Guidance for Industry Exposure-Response Relationships — Study Design, Data Analysis, and Regulatory Applications

III. DRUG DEVELOPMENT AND REGULATORY APPLICATIONS

This section describes the potential uses of exposure-response relationships in drug development and regulatory decision-making. The examples are not intended to be all-inclusive, but rather to illustrate the value of a better understanding of exposure-response relationships. We recommend that sponsors refer to other ICH and FDA guidances for a discussion of the uses of exposure-response relationships (see Appendix A).

A. Information to Support the Drug Discovery and Development Processes

Many drugs thought to be of potential value in treating human disease are introduced into development based on knowledge of in vitro receptor binding properties and identified pharmacodynamic effects in animals. Apart from describing the tolerability and PK of a drug in humans, Phase 1 and 2 studies can be used to explore the relationship of exposure (whether dose or concentration) to a response (e.g., nonclinical biomarkers, potentially valid surrogate endpoints, or short-term clinical effects) to (1) link animal and human findings, (2) provide evidence that the hypothesized mechanism is affected by the drug (proof of concept), (3) provide evidence that the effect on the mechanism leads to a desired short-term clinical outcome (more proof of concept), or (4) provide guidance for designing initial clinical endpoint trials that use a plausibly useful dose range. Both the magnitude of an effect and the time course of effect are important to choosing dose, dosing interval, and monitoring procedures, and even to deciding what dosage form (e.g., controlled-release dosage form) to develop. Exposure-response and PK data can also define the changes in dose and dosing regimens that account for intrinsic and extrinsic patient factors.

B. Information to Support a Determination of Safety and Efficacy

Apart from their role in helping design the well-controlled studies that will establish the effectiveness of a drug, exposure-response studies, depending on study design and endpoints, can:

• 
  Represent a well-controlled clinical study, in some cases a particularly persuasive one, contributing to substantial evidence of effectiveness (where clinical endpoints or accepted surrogates are studied)
  

• 
  Add to the weight of evidence supporting efficacy where mechanism of action is well understood (e.g., when an effect on a reasonably well-established biomarker/surrogate is used as an endpoint)
  

• 
  Support, or in some cases provide primary evidence for, approval of different doses, dosing regimens, or dosage forms, or use of a drug in different populations, when effectiveness is already well-established in other settings and the study demonstrates a PK-PD relationship that is similar to, or different in an interpretable way from the established setting
  

In general, the more critical a role that exposure-response information is to play in the establishment of efficacy, the more critical it is that it be derived from an adequate and well-controlled study (see 21 CFR 314.126), whatever endpoints are studied. Thus, we recommend that critical studies (1) have prospectively defined hypotheses/objectives, (2) use an appropriate control group, (3) use randomization to ensure comparability of treatment groups and to minimize bias, (4) use well-defined and reliable methods for assessing response variables, and (5) use other techniques to minimize bias.

In contrast, some of the exposure-response studies considered in this document include analyses of nonrandomized data sets where associations between volunteer or patient exposure patterns and outcomes are examined. These analyses are often primarily exploratory, but along with other clinical trial data may provide additional insights into exposure-response relationships, particularly in situations where volunteers or patients cannot be randomized to different exposures, such as in comparing effects in demographic subgroups.

1. Contributing to Primary Evidence of Effectiveness and/or Safety

A dose-response study is one kind of adequate and well-controlled trial that can provide primary clinical evidence of effectiveness. The dose-response study is a particularly informative design, allowing observations of benefits and risks at different doses and therefore providing an ability to weigh the benefits and risks when choosing doses. The dose-response study can help ensure that excessive doses (beyond those that add to efficacy) are not used, offering some protection against unexpected and unrecognized dose-related toxicity. Captopril, for example, was a generally well-tolerated drug that caused dose and concentration-related agranulocytosis. Earlier recognition that daily doses beyond 75-150 milligrams were not necessary, and that renal impairment led to substantial accumulation, might have avoided most cases of agranulocytosis.

Dose-response studies can, in some cases, be particularly convincing and can include elements of internal consistency that, depending on the size of the study and outcome, can allow reliance on a single clinical efficacy study as evidence of effectiveness. Any dose-response study includes several comparisons (e.g., each dose vs. placebo, each dose vs. lower doses). A consistent ordering of these responses (most persuasive when, for example, several doses are significantly different from placebo and, in addition, show an increasing response with dose) represents at least internal (within-study) replication, reducing the possibility that an apparent effect is due to chance. In principle, being able to detect a statistically significant difference in pairwise comparisons between doses is not necessary if a statistically significant trend (upward slope) across doses can be established, as described in the ICH E4 guidance on dose-response. It may be advisable, however, if the lowest dose tested is to be recommended, to have additional data on that dose.
In some cases, measurement of systemic exposure levels (e.g., plasma drug concentrations) as part of dose-response studies can provide additional useful information. Systemic exposure data are especially useful when an assigned dose is poorly correlated with plasma concentrations, obscuring an existing concentration-response relationship. This can occur when there is a large degree of interindividual variability in pharmacokinetics or there is a nonlinear relationship between dose and plasma drug concentrations. Blood concentrations can also be helpful when (1) both parent drug and metabolites are active, (2) different exposure measures (e.g., Cmax, AUC) provide different relationships between exposure and efficacy or safety, (3) the number of fixed doses in the dose-response studies is limited, and (4) responses are highly variable and it is helpful to explore the underlying causes of variability of response.

2. Providing Support for Primary Efficacy Studies

Exposure-response information can support the primary evidence of safety and/or efficacy. In some circumstances, exposure-response information can provide important insights that can allow a better understanding of the clinical trial data (e.g., in explaining a marginal result on the basis of knowledge of systemic concentration-response relationships and achieved concentrations). Ideally, in such cases the explanation would be further tested, but in some cases this information could support approval. Even when the clinical efficacy data are convincing, there may be a safety concern that exposure-response data can resolve. For example, it might be reassuring to observe that even patients with increased plasma concentrations (e.g., metabolic outliers or patients on other drugs in a study) do not have increased toxicity in general or with respect to a particular concern (e.g., QT prolongation). Exposure-response data thus can add to the weight of evidence of an acceptable risk/benefit relationship and support approval. The exposure-response data might also be used to understand or support evidence of subgroup differences suggested in clinical trials, and to establish covariate relationships that explain, and enhance the plausibility of, observed subgroup differences in response.

Exposure-response data using short-term biomarkers or surrogate endpoints can sometimes make further exposure-response data from clinical endpoint exposure-response studies unnecessary. For example, if it can be shown that the short-term effect does not increase past a particular dose or concentration, there may be no reason to explore higher doses or concentrations in the clinical trials. Similarly, short-term exposure-response studies with biomarkers might be used to evaluate early (e.g., first dose) responses seen in clinical trials.

3. Supporting New Target Populations, Use in Subpopulations, Doses/Dosing Regimens, Dosage Forms, and Routes of Administration

Exposure-response information can sometimes be used to support use, without further clinical data, of a drug in a new target population by showing similar (or altered in a defined way) concentration-response relationships for a well-understood (i.e., the shape of the exposure-response curve is known), short-term clinical or pharmacodynamic endpoint. Similarly, this information can sometimes support the safety and effectiveness of alterations in dose or dosing interval or changes in dosage form or formulation with defined PK effects by allowing assessment of the consequences of the changes in concentration caused by these alterations. In some cases, if there is a change in the mix of parent and active metabolites from one population (e.g., pediatric vs. adult), dosage form (e.g., because of changes in drug input rate), or route of administration, additional exposure-response data with short-term endpoints can support use in the new population, the new product, or new route without further clinical trials.

a. New target populations

A PK-PD relationship or data from an exposure-response study can be used to support use of a previously approved drug in a new target patient population, such as a pediatric population, where the clinical response is expected to be similar to the adult population, based on a good understanding of the pathophysiology of the disease, but there is uncertainty as to the appropriate dose and plasma concentration. A decision tree illustrating the use of a PK-PD relationship for bridging efficacy data in an adult population to a pediatric population is shown in Appendix B. Possible use of PK-PD bridging studies assessing a well-described PD endpoint (e.g., beta-blockade, angiotensin I or II inhibition) to allow extension of clinical trial information performed in one region to another region is discussed in the ICH E5 guidance on Ethnic Factors in the Acceptability of Foreign Clinical Data.

b. Adjustment of dosages and dosing regimens in subpopulations defined on the basis of intrinsic and extrinsic factors

Exposure-response information linking dose, concentration, and response can support dosage adjustments in patients where pharmacokinetic differences are expected or observed to occur because of one or more intrinsic (e.g., demographic, underlying or accompanying disease, genetic polymorphism) or extrinsic (e.g., diet, smoking, drug interactions) factors. In some cases, this is straightforward, simply adjusting the dose to yield similar systemic exposure for that population. In others, it is not possible to adjust the dose to match both Cmax and AUC. Exposure-response information can help evaluate the implications of the different PK profiles. In some cases, exposure-response information can support an argument that PK changes in exposure would be too small to affect response and, therefore, that no dose or dose regimen adjustments are appropriate.

c. New dose regimens, dosage forms and formulations, routes of administration, and minor product changes.

A known exposure-response relationship can be used to (1) interpolate previous clinical results to new dosages and dosing regimens not well studied in clinical trials, (2) allow marketing of new dosage forms and formulations, (3) support different routes of administration, and (4) ensure acceptable product performance in the presence of changes in components, composition, and method of manufacture that lead to PK differences. Generally, these uses of exposure-response information are based on an understanding of the relationship between the response and concentration, and between dose and concentration.

Exposure-response data can sometimes be used to support a new dose or dosing schedule (e.g., twice a day to once a day) that was not studied in safety and efficacy clinical trials. Exposure-response information can provide insight into the effect of the change in concentrations achieved with these changes and whether or not this will lead to a satisfactory therapeutic response. The new regimen would usually be within the range of total doses studied clinically, but in certain circumstances could be used to extend an approved dose range without additional clinical safety and efficacy data. For example, a once-daily dosing regimen could produce a higher Cmax and a lower Cmin than the same dose given as a twice-daily regimen. If exposure-response data were available, it might be considered reasonable to increase the recommended daily dose to maintain a similar Cmin, even without further studies. Exposure-response data are not likely to be useful in lieu of clinical data in supporting new dosing schedules unless the relationship of the measured responses to relevant safety and efficacy outcomes is well understood.
In some cases, exposure-response data can support the approval of a new drug delivery system (e.g., a modified-release dosage form) when the PK profile is changed intentionally relative to an approved product, generally an immediate-release dosage form. A known exposure-response relationship could be used to determine the clinical significance of the observed differences in exposure, and to determine whether additional clinical efficacy and/or safety data are recommended.
Exposure-response data can also support a new formulation that is unintentionally pharmacokinetically different from the formulation used in the clinical trials to demonstrate safety, or efficacy and safety. In the case of new drugs, in vitro and/or in vivo bioequivalence testing alone is usually used to show that the performance of a new formulation (e.g., to-be-marketed formulation) is equivalent to that used to generate the primary efficacy and safety data. It is possible to demonstrate differences in exposure that are real but not clinically important, even when the 90% confidence interval for the bioequivalence measures fall within the standard of 80-125%. It is possible for these bioequivalence studies to fail to meet the standard bioequivalence acceptance intervals of 80-125%. Rather than reformulating the product or repeating the bioequivalence study, a sponsor may be able to support the view that use of a wider confidence interval or accepting a real difference in bioavailability or exposure would not lead to a therapeutic difference. In other cases, where the altered bioavailability could be of clinical consequence, adjustment of the marketed dosage strength might be used to adjust for the PK difference.
In the case of biological drugs, changes in the manufacturing process often lead to subtle unintentional changes in the product, resulting in altered pharmacokinetics. In cases in which the change in product can be determined not to have any pharmacologic effects (e.g., no effect on unwanted immunogenicity), exposure-response information may allow appropriate use of the new product. Exposure-response data are not likely to obviate the need for clinical data when formulation or manufacturing changes result in altered pharmacokinetics, unless the relationships between measured responses and relevant clinical outcomes are well understood.
Exposure-response information could also be used to support a change in route of administration of a drug. An established exposure-response relationship would allow interpretation of the clinical significance of the difference in PK related to the different route. Such information about active metabolites could also be important in this situation.