Featured Report:
Generic Drugs (A-02) Full Text

Introduction and Background
Generic Drug Approval Process
Pre-1984
Drug Price Competition and Patent Term Restoration Act of 1984 (Hatch-Waxman Act)
Requirements for Establishing the Bioequivalence of Generic Products
Determination of Bioequivalence
Current FDA Guidance
Individual Bioequivalence
Does Average Bioequivalence Equal Therapeutic Equivalence?
Narrow Therapeutic Index Drugs
Antiepileptic Drugs
Antiarrhythmic Drugs
Warfarin
Cyclosporine
Summary and Conclusion
Recommendations (Adopted AMA Policy)
References
Appendix

Introduction

Resolution 222, introduced by the California Delegation and referred to the Board of Trustees at the 2001 Annual Meeting, asks:

That the American Medical Association (AMA) support legislation to increase the availability of generic drugs, and specifically support legislation to prohibit anti-competitive agreements that delay market entry for generic drugs.

Testimony provided at the Reference Committee focused on questions about the current regulatory framework governing the approval of generic drugs by the Food and Drug Administration (FDA), the safety and effectiveness of these drugs compared with brand-name innovator products, and potential harms to patients associated with generic substitution.  This report analyzes several issues relevant to the approval and clinical use of orally administered generic drugs since passage of the Drug Price Competition and Patent Term Restoration Act of 1984 (commonly referred to as the Hatch-Waxman Act). It is intended to serve as a primer for physician members of the AMA’s House of Delegates and as a resource document for any legislative or regulatory efforts deemed advisable by the AMA that may be related to the availability and/or interchangeability of multisource drug products. This report does not address patent term restoration or patent extension issues, or economic/legal strategies that have been employed by manufacturers to retain market exclusivity for innovator products.

Methods. Literature searches were conducted in the MEDLINE database for English-language articles published between 1992 and February 2002 using the search terms drugs, generic in combination with therapeutic equivalency, human, drug approval, health care costs,drug monitoring, and pharmacokinetics. Searches were also conducted for articles published between 1984 and February 2002 using the terms generic substitution or therapeutic equivalency, excluding the terms drug utilization, pharmacy and therapeutic committee, decision making, formularies, hospital, and  dose-response relationship . A total of 783 articles were retrieved for analysis. Articles on therapeutic alternates, patents or other legal issues, or involving generic drugs not approved for marketing in the United States were excluded, leaving 253 articles that were reviewed for information pertinent to this report. Additional references were culled from the bibliographies of these references.

Background

Current Status of Generic Drug Use and Costs.  The rationale for the development and clinical use of generic drugs is to reduce drug-related health care costs. Generic drugs accounted for approximately 42% of all prescriptions at the retail level in the year 2000, but consumed only 8% of the $141 billion spent on prescription drugs.¹ The average price of a generic prescription was $19.33 in 2000, while the average price of a prescription dispensed with a brand-name drug was $65.29.² A recent study conducted by the Managed Care Institute of Samford University and released by the Generic Pharmaceutical Association estimated that $1.16 billion could be saved for every percentage-point increase in the use of generic medications.³ Another report estimated that beneficiaries of a Medicare prescription drug program could save more than $350 per person annually in 2003 with the use of generic incentives, for total savings of $14 billion.⁴

The relative proportion of generic prescriptions began increasing in the early 1990s but has since stabilized.⁵  The proportion of total pharmaceutical sales represented by generic drugs has decreased, reflecting the stable percentage of generic prescriptions coupled with increases in total pharmaceutical sales attributed to the successful marketing of higher priced, brand-name (innovator) products.⁵

AMA Policy.  The AMA has extensive policy on issues related to generic drugs.⁶ These policies: (1)   support the ability of physicians to use either generic or brand-name drugs;  (2) encourage physicians to consider relative cost when making their decision; (3) recognize that only  “A”-rated generic drugs are suitable for substitution; and (4) suggest several steps that can be taken by physicians and pharmacists to avoid confusion among patients when generic substitution or switching among generic products occurs.

Various specialty societies also have taken formal positions on the issue of generic substitution; these statements or policies are summarized in the Appendix. Of note, every state has a mechanism in place by which physicians can prevent substitution.

Generic Drug Scandal and FDA Reaction. In 1989 federal investigators implicated several generic industry officials in the conduct of fraud, obstruction of justice, and noncompliance with various manufacturing procedures.⁷ The investigations also revealed that several FDA employees had accepted illegal gratuities or other compensation in exchange for information and assistance that gave certain firms an advantage in the approval process. Investigators also discovered that 10 or more generic companies had submitted fraudulent data related to bioequivalency, stability testing, and manufacturing protocols for some of their products. 

The FDA reacted to these findings by reorganizing its generic drug operations and conducting comprehensive inspections. FDA investigators reevaluated data from hundreds of generic drug applications. More than 2,550 samples of the top 30 prescribed generic drugs--or about 30%  of all generic drugs on the market--were collected and laboratory-tested, and the agency conducted intensive inspections of 36 of the largest generic drug firms and 12 contract laboratories.⁸ The agency determined that only 27 samples, or approximately 1% of those tested, did not comply with standards of potency, dissolution, content uniformity, product identification, moisture determination, or purity. 

The FDA also tested 429 samples representing at least three different batches of so-called narrow-therapeutic-range drugs that were currently marketed.⁸ These 24 drugs, made by 73 brand-name and generic drug manufacturers, were selected because of their potential for adverse reactions or therapeutic failure if they lacked bioequivalency. Only five of the samples (all aminophylline tablets) failed to meet United States Pharmacopoeia standards. None of the defects in the generic drugs were judged to pose a public health hazard.

Based on these results and the fact that brand-name products demonstrated similar failure rates, the agency recommended that doctors continue to consider prescribing generic drugs when appropriate in order to offer products at lower cost to consumers.  However, as a result of the negative publicity and fraud associated with these events, the feeling was reinforced among many physicians that generic drugs were inferior and potentially harmful.   Back to top

Generic Drug Approval Process

Pre-1984.  The process for generic drug approval has evolved along with changes in federal drug law and regulations. Before enactment of the Food, Drug, and Cosmetic Act (FDCA) in 1938, significant regulatory barriers to generic competition in the market did not exist.⁹ Manufacturers of such products (eg, codeine sulfate, phenobarbital) could formulate, manufacture, and sell their products without submitting bioequivalence or efficacy data to the FDA. The 1938 Act established a “new drug” category, requiring manufacturers to document the safety of a product to the FDA and established a 60-day delay before marketing could proceed, absent FDA objection.⁹ Until 1962, generic versions of post-1938 drugs were marketed based on a  “general recognition” of safety. Typically, this designation was based on a history of safe use of the innovator product. Such generic products were designated as “not new drugs.”⁹ 

Amendments to the FDCA in 1962 added requirements for “substantial evidence of both safety and efficacy, obtained in adequate and well-controlled studies,” and affirmative FDA approval of the New Drug Application (NDA); these criteria also applied to generic drugs. These amendments also contained a provision for retroactive evaluations of pre-1962 drugs that had been recognized as safe.  The Drug Efficacy Study Implementation (DESI) Review established expert panels to review data on all drugs marketed between 1938 and 1962 and to make recommendations on their efficacy. Drugs found to be ineffective were to be withdrawn by the FDA. The agency also was to notify potential generic manufacturers of requirements for developing generic versions of these approved drugs. The result was the Abbreviated New Drug Application (ANDA), for which approval was based on active ingredients and bioequivalence, rather than on safety and efficacy data (see below).

The original ANDA procedure did not apply to products that were chemically equivalent to drugs first marketed after 1962. As patents on post-1962 drugs began to expire in the 1970s, the FDA created a paper NDA policy that permitted generic versions of new drugs to be approved based on submission of safety and efficacy information obtained from the published medical literature rather than new clinical data.¹⁰ This approach did not establish a viable generic drug approval process and ultimately the Drug Price Competition and Patent Term Restoration Act was enacted in 1984, under which generic drugs are approved today.

Drug Price Competition and Patent Term Restoration Act of 1984 (Hatch-Waxman Act).  The dual purposes of the Hatch-Waxman Act were to encourage the development of new innovator drugs by extending patent rights and to establish procedures facilitating the approval of low-cost generic drugs.¹¹ These amendments to the FDCA codified in statute an abbreviated process (ANDA) for post-1962 drugs whereby a generic company could gain approval of its version of a drug without repeating the expensive and lengthy clinical trials used to establish safety and efficacy of the innovator drug. Under certain circumstances relating to patent challenges, the first generic version of a brand-name innovator medication receives a 180-day period of market exclusivity.

Products approved under an ANDA must be pharmaceutical equivalents (ie, have the same active ingredient(s), route of administration, dosage form, and strength) as the reference drug.  They must also be bioequivalent and the manufacturer must supply other basic technical information related to manufacturing of the product that is normally required of an NDA.¹² Generic drugs are pharmaceutical equivalents only with respect to their active ingredients. The binders, diluents, and excipients (filler) in the formulation, as well as the method of manufacture, may vary.

In contrast to FDA regulations for NDAs, which require submission of each study and a description and analysis of any other data or information relevant to an evaluation of the safety and effectiveness of the drug product, the regulations for ANDAs only require that the submission include information that shows that the drug product is bioequivalent to the reference listed drug upon which the applicant relies. The FDA’s policy of not requiring failed bioequivalence studies to be submitted in ANDAs along with the passing study may lead to an incomplete evaluation of the data. This policy applies to both innovator and generic firms submitting an ANDA. In a recent example involving cyclosporine, the revelation of a failed bioequivalence study resulted in a Class II recall of SangCya Oral Solution.^13,14  The company subsequently discontinued this product.¹⁵

The FDA considers drug products to be therapeutic equivalents if they are pharmaceutical equivalents and are bioequivalent.

The publication, Approved Drug Products with Therapeutic Equivalence Evaluations (the “Orange Book”), identifies drug products approved by the FDA on the basis of safety and effectiveness and includes therapeutic equivalence evaluations for approved multisource prescription drug products.¹⁶ Products on the list are identified by the names of the holder of approved applications (applicants) who may not necessarily be the manufacturer of the product. This publication satisfies the requirements of the Hatch-Waxman Act that the FDA make publicly available a list of approved drug products that is updated monthly.

For every multiple-source product, the Orange Book cites a letter code that indicates the FDA’s evaluation regarding the therapeutic equivalence of the product relative to the reference innovator or brand-name product. These drugs are placed in one of two categories as follows: “A”-rated products are considered to be therapeutically equivalent to other pharmaceutically equivalent products; “B”-rated products are considered not to be therapeutically equivalent to other pharmaceutically equivalent products. Class AB is a subset of “A” and includes DESI drug products and post-1962 drug products for which actual or potential bioequivalence problems have been resolved and that the FDA now considers to be therapeutically equivalent. Most new generic products are defined as having “potential” problems until data is submitted to establish their bioequivalence.

Requirements for Establishing the Bioequivalence of Generic Products. The primary substantive requirement for approval of an ANDA is that the manufacturer seeking approval to market a generic drug product must submit data demonstrating that the drug product is bioequivalent to the  innovator brand-name drug product.  A major premise underlying the 1984 law is that bioequivalent drug products are therapeutically equivalent and, therefore, interchangeable.

Bioequivalent drug products must display comparable bioavailability when studied under similar experimental conditions. Bioavailability refers to the “rate and extent to which the active ingredient or active moiety is absorbed from a drug product and becomes available at the site of action.” ^12,17 The bioequivalence requirement means that an ANDA must contain either results of human studies showing bioequivalence, or other “information” enabling the FDA to conclude that the ANDA product will be bioequivalent to its reference listed drug.¹⁷ For example, bioequivalence for certain highly soluble, rapidly dissolving oral dosage forms may be inferred by using an in vitro dissolution standard when such an in vitro test has been correlated with human in vivo bioavailability data. Alternate study methods, using clinical or pharmacodynamic endpoints, are used for drug products when plasma concentrations are not useful, such as with oral inhalers, nasal sprays, and topical products applied to the skin. This report does not address such products.

Determination of Bioequivalence. Originally, bioequivalence was based on a demonstration that simple mean bioavailability parameters differed by less than 20% from the brand-name product. In 1977 this was modified to include a “power” approach that tested the null hypothesis that the rate and extent of bioavailability of the generic product was similar to the innovator product, and the power of the study was sufficient to detect at least a 20% difference.¹⁸  In 1986, the FDA adopted the currently used average bioequivalence approach, which involves a comparison of means.

For immediate-release oral dosage forms, the standard average bioequivalence determination employs a single-dose crossover study, typically conducted in a limited number of healthy volunteers (usually 24 to 36 adults).¹⁶ For drugs with long half-lives, parallel design studies may be used. Both the rate and extent of absorption are evaluated.  The former includes the maximum plasma concentration (Cmax) and the time required to achieve this value (Tmax). The extent of absorption is measured by the area under the plasma concentration-time curve (AUC).

Results are analyzed according to whether the generic product (test), when substituted for the brand-name product (reference), is significantly less bioavailable, and alternatively, whether the brand-name product, when substituted for a generic product, is significantly less bioavailable (the two 1-sided tests).¹⁶ The core of the bioequivalence concept is an “absence of a significant difference.” A difference of  20% is viewed by the FDA as significant.  By convention, all data are expressed as a ratio of the average response (AUC and Cmax) for test/reference, so the limit expressed in the second analysis is 125% (reciprocal of 80%).

Tests are carried out using an analysis of variance and calculating a 90% confidence interval (CI) for the average of each pharmacokinetic parameter, which must be entirely within the 80% to 125% boundaries.¹⁶ The width of the CI reflects, in part, the within-subject variability of the test and reference products.  A common misconception is that the average values between the reference and test product can vary by -20/+25%, which could lead to large differences between multisource products.  In fact, when applying these statistical criteria to studies involving 20 to 40 subjects, generic products whose mean arithmetic bioavailability parameters differ by more than 5% to 10% from the reference product begin failing the CI requirement.  The FDA’s Office of Generic Drugs has conducted two large surveys to quantify the differences between generic and brand-name products. The first, conducted on 224 bioequivalence studies submitted in approved applications during 1985 and 1986, found an average difference in AUC measures between reference and generic products of 3.5%.¹⁹ The second, involving 127 bioequivalence studies submitted in 1997 found average differences of 3.47% for AUC and 4.29% for Cmax.²⁰

Current FDA Guidance. After careful consideration of all the recommendations from an expert panel as well as public comments, the FDA in 2000 issued a final guidance for industry entitled Bioavailability and Bioequivalence Studies for Orally Administered Drug Products – General Considerations.²¹  In this guidance, the agency recommends non-replicated bioequivalence study designs for most orally administered immediate-release drug products and replicated bioequivalence study designs for modified-release dosage forms. The guidance maintains the average bioequivalence criterion but allows the option for a sponsor to provide rationale a priori for using another criterion to declare bioequivalence, such as the individual bioequivalence criterion (see following section) for highly variable drug products. An additional concern with bioequivalence testing is that, with certain drugs, it is the peak effect and not AUC that is important for the therapeutic response or ADRs. This guidance  recommends use of the partial AUC as an early exposure measure instead of Tmax to address these concerns.

Individual Bioequivalence. Concerns have been raised about use of the average bioequivalence approach to assure interchangeability for multisource products. It has been suggested that this approach may not be adequate for all drugs and that modified procedures and additional data may be necessary.^22-24  Measures of average bioequivalence lack any measure of intrasubject variability, and no such information is provided to physicians in the package inserts.  An individual bioequivalence approach has been advocated as a more appropriate measure to ensure “interchangeability.”^25,26 Because of the within-subject focus, individual bioequivalence assessments are usually based on a replicate study design in which each subject receives both the test and reference products on at least two occasions. This approach requires a criterion and statistical analysis of within-individual variance for both test and reference products, and also estimates if two pharmaceutically equivalent products exhibit a subject-by-formulation interaction.  Presence of a subject-by-formulation interaction means that the difference between formulations is not the same from subject to subject.

There is a paucity of data to verify that subject-by-formulation interactions are clinically important.  However, FDA post-hoc analysis of data sets from replicate design studies in NDAs and ANDAs (involving predominantly healthy male subjects) found some evidence of subject-by-formulation interactions involving Cmax and AUC.²⁷  Additionally, two studies of generic verapamil have been published showing subject-by-formulation interactions in elderly subjects and in the presence of food.^28,29  In both cases the generic product was better absorbed than the reference product. 

There is disagreement as to whether these data are adequate to validate individual bioequivalence as a necessary approach for the approval of generic drug products, and whether the additional expense is justified. However, one consequence of the individual bioequivalence discussion has been more focus on the appropriate subjects for bioequivalence trials.  The 2000 FDA guidance on bioavailability and bioequivalence recommends inclusion of women, and, for products intended for use primarily in geriatric patients, inclusion of elderly subjects as well. If this guidance is followed, it may allow for a better understanding of the general applicability of bioequivalence results and the potential value of individual bioequivalence because studies involving subjects more representative of the general population would likely use a replicate study design. Additionally, bioequivalence testing has not been well-validated in pediatric populations and this deficiency should be addressed. Back to top 

Does Average Bioequivalence Equal Therapeutic Equivalence?

The critical question is whether assessment of bioequivalence assures therapeutic equivalence. Pharmacokinetic bioequivalence studies are a surrogate for clinical outcomes. Clinical studies comparing pioneer and generic drugs are rarely performed, and studies comparing one generic product with another are almost never performed. However, in the 1970s it was recognized that differences in the formulation of products containing the same amount of active ingredient could result in significant differences in bioavailability, and several cases of therapeutic inequivalence involving generic products were reported.^30,31  These involved products that were never judged by the FDA to be bioequivalent, leading to development of the clinical and statistical approaches noted above to establish bioequivalence among pharmaceutically equivalent products.  Similarly, several more recent reports involving clinical differences or serious bioequivalence problems with generic products have involved “B”-rated products.^32-40 These types of reports fueled perceptions that generic products were not equivalent when in fact these products were never rated as equivalent.

However, numerous case reports have also noted problems temporally related to generic switches for a number of  “A”-rated products.^41-52 In response to such reports of possible therapeutic inequivalence, the FDA established the Therapeutic Inequivalence Action Coordinating Committee (TIACC) housed within the FDA Center for Drug Evaluation and Research. This Committee was to identify, evaluate, and when appropriate, investigate reports of apparent therapeutic inequivalence and take appropriate corrective action.

Since the formation of the TIACC in 1988, the FDA has investigated more than 60 reports of potential generic product inequivalence. The agency has been unable to document a single example of  therapeutic failure when an FDA-designated therapeutically equivalent generic product, which was manufactured to meet its approved specifications, was substituted for the corresponding brand-name drug listed in the FDA’s Approved Drug Products with Therapeutic Equivalence Evaluations.⁵³ However, when reports reveal other problems (eg, regarding quality) that are substantiated through investigation, appropriate actions are taken; these may include recommendations for product recall, withdrawal, or reclassification of its therapeutic equivalence code.

Even though other independent studies have confirmed the bioequivalence and/or therapeutic equivalence of many other “A”-rated generic products,^54-71 the perception persists that the current bioequivalence approach for approving generic products does not adequately account for intraindividual variation in drug disposition.  Some investigators believe that the extrapolation of  results from single dose studies to steady-state conditions can be problematic, particularly when active metabolites are involved. Also, because bioequivalence tests have been typically carried out in young, healthy male volunteers, some clinicians doubt that this approach assures bioequivalence in the actual target population in which drugs are administered, such as the elderly; individuals with gastrointestinal tract, kidney, or liver disease; and in those taking other medications.

However, few physicians appreciate that a similar situation exists for the majority of innovator new drug products on the market. That is, the  formulation that is finally approved is often not the one used in clinical efficacy studies that supported the NDA.  For example, a survey of new molecular entities approved from 1981 to 1990, found that nearly 60% of the final marketed formulations were different from those used in the clinical trial.²²  Bioequivalence standards, although primarily used for the evaluation of generic products, are also employed to evaluate innovator products when they are reformulated or when other significant manufacturing changes are made. Therefore, imposing tighter controls has implications for both the generic and innovator drug industries.

Narrow Therapeutic Index Drugs.  The process of generic substitution or switching among multisource products requires an act of faith by the prescriber or pharmacist that each product will be therapeutically equivalent.  As mentioned, certain segments of the medical community have emphasized their concern about the therapeutic equivalence of generic drugs, especially for drugs with a narrow therapeutic index (NTI). Currently, the NTI designation is not formally recognized by the FDA, although an internal working list was developed in the 1980s. Additionally, there are state-to-state variations in how these drugs are defined. Previously,  drugs characterized by a narrow therapeutic ratio were defined as follows:⁷²

• There is less than a twofold difference in median lethal does (LD50) and the median effective dose (ED50) values; or
• There is less than a twofold difference in the minimum toxic concentrations and minimum effective concentrations in the blood; and
• Safe and effective use of the drug products requires careful titration and patient monitoring.

An updated definition was provided in the 2000 guidance, which defined narrow therapeutic range drug products as those “containing certain drug substances that are subject to therapeutic drug concentration or pharmacodynamic monitoring, and/or where product labeling indicates a narrow
therapeutic range designation.”²¹ Examples include digoxin, lithium, phenytoin, theophylline, and warfarin. Recent surveys and guidelines confirm that many physicians remain concerned about the potential therapeutic inequivalence of generic NTI products including antiepileptic drugs, antiarrhythmics, warfarin, and cyclosporine.^50,73-77

Antiepileptic Drugs. The effects of therapeutic nonequivalence can be catastrophic in a previously well-controlled patient with epilepsy.  The presence of low water solubility, nonlinear pharmacokinetics, and narrow therapeutic ranges for certain antiepileptic drugs exacerbates theoretical concerns.  Expert panels have established practice guidelines opposing generic substitution except when medically necessary, especially for phenytoin and carbamazepine.⁷⁶

The occurrence of breakthrough seizures or toxicity has been attributed in case reports to generic primidone, carbamazepine, and valproic acid.^44-49 Recalls of generic phenytoin and carbamazepine in 1988 were prompted by such reports; these products had different dissolution profiles than the formulation used in bioequivalence studies, and yielded lower serum values.^78,79 Subsequently, several double-blind, crossover trials in patients with epilepsy (including children) found no bioinequivalence or therapeutic differences for carbamazepine between “A”-rated generics and the innovator brand-name product in either pharmacokinetic parameters or seizure frequency.^64-67  Switching to generic valproic acid in mentally retarded individuals with seizures also resulted in comparable blood levels and seizure control.⁸⁰ Of note, studies have also demonstrated clinical sequelae from changes in the formulation of brand-name compounds.⁸¹ Additionally, in the last 10 years, FDA Class II recalls of carbamazepine and phenytoin for failed dissolution or stability testing have more often involved innovator products.⁸²

For phenytoin, a randomized, double-blind crossover study showed a generic product to be associated with higher blood levels of phenytoin.⁸³ Controlled studies of different phenytoin lots found no significant subject-by-formulation interaction, although women had 30% lower AUCs than men.⁸⁴ The effects of food on extended phenytoin preparations may involve some subject-by formulation interactions.⁸⁵
 
For the most part, controlled studies based on average bioequivalence measures, as well as therapeutic measures, support the contention that “A”-rated generics for carbamazepine, phenytoin, and valproate are equivalent to their brand-name counterparts. However, many neurologists remain unconvinced by average bioequivalence studies, and hold to the view that interchangeability of products in previously stabilized individual patients can be problematic.⁸⁶

Antiarrhythmic Drugs.  Similar to the situation with antiepileptic drugs, many cardiologists believe that generic substitution of antiarrhythmic drugs poses an additional hazard to patients with arrhythmias, and is a risk factor for recurrence of arrhythmias and/or proarrhythmic events.⁵⁰ Lack of clinical efficacy data on generic products has been emphasized.⁸⁷ One widely quoted case report involving generic procainamide was subsequently retracted.^88,89   Another involving controlled-release quinidine involved non-“A”-rated products.³³  Several randomized trials in the 1980s supported the therapeutic equivalence of both immediate- and controlled-release generic formulations of procainamide and immediate-release quinidine, but found a lack of bioequivalence for sustained-release preparations of the latter.⁹⁰  Many of these products were discontinued, so the relevance of these studies to contemporary preparations is questionable. More recently, generic substitution of amiodarone has been associated with altered metabolite profiles, and at least 3 cases involving suspected therapeutic inequivalence for generic amiodarone have been forwarded to the FDA for review.^50,91
 
As with antiepileptic drugs, therapeutic failure with an antiarrhythmic drug can be catastrophic, and many physicians are reluctant to endorse product interchange for stabilized patients.

Warfarin.  Few would disagree that warfarin is an NTI drug requiring special skill to use appropriately. Effective clinical use must deal with the many vagaries of this drug, including nonlinear pharmacokinetics, intra- and interpatient variability, possible drug interactions, vitamin K and ethanol intake, effects of various comorbidities on coagulation profiles, patient age and compliance. Specialized anticoagulation therapy management services have been developed to improve patient care through optimal warfarin dosing, patient education, and follow-up monitoring.

Optimal patterns of clinical use have lagged because some physicians are reluctant to use the drug due to concerns, perceived or real, that the potential for major bleeding outweighs the risk of stroke prevention. Concern about the equivalence of generic products adds another element of uncertainty.

Three generic warfarin equivalents are currently on the market; the first was approved in 1997.  Some articles purporting to highlight the dangers of generic substitution have relied on studies done in the 1970s and 1980s with products never judged bioequivalent by the FDA.92-95 Three case reports suggesting therapeutic inequivalence of “A”-rated generic products have been published.^51,52

Studies submitted by one generic manufacturer demonstrated bioequivalence in single-dose studies in healthy volunteers.⁹⁶ Generic warfarin also was compared with Coumadin in a clinical setting in which clotting time was measured in a blinded, replicate-design, crossover study.⁶⁸ Results showed average therapeutic equivalence with smaller intrasubject variability for the generic product, although individual data were not supplied. Equivalence was also demonstrated in other randomized, crossover studies.^69,71 Recently, an observational study using a parallel cohort of 210 patients in a managed care organization showed that use of generic warfarin in patients previously stabilized on Coumadin did not change coagulation profiles more than continued use of Coumadin in another group of patients.⁷⁰ Confusion surrounding generic warfarin stems from the fact that the tablet itself is not a variable product but that the pharmacodynamic (receptor variation) and pharmacokinetic (metabolic enzyme variants) responses are  variable.

Cyclosporine.  The immunosuppressant cyclosporine is generally considered an NTI drug. The availability of generic products raised concern in the transplant community about the relevance to transplant recipients of studies done in healthy volunteers. The validity of standard FDA criteria to establish bioequivalence between cyclosporine formulations has also been challenged.^77,97,98 Recommendations have included establishing individual bioequivalence rather than average bioequivalence, establishing bioequivalence in transplant patients and in subgroups known to be poor absorbers, and requiring long-term safety and efficacy studies in transplant patients.

However, at present individual bioequivalence is a theoretical concept, the practical benefits of which have not been statistically proven.⁹⁹ The common practice of blood-concentration-guided dosing of cyclosporine has sufficiently compensated for interindividual and intraindividual variability in response, and previously allowed for the safe switching of cyclosporine formulations that were not bioequivalent.99  Published studies comparing the first generic cyclosporine oral solution formulation with the innovator product, evaluating individual bioequivalence, evaluating bioequivalence in transplant patients, and monitoring of long-term safety after switching appeared to confirm the validity of the standard average bioequivalence criteria for generic cyclosporine, even though this product (as described above) was later withdrawn from the US market.^99-103  Back to top

Summary and Conclusion

The approval of generic drug products was greatly facilitated by passage of the Drug Price Competition and Patent Term Restoration Act of 1984 (Hatch-Waxman Act).  The main assumption inherent in this Act is that average bioequivalence data obtained in healthy adults are an effective surrogate for safety and efficacy in the target patient population. 

A sequence of events over the last 30 years has tended to undermine physician confidence in generic drugs. Several case reports appeared in the 1970s and 1980s suggesting problems with the therapeutic equivalence of generic products.  Many of these products had never been judged to be bioequivalent by the FDA, and the criteria used to define bioequivalence were, at that time, not sufficient. At a time when the potential benefits of the Hatch-Waxman Act were being beginning to be realized, a series of criminal events involving both the generic industry and the FDA reinforced concerns about the quality of these products. The availability of generic versions of several NTI drugs has added another layer of concern and enhanced the scrutiny of generic products.

The criteria used by the FDA to ensure bioequivalence among multisource products are widely misunderstood. These criteria do not allow for -20% to +25% difference in bioavailability between products. Rather, these parameters represent the statistical universe in which measures of variance must reside. In practice, the mean differences in pharmacokinetic parameters for most orally administered generic products are closer to 3% or 4%. Additionally, the same criteria for bioequivalence are applied to brand name products when they undergo formulation changes, which often occurs prior to marketing. These re-formulated brand name products also are never tested in a clinical population.

While case reports continue to surface questioning the therapeutic equivalence of selected generic products, the FDA firmly believes that any differences that could exist are no greater than would be expected if one lot of the innovator’s product was substituted for another.  Nevertheless, some investigators believe that different approaches may be needed to ensure “switchability” among multisource products, such as testing NTI drugs under clinical conditions in the target population or narrowing the confidence interval allowed for average bioequivalence, or by applying individual bioequivalence criteria. Tighter acceptance criteria or narrower confidence intervals have been proposed for NTIs and are required by some drug regulatory agencies (eg, Canada).  However, the FDA believes that the present requirements to prove bioequivalence are rigorous enough to prevent the possibility that dosage forms meeting regulatory criteria could lead to therapeutic problems, even for NTI drugs.

Another approach would be to develop individual guidances that are drug-specific, in which acceptance limits would be based on potency, dose-response relationships, and the intrasubject variability of the drug product. Presently, no evidence exists that the latter would result in safer products, but it would certainly be more costly for product development.  However, use of an individual bioequivalence approach may be required for the clinician and patient to be confident with the interchange of different formulations, and the FDA is open to sponsors requesting this approach.

Theoretical assumptions of the possibility of inequivalence are not a sufficient basis for presuming its presence and acting on that assumption. Anecdotal reports are similarly unhelpful, since one is often unable to distinguish product failure from a natural change in disease process or patient response.  The FDA continues to seek evidence of unequal therapeutic effects between  “A”-rated generics and brand-name products.  The agency is reluctant to modify its procedures without scientific evidence. Nevertheless, little is known about the effects on bioequivalence of lot-to-lot variability during the manufacturing process.  All bioequivalence data for a given drug are based on a single point in time, as there is no requirement for repeated bioequivalence testing to ensure that production batches of subsequent lots of both brand-name and generic products remain bioequivalent. However, in the relative absence of data to verify that a problem exists with the interchange of multisource products, little is to be gained from this approach.

Steps taken recently by the FDA reveal a willingness to enhance what appears to be a sound approach to generic drug approval.  Physicians are encouraged to be alert for the possible occurrence of therapeutic inequivalence resulting from the substitution of multisource drug products in order to provide the scientific evidence that may be required to further enhance this process.  Back to top

RECOMMENDATIONS (Adopted AMA Policy)

The following statements, recommended by the Council on Scientific Affairs, were adopted by the AMA House of Delegates as AMA policy at the 2002 AMA Annual Meeting:

The AMA believes that:

1. Physicians should be free to use either the generic or brand name in prescribing drugs for their patients, and physicians should supplement medical judgments with cost considerations in making this choice. (Policy )
2. It should be recognized that generic drugs frequently can be less costly alternatives to brand-name products. (Policy )
3. Substitution with Food and Drug Administration (FDA) “B”-rated generic drug products (i.e., products with potential or known bioequivalence problems) should be prohibited by law, except when there is prior authorization from the prescribing physician. (Policy )
4. That AMA Policy H-115.974 be reaffirmed. This policy states, in part, that when a physician desires to prescribe a brand name drug, he or she do so by designating the brand name drug product and the phrase "Do Not Substitute" (or comparable phrase or designation, as required by state law or regulation) on the prescription. Every state has HOD Policy) CHECK
5. Physicians should report serious adverse events that may be related to generic substitution, including the name, dosage form, and the manufacturer, to the FDA’s MedWatch program. (Policy )
6. The FDA, in conjunction with the AMA and the United States Pharmacopoeia, should explore ways to more effectively inform physicians about the bioequivalence of generic drugs, including decisional criteria used to determine the bioequivalence of individual products. (Policy )
7. The FDA should fund or conduct additional research in order to identify the optimum methodology to determine bioequivalence, including the concept of individual bioequivalence,  between pharmaceutically equivalent drug products (i.e., products that contain the same active ingredient(s), are of the same dosage form, route of administration, and are identical in strength). (Policy )
8. That Congress should provide adequate resources to the FDA to continue to support an effective generic drug approval process. (Policy )

Back to top

CSA home page
CSA reports by topic

References

1. Generic price and market share. Available at: www.gphaonline.org/news. Accessed: February 6, 2002.
2.  IMS Health, Inc, National Prescription Audit Plus, April 2001.
3. The Managed Care Institute-Samford University. An economic prescription for America. September 2000. Available at: www.gphaonline.org/pdf/covington.pdf.  Accessed: February 8, 2002.
4. The Schneider Institute for Health Policy-Brandeis University. Greater use of generics. A prescription for drug cost savings. 2002. Available at: www.gphaonline.org/pdf/brandeis-text.pdf. Accessed: February 8, 2002.
5. Prescription Drug Trends. Henry J. Kaiser Family Foundation. Available at: www.kff.org/content/2001/3112/RxChartbook.pdf. Accessed: February 13, 2002.
6. Policy FinderAmerican Medical Association. Available at: www.ama-assn.org/apps/pf_online/pf_online.  Accessed: March 12, 2002.
7. Major events during the generic drug investigations. Am Pharm. 1990;30(7):38-39.
8. York J. FDA ensures equivalence of generic drugs. FDA Consumer. January 1995. Available at: www.fda.gov/fdac/special/newdrug/generic.html. Accessed: February 27, 2002.
9. The drugs/biologic approval process. In:  Pina KR, Pines, WL, eds  A Practical Guide to Food and Drug Law Regulation. Washington, DC; Food and Drug Law Institute; 1998.
10. Department of Health, Education, and Welfare. Food and Drug Administration. Abbreviated new drug applications for new drugs approved after October 10, 1962, for human use. Fed Register. (November 24) 1980;45:77807-77808.
11. Drug Price Competition and Patent Term Restoration Act of 1984. Public Law No. 98-417, 98 Stat. 1585(1984), 21 U.S.C. 335(j)(7)(B)(i).
12. Code of Federal Regulations.  Title 21, Section 320, Part 24. Types of evidence to establish bioavailability or bioequivalence.
13. Food and Drug Administration. Talk Paper. Nationwide recall of SangCya oral solution. Available at: www.fda.gov/bbs/topics/ANSWERS/ANS01025.html. Accessed: March 7, 2002.
14. Food and Drug Administration. Approval letter SangCya.  Available at: www.fda.gov/cder/foi/appletter/1998/64195ltr.pdf. Accessed: March 18, 2002.
15. New drug applications. SangStat Medical Corporation. Approval withdrawn. Fed Register. 2000;65:75717-75718.
16. Approved drug products with therapeutic equivalence evaluations. Food and Drug Administration, Center for Drug Evaluation and Research.  Available at: www.fda.gov/cder/ob/default.htm. Accessed: March 6, 2002.
17. Federal Food, Drug and Cosmetic Act, section 505(j)(8).
18. Bioequivalence requirements and in vivo bioavailability procedures. Fed Register. 1977:42:1624-1653.
19. Nightingale SL, Morrison JC. Generic drugs and the practicing physician. JAMA. 1987;258;1200.
20. Henney JE. From the Food and Drug Administration. JAMA. 1999;282:1995.
21. Food and Drug Administration. Guidance for Industry. Bioavailability and bioequivalence studies for orally administered drug productsgeneral considerations. October 2000. Available at: www.fda.gov/cder/guidance/3651fnl.htm. Accessed: February 27, 2002.
22. Benet LZ, Goyan J.  Bioequivalence and narrow therapeutic index drugs. Pharmacotherapy. 1995;15:433-440.
23. Schall R, Luus HG. On population and individual bioequivalence. Stat Med. 1993;12:1109-1124.
24. Blume HH, McGilveray IJ, Midha KK. BIO-international 94. Conference on bioavailability, bioequivalence and pharmacokinetic studies. Eur J Pharm Sci. 1995;3:113-124.
25. Williams RL, Patnaik RN, Chen ML, et al. The basis of individual bioequivalence. Eur J Drug Metab Pharmacokinet. 2000;25:13-17.
26. Williams RL, Adams W, Chen ML, et al. Where are we now and where do we go next in terms of the scientific basis for regulation. Eur J Drug Metab Pharmacokinet. 2000;25:7-12.
27. Food and Drug Administration Advisory Committee for Pharmaceutical Sciences Meeting. November 28-29, 2001.Available at: www.fda.gov/ohrms/dockets/ac/01/slides/3804s2.htm.
28. Carter BL, Noyes MA, Demmler RW. Differences in serum concentrations of and responses to generic verapamil in the elderly.  Pharmacotherapy. 1993;13:359-368.
29. Waldman SA, Morganroth J. Effects of food on the bioequivalence of different verapamil sustained-release formulations.  J Clin Pharmacol. 1995;35:163-169.
30. Gleiter CH, Gundert-Remy U. Bioinequivalence and drug toxicity. How great is the problem and what can be done?  Drug Saf. 1994;11:1-6.
31. Final task force report on bioavailability and bioequivalence of psychotropic drugs. Psychopharmacol Bull. 1980;16:9-13.
32. Dubovsky SL. Severe nortriptyline intoxication due to change from a generic to a trade preparation.  J Nerv Ment Dis. 1987;175:115-117.
33. Meyer MC, Straugh AB, Lieberman P, et al. Serious bioavailability problems with a generic prolonged-release quinidine gluconate product. J Clin Pharmacol. 1982;22:131-134.
34. Lund L. Clinical significance of generic inequivalence of three different pharmaceutical preparations of phenytoin. Eur J Clin Pharmacol. 1974;7:119-124.
35. Campagna FA, Cureton G, Mirigian RA, et al. Inactive prednisone tablets USP XVI. J Pharm Sci. 1963;52:605-606.
36. Lander H. L-DOPAvariation in response with different pharmaceutical preparations (Syndopa vs Larodopa). Med J Aust. 1971;2:984.
37. Bablenis E. Weber SS, Dufresne RL. Effects of use of generic thioridazine products on clinical outcome and quality of life. Clin Pharm. 1991;10:938-941.
38. Baker JR, Moessner H, Gonzalez U, et al. Clinical relevance of the substitution of different brands of sustained release theophylline.  J Allergy Clin Immunol. 1988;81:664-673.
39. Ansbacher R.  Conjugated estrogens: do not substitute.  Am Pharm. 1990;NS30:27-28.
40. Alvarez N, Hartford E, Cavalleri E. Low blood levels of phenobarbital due to poor gastrointestinal solubility of phenobarbital tablets. Ann Neurol. 1981;9:309-310.
41. Rosenbaum JF, Falk WE, Gastfriend DR, et al. Acute distress after switch from Norpramin to generic desipramine.  Am J Psychiatry. 1989;146:122.
42. Sajbel TA, Carter GW, Wiley RB. Converting patients from brand-name clozapine to generic clozapine.  Ann Pharmacother. 2001;35:281-284.
43. Kluznik JC, Walbek NH, Farnsworth MG, Melstrom K. Clinical effects of a randomized switch of patients from clozaril to generic clozapine.  J Clin Psychiatry. 2001;62(.Suppl.5):14-17.
44. MacDonald JT. Breakthrough seizure following substitution of Depakene capsules (Abbott) with a generic product.  Neurology. 1987;37:1885.
45. Sachdeo RC, Belendiuk G. Generic versus branded carbamazepine.  Lancet 1987;1:1432.
46. Koch G, Allen JP. Untoward effects of generic carbamazepine therapy.  Arch Neurol. 1987;44:578-579.
47. Welty TE, Pickering PR, Hale BC, Arazi R. Loss of seizure control associated with generic substitution of carbamazepine.  Ann Pharmacother. 1992;26:775-777.
48. Wyllie E, Pippenger CE, Rothner AD.  Increased seizure frequency with generic primidone.  JAMA 1987;258:1216-1217.
49. Gilman JT, Alvarez LA, Duchowny M. Carbamazepine toxicity resulting from generic substitution.  Neurology. 1993;43:2696-2697.
50. Reiffel JA, Kowey PR. Generic antiarrhythmics are not therapeutically equivalent for the treatment of tachyarrhythmias.  Am J Cardiol. 2000;85:1151-1153.
51. Wagner JL, Dent LA. Epistaxis associated with elevation of INR in a patient switched to generic warfarin. Pharmacotherapy 2000;20:240-243.
52. Hope KA, Havrda DE. Subtherapeutic INR values associated with a switch to generic warfarin. Ann Pharmacother. 2001;35:183-187.
53. Letter from Jane E. Henney, MD, Commissioner, Food and Drug Administration to E. Ratcliffe Anderson, Jr, Executive Vice President and Chief Executive Officer, American Medical Association. June 27, 2000.
54. De VJ, McGuire BW, De LR, Weiss AI. Bioequivalence of two leucovorin calcium tablet formulations.  DICP. 1989;23:153-154.
55. Diamond MP, Herbert CM, Maxson WS, Wentz AC. Comparison of two brands of clomiphene citrate for stimulation of follicular development in a program for in vitro fertilization.  Fertil Steril. 1986;45:522-525.
56. Eldon MA, Kinkel AW, Daniel JE, Latts JR. Bioavailability of propranolol hydrochloride tablet formulations: application of multiple dose crossover studies.  Biopharm Drug Dispos. 1989;10:69-76.
57. Francisco GE, Honigberg IL, Stewart JT, et al. In vitro and in vivo bioequivalence of commercial prednisone tablets.  Biopharm Drug Dispos. 1984;5:335-344.
58. Haroldson JA, Somerville KT, Carlson S, Hanson J, Emery RW, Lake KD. A retrospective assessment of safety, efficacy, and pharmacoeconomics of generic azathioprine in heart-transplant recipients.  J Heart Lung Transplant. 2001;20:372-374.
59. Midha KK, Hawes EM, Korchinski ED, et al. A bioequivalency study of two trifluoperazine tablet formulations using RIA and GC-MS.  Biopharm Drug Dispos. 1984;5:25-32.
60. Midha KK, Chakraborty BS, Schwede R, et al.  Comparative bioavailability of two tablet formulations of fluphenazine dihydrochloride in drug-free psychiatric patients.  J Pharm Sci. 1990;79:3-8.
61. Sharoky M, Perkal M, Tabatznik B, Cane RCJ, Costello K, Goodwin P.  Comparative efficacy and bioequivalence of a brand-name and a generic triamterene-hydrochlorothiazide combination product.  Clin Pharm. 1989;8:496-500.
62. Teresi ME, Riggs CE, Webster PM, Adams MJ , Noonan PK, O'Donnell JP.  Bioequivalence of two methotrexate formulations in psoriatic and cancer patients.  Ann Pharmacother. 1993;27:1434-1438.
63. Zaman R, Honigberg IL, Francisco GE, et al. Bioequivalency and dose proportionality of three tableted promethazine products.  Biopharm Drug Dispos. 1986;7:281-291.
64. Jumao-as A, Bella I, Craig B, Lowe J, Dasheiff RM. Comparison of steady-state blood levels of two carbamazepine formulations.  Epilepsia. 1989;30:67-70.
65. Oles KS, Gal P. Bioequivalency revisited: Epitol versus Tegretol.  Neurology. 1993;43:2435-2436.
66. Hartley R, Aleksandrowicz J, Bowmer CJ, Cawood A, Forsythe WI. Dissolution and relative bioavailability of two carbamazepine preparations for children with epilepsy.  J Pharm Pharmacol. 1991;43:117-119.
67. Meyer MC, Straughn AB, Mhatre RM, Shah VP, Williams RL, Lesko LJ. The relative bioavailability and in vivo-in vitro correlations for four marketed carbamazepine tablets.  Pharm Res. 1998;15:1787-1791.
68. Neutel JM, Smith DHG.  A randomized crossover study to compare the efficacy and tolerability of Barr warfarin sodium to the currently available Coumadin. Cardiovasc Rev Rep. 1998;19:49-89.
69. Handler J, Nguyen TT, Rush S, Pham NT. A blinded, randomized, crossover study comparing the efficacy and safety of generic warfarin sodium to Coumadin. Prev Cardiol. 1998;4:13-20.
70. Swenson CN, Fundak G.  Observational cohort study of switching warfarin sodium products in a managed care organization.  Am J Health Syst Pharm. 2000;57:452-455.
71. Plowman BK, Morreale AP, Mehdigholi E.  A comparative analysis of Coumadin and generic warfarin. Pharmacotherapy. 2000;20:346.
72. Code of Federal Regulations. Title 21, Section 320, Part 33. Criteria and evidence to assess actual or potential bioequivalence problems.
73. Banahan BF, Kolassa EM. A physician survey on generic drugs and substitution of critical dose medications.  Arch Intern Med. 1997;157:2080-2088.
74. Guberman A, Corman C. Generic substitution for brand name antiepileptic drugs: a survey.  Can J Neurol Sci. 2000;27:37-43.
75. Crawford P, Hall WW, Chappel B, et al. Generic prescribing for epilepsy. Is it safe? Seizure. 1996;5:1-5.
76. Therapeutic and Technology Assessment Subcommittee. American Academy of Neurology. Assessment: generic substitution for antiepileptic medication. Neurology. 1990;40:1641-1643.
77. Sabatini S, Ferguson RM, Helderman JH, Hull AR, Kirkpatrick BS, Barr WH. Drug substitution in transplantation: a National Kidney Foundation White Paper.  Am J Kidney Dis. 1999;33:389-397.
78. FDA recalls and court actions, extended sodium phenytoin capsules, USP, 100 mg Sidmark. FDC Reports. January 18, 1988: T& G 8.
79. FDA recalls and court actions, carbamazepine tablets. USP, 200 mg, Pharmaceutical Basics, Inc. FDC Reports. September 5, 1988; T&G 11.
80. Vadney VJ, Kraushaar KW: Effects of switching from Depakene to generic valproic acid on individuals with mental retardation.  Ment Retard. 1997;35:468-472.
81. Food and Drug Administration memo. October 1987
82. Food and Drug Administration Enforcement Report Index.  Available at www.fda.gov/opacom/Enforce.html. Accessed: March 14, 2002.
83. Meyer MC, Straughn AB, Mahtree RM, et al. Variability in the bioavailability of phenytoin capsules in males and females. Pharm Res. 2001;18:394-397.
84. Mikati M, Bassett N, Schachter S. Double-blind randomized study comparing brand-name and generic phenytoin monotherapy. Epilepsia. 1992;33:359-365.
85. Wilder BJ, Leppik I, Hietpas TJ, Cloyd C, Randinitis EJ, Cook J. Effect of food on absorption of Dilantin Kapseals and Mylan extended phenytoin sodium capsules. Neurology. 2001;57:582-589.
86. Nuwer MR. Generic substitutions for antiepileptic drugs-reply. Neurology. 1991;41:1165.
87. Jardine RM, Milne DA, Obel IW, et al. Generic substitution of anti-arrhythmic drugs.  S Afr Med J. 2001;91:785.
88. Grubb BP. Recurrence of ventricular tachycardia after conversion from proprietary to generic procainamide.  Am J Cardiol. 1989;63:1532-1533.
89. Grubb BP. Therapeutic equivalency of procainamide--a correction.  Am J Cardiol. 1991;67:334.
90. Nolan PE. Generic substitution of antiarrhythmic drugs. Am J Cardiol. 1989; 64;1371-1373.
91. Pollack PT. Altered metabolite concentrations with amiodarone generic substitution cannot be observed without monitoring.  Can J Cardiol. 2001;17:1159-1163.
92. Wittkowsky AK.  Generic warfarin: implications for patient care.  Pharmacotherapy. 1997; 17:640-643.
93. Wagner JG, Welling PG, Lee KP, et al. In vivo and in vitro availability of commercial warfarin
tablets. J Pharm Sci. 1971;60:666-677.
94. Ruedy J, Davies RO, Gagnon MA, et al.  Drug bioavailability. Can Med Assoc J. 1976;115:105.
95. Richton-Hewett S, Foster E, Apstein CS. Medical and economic consequences of a blinded oral anticoagulant brand change at a municipal hospital. Arch Intern Med. 1988;148:806-808.
96. Haines ST. Reflections on generic warfarin. Am J Health-Syst Pharm. 1998;55:729-733.
97. Kahan BD.  Recommendations concerning the introduction of generic formulation of cyclosporin. Transplant Proc. 1999;31:1634.
98. Haug M, Wimberley SL. Problems with the automatic switching of generic cyclosporine oral solution for the innovator product.  Am J Health Syst Pharm. 2000;57:1349-1353.
99. Christians U, First MR, Benet LZ.  Recommendations for bioequivalence testing of cyclosporine generics revisited. Ther Drug Monitor. 2000;22:330-345.
100. First MR, Alloway R, Schroeder TJ.  Development of Sang-35: a cyclosporine formulation   bioequivalent to Neoral. Clin Transplant. 1998;12:318-324.
101. Schroeder TJ, Cho MJ, Pollack GM, et al.  Comparison of two cyclosporine formulations in healthy volunteers. Bioequivalence of the new Sang-35 formulation and Neoral. J Clin Pharmacol. 1998;38:807-814.
102. Alloway RR. Generic immunosuppressants used in solid organ transplantation. Transplant Proc. 1999;31(Suppl 3A):2S-5S.
103. Canafax DM, Irish DM, Berger MH, et al. An individual bioequivalence approach to compare the intrasubject variability of two ciclosporin formulations, SangCya and Neoral. Pharmacology. 1999;59:78-88.

Back to top

Back to top

CSAPH home page
Reports by topic