By Kevin E. Noonan --
On December 29, the U.S. Food and Drug Administration released its latest Guidance for Industry relating to the biosimilar application process set forth in the Biologic Price Competition and Innovation Act of 2009 (BCPCIA). This Guidance, entitled Clinical Pharmacology Data to Support a Demonstration of Biosimilarity to a Reference Product, is directed to products for which "pharmacokinetic (PK) and pharmacodynamic (PD) data are needed to support a demonstration of biosimilarity" (presumably all biosmilar products). The rubrics set forth in this Guidance are intended to be relevant for establishing that there are no "clinically meaningful differences" between the biosimilar and the reference biologic drug product. In particular, the Guidance is to be used to "address residual uncertainties" after comparative analytical data is acquired, and for providing direction and the need for additional clinical studies, as well as adding to the "totality of the evidence" in favor of biosimilarity and "extrapolation of data" regarding additional indications.
While acknowledging that the extent and types of clinical pharmacological data required will depend on each particular biosimilar, the Guidance provides "critical considerations" for using such studies to support a determination of biosimilarity. These are summarized as "three key concepts": "a PK and PD response assessment, an evaluation of residual uncertainty, and assumptions about analytical quality and similarity," which are especially relevant to biosimilarity determinations.
According to the Guidance, pharmacokinetic and pharmacodynamics studies "should" include exposure and, preferably, exposure-response data (although the Guidance recognizes that informative exposure-response data may be difficult to obtain in view of the complexity and heterogeneneity of biological products). For pharmacodynamics (PD) studies, a single biomarker is preferred although a composite of more than one biomarker can be used (indeed, the Guidance states that "[u]sing broader panels of PD biomarkers (e.g., by conducting a protein or mRNA microarray analysis) that capture multiple pharmacological effects of the product can be of additional value." There are five characteristics set forth in the Guidance that should be considered when choosing a PD-relevant biomarker:
• The time of onset of change in the PD biomarker relative to dosing and its return to baseline with discontinuation of dosing;
• The dynamic range of the PD biomarker over the exposure range to the biological product;
• The sensitivity of the PD biomarker to differences between the proposed biosimilar product and the reference product;
• The relevance of the PD biomarker to the mechanism of action of the drug (to the extent that the mechanism of action is known for the reference product); and
• The analytical validity of the PD biomarker assay.
Even when there are no identified biomarkers that meet these criteria, the Guidance encourages applicants to use PD biomarkers "that achieve a large dynamic range over the concentration range in the PK evaluation because these PD biomarkers represent potential orthogonal tests that can support similarity." And in the absence of sufficiently sensitive PD markers, the Guidance asserts that "derived" PK markers can be used to provide the "primary basis" for evaluating biosimilarity with PD markers being used to augment this data.
The Guidance specifically reasserts the FDA's fundamental policy decision to base approval of biosimilars vel non on the totality of the circumstances, using "a risk-based approach" that includes all relevant data (enumerated as "data from the structural and functional characterizations, nonclinical evaluations, clinical PK and PD studies, clinical immunogenicity testing and an investigation of clinical safety, and, when appropriate, clinical effectiveness"). The protocol FDA has adopted is to require such information in a "stepwise" approach, wherein at each step the agency reviews the amount of residual uncertainty that exists on the question of biosimilarity and fashions what is required to resolve such uncertainty.
While being directed to PD and PK studies the Guidance also reiterates the importance of "extensive and robust comparative structural and functional studies" using "available state-of-the-art" assays for evaluating a variety of physical and biochemical parameters (such as molecular weight, post-translational modifications, heterogeneity, impurity profiles, and others). FDA requires the biosimilar applicant to identify the "type, nature, and extent" of any differences with the reference biologic drug product. The Guidance notes the utility of using a "meaningful fingerprint-like analysis algorithm in this regard, specifying that by "fingerprint-like" is meant "[i]ntegrated, multi-parameter approaches that are extremely sensitive in identifying analytical differences." The Guidance also identifies four categories of analytical results:
• Insufficient analytical similarity (wherein further development is not recommended);
• Analytical similarity with residual uncertainty (requiring additional studies (such as PD or PK studies));
• Tentative analytical similarity (which may require "targeted and selective animal and/or clinical studies); and
• Fingerprint-like analytical similarity (which may still require targeted and selective animal and/or clinical studies).
With regard to specifics of PD and PK study requirements, this Guidance references an earlier one, Bioanalytical Method Validation, and also sets forth both general and specific recommendations. For PK studies, the Guidance recommends that the selected assay be chosen in view of "a thorough understanding of the mechanism of action" or the structural elements understood to be "critical" for biological activity, and ones that produce concentration data with which the biosmilar and reference product sponsor can be compared. For PD studies, the Guidance recommends (in addition to choosing the "most suitable" assays) that the sponsor give the agency a rationale for its assay choice and relevance of the assay to biological activity. The Guidance then sets forth specific considerations for ligand binding assays, concentration and activity assays, and PD assays.
Turning to safety and immunogenicity considerations, the Guidance specifically recites instances where neutralizing antibodies or immune-related toxicity reduces or loses PD efficacy, wherein such data should be disclosed to the agency and perhaps supplemented where appropriate, and should include "relevant patient populations that are not immunocompromised." The measurements performed on the biosimilar should be chosen using "[p]ublicly available information on the safety and immunogenicity profile of the reference product." For specifics on immunogenicity assay development the Guidance references a separate Guidance, Immunogenicity Assessment for Therapeutic Protein Products.
This Guidance repeats FDA's prior encouragement for biosimilar applicants to consult with the agency, here for developing its plan for clinical pharmacological evaluation of the biosimilar product. The Guidance sets forth "[c]ritical topics" what should be discussed with FDA, including specifics on crossover and parallel study designs, which the Guidance characterizes as having "particular relevance." With regard to crossover study design the Guidance states that "a single-dose, randomized, crossover study is generally preferred" for PK similarity assessments. The Guidance recommends such studies for products having a half-life shorter than 5 days, or a rapid PD response (i.e., related to "the time of onset, maximal effect, and disappearance in conjunction with drug exposure"), and a low incidence of immunogenicity. Conversely, a multiple-dose design is recommended for PD similarity assessments. In either case, the Guidance recommends that "time course of appearance and disappearance of immunogenicity and its relation to the washout period should be considered." Parallel design studies, according to the Guidance, are more appropriate for the many biologic drug products having a long half-life that are capable of eliciting an immune response. In any case, the biosimilar applicant should also submit equivalent data on the U.S-licensed reference product. Non-U.S. licensed reference products can also be used for this comparison, provided that the requirements of PHS Act section 351(k)(2)(A) are met and the biosimilar sponsor provides sufficient evidence to "scientifically justify the scientific relevance of these comparative data to an assessment of biosimilarity and establish an acceptable bridge to the U.S.-licensed reference product."
The Guidance also discusses the patient population, which preferably comprises healthy individuals if the biosimilar can be safely administered to them. With regard to population demographics, the Guidance recommends selection of such a population what would be "most likely to provide a sensitive measure of differences between the proposed biosimilar product and the reference product" with evidence of which population was selected on this basis. And of course the total number of patients should be sufficient to provide "adequate statistical power" for PK and PD assessments. Similarly, dose selection should be chosen the be the most sensitive dose for identifying differences between the biosimilar product and the reference product biologic drug (which may be the approved dose for the reference product), and the route of administration should be the same route as used for the biologic drug reference product (and if there is more than one approved route then the one "most sensitive for detecting clinically meaningful difference" should be chosen").
With regard to PK measures, the Guidance specifically states:
All PK measures should be obtained for both the proposed biosimilar product and the reference product. The sponsor should obtain measures of peak concentration (Cmax) and total area under the curve (AUC) in a relevant biological fluid. For single-dose studies, AUC should be calculated as the area under the biological product concentration-time curve from time zero to time infinity (AUC0-∞), where AUC0-∞ = AUC0-t + Ct/kel (or Ct (concentration at the last measurable timepoint) divided by kel (elimination rate constant)) is calculated based on an appropriate method. Cmax should be determined from the data without interpolation. For intravenous studies, AUC0-∞ will be considered the primary endpoint. For subcutaneous studies, Cmax and AUC will be considered coprimary study endpoints. For multiple dose studies, the measurement of total exposure should be the area under the concentration-time profile from time zero to the end of the dosing interval at steady-state (AUC0-tau), and is considered the primary endpoint. Both the concentration prior to the next dose during multiple dosing (Ctrough ss) and Cmax are considered secondary endpoints. Population PK data will not provide an adequate assessment for PK similarity.
For PK measurements, the Guidance recognizes that sometimes "clinical PK and PD data that demonstrate similar exposure and response between a proposed biosimilar product and the reference product can be sufficient to completely assess whether there are clinically meaningful differences between products" (not taking into consideration the need for separate assessment of immunogenicity). But in instances where this data is not sufficient to establish biosimilarity, the Guidance states that the human PD data can be used to lead to a more "targeted" approach for developing further clinical comparison studies. And "sampling strategies" for both PD and PK measurements should be optimized for each, because "the PD-time profile might not mirror the PK-time profile."
For making statistical comparisons of PK and PD results between the biosimilar and reference products, the Guidance recommends a "similarity assessment" that relies on "(1) a criterion to allow the comparison, (2) a confidence interval for the criterion, and (3) an acceptable limit for the biosimilarity assessment." Applicants should use an "average equivalence" approach (further described in FDA Guidance entitled Statistical Approaches to Establishing Bioequivalence) for comparing AUC and Cmax measurements, based on a "two one-sided test procedure that involves calculation of the 90% confidence interval for the ratio of logarithmically transformed averages of the measurements of reference product and biosimilar drugs."
Finally, the Guidance contains a discussion of "simulation tools" for PK/PD study design, wherein these tools are used to select inter alia optimally informative doses for evaluating PD similarity. However, in such cases the biosimilar applicant should chose informative dosages (on the "steep" part of the dose-response curve) and provide reference to publicly available information for such curves or propose a small-scale study to generate this information, particularly with regard to the 50% maximal response for the biologic drug.
The Guidance ends by stating that:
Clinical pharmacology studies play a critical role in the development of biosimilar products. These studies are part of a stepwise process for demonstrating biosimilarity between a proposed biosimilar product and the reference product. These studies may support a demonstration that there are no clinically meaningful differences between the products. These studies may address residual uncertainties that remain after the analytical evaluation, may add to the totality of the evidence supporting a demonstration of biosimilarity, and may also support a selective and targeted approach to the design of any recommended subsequent clinical studies to support a demonstration of biosimilarity.
And like all such Guidances, it contains an express disclaimer:
This guidance represents the current thinking of the Food and Drug Administration (FDA or Agency) on this topic. It does not establish any rights for any person and is not binding on FDA or the public. You can use an alternative approach if it satisfies the requirements of the applicable statutes and regulations. To discuss an alternative approach, contact the FDA office responsible for this guidance as listed on the title page.