In cases where a new drug does not have adequate safety and efficacy established for either IR or ER dosage forms, safety and efficacy trials are required for an ER product. An example of such a case is where an ER product is being developed as the first dosage form of a new drug without prior approval or study of an IR product. As noted below, PK and PK/PD approaches may alleviate the need to conduct all of the usual safety and efficacy studies (i.e., a complete clinical trial program with two clinical efficacy and safety trials) for an ER product when an IR product is already approved.
The general approaches for studying and evaluating ER products are described below:
Demonstration of Safety and Efficacy Primarily based on Clinical Trials
• In general, for drugs where the concentration-response relationships are not established or are unknown, applications for an ER product where an IR product already exists will require the demonstration of the safety and efficacy of the product in the target patient population. In these cases, the PK and biopharmaceutics studies conducted to address the CFR requirements (described in the previous section) while necessary are mostly supportive and are usually for descriptive and labeling purposes. These studies may also help in the initial-dose selection.
When a new molecular entity is developed as an ER formulation, additional studies to characterize its clinical pharmacology and ADME characteristics will be necessary.
Demonstration of Safety and Efficacy based on PK, PK/PD, and Supportive Clinical Trials
The FDA "Guidance for Industry—Providing Clinical Evidence of Effectiveness for Human Drug and Biologic Products"  indicates that in certain cases, the clinical efficacy of modified-release dosage forms or different dosage forms can be extrapolated from existing studies, without the need for additional well-controlled clinical trials. This may be possible because other types of data such as PK studies (BA/BE studies) and/or PK/ PD studies allow the application of known effectiveness to the new dosage form.
• "Where blood levels and exposure are not very different, it may be possible to conclude that a new form is effective on the basis of PK data alone."
• "Where blood levels are quite different, if there is a well-understood relationship between blood concentration and response, including an understanding of the time course of that relationship, it may be possible to conclude that the new dosage form is effective on the basis of pharmacokinetic data without an additional clinical efficacy trial."
The types of studies generally necessary in such cases will depend on the existence and nature of exposure-response relationships, and whether a therapeutic window has been established. The following cases provide some general ideas as to what studies and criteria may need to be met.
There is no prior knowledge of a concentration or exposure—response relationship or of a therapeutic window; approval is based solely on plasma profile comparisons and BE comparisons of PK parameters. Generally clinical trial(s) are necessary for approval in the case where there is no exposure-response relationship or a therapeutic window. An approach based solely on pharmacokinetic data with minimum or no information on PK/PD relationships is not generally encouraged. If it is agreed that the approval will be entirely based on PK data (e.g., based on prior knowledge of drug or its extensive use, or another appropriate reason agreed with FDA), bioequivalence between the IR and ER product is required in terms of Cmax, Cminj and AUC at steady state. The overall plasma profile over the ER product's dosage interval must also be quite similar to the IR product's profile over the same time period. Differences in shapes of the plasma profiles may affect the efficacy and safety profiles of the drug. In such cases, the differences in shapes may outweigh findings of BE based on Cmax, Cmin, and AUC. If deviations in the steady-state PK profiles are seen between the ER and IR product regimens, additional PK/PD information or clinical studies may be required.
In certain cases, it may also be important to assess differences in steady-state tmax between the ER and IR products for approval purposes. Additional BA studies as previously outlined would also be required.
There is no quantitative concentration or exposure-response relationship but a well-defined therapeutic window in terms of safety and efficacy exists.
1. Case where the rate of input is known not to influence the safety and efficacy profile: When a therapeutic window that is well accepted exists and rate of input does not affect the safety/ efficacy profile of the drug, the following criteria may be appropriate for comparing extended-release products to its reference (Note: there is no specific FDA Guidance that addresses this):
• For AUCss, the 90% confidence interval for the log-transformed ratio should be between 80-125
• The Cmax ss should be equal to or below the upper limit of the defined therapeutic window and the absolute Cmin ss should be equal to or above the lower limit of the defined therapeutic window.
Additional BA studies as previously outlined would also be necessary.
2. Case where it is unknown whether the rate of drug input influences the safety or efficacy profiles of the drug:
Criteria can be the same as subcase 1, but in addition, studies investigating the impact of the rate of input on the pharmacodynamics of the drug in terms of safety and efficacy should be conducted and shown to have no rate effect.
Additional BA studies as previously outlined would also be necessary.
There is a well-defined quantitative exposure-response relationship shown using different input rates or developed using the ER product.
1. If a concentration, or exposure-response relationship is established with the intended clinical endpoint and the safety profile of the drug is well understood, clinical safety and efficacy studies on the ER product may not generally be necessary. Acceptance criteria can be based on predictions of the clinical response from the steady-state plasma concentration time profile. Additional BA studies as previously outlined would also be required.
2. If a concentration, or exposure-response relationship is established with a validated surrogate measure, which is accepted as a validated marker for clinical efficacy, and the safety profile of the drug is well understood, clinical safety and efficacy studies may not generally be necessary. Acceptance criteria can be based on predictions of the clinical response from the plasma concentration profile. Additional BA studies as previously outlined would also be required.
GENERAL CONSIDERATIONS IN EVALUATING PK/PD RELATIONSHIPS.
In assessing PK/PD relationships, it is important to establish concentration-effect relationships and to determine the significance of differences in the shape of the steady-state concentration vs. time profile for an ER product regimen as compared to the approved IR product regimen. In this regard, any differential effects based on the rate of absorption and/or the fluctuation within a profile as related to safety and/or efficacy should be determined. Issues of tolerance to therapeutic effects and toxic effects related to drug exposure, concentration, absorption rate, and fluctuation should also be examined as part of the PK/PD assessment. In certain cases minimizing fluctuation in a steady-state profile for an ER product may be desirable to reduce toxicity while maintaining efficacy as compared to the IR product regimen (e.g., theophylline products). In other cases, minimizing fluctuation in a steady-state profile for an ER product may reduce efficacy (e.g., nitroglycerin—due to tolerance) as compared to the IR product regimen's profile where higher fluctuation is observed. It is therefore important to know the profile shape vs. PD relationships.
Studies to assess the safety of the ER dosage form are generally necessary. An example of dosage unit or dosage unit/drug safety problems could be bezoar formation from some ER formulations.
In vivo Studies Needed for Approval of ER ANDAs (Generics) 
• A single-dose nonreplicate design fasting study comparing the test and reference-listed drug product. Since single-dose studies are considered to be most sensitive in addressing the primary question of bioequivalence  i.e., release of the drug at the same rate and to the same extent, multiple-dose BE studies are no longer necessary. For extended-release products marketed in multiple strengths, a single-dose bioequivalence study under fasting conditions is required only on the highest strength if all the strengths are proportionally similar and all strengths are manufactured under the same conditions. Bioequivalence studies on the lower strengths may be waived based on in vitro dissolution profiles. If the strengths are not proportionally similar, a single-dose bioequivalence study is required for each strength. This requirement can, however, be waived in the presence of an acceptable in vitro/in vivo correlation .
• A fed state nonreplicate design bioequivalence study comparing the highest strength of the test and reference product .
In vitro Studies Needed (Dissolution)
Dissolution testing should be conducted on the ER product batches that were used in the pivotal BA/BE studies. The dissolution method should be appropriately selected after evaluation of several dissolution media (different pH) and agitation speeds, and should have adequate discriminatory power to differentiate between optimal and suboptimal batches. The sponsors are encouraged to develop dissolution methods that correlate with in vivo performance. If bio waivers for lower strengths are requested, adequate dissolution data needs to be submitted. Details of dissolution testing for ER products [2-4] can be found in the FDA "Guidance for Industry—Extended Release Oral Dosage Forms: Development, Evaluation, and Applications of in vitro/in vivo Correlations."
Refer to SUPAC-MR guidance, IVIVC (next section), and biowaivers chapter for details. In general, when manufacturing changes are made to an approved extended-release product, e.g., changes in composition, manufacturing site, batch size, equipment, process, etc., the requirements are defined under the FDA guidance "Scale-up and post approval changes for modified release dosage forms" . In cases when the SUPAC-MR Guidance recommends a biostudy to support the change, an adequate in vitro/in vivo correlation can be used as justification. These are clearly explained in the FDA guidance on IVIVC (Extended release oral dosage forms: Development, evaluation and applications of in vitro/in vivo correlations ).
IVIVC [IN VITRO (DISSOLUTION)/IN VIVO (BIOAVAILABILITY) CORRELATIONS] [4, 9, 12, 13]
Why are IVIVCs Important?
In vitro dissolution has been extensively used as a quality control tool for solid oral dosage forms. Many times, however, it is not known whether one can predict the in vivo performance of these products from in vitro dissolution data. In an effort to minimize unnecessary human testing, investigations of in vitro/in vivo correlations between in vitro dissolution and in vivo bioavailability are increasingly becoming an integral part of extended-release drug product development. This increased activity in developing IVIVCs indicates the value of IVIVCs to the pharmaceutical industry. Because of the scientific interest and the associated utility of IVIVC as a valuable tool, the U.S. Food and Drug Administration has published a Guidance in September 1997, titled Extended Release Oral Dosage Forms: Development, Evaluation and Applications of in vitro/in vivo Correlations. A predictive IVIVC enables in vitro dissolution to serve as a surrogate for in vivo bioequivalence testing. In vitro/in vivo correlations can be used in place of biostudies that may otherwise be required to demonstrate bioequivalence, when certain changes are made in formulation, equipment, manufacturing process, or the manufacturing site. In vitro/in vivo correlation development could lead to improved product quality (more meaningful dissolution specifications) and decreased regulatory burden (reduced biostudy requirements).
In order to successfully develop an IVIVC, dissolution or release from the formulation has to be the rate-limiting step in the sequence of steps leading to absorption of the drug into the systemic circulation. Further, to utilize this dissolution test as a surrogate for bioequivalence (where a relatively simple in vitro test is used in place of human testing), the IVIVC must be predictive of in vivo performance of the product.
Four categories of IVIVCs (levels A, B, C, and multiple level C) have been described in the FDA guidance. In addition, a qualitative rank order correlation (level D) has also been described in the U.S. Pharmacopoeia.
A level "A" correlation represents a point-to- point relationship between in vitro dissolution and the in vivo input rate (e.g., the in vivo dissolution of the drug from the dosage form). Level A correlation refers to a predictive mathematical model for the relationship between the entire in vitro dissolution/release time course and the entire in vivo response time course, e.g., fraction absorbed vs. fraction dissolved (see Fig. 1). Generally these correlations are linear; however, nonlinear correlations are also acceptable. A level "A" correlation is considered to be the most informative and very useful from a regulatory point of view.
A level "B" correlation uses the principles of statistical moment analysis . Level B correlation is a predictive mathematical model of the
0 20 40 60 80 100 120
0 20 40 60 80 100 120
FIGURE 1 Level "A" correlation.
relationship between summary parameters (Fig. 2) that characterize the in vitro and in vivo time courses, e.g., a. mean in vitro dissolution time versus mean in vivo dissolution time b. mean in vitro dissolution time versus mean residence time in vivo
Although this type of correlation uses all of the in vitro and in vivo data, it is not considered very useful since many different dissolution and plasma
MDT in vitro FIGURE 2 Level "B" correlation.
concentration profiles and shapes can give the same mean summary parameters. Since it does not uniquely reflect the actual in vivo plasma level curve, this is not very useful from a regulatory point of view.
A level "C" correlation establishes a single-point relationship between a dissolution parameter (e.g., time for 50% dissolved or % dissolved in six hours) and a pharmacokinetic parameter (AUC and Cmax) (Fig. 3).
A level "C" correlation does not reflect the complete shape of the plasma concentration time curve, therefore is not the most useful correlation from a regulatory point of view. However, this type of correlation can be useful in early formulation development.
A multiple level "C" correlation relates one or several pharmacokinetic parameters of interest to the amount of drug dissolved at several time points of the dissolution profile (e.g., Cmax vs. % dissolved in two hours, six hours, and 12 hours)—see Fig. 4 below demonstrating a multiple level C correlation using formulations I to P . This might be accomplished via linear regression. Multiple level "C" correlation can be as useful as level "A" IVIVC from a regulatory point of view. However, if one can develop a multiple level "C" correlation, it is likely that a level "A" correlation can be developed as well.
2800 2600 O 2400 < 2200 2000 1800
FIGURE 3 Level "C" correlation.
FIGURE 3 Level "C" correlation.
When is an IVIVC Likely?
In vitro/in vivo correlations are generally seen when the dissolution or release of the drug from the dosage form is the rate-limiting step in the absorption and appearance of the drug in in vivo circulation.
FDA Guidance, "Extended Release Oral Dosage Forms: Development, Evaluation and Applications of in vitro/in vivo Correlations" 
This guidance has been developed (1) to reduce the regulatory burden by decreasing the number of biostudies needed to approve and maintain an extended-release product on the market and (2) to set clinically more meaningful dissolution specifications. It is anticipated that with a predictive IVIVC, the biostudies that are generally required for major manufacturing changes are replaced by a simple in vitro dissolution test.
The following general considerations apply in the development of an IVIVC:
• Human data are necessary for regulatory consideration of an IVIVC.
• Bioavailability studies for IVIVC development should be performed with enough subjects to characterize adequately the performance of the drug product under study. The number of subjects in some established IVIVCs has ranged from 6 to 36. Although crossover studies are preferred, parallel studies or cross-study analyses (with appropriate normalization with a common reference) may be acceptable. The reference product in developing an IVIVC may be an intravenous solution, an aqueous oral solution, or an immediate-release product.
• In vitro/in vivo correlations should usually be developed in the fasted state, unless the drug is not tolerated in fasted state and is indicated to be taken only in fed state due to tolerability concerns.
• Any in vitro dissolution method may be used to obtain the dissolution characteristics of the ER dosage form. The most common dissolution apparatus is USP apparatus I (basket) or II (paddle), used at compendially recognized rotation speeds (e.g., 100 rpm for the basket and 50-75 rpm for the paddle). An aqueous medium, either water or a buffered solution preferably not exceeding pH 6.8, is recommended as the initial medium for development of an IVIVC. For poorly soluble drugs, addition of surfactant (e.g., sodium lauryl sulfate) may be appropriate. Nonaqueous and hydroalcoholic systems are generally discouraged. The dissolution profiles of at least 12 individual dosage units from each lot should be determined.
• Generally, IVIVC should be developed using two or more formulations with different release rates. When two or more drug product formulations with different release rates are developed, their in vitro dissolution profiles should be generated using an appropriate dissolution methodology. The dissolution method used should be the same for all the formulations. The IVIVC relationship should be demonstrated consistently with two or more formulations with different release rates to result in corresponding differences in absorption profiles. [9, 12]. When in vitro dissolution is independent of the dissolution test conditions (e.g., medium, agitation, pH), development of IVIVC using one release rate formulation may be sufficient.
• An important factor is the range of release rates to study. The in vitro and in vivo profiles of the formulations used to develop IVIVC should be adequately different.
• Dissolution testing can be carried out during the formulation screening stage using several methods. Once a discriminating system is developed, dissolution conditions should be the same for all formulations tested in the biostudy for development of the correlation and should be fixed before further steps towards correlation evaluation are undertaken.
• It is important to note that the relationship between in vitro dissolution and in vivo dissolution, or absorption, should be the same for all the formulations studied. If one or more of the formulations (highest or lowest release rate formulations) does not show the same relationship between in vitro dissolution and in vivo performance compared with the other formulations, the correlation may still be used within the range of release rates encompassed by the remaining formulations.
The initial stage of establishing an IVIVC is an exploratory modeling process. One method to develop a level "A" correlation is to estimate the in vivo absorption or dissolution time course using an appropriate deconvolution technique for each formulation and subject (using WagnerNelson method, numerical deconvolution, etc.). The in vivo absorption profile is plotted against the in vitro dissolution profile to obtain a correlation (see Figs. 5 and 6).
A Level "A" correlation is usually estimated by a two-stage procedure: deconvolution followed by comparison of the fraction of drug absorbed to the fraction of drug dissolved . Details of the deconvolution/ convolution methodology can be found in several literature articles [14-17] and will not be discussed here. One alternative is based on a convolution procedure that models the relationship between in vitro dissolution and plasma concentration in a single step. Plasma concentrations predicted from the model and those observed are compared directly. For these methods, a reference treatment is desirable, but the lack of one does not preclude the ability to develop an IVIVC . Whatever the method used to develop a Level "A" IVIVC, the IVIVC model should predict the entire in vivo time
Time (hours) Time (hours)
FIGURE 5 In vitro dissolution and in vivo profiles.
Time (hours) Time (hours)
course from the in vitro data. Here the model refers to the relationship between in vitro dissolution of an ER dosage form and an in vivo response such as plasma drug concentration or amount of drug absorbed.
One could use alternative approaches than the ones mentioned to develop correlations. Also, if there is no one-to-one relationship, then dissolution conditions may be altered (prior to evaluation of predictability), or time-scaling approaches  may be used to develop the correlation. However, the time-scaling factor should be the same for all formulations tested. Different time scales for each of the formulations indicate absence of an IVIVC.
Evaluation of Predictability of IVIVC (IVIVC Validation)
An IVIVC should be evaluated to demonstrate that the predictability of the in vivo performance of a drug product, from the in vitro dissolution characteristics of the drug product formulations, is maintained over a range of in vitro release rates. A correlation should predict the in vivo performance accurately and consistently. When such an IVIVC has been established, in vitro dissolution can be used confidently as a surrogate for in vivo bioavailability/bioequivalence of ER drug products. Since the focus of IVIVC evaluation is on the predictive performance of the model, prediction error is evaluated and used as the criteria for IVIVC evaluation in the FDA Guidance (Figs. 7 and 8). Depending on the intended application of an IVIVC and the therapeutic index of the drug, evaluation of predictability internally and/or externally may be appropriate. Evaluation of internal predictability is based on the initial data used to develop the IVIVC. Evaluation of external predictability is based on additional data sets. External predictability evaluation is not necessary unless the drug is a
Was this article helpful?