1088 in vitro & in vivo evaluation of dosage forms - usp 36.pdf

720 ⟨1087⟩ Intrinsic Dissolution / General Information USP 36

data is typically indicative of a systematic experimentalproblem. Possible problems include physical degradation ofthe compact by cracking, delaminating, or disintegration.Downward (negative second derivative) curvature of the dis-solution profile is often indicative of a transformation of thesolid form of the compact at the surface or when saturationof the dissolution medium is inadvertently being ap-proached. This often occurs when a less thermodynamicallystable crystalline form converts to a more stable form. Ex-amples include conversion from an amorphous form to acrystalline form or from an anhydrous form to a hydrateform, or the formation of a salt or a salt converting to thecorresponding free acid or free base. If such curvature isobserved, the crystalline form of the compact may be as-sessed by removing it from the medium and examining itby powder X-ray diffraction or another similar technique todetermine if the exposed surface area is changing.

The constant surface area dissolution rate is reported inunits of mass sec–1, and the dissolution flux is reported inunits of mass cm–2 sec–1. The dissolution flux is calculated byFigure 3 dividing the dissolution rate by the surface area of the com-pact. Test conditions, typically a description of the appara-tus, rotation speed, temperature, buffer species andstrength, pH, and ionic strength should also be reportedwith the analyses.

⟨1088⟩ IN VITRO AND IN VIVOEVALUATION OF DOSAGE FORMS

PURPOSE

This chapter provides an overview of the methodology forcharacterizing the physicochemical properties of a drug sub-stance as well as its associated drug product and discussesthe relationship of these methods and properties to thepharmacokinetic and pharmacodynamic properties of the

Figure 4 drug product. Results of in vitro methods are linked withinformation from in vivo evaluations through an in vitro–invivo correlation (IVIVC).The dissolution rate depends on the rotation speed and

precise hydrodynamics that exist. Typically, the concentra-tion of the test specimen is measured as a function of time, SCOPEand the amount dissolved is then calculated. The samplinginterval will be determined by the speed of the dissolution The ultimate goal of these characterization studies is anprocess (see Rotating Disk). If samples are removed from the understanding of the relationship between the physico-dissolution medium, the cumulative amount dissolved at chemical and pharmacological properties of the drug sub-each time point should be corrected for losses due to stance to the pharmacokinetic properties and in vitro perfor-sampling. mance of the drug product. This chapter outlines the in

vitro and in vivo testing that goes into the development ofthe body of data that informs decision making relating toDATA ANALYSIS AND INTERPRETATION the formulation, manufacturing, and related regulatory ac-tivities necessary for the development, regulatory approval,The dissolution rate is determined by plotting the cumula- and marketing of any drug product. The chapter comple-tive amount of solute dissolved against time. Linear regres- ments the information in general chapters, Assessment ofsion analysis is performed on data points in the initial linear Drug Product Performance—Bioavailability, Bioequivalence, andregion of the dissolution curve. The slope corresponds to Dissolution ⟨1090⟩ and The Dissolution Procedure: Develop-the dissolution rate (mass sec–1). (More precise estimates of ment and Validation ⟨1092⟩ by detailing the essential in vitroslope can be obtained using a generalized linear model that and in vivo data elements underlying an understanding oftakes into account correlations among the measurements of bioequivalence and bioavailability. The chapter text recog-the cumulative amounts dissolved at the various sampling nizes that regulatory guidances and a wealth of text bookstimes.) are available to elaborate on the content provided, and it isThe amount versus time profiles may show curvature. not the purpose to provide an exhaustive disquisition on theWhen this occurs, only the initial linear portion of the profile subjects presented but rather to provide a guide and listingis used to determine the dissolution rate. Upward curvature of the issues of interest.(positive second derivative) of the concentration versus time

Official from December 1, 2013Copyright (c) 2014 The United States Pharmacopeial Convention. All rights reserved.

Accessed from 112.205.134.201 by schering on Tue Jan 14 09:46:47 EST 2014

USP 36 General Information / ⟨1088⟩ In Vitro and In Vivo Evaluation 721

integration. Dissolution testing should be conducted onBACKGROUND INFORMATIONequipment that conforms to the requirements in Dissolution⟨711⟩ and on which a performance verification test has beenEstablishing a meaningful relationship between dissolutionconducted when one is available. On its website, USP pro-behavior and in vivo drug performance (i.e., IVIVC) has longvides a guidance for optimizing dissolution instrument per-been sought from the perspectives of both bioavailabilityformance by mechanical calibration and performance verifi-(BA) and bioequivalence (BE) and quality control considera-cation testing (http://www.usp.org/pdf/EN/tions. In setting dissolution acceptance criteria for a productdissolutionProcedureToolkit2010-03.pdf).monograph, USP’s policy has been to give predominant

In vitro dissolution testing generally should attempt toconsideration to valid BA or BE studies, when available.mimic in vivo dissolution, but such in vitro conditions can-The earliest achievable in vitro characteristic thought tonot be selected reliably a priori. A range of in vitro dissolu-predict an acceptable in vivo performance was tablet andtion test conditions (e.g., media of varying pH, surfactant,capsule disintegration. A test for disintegration was adoptedand apparatus rotational speed) should be evaluated.in USP XIV (1950). At that time, no quantitative work wasKnowledge of drug substance properties, product formula-done to attempt to demonstrate such a relationship, espe-tion, gastrointestinal physiology, in vitro dissolution, and incially with regard to in vivo product performance. Advancesvivo pharmacokinetics will aid in the selection of in vitroin instrumental methods and analytical precision ultimatelydissolution test conditions and specifications.opened up prospects for this work. The USP–NF Joint Panel

For products that contain more than a single active ingre-on Physiologic Availability recognized that the disintegrationdient, dissolution typically should be determined for eachtest was insufficiently sensitive and in 1968 directed theactive ingredient. When a dissolution test is added to anidentification of candidate articles for the first 12 official dis-existing monograph, the disintegration test is deleted, butsolution tests that used Apparatus 1.in the case of sublingual preparations and orally disintegrat-USP requires drug release testing via the USP performanceing tablets, disintegration may be a critical quality attributetest in the majority of monographs for non-solution oral,in addition to dissolution. In such cases one or both testssublingual, and transdermal dosage forms. In the currentcan be included in the monograph.state of science, in vivo testing is necessary during the de-

When a single set of specifications cannot be establishedvelopment and evaluation of both immediate-release andfor multisource products described in monographs, multiplemodified-release dosage forms. In some cases, dependingdissolution tests are allowed, and labeling is required to in-on the Biopharmaceutics Classification System (BCS) classifi-dicate the appropriate dissolution test for the specificcation of the drug, and depending on regulatory policy, inproduct.vivo testing may not be necessary. The special sensitivity of

Detailed information about method development and val-the dissolution test to changes in composition or method ofidation can be found in The Dissolution Procedure: Develop-manufacturing that do not result in significant changes inment and Validation ⟨1092⟩.performance in vivo is well recognized. An understanding of

the full complement of information given by in vitro and invivo evaluation of the drug substance and product is the IMMEDIATE-RELEASE DOSAGE FORMSstarting point in the development of a meaningful in vitroperformance test. For immediate-release dosage forms the in vitro dissolu-

tion process typically requires no more than 60 min, and inmost cases a single time-point specification is adequate forIN VITRO EVALUATIONPharmacopeial purposes. To allow for typical disintegrationtimes, test times of less than 30 min should be based ondemonstrated need.

Physicochemical Properties—Drug Substance

Physicochemical information typically includes polymor- EXTENDED-RELEASE DOSAGE FORMSphism, stability, particle size distribution, solubility, dissolu-tion rate, lipophilicity, permeability, and other release-con- For extended-release products in vivo dissolution generallytrolling variables of the drug substance under conditions is rate limiting, which results in protracted drug absorptionthat may mimic the extremes of the physiologic environ- and thus facilitates the identification of in vitro test condi-ment experienced by the dosage form. tions that may be predictive of in vivo dissolution. Multiple

sampling time points, therefore, are necessary to define adissolution profile for a modified-release dosage form.Physicochemical Properties—Drug Product The choice of apparatus should be based on knowledgeof the formulation and actual dosage form performance inThe variables tested to characterize the physicochemical the in vitro test system. Apparatus 1 (basket) or Apparatus 2properties of the drug product should be the same as those (paddle) may be more useful at higher rotation rates (e.g.,that are tested to characterize the drug substance. Dissolu- the paddle at 100 rpm). Apparatus 3 (reciprocating cylinder)tion profiles over a relevant pH range, usually from pH has been especially useful for bead-type modified-release1–6.8, should be obtained with particular attention to for- dosage forms. Apparatus 4 (flow cell) may offer advantagesmulation effects. Characterization of formulations that are for modified-release dosage forms that contain active ingre-insoluble in aqueous systems may require the addition of dients that have limited solubility. Apparatus 7 (reciprocatingsodium lauryl sulfate or another surfactant. The BCS classifi- disk) is applicable to nondisintegrating oral modified-releasecation of the drug substance should be determined, espe- dosage forms, as well as to transdermal dosage forms. Appa-cially for immediate-release dosage forms. ratus 5 (paddle over disk) and Apparatus 6 (cylinder) alsoare useful for evaluating and testing transdermal dosageforms.Dissolution Testing

At least three timepoints are chosen to characterize the invitro drug release profile of an extended-release dosageDissolution testing is required for all non-solution oral, in-form for Pharmacopeial purposes. Additional sampling timescluding sublingual, Pharmacopeial dosage forms in whichmay be required for drug approval purposes. An early timeabsorption of the drug is necessary for the product to exertpoint, usually 1–2 h, is chosen to show that dose dumpingthe desired therapeutic effect. Exceptions include tabletsis not probable. An intermediate time point is chosen tothat meet a requirement for completeness of solution, prod-define the in vitro release profile of the dosage form, and aucts that contain radiolabeled drugs, or products that con-

tain a soluble drug and demonstrate rapid (10–15 min) dis-



722 ⟨1088⟩ In Vitro and In Vivo Evaluation / General Information USP 36

final time point is chosen to show essentially complete re- this formulation serves as a reference to evaluate the inputlease of the drug. profile of the modified-release dosage form. This informa-

tion, together with the pharmacokinetics of the active drugentity, can characterize drug absorption and can predict

IN VIVO EVALUATION OF DOSAGE FORMS changes in drug BA when input is modified as in modified-release dosage forms. For example, if the active drug entity

In evaluating a drug product‘s performance, analysts fun- exhibits saturable first-pass hepatic metabolism, a reductiondamentally must ask what type of study should be per- in systemic availability could result after oral administrationformed to give reasonable assurance of BE of a marketed if the input rate is decreased.product to the clinical trial product that demonstrated safety In designing an oral modified-release dosage form, ana-and efficacy. Although they provide important information lysts may find it useful to determine the absorption of theconcerning the release characteristics of the drug from the active drug entity in various segments of the gastrointestinaldosage form, in vitro dissolution studies at present are used tract, particularly in the lower gastrointestinal tract (colon)primarily for setting or supporting specifications for drug for delayed-release dosage forms that release drug in thisproducts (e.g., shelf life) and manufacturing process control region. Food effects also may be important and should be(e.g., scale-up or postapproval changes). Normally BE is best investigated.demonstrated by in vivo evaluation but can sometimes bereplaced by in vitro studies.1 BE assessment of modified-re-

Drug Dispositionlease dosage forms is best achieved by observing in vivodrug pharmacokinetic and/or pharmacodynamic behavior

The information required to characterize drug dispositionby means of well-designed clinical studies. Multiple guid-may include the following.ances for the conduct of such studies are provided by regu-

1. Disposition parameters—clearance, area under thelatory agencies. Moreover, when a well-defined, predictivetime—plasma concentration curve (AUC), maximumrelationship exists between plasma concentrations of a drugplasma concentration (Cmax), time to maximumor its active metabolites and the clinical response (therapeu-plasma concentration (Tmax), volume of distribution,tic and adverse), it is possible to use plasma drug concentra-half-life, mean residence time, or model-dependenttion data alone as a basis for the approval of a modified-parameters.release dosage form that is designed to replace an immedi-

2. Linearity or characterization of nonlinearity over theate-release dosage form.dose or concentration range that could beAlthough human pharmacokinetic studies often are usedencountered.to assess BE of immediate-release solid oral dosage forms, in

3. Drug/metabolite accumulation.some cases in vitro studies can be used to assess BE. The4. Metabolic profile and excretory pathway, with specialprincipal advantage of in vitro studies is that they reduce

attention to the active metabolites and active enanti-development costs. For example, an in vitro test is prefera-omers of racemic mixtures.ble when one is testing BCS Class I drugs with rapid dissolu-

5. Enterohepatic circulation.tion. Some regulatory agencies permit this type of testing in6. Protein-binding parameters and effect of dialysis.lieu of in vivo testing.7. The effects of age, gender, race, and relevant diseaseThe following discussions are intended to provide guid-

states.ance for drug substance evaluation and the design, con-8. Plasma: blood ratios.duct, and evaluation of studies involving dosage forms. Al-9. A narrow therapeutic index or a clinical response thatthough these guidelines focus on oral drug delivery systems,

varies significantly as a function of the time of daythe principles may be applicable to other routes of drug(chronopharmacokinetics).administration (e.g., transdermal, subcutaneous, intramuscu-

lar, etc.).Pharmacodynamic Properties

CHARACTERIZATION OF DRUG SUBSTANCE Before developing a dosage form, analysts should obtainconcentration–response relationships over a dose range suffi-ciently wide to encompass important therapeutic and ad-

The Biopharmaceutics Classification System verse responses. In addition, the equilibration-time2 charac-teristics between plasma concentration and effect should be(BCS)evaluated. For modified-release products that typically havelarger drug doses in the dosage form, these concentra-FDA has issued a guidance titled “Waiver of In Vivotion–response relationships should be sufficiently character-Bioavailability and Bioequivalence Studies for Immediate-ized so that a reasonable prediction of the safety marginrelease Solid Oral Dosage Forms Based on a Biopharma-can be made if dose dumping should occur. If there is aceutics Classification System” (www.fda.gov/downloads/well-defined relationship between the plasma concentrationDrugs/GuidanceComplianceRegulatoryInformation/of the active drug substance or active metabolites and theGuidances/UCM070246.pdf). A key assumption in the ap-clinical response (therapeutic and adverse), the clinical per-proach is that drug release and dissolution is sufficientlyformance of a new modified-release dosage form could berapid so that an in vitro–in vivo correlation is not possiblecharacterized by plasma concentration–time data. If suchand/or useful. When applicable, the BCS allows dissolutiondata are not available, clinical trials of the modified-releaserate data in lieu of BA or BE studies for product approval.dosage form should be carried out with concurrentpharmacokinetic and pharmacodynamic measurements.

Pharmacokinetic Properties2 Equilibration time is a measure of the time-dependent discontinuity betweenmeasured plasma concentrations and measured effects. The discontinuity isAnalysts should thoroughly characterize the input absorp- more often characterized by the degree of hysteresis observed when the ef-

tion profile of the active drug entity from a formulation that fect-concentration plot for increasing concentrations is compared with thatfor decreasing concentrations. Where the equilibration time is very short (i.e.,shows rapid BA (an intravenous solution, oral solution, or arapid equilibration with no active metabolites generated), there will be littlewell-characterized immediate-release drug product). In turn, or no hysteresis. That is, the same effect will be observed for a given concen-tration independent of the interval between the time of dosing and the time121 CFR 320.22 Criteria for waiver of evidence of in vivo bioavailability or that measurements are made.bioequivalence.




be characterized on the basis of comparative in vitro dissolu-CHARACTERIZATION OF THE DOSAGE FORMtion data.

The pharmacokinetic studies described below are neededfor most modified-release dosage forms. These studies may

Pharmacokinetic Properties: Immediate- be the basis for characterization of the dosage form. If regu-Release Products latory approval is sought without conducting clinical trials,

manufacturers should consult with the regulatory authoritiesThe types of pharmacokinetic studies that should be con- to ensure that an adequate database exists for the approval.

ducted are based on how much is known about the active The types of pharmacokinetic studies generally conducteddrug substance, its clinical pharmacokinetics, and its BCS can be categorized as follows.Class. For example, a new chemical entity requires greaterpharmacokinetic characterization than does an FDA-ap-

CASE Aproved formulation that is undergoing scale-up and postap-proval changes (SUPAC) evaluation.

Case A applies to an original modified-release oral dosageThe latter is seen when an FDA-approved drug productform for a drug already marketed in an immediate-releaseundergoes changes in the manufacturing of the product af-dosage form and for which extensive pharmacokinetic/phar-ter the product has been approved. Such changes are com-macodynamic data exist.mon and can be caused by expansion in the size of the lots

manufactured, new manufacturing locations, or the intro- Single-dose crossover study: A single-dose crossoverduction of new technology. Necessary in vitro dissolution study should include the following treatments: the modified-tests and/or in vivo BE tests are described in the FDA “Guid- release dosage form administered under fasting conditions;ance for Industry: Immediate-release Solid Oral Dosage a dosage form that is rapidly available administered underForms: Scale-up and Postapproval Changes: Chemistry, fasting conditions; and the modified-release dosage formManufacturing, and Controls, In Vitro Dissolution Testing, administered immediately after a high-fat standardizedand In Vivo Bioequivalence Documentation” (www.fda.gov/ meal. The food effects study should control the ambient-downloads/Drugs/GuidanceComplianceRegulatoryInforma- temperature fluid intake (e.g., 6–8 oz.) at the time of drugtion/Guidances/UCM070636.pdf). administration. The dosage form should be administered

Similar requirements apply to a generic equivalent of an within 5 min after completion of the meal. Ideally all sub-approved immediate-release dosage form that must be BE jects should consume the meal in approximately 15 min. Ifto the innovator drug, known as the reference listed drug. there are no significant differences in the rate or extent ofThe two most frequently used methods for meeting bioe- bioavailability (AUC, Cmax, and Tmax) as a function of thequivalence requirements are in vivo pharmacokinetic studies meal, then additional food effect studies are not necessary.and BCS-based in vitro studies. If significant differences in bioavailability are found, research-

ers must define how food affects the modified-release dos-age form,3 as well as how the food–drug effect relates toPharmacokinetic Properties: Modified-Release time.

Products Use the following guidelines in evaluating food effects.1. If no well-controlled studies have previously defined

Like the approaches for immediate-release products, the the effects of a concurrent high-fat meal on an imme-types of pharmacokinetic studies that should be conducted diate-release dosage form, studies should be per-for modified-release products are based on how much is formed to determine whether a food effect is a resultknown about the drug substance, its pharmacokinetics, bi- of problems with the dosage form. Does the dosageopharmaceutics, and whether pharmacokinetic studies are form show food-related changes in release, or areintended to be the sole basis for product approval. At a there food effects that are unrelated to the dosageminimum, two studies are required to characterize the prod- form, e.g., changes in the drug’s absorption from theuct when no reference modified-release product exists: (1) a gastrointestinal tract or changes in the drug’s disposi-single-dose crossover study for each strength of a modified- tion that are independent of absorption? The causerelease dosage form and (2) a multiple-dose, steady-state of the food effect should be determined by a single-study using the highest strength of a modified-release dos- dose crossover study comparing the solution (or im-age form. A food effects study to evaluate the potential for mediate-release dosage form) under fed and fastingdose dumping from extended-release dosage forms also is conditions. If there is no food effect, then one con-required as a separate study or is included as an arm of a cludes that there are problems with the dosage form.crossover study. In the demonstration of interchangeability, If there is a food effect, then one concludes that thea single-dose, fasting crossover study vs. the reference prod- effect is unrelated to the dosage form.uct usually will suffice. In some cases, a food-effects study is 2. The influence of timing on the food effect should berequired if the reference product has demonstrated a food tested by a four-way crossover study, in which theeffect on BA. Some appropriate single-dose crossover and modified-release dosage form is administered undermultiple-dose steady-state studies are described below. the following treatment conditions: fasting, taken

For modified-release products, intravenous solutions, oral with a high-fat meal, 1 h before a high-fat meal, andsolutions, or well-characterized immediate-release drug 2 h after a high-fat meal.products are possible reference products to evaluate a modi- 3. If the food effect on an immediate-release dosagefied-release formulation. For example, if the active drug en- form is determined to result from changes in the dis-tity exhibits saturable first-pass hepatic metabolism from the solved drug’s absorption from the gastrointestinalsmall intestine, a reduction in systemic availability could re- tract or from changes in drug disposition, studiessult after oral administration if the input rate is decreased. should define the appropriate relationship betweenAn increase in systemic availability could be observed if a drug dosing and meals.drug is absorbed from the colon from a delayed-release dos- 4. Alternative appropriate studies can be conducted ifage form that targets the colon, thus avoiding a first-pass the applicant labels the drug for administration witheffect. a meal that is not fat loaded. In this case, an alterna-

In some modified-release capsule dosage forms, the tive meal composition should be considered.strengths differ from each other only in the amount of iden- 5. Analysts should monitor the entire single-dose, modi-tical beaded material contained in each capsule. In this case, fied-release absorption profile. Where appropriatesingle-dose and multiple-dose steady-state studies at the

3 Wagner–Nelson, Loo–Riegelman, and other deconvolution methods arehighest dosage strength are sufficient. Other strengths can found in textbooks on biopharmaceutics.




(e.g., in a multiple-dose study) for specific drugs and on the therapeutic use of the drug and the types of individ-drug delivery systems, blood samples should be taken uals for whom the modified-release dosage form will be rec-following breakfast on the second day, before the ommended. For drugs that have narrow therapeutic indices,second dose is administered. This sampling schedule it may be necessary to perform more extensive plasma con-is particularly important for once-a-day products. centration measurements to determine the potential for un-

6. For delayed-release (enteric-coated) dosage forms, usual drug-release patterns in certain subpopulations. Inanalysts should perform BA studies to characterize such studies, researchers should perform more than onefood effect and to support the dosing claims stated in AUC measurement per patient to assess variability with boththe labeling. the modified-release and the immediate-release dosage

The purpose of these studies is twofold: first, to determine forms.whether a need exists for labeling instructions describingspecial conditions for administration with respect to meals;

CASE Band second, to provide information concerning the patternof absorption of the modified-release dosage form com-

Case B applies to a non-oral, modified-release dosagepared to that of the immediate-release dosage form. Drugform of an already marketed active drug entity for whichinput function should be defined for modified-release dos-extensive pharmacokinetic and pharmacodynamic dataage forms. This will aid in the development of an appropri-exist.ate in vitro dissolution test. For dosage forms that exhibit

Case A studies (omitting the food effects studies) are ap-high variability, a replicate study design is recommended.propriate for the evaluation of a modified-release dosageMultiple-dose, steady-state studies form designed for a non-oral route of administration if the

Study I—When data demonstrating linear pharmacokinetics pattern of biotransformation to active metabolites is identi-exist for an immediate-release dosage form, a steady-state cal for the two routes. If the biotransformation patterns arestudy should be conducted with the modified-release dos- different, then clinical efficacy studies should be performedage form at one dose rate (preferably at the high end of the with the modified-release dosage form. In addition, specialusual dosage regimen) using a comparable total daily dose studies may be necessary to assess specific risk factors re-of an immediate-release dosage form as a control. At least lated to the dosage form (e.g., irritation and/or sensitizationthree trough plasma drug concentration (Cmin) determina- at the site of application of a transdermal drug delivery sys-tions at the same time of day should be made to demon- tem).strate that steady-state conditions have been achieved.Plasma drug concentration determinations, over at least onedosing interval of the modified-release dosage form, should CASE Cbe made in each phase of the crossover study. It may bepreferable (as in the case of rhythmic variation in absorption Case C applies to a generic equivalent of an approvedor disposition of the drug) to measure concentrations over modified-release dosage form, which should be BE to thean entire day in each phase. The presence or absence of reference drug in its rate and extent of drug exposure (i.e.,circadian variation should be verified. The modified-release AUC, Cmax, Cmin, and degree of fluctuation) in crossover sin-dosage form should produce an AUC that is equivalent to gle-dose studies. For an oral modified-release dosage form,that of the immediate-release dosage form if the extent of the food studies described under Case A also should beabsorption from the modified-release dosage form is compa- performed.rable to the immediate-release dose. The degree of fluctua-tion for the modified-release product should be the same as,

CASE Dor less than, that for the immediate-release dosage formgiven by the approved regimen. Appropriate concentration

Case D applies to an FDA-approved product that has un-measurements should include unchanged drug and majordergone SUPAC. Necessary in vitro dissolution tests and/oractive metabolites. For racemic drug entities, analysts shouldin vivo bioequivalence tests are described in the FDA guid-consider measurement of the active enantiomers.ance, SUPAC-MR: Modified Release Solid Oral Dosage Forms;Study II—When comparisons of the pharmacokinetic proper-Scale-Up and Postapproval Changes: Chemistry, Manufactur-ties of an immediate-release dosage form at different dosesing, and Controls, In Vitro Dissolution Testing, and In Vivoare not available, or when the data demonstrate non-Bioequivalence Documentation (www.fda.gov/downloads/linearity, steady-state crossover studies comparing effects ofDrugs/GuidanceComplianceRegulatoryInformation/the modified-release dosage form and those of the immedi-Guidances/UCM070640.pdf).ate-release dosage form should be conducted at two differ-

ent dose rates: one at the low end of the recommendeddosing range and the second at the high end of the dosing Statistical Analysis of In Vivo Bioequivalencerange. In each case, the modified-release dosage form mustmeet the criteria described in Study I with respect to AUC An appropriate statistical method should be selected. (Seeand fluctuations in plasma drug concentrations. If there are Assessment of Drug Product Performance—Bioavailability, Bioe-significant differences between the modified-release dosage quivalence, and Dissolution ⟨1090⟩).form and the immediate-release dosage form at either thelow or the high dosing rate, these data alone are not ade-quate to characterize the product. Data can be misleading IN VITRO–IN VIVO CORRELATIONSwhen obtained from subjects with atypical drug dispositionor physiologic characteristics relative to the target popula- The term IVIVC first appeared in the pharmaceutical litera-tion. Therefore, subject selection should be from an appro- ture as a result of the awareness of the importance ofpriate target population with randomized assignment to bioavailability concepts and in vitro dissolution rate determi-dosage form population. If the modified-release dosage nations. IVIVC refers to the establishment of a rational rela-form is for use in a specific subpopulation (e.g., for chil- tionship between a biological property, or a parameter de-dren), it should be tested in that population. Whether a rived from drug plasma concentrations produced by adrug exhibits linear or nonlinear pharmacokinetics, the basis dosage form, and a physicochemical property or characteris-for characterization is equivalence of AUC and of the relative tic of the same dosage form. The biological properties mostdegree of fluctuation of concentrations of the modified- commonly used are one or more pharmacokinetic parame-release and immediate-release dosage forms. ters such as Cmax or AUC, obtained following the administra-

Steady-state studies in selected patient populations or tion of the dosage form. The physicochemical propertydrug interaction studies may also be necessary, depending most commonly used is a dosage form’s in vitro dissolution




behavior (e.g., percent of drug released under a given set of tion and the in vivo input rate (absorption rate of the drugconditions). The quantitative relationship between the two from the dosage form). For a Level A correlation, a product’sproperties, biological and physicochemical, is an IVIVC. The in vitro dissolution curve is compared to its in vivo inputmost important use of an IVIVC is for predictability. In many curve, i.e., the curve produced by deconvolution of thecases the actual drug plasma concentration profile can be plasma profile. Deconvolution can be accomplished usingpredicted from in vitro dissolution data. mass balance model-dependent methods, such as the Wag-

Historically, IVIVC analysis has been more successful for ner–Nelson or Loo–Riegelman methods, or by model-inde-extended-release products than for immediate-release prod- pendent, mathematical deconvolution. In an ideal correla-ucts. This difference probably reflects the application of spe- tion, the in vitro dissolution and in vivo absorption ratecific data analysis techniques and interpretations that require curves are superimposable or can be made superimposed bydissolution rate–limited drug absorption. However some cor- the use of a constant offset value of the time scale. Therelations with immediate-release products have been dem- equations describing each curve are the same. This proce-onstrated using methods that rely on the current, broad dure often is found with modified-release dosage systemsavailability of computers and nonlinear regression software, that demonstrate an in vitro release rate that is essentiallyalong with new correlation methods. independent of the dissolution media and stirring speeds

used in a dissolution apparatus. Superimposition is not anabsolute requirement for a Level A correlation. If the dissolu-General Considerations tion and absorption curves are different and a mathematicalrelationship can be developed to relate the two, the plasmaWith the proliferation of modified-release products, it be- level profile still is predictable from the in vitro dissolutioncomes necessary to examine IVIVC in greater detail. Unlike data. This relationship must be true not only at that singleimmediate-release dosage forms, modified-release products, input rate but also over the entire quality control dissolutionparticularly extended-release dosage forms, cannot be char- range for the product. Furthermore, when the dissolutionacterized using a single time point dissolution test. These rate depends on mixing speed, the two curves can be madeproducts are designed to deliver drug so that a patient has to superimpose by either increasing or decreasing the ina specific plasma level profile over a prolonged period, usu- vitro mixing speed or some other alteration of the dissolu-ally 12–24 h. Analysts require an in vitro means of ensuring tion method.that each batch of the product will perform identically in The advantages of a Level A correlation are as follows.vivo. An IVIVC satisfies this requirement. Initially, it was 1. It develops a point-to-point correlation. This is notthought that developing a meaningful correlation for imme- found with any other correlation level. It is developeddiate-release dosage forms would be an easier task than for using every plasma level and dissolution point col-extended-release products. However, because of the nature lected at different time intervals, so it reflects theof the principles on which each type is based, analysts now complete plasma level curve. As a result, in the casebelieve that an IVIVC is more readily achieved for modified- of a Level A correlation an in vitro dissolution curverelease dosage forms. can serve as a surrogate for in vivo performance. AOne expects all extended-release products to be dissolu- change in manufacturing site, method of manufac-tion rate limited. For these products, the formulation signifi- ture, raw material supplies, minor formulation modifi-cantly contributes to the prolongation of drug release from cations, and even product strength using the samethe dosage form. Because of the impact of formulation on formulation can be justified without the need for ad-BA from an extended-release product, numerous attempts ditional BA-BE studies.4,5

have been made to correlate one or more pharmacokinetic 2. A truly meaningful quality control procedure that in-parameters determined from in vivo studies with the dicates in vivo performance and is predictive of aamount released in a given time during an in vitro dissolu- dosage form’s performance is defined for the dosagetion test. Single-point correlations can indicate that increas- form.ing or decreasing the in vitro dissolution rate of the modi- 3. The extremes of the in vitro quality control standardsfied-release dosage form would result in a corresponding can be justified either by convolution (simulating thedirectional change in the product’s performance. However, plasma level profile from the dissolution curve) or bysuch single-point correlations reveal little about the overall deconvolution (using the upper and lower confidenceplasma level curve, which is a major factor for drug perfor- interval limits).mance in the patient. Rather, correlation methods that util-ize all plasma drug concentration data and all in vitro disso-lution data are preferred. Three correlation procedures are LEVEL Bavailable that use all dissolution and plasma data, alongwith statistical moment analysis. Each procedure displays im- This correlation uses the principles of statistical momentportant differences in the quality of the correlation. These analysis. The mean in vitro dissolution time is compared tomethods are discussed in terms of the advantages of each either the mean residence time or the mean in vivo dissolu-along with its potential utility as a predictive tool for phar- tion time. As with a Level A correlation, Level B uses all ofmaceutical scientists. the in vitro and in vivo data but is not considered a point-

to-point correlation. It does not correlate the actual in vivoplasma profiles but rather a parameter that results from sta-Correlation Levelstistical moment analysis of a plasma profile component suchas mean residence time. Because a number of differentThree correlation levels have been defined and catego-plasma profiles can produce similar mean residence timerized in descending order of quality. The concept of correla-values, one cannot rely on a Level B correlation alone totion level is based on the ability of the correlation to reflectpredict a plasma profile from in vitro dissolution data. Inthe entire plasma drug concentration–time curve that resultsaddition, in vitro data from such a correlation cannot befrom administration of the given dosage form. The relation-

ship of the entire in vitro dissolution curve to the entire 4 FDA Guidance SUPAC-MR: Modified Release Solid Oral Dosage Forms—Scale-Up and Postapproval Changes: Chemistry, Manufacturing, and Controls;plasma concentration–time profile defines the strength ofIn Vitro Dissolution Testing, and In Vivo Bioequivalence Documentationthe correlation and, therefore, the predictability. (1997).5 FDA Guidance Extended-Release Solid Oral Dosage Form—Development,Evaluation, and Application of In Vitro/In Vivo Correlations, “If an IVIVC is

LEVEL A developed with the highest strength, waivers for changes made on the high-est strength and any lower strengths may be granted if these strengths arecompositionally proportional or qualitatively the same, the in vitro dissolutionThis level is the highest category of correlation. It repre- profiles of all the strengths are similar, and all strengths have the same release

sents a point-to-point relationship between in vitro dissolu- mechanism.”




used to justify values at the extremes of quality control stan- of the conditions indicated. The number of condi-dards. tions investigated depends largely on whether a cor-

relation can be developed with the in vitro resultsobtained under the more commonly investigated

LEVEL C conditions such as apparatus, agitation intensity, ordissolution medium and pH value. Each formulation

This category relates one dissolution time point (t50%, t90%, and every drug represents an individual challenge.etc.) to one pharmacokinetic parameter such as AUC, Cmax, The resulting dissolution profiles from the use of dif-or Tmax. It represents a single-point correlation and does not ferent dissolution media are illustrated in Figures 1reflect the complete shape of the plasma profile, which best and 2 in which the same formulations were tested indefines the performance of modified-release products. Be- water and an acid buffer.cause this type of correlation is not predictive of actual invivo product performance, generally it is useful only as aguide in formulation development or as a production qualitycontrol procedure. Because of its obvious limitations, a LevelC correlation has limited usefulness in predicting in vivodrug performance and is subject to the same caveats as aLevel B correlation in its ability to support product and sitechanges as well as justification of the extreme values inquality control standards. The FDA Guidance “Extended-Release Solid Oral Dosage Forms—Development, Evaluation,and Application of In Vitro/In Vivo Correlations”(www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM070239.pdf) statesthat manufacturers can obtain biowaivers based on multipleLevel C correlations. The guidance shows how manufacturerscan achieve this correlation. The FDA also indicates that ifsuch a correlation is achievable, it is likely that the develop-ment of a Level A correlation is also feasible.

Figure 1. Mean dissolution profiles of three modifications ofDEVELOPING A CORRELATION a new modified-release formulation (USP Apparatus 2, 50

rpm, 0.9 L water, 37°).This chapter does not define the only procedures for de-

veloping an IVIVC, and any well-designed and scientificallyvalid approach is acceptable. To assist the pharmaceuticalscientist, one possible procedure for developing a Level Acorrelation is described below:

1. In order to perform deconvolution properly, analystsshould be familiar with the pharmacokinetics of thedrug itself as well as when it is incorporated into amodified-release dosage form. For example, if a drugis known to be fully absorbed but demonstrates satu-rable first-pass kinetics, it is best to assume 100%bioavailability for purposes of absorption rate calcula-tion. This is based upon the fact that the drug is fullyabsorbed, but because of liver metabolism, one seesless than if the drug were administered as an immedi-ate-release bolus. If one utilizes the extent of absorp-tion relative to an immediate-release or solution dos-age form, the input profiles will not superimpose withthat calculated assuming 100% absorption. However,point-to-point correlations most likely will be possible.

Figure 2. Mean dissolution profiles of a new modified-2. Different dissolution profiles of a formulation shouldrelease formulation (USP Apparatus 2, 50 rpm, 0.9 L, pH 4.5be obtained as illustrated in Figure 1. The formulation

buffer, 37°).should be modified only sufficiently to produce differ-ent dissolution profiles so that the formulation hasthe same excipients in all the lots that will be tested. 3. The plasma level or urinary excretion data obtained inThe formulation modifications used in these batches the definitive bioavailability study of the modified-should be based on factors that would be expected release dosage form are treated by a deconvolutionto influence the product’s modified-release rate and procedure. The resulting data may represent the drugcould occur during normal product manufacture. In input rate of the dosage form. They also represent invitro drug release is performed on the batches that vivo dissolution when the rate-controlling step of thewill be used in the bioavailability study, and the effect dosage form is its dissolution rate (i.e., drug absorp-of varying the dissolution conditions is investigated. tion after dissolution is considered to be instantane-Some of the variables that should be studied are the ous). Any deconvolution procedure (e.g., mass bal-apparatus (it is preferable to use official dissolution ance or mathematical deconvolution) will produceequipment), mixing intensity, and dissolution me- acceptable results. Figure 3 illustrates the results ofdium (i.e., pH value, enzymes, surfactants, osmotic numerical deconvolution of the plasma profiles ob-pressure, ionic strength, etc.). The dissolution behav- tained for the batches in Figures 1 and 2.ior of the dosage form need not be studied under all




Figure 5. IVIVC attempt: pH 4.5 buffer.Figure 3. Mean absorption profiles from numerical decon-

volution of plasma concentration–time plots.5. If from the studies indicated in the in vitro dissolution

evaluation, given above, the modified-release dosage4. The in vitro dissolution curve is then compared to the form exhibits dissolution behavior that is independent

drug absorption rate curve. This can be performed by of the variables studied and a Level A correlation isvarious methods. Simply positioning one curve on demonstrated when the in vitro dissolution curve isthe other often can indicate the existence of a corre- compared to the drug input rate curve, then it islation. This may then be quantified by defining the likely that the correlation is general and can be extra-equation for each curve and comparing the corre- polated within a reasonable range for that formula-sponding constants. The simplest way to demonstrate tion of the active drug substance. If the dosage forma correlation is to plot the fraction absorbed in vivo exhibits dissolution behavior that varies with the invs. the fraction released in vitro, as illustrated in vitro conditions, analysts must determine which set ofFigures 4 and 5. With a Level A correlation, this rela- dissolution conditions best correlates with in vivo per-tionship is often linear with a slope approaching 1. As formance. One can then establish whether the corre-illustrated in Figures 4 and 5, a correlation may be lation is real or an artifact. This is achieved by prepar-curvilinear. The intercept may or may not be zero ing at least two formulations that have significantlydepending upon whether there is a lag time before different in vitro behavior. One should demonstrate athe system begins to release drug in vivo, or the ab- more rapid release and the other a slower releasesorption rate is not instantaneous, resulting in the than the clinical bioavailability lot (biobatch). A pilotpresence of some finite quantity of dissolved but un- BA-BE study should be performed with these formula-absorbed drug. In either case, it is a point-to-point or tions, and the previously established correlationa Level A correlation when the least-squares fit of the should be demonstrated for both. The formulationline approaches a coefficient of determination, R2, of modifications of these batches should be based upon1. For the correlations illustrated in Figures 4 and 5, formulation factors that would be expected to influ-the IVIVC using the acid buffer dissolution profiles ence the product’s modified-release mechanism, andwas superior to that obtained from water. modification of these formulation factors are ex-

pected to influence the dosage form’s release rate.6. Alternatively, the in vivo performance of the biobatch

formulation can be simulated based on the correla-tion developed with these formulations that wereused in the BA-BE study. Analysts then can comparethe predicted and experimentally determined values,the prediction error. The exercise illustrated in Figures6 and 7 serves as an internal validation of the Level Acorrelation. An external validation would involvesimulating data for a formulation batch that was notincluded in the Level A correlation calculations. Such avalidation was performed using the in vivo data fromthe medium lot of the formulation, and the resultsare illustrated in Figure 8.

Figure 4. IVIVC attempt: water (using slow and fastformulations).




Establishment of Dissolution SpecificationRanges

It is relatively easy to establish a multipoint dissolutionspecification for a modified-release dosage form. The disso-lution behavior of the biobatch can be used to define theamount that will be released at each time point. The diffi-culty arises in the variation that will be allowed around eachtime point. In the case of a Level A correlation, this can bedone in two ways, both of which use IVIVC: convolutionand deconvolution.

CONVOLUTION

Reasonable upper and lower dissolution values are se-lected for each time point established from the biobatch.Figure 6. Observed and predicted mean plasma profiles: Historically, dissolution specifications have been selected byslow formulation. using the average dissolution of the development batches,with a range of ±2.5–3 standard deviations. It is now ex-pected that the average dissolution values be approximatelythe same as those of the biobatch. The dissolution curvesdefined by the upper and lower extremes are convoluted toproject the anticipated plasma level curves that would resultfrom administration of these formulations to the same pa-tients to whom the biobatch was administered. If the result-ing plasma level data fall within the 95% confidence inter-vals obtained in the definitive BA-BE study, these ranges canbe considered acceptable. An alternative acceptance ap-proach that can be used after the therapeutic window for adrug has been defined, is to establish whether the upperand lower limits of the convolution results fall within thetherapeutic window, even if they fall outside the confidenceinterval. If they fall outside the window, a more limitedrange must be established. This procedure should be contin-ued until the predicted values meet the desired ranges.

DECONVOLUTIONFigure 7. Observed and predicted mean plasma profiles: fast An acceptable set of plasma-level data is established bothformulation. for a batch of material demonstrating a more rapid release

and for one demonstrating a slower release than that of thebiobatch. These can be selected by using the extremes ofthe 95% confidence intervals or ±1 standard deviation ofthe mean plasma level. These curves are then deconvoluted,and the resulting input rate curve is used to establish theupper and lower dissolution specifications at each timepoint. In the case of Level B and C correlations, batches ofproduct must be made at the proposed upper and lowerlimits of the dissolution range, and it must be demonstratedthat these batches are acceptable by a BA-BE study.

Immediate-Release Dosage Forms

GENERAL CONSIDERATIONS

Because the mechanisms for drug release from modified-release dosage forms are more complex and variable thanthose associated with immediate-release dosage forms, onewould anticipate that an IVIVC would be easier to developwith the latter formulations. Unfortunately, most of the cor-Figure 8. Observed and predicted mean plasma profiles:relation efforts to date with immediate-release dosage formsmedium formulation.have been based on the correlation Level C approach, al-though there also have been efforts employing statistical

7. Once a Level A correlation is established, in vitro test- moment theory (Level B). Although it is conceivable that theing can be used to establish dissolution specifications, same Level A correlation approach can be used with imme-biowaivers to facilitate SUPAC, and changes in dosage diate-release dosage forms, until data have been gatheredform strength for the same formulation. It is ques- to support this concept, Level B and Level C are the besttionable whether such an extrapolation with Level Band C correlations is possible.



USP 36 General Information / ⟨1090⟩ Assessment of Drug Product Performance 729

approaches that can be recommended with these dosage This chapter provides general information about the con-forms. duct of bioequivalence (BE) studies as a surrogate measure

of in vivo drug product performance and dissolution profilecomparisons as a measure of in vitro drug product perfor-mance. The chapter also discusses conditions when an invivo BE requirement may be waived (biowaiver) for certaindrug products and shows how the Biopharmaceutics Classi-fication System (BCS) can be used as a predictor of a drugproduct’s performance. An appendix to this chapter defines⟨1090⟩ ASSESSMENT OF DRUGkey scientific terminology and provides a comparison be-tween FDA and WHO in drug product performancePRODUCT PERFORMANCE—assessment.

BIOAVAILABILITY,BIOAVAILABILITY, BIOEQUIVALENCE, ANDBIOEQUIVALENCE, AND DISSOLUTION

DISSOLUTION Bioavailability (BA) studies focus on determining the pro-

cess and time frame by which a drug is released from theoral dosage form and moves to the site of action [see FDAGuidance Guidance for Industry—Bioavailability and Bioe-quivalence Studies for Orally Administered Drug Products—BACKGROUND General Considerations (2003)]. BA is an indirect or surrogatemeasure of the rate and extent to which the API or activeThis chapter provides recommendations for the in vivo moiety is absorbed from a drug product and becomes avail-and in vitro assessment of drug product performance. The able at its target sites of action. BA data provide an estimatechapter is intended as a guide to scientists and clinicians of systemic drug exposure, including fraction of drug ab-seeking to evaluate drug product performance by surrogate sorbed. For drug products that are not intended to be ab-procedures correlative and/or antecedent to clinical trials in sorbed into the bloodstream, availability may be assessed byhumans. USP–NF provides quality standards for drug sub- measurements that reflect the rate and extent to which thestances, excipients, and finished preparations. A USP–NF active ingredient or active moiety becomes available at themonograph for an official substance or preparation includes sites of action. Drug products are considered BE if a testthe article’s definition; packaging, storage, and other re- drug product does not show a significant difference in ratequirements; and a specification. The specification consists of and extent of absorption by comparison with a designateda series of universal tests (description, identification, impuri- reference drug when administered at the same molar doseties, and assay) and specific tests, one or more analytical of the same active moiety in the same dosage form underprocedures for each test, and acceptance criteria. Quality similar experimental conditions in either a single dose or instandards are important attributes that must be built into multiple doses.the drug product. Meeting USP–NF standards is accepted BA and BE generally can be obtained by serially measur-globally as assurance of high quality and is part of the re- ing drug and/or metabolite concentrations in the systemicquirements necessary for approval of a bioequivalent (BE), circulation over a prescribed period. BE studies can useinterchangeable multisource drug product. Multisource drug other approaches when systemic drug concentrations can-products must meet certain in vivo and/or in vitro perfor- not be measured or are not appropriate. For these cases,mance standards to be considered therapeutically equivalent more indirect approaches to BE determination include acuteand interchangeable. Regulatory approval for interchangea- pharmacodynamic endpoints, clinical endpoints, and in vitroble multisource products may differ somewhat in each studies that typically involve comparisons of the dissolutioncountry (see the forthcoming chapter Essentials for Drug profiles of test and reference drug products.Product Selection ⟨1096⟩ for further discussion). Drug prod- BA and BE information are important in regulatory sub-uct performance may be defined as the release of the active missions. BA information broadly addresses the absorption,pharmaceutical ingredient (API) from the drug product dos- distribution, metabolism, and excretion of the API. For anage form, leading to systemic availability of the API neces- innovator product, BE studies establish the performance ofsary for achieving a desired therapeutic response. This chap- the product intended for marketing by comparing theter discusses in vivo and in vitro approaches to determining bioavailability of the product as developed for marketing ap-drug product performance. The focus of the chapter is pri- proval to the clinical trial material, the drug product used inmarily on the performance of solid oral drug products. safety/efficacy trials. For the development and regulatory ap-The chapter references a Food and Drug Administration proval of a generic drug product, the test drug product(FDA) guidance, Guidance for Industry—Bioavailability and must be BE to the reference listed drug (RLD) product (usu-Bioequivalence Studies for Orally Administered Drug Products— ally the brand or innovator drug product that is designatedGeneral Considerations (2003) (http://www.fda.gov/; search by the applicable regulatory authority).by document title) and a World Health Organization (WHO) The ICH document titled Guidance on Q6A Specifications:document titled Annex 7 Multisource (Generic) Pharmaceutical Test Procedures and Acceptance Criteria for New Drug Sub-Products: Guidelines on Registration Requirements to Establish stances and New Drug Products: Chemical Substances (2000)Interchangeability (2006) (http://who.int/en/; search by doc- (http://www.fda.gov/; search by document title) providesument title). FDA guidances are used in the United States; approaches for setting acceptance criteria for drug productand WHO, FDA, and national/regional guidelines may be performance. This approach relies on dissolution or disinte-used by national/regional drug regulatory authorities. Fol- gration based on clinically acceptable batches, as doeslowing approval, control of the quality of a drug product FDA’s. BE studies focus on the performance of the drugcan be achieved in part by the private and/or public specifi- product and usually involve comparisons of two drug prod-cation, which can include a performance test. USP provides ucts: the test (T) and reference (R) or comparator product.the general chapters Disintegration ⟨701⟩, Dissolution ⟨711⟩, The required studies and the determination of BE are theDrug Release ⟨724⟩, In Vitro and In Vivo Evaluation of Dosage province of regulatory agencies. In the United States, R isForms ⟨1088⟩, and The Dissolution Procedure: Development termed the reference listed drug (RLD) and is so noted inand Validation ⟨1092⟩, which describe these tests and FDA’s Approved Drug Products with Therapeutic Equivalenceprocedures. Ratings [Orange Book (2008) (http://www.fda.gov/cder/ob)].To assist countries and regions where the R product may



1088 in vitro & in vivo evaluation of dosage forms - usp 36.pdf

Documents

dissolution flux

physicodissolution medium

amorphous form

acrystalline form

anhydrous form

stable form

associated drug product

drug substance