), or CD (controlled delivery) Delayed-release (DR) formulations were also likely included in this analysis, which is a technology that releases the drug over time rather than immediately after administration. Other MR forms that were not included are the oral-disintegrating tablets (ODT) and the targeted-release drug products, which release the drug at or near the physiologic site of action. In this report, similar area under the curve (AUC) and peak concentration (Cmax) values were found between generic and reference products, but there was a significant minority for which these parameters in generics varied by more than 15% from brand; 11% and 26% of all studies for AUC and for Cmax respectively. Further, time to Cmax (Tmax) was markedly delayed by being fed compared with fasting, with more than 30% difference found in one-third of studies for both generic and brand formulations.

The most interesting and clinically applicable finding was that of the intersubject variability found in IR versus MR products across all antiseizure medications. Lamotrigine had higher intersubject variability in AUC than either levetiracetam or carbamazepine for both the IR and the MR formulations. This leads to the clinical dilemma of understanding both the risks and the benefits of the long-acting MR formulations. As the authors point out, MR formulations are more convenient, and therefore adherence could improve. However, because of the profound variability between MR formulations (>20% variability in AUC for 35% of levetiracetam MR formulations, 65% of carbamazepine MR formulations and all of lamotrigine MR formulations), switching between MR generic formulations may actually pose some clinical risk. Perhaps clinicians should be provided specifically the pharmacokinetic intersubject variability from clinical trial data. While it is not clear that this level of transparency will be forthcoming in the new reporting system, the ready availability of more clinical trial data may help us understand and guide our patients' experiences.