HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Correspondence to:

ARTICLES: B. J. Wilder, I. Leppik, T. J. Hietpas, J. C. Cloyd, E. J. Randinitis, and J. Cook Effect of food on absorption of Dilantin Kapseals and Mylan extended phenytoin sodium capsules Neurology 2001; 57: 582-589 [Abstract] [Full text] [PDF]

Correspondence: Submit a response to this article

Correspondence published:

Reply from editorialists to Davit et al.

RP Lesser, G Krauss   (10 December 2001)

Reply from authors to Davit et al.

BJ Wilder, "J C Cloyd, J Cook"   (10 December 2001)

Effect of food on absorption of Dilantin Kapseals and Mylan extended phenytoin sodium capsules

B Davit, "GJP Singh, DP Conner"   (10 December 2001)

Reply from editorialists to Davit et al. 10 December 2001
RP Lesser John Hopkins University School of Medicine Baltimore, MD, G Krauss Send Correspondence to journal: Re: Reply from editorialists to Davit et al. rl{at}jhmi.edu RP Lesser, et al.	We agree with many of the comments from the Food and Drug Administration (FDA), but still wonder whether these comments go to the core of the problem. The fundamental question is not whether Mylan’s product meets the FDA standard for average bioequivalence, but whether this standard is good enough [2]. Current FDA regulations permit 20-25% variability. Phenytoin is a narrow therapeutic index drug, which displays zero order kinetics at higher blood levels. As the FDA has said, “subject-by-formulation” interactions occur [3]. Individual patients may experience seizures or toxicity after a change in phenytoin formulation even though “average” ioequivalence has been shown. The question also isn’t whether Mylan phenytoin is a good drug. It is. However, blood levels tended to be lower when taking this formulation. Whether the dose was 100mg[1] or 300mg,[4] the 90% CI were generally below 1, so levels would likely be slightly lower on the Mylan preparation, at least in some. However, this could be beneficial. The level on 300mg Dilantin could be too high, and the level on 300mg Mylan phenytoin just right. We suggest the FDA continue moving from “average” bioequivalence standards for drugs such as phenytoin and implement its proposed “individual” bioequivalence standard[5], after appropriate modifications.[6] The test strategy is relatively simple for evaluating “individual” bioequivalence: measure individual variability in AUC and Cmax by crossing-over randomized subjects between test (T; i.e. the generic equivalent drug) and reference (R; i.e. the trade name drug)doses in T-R-T-R and R-T-R-T sequences. The FDA has noted that this standard would motivate both generic and brand name drug firms to produce less “variable” drugs. This would ensure safer “substitution” of generic products. It would be a marketing advantage for a manufacturer and a benefit for patients if a formulation of a drug showed more uniform bioavailability. Dilantin has been a difficult drug to manufacture.[6, 7] If Mylan extended-release phenytoin had less individual variability than Dilantin, many would use it both as initial and replacement therapy. We like to know what we’re getting. Finally, we find data based on reports to the FDA unconvincing. Everybody knows that many adverse events go unreported, although the reporting mechanisms are reasonably straightforward.[8] Similarly, although the Wilder study[1] was criticized[4] for having used a subject number insufficient to perform simulations, this cuts both ways. Does the FDA mean to advocate that generic manufacturers provide data on 1000 subjects prior to approval of every new drug? References 1. Wilder BJ, Leppik,I, Hietpas,TJ, Cloyd,JC, Randinitis,EJ, Cook,J. Effect of food on absorption of Dilantin Kapseals and Mylan extended Phenytoin sodium capsules. Neurology 2001;57:582-589. 2. Lesser RP, Krauss,G. Buy some today. Can generics be safely substituted for brand-name drugs? Neurology 2001;57:571-573. 3. Williams RL. Therapeutic Equivalence of Generic Drugs. Response to National Association of Boards of Pharmacy. http://www.fda.gov/cder/news/ntiletter.htm Accessed 6/04/01 4. Davit BM, Singh,GJP, Conner,DP. Comment: effect of food on absorption of Dilantin Kapseals and Mylan Extended Phenytoin Sodium Capsules. Neurology 2001;57. 5. Chen ML, Patnaik R, Hauck WW, Schuirmann DJ, Hyslop T, Williams R. An individual bioequivalence criterion: regulatory consideration. Statist. Med. 2000:19:2821-2842. 6. Chow S-C, Liu,J. Special Issue: Individual Bioequivalence. Statistics in Medicine 19[20], 2719-2897. 2001. 7. 171 Opposition to Defense Motions re Due Process, Vagueness, "Van Liew" and "Minarik" Problems http://www.usdoj.gov/usao/eousa/foia_reading_room/usam/title4/civ00171.htm Accessed 10/23/01. 8. NEWS 08/16/1993. http://www.fda.gov/bbs/topics/NEWS/NEW00430.html Accessed 10/23/01. 9. MedWatch Reporting Forms. http://www.fda.gov/medwatch/getforms.htm. Accessed 10/17/01. Conflict of Interest Statement Dr. Lesser has been on the Speaker's Bureaus and have given lectures supported by, or have been a consultant for the following companies: Abbott Laboratories, Novartis, Ortho-McNeil, Wallace Laboratories, Warner- Lambert/Parke-Davis, Wyeth, Medtronic. He also has received funding from and is entitled to sales royalty from Bio-logic Systems, Inc. which is developing products related to his research involving a computerized method for analyzing physiologic data. The terms of this arrangement with Bio-logic Systems have been reviewed and approved by the Johns Hopkins University in accordance with its conflict of interest policies. Dr. Lesser has been asked to sign a onfidentiality agreement with Bertek Pharmaceuticals, Inc., a division of Mylan. Drs. Lesser and Krauss are both Investigators for studies sponsored by Pfizer. Ronald P. Lesser, M.D. Gregory Krauss, M.D. Johns Hopkins Medical Institutions Baltimore, Maryland
Reply from authors to Davit et al. 10 December 2001
BJ Wilder University of Florida Gainesville, FL, "J C Cloyd, J Cook" Send Correspondence to journal: Re: Reply from authors to Davit et al. Epilepsy{at}worldnet.net BJ Wilder, et al.	Davit et al. have raised concerns regarding our recent publication comparing the effect of a high fat meal on absorption of Dilantin Kapselas and Mylan Extended Release Phenytoin Sodium Capsules. [1] One might interpret Davit et al.’s response to mean that we implied that the differences in absorption between the products would affect every patient who was switched from one to the other product. We stated in our paper that the absorption was the same for both products when taken on an empty stomach. Differences will only occur in those patients who take their medication with food and switch from Dilantin to Mylan phenytoin or from the Mylan product to Dilantin. In the former phenytoin levels will fall and break-thru seizures may occur. In the latter, blood levels will rise and toxicity may occur. We based our simulations on the discovery of a significant value (<0.05) 13% reduction in the bioavailability of the Mylan product compared to Dilantin when both are taken with a high fat meal. Such a small change in bioavailability would produce proportionally small changes in steady-state concentrations of drugs if the drug follows linear kinetics. However, bioavailability differences of 10 to 15% can produce substantial, clinically important, changes in steady-state concentrations of drugs, which follow Michaelis-Menten nonlinear pharmacokinetics. Phenytoin is one of the best examples. We consider the results of our simulation to be valid. Davit et al. asked how Vmax and Km were determined in our patient group, if doses were high enough for accurate estimation of Km and Vmax and if steady-state conditions existed. Km and Vmax were derived using a linearized version of the Michaelis-Menten equation proposed by Ludden et al. [5], a method now considered standard when determining phenytoin pharmacokinetic parameters. As described in our paper, the daily dose and serum concentration data were obtained in a prospective study from epilepsy patients taking phenytoin who required one or more doses to optimize treatment. Daily doses and serum concentration data from each patient were used to calculated Vmax and Km. Initial daily doses used in the calculations ranged from 120 to 600mg/day with corresponding serum levels of 6.0 to 16.7 ug/ml; the final daily dose was from 200 to 860 mg/day and the serum phenytoin levels from 9.6 to 37 ug/ml. Each patient remained on each dosage regimen for at least 11 days before blood samples were collected. That amount is sufficient to achieve steady-state conditions. It was also pointed out that simulations should include appropriate levels of between and within-subject variability and includes at least 1000 subjects. Our simulation did take into account between subject variability as we used individual Vmax and Km data obtained from our patient population. Ludden et al. [5] previously performed exactly such a simulation when they examined the effects of bioavailability on steady- state phenytoin concentration. Their results were consistent with ours and they suggested that a 10% reduction in at least 14% of individuals with plasma phenytoin concentrations <10ug/ml; a 10% increase in bioavailability would result in 61% of individuals with phenytoin concentrations of >20ug/ml. The results of the Ludden paper led to the eventual tightening of the USP specifications for phenytoin content from 7% to 5%. Concern was raised about the patient data used to extrapolate the results observed in our bioequivalence study performed in healthy volunteers. I would point out that Vmax, and Km values for our patient population (Vmax: 9.63mg/kg/d; Km: 7.32mg.L) are in agreement with those reported in previous studies: Vmax: 5.93; Km: 5.7 [2], Vmax: 6.0; Km: 5.4- 5.8 [3], and Vmax:7.22; Km: 4.44 [4]. The slightly higher Km values in our group would make our patient population less sensitive to formulation differences compared with those previously reported. Davit et al. state that there is evidence that the Mylan phenytoin does not differ when taken with food using extrapolation of unpublished data on an investigational formulation that was content proportional to their marketed product. [6] Our study did not involve extrapolation across dosage forms. It is impossible for us to comment on the unpublished Mylan data. Differences in manufacturing processes may account for the different food effects associated with the Mylan product and their other investigation formulation. The FDA suggests the absence of reported lack-of-effect demonstrate the absence of a problem. Hopefully, this is the case. There are a number of reasons why lack-of-effect cases have not been brought to the attention of the FDA. There is a tendency for under reporting unless they result in death. Patients may not report break-thru seizures for fear of losing their drivers license, employment, becoming ineligible for insurance and other personal reasons. I hope we have adequately answered the questions raised by Davit et al. As a physician who has treated thousands of epileptic patients over the past forty-two years, I am keenly aware of the non-linear or rate limited metabolism of phenytoin. Small changes in doses or bioavailability may result in profound changes in blood levels and clinical effects. Neurologist and certainly epileptologist are extremely careful when manipulating doses of phenytoin, hence our concern for patients subject to changes in phenytoin formulations and the changes in bioavailability caused by food effect. References: 1)Wilder BJ Leppik I, Hietpas TJ, Cloyd JC, Randinitis EJ, Cook J. Effect of food on absorption of Dilantin Kapscals and Mylan extended release phenytoin sodium capsules. Neurology 2001;57:582-589. 2)Godley PJ, Ludden TM, Clementi WA, Godley SE, Ramsey RR. Evaluation of a baysian regression-analysis computer program using non-steady-state phenytoin concentrations. Clin Pharm 1987;6:634-639. 3)Bauer LA, Blouin RA. Age and phenytoin kinetics in adult epileptics. Clin Pharmacol Ther 1982;31:301-304. 4)Vozeh S, Muir KT, Sheiner LB, Follath F. Predicting individual phenytoin dosage. J Pharmacokinet Biopharm 1981;9:131-146. 5)Ludden TM, Allerheiligen SRB, Browne TR, Koup Jr. Sensitivity analysis of the effect of bioavailability or dosage from content on mean steady state phenytoin concentrations. Ther Drug Mont 1991;13:120-125. 6)Mylan Pharmaceuticals, Inc. 2001 Annual Report. Conflict of Interest Statement I have no financial affiliation with Pfizer other than being on their Neurontin Speakers Bureau. I am also a member of additional five speakers bureaus for other pharmaceutical companies. BJ Wilder, MD
Effect of food on absorption of Dilantin Kapseals and Mylan extended phenytoin sodium capsules 10 December 2001
B Davit Center for Drug Evaluation and Research FDA Rockville, MD, "GJP Singh, DP Conner" Send Correspondence to journal: Re: Effect of food on absorption of Dilantin Kapseals and Mylan extended phenytoin sodium capsules DavitB{at}cder.fda.gov B Davit, et al.	Wilder et al. [1] conclude that, based on simulations, switching between Parke-Davis’ Dilantin Kapseals and Mylan’s Phenytoin Sodium Extended Capsules will alter steady-state phenytoin plasma concentrations causing either toxicity or lack of effect. We disagree with the assumptions underlying the simulations and with the authors’ conclusions. Mylan’s product approved in 1998, met Agency requirements for bioequivalence (90% confidence intervals of generic/reference AUC and Cmax geometric mean ratios between 0.8 and 1.25). For AUC, the 90% CI was 0.92 to 1.01 and the ratio of means was 0.97. [2] In 1999, Parke-Davis, who manufactures Dilantin Kapseals, submitted to the Agency a study showing that Mylan’s product and Dilantin Kapseals were bioequivalent under fed conditions. [2] The ratio of AUC means was 0.87. Wilder et al. [1] used Vmax and Km estimations from an unrelated study [3] to simulate effects on steady-state plasma concentrations. When the authors assumed either a 13% decrease or 15% increase in bioavailability, 46% and 84% of simulated values were outside the therapeutic range. The authors’ simulations used a mean AUC ratio from one study and 30 Km and Vmax values from another study, published as an abstract. Mean data does not capture the extent of pharmacokinetic variability, the major factor contributing to frequency of subtherapeutic or toxic concentrations. Meaningful simulations should include appropriate levels of between- and within-subject variability and a sufficient number (at least 1000) of subjects. The reader cannot evaluate how Km and Vmax were determined. Were doses high enough for accurate estimation, or were dose- durations sufficient to achieve steady state? The authors incorrectly conclude that small differences in phenytoin bioavailability following a 100-mg dose would greatly expand when higher doses are taken with food. However, in a fed bioequivalence study of 300- mg doses of Dilantin Kapseals and Mylan’s product (an investigational formulation proportional to their 100-mg product), the 90% CI was 0.90 to 0.98 and the ratio of means was 0.94 for AUC. Thus, at 300 mg, Mylan’s product was bioequivalent to Dilantin Kapseals under fed conditions. [4] Wilder et al. [1] predict up to a 46% therapeutic failure rate when switching from Dilantin Kapseals to Mylan’s product. Actually, from launch to May 2001, over 3,000,000 prescriptions of Mylan’s product were dispensed [5] and 63 lack-of-effect cases, of which 23 were likely due to other factors, reported to the Agency. [2] Clearly, the data submitted by Mylan and Parke-Davis as well as post-marketing reports show that Mylan’s product is equivalent to Dilantin Kapseals. References 1)Wilder BJ, Leppik L, Hietpas TJ et al. Effect of food on absorption of Dilantin Kapselas and Mylan extended pheynytoin sodium capsules. Neurology 2001;57;582-589. 2)Division of Information Disclosure Policy, Center for drug evaluation and research, FDA, Rockville, MD 20957. 3)Cloyd JC, Sawchuk, RJ, Leppik IE et al. The direct linear plot: use in estimating Michaelis-Menton parameters and individualizing phenytoin dosage regimens in epileptic patients. Epilepsy International Symposium 1978:110-111. Abstract. 4)Mylan Pharmaceuticals, Inc., 2001 Annual Report. 5)IMS Health National Prescription Audit Plus 7 Weekley.