This section discusses scientific considerations in the stepwise approach to developing data and information needed to support a demonstration of biosimilarity. To demonstrate biosimilarity, a sponsor must provide sufficient data and information to show that the proposed product and the reference product are highly similar notwithstanding minor differences in clinically inactive components and that there are no clinically meaningful differences between the two products in terms of safety, purity, and potency. The type and amount of analyses and testing that will be sufficient to demonstrate biosimilarity will be determined on a product-specific basis.
The PHS Act requires that a 351(k) application include information demonstrating biosimilarity based on data derived from, among other things, analytical studies that demonstrate that the biological product is highly similar to the reference product notwithstanding minor differences in clinically inactive components, unless FDA determines that an element is unnecessary in a 351(k) application. FDA expects that first, a sponsor will extensively characterize the proposed product and the reference product with state-of-the-art technology, because extensive characterization of both products serves as the foundation for a demonstration of biosimilarity. It is expected that the expression construct for a proposed product will encode the same primary amino acid sequence as its reference product. However, minor modifications such as N- or Cterminal truncations that are not expected to change the product performance may be justified and should be explained by the sponsor. Additionally, sponsors should consider all relevant characteristics of the protein product (e.g., the primary, secondary, tertiary, and quaternary structure; posttranslational modifications; and biological activities) to demonstrate that the proposed product is highly similar to the reference product notwithstanding minor differences in clinically inactive components. The more comprehensive and robust the comparative structural and functional characterization is, the stronger the scientific justification for a selective and targeted approach to animal and/or clinical testing.
PHS Act要求 351(k) 申请提交的证明生物相似性性的信息，除了其他必需信息外，还应包括能够证明即使临床非活性成分存在微小差异，类似药和参比制剂仍然高度相似的分析试验数据，除非FDA认为某一数据在351（k）申请中不是必要的。 FDA 希望申请者首先采用最先进的技术对类似药和参比制剂进行广泛研究，因为这些研究结果是论证生物相似性的基础。FDA 期望类似药和参比制剂的氨基酸序列表达构造相同。然而，如N-或C-端断裂等微小差异，申请者可通过科学论证其不会影响安全有效性。另外，申请人还必须考虑到蛋白制品所有相关性质（包括一、二、三、四级结构、翻译后修饰、生物活性），以证明即使在临床非活性成分存在微小差异的情况下，其与参比制剂仍具有高度生物相似性。对结构功能比较得越全面，越有利于动物和/或临床试验的选择和针对性的科学论证。
Sponsors should use appropriate analytical methodologies with adequate sensitivity and specificity for structural characterization of the proteins. Generally, such tests include the following comparisons of the proposed product and the reference product:
（1）Primary structures, such as amino acid sequence 一级结构，如氨基酸序列；
（2）Higher order structures, including secondary, tertiary, and quaternary structure (including aggregation) 高级结构，包括二级、三级和四级结构（包括聚合体）；
（3）Enzymatic posttranslational modifications, such as glycosylation and phosphorylation 酶转录后的修饰，如糖基化和磷酸化；
（4）Other potential variations, such as protein deamidation and oxidation 其他可能的变化，如蛋白脱酰基化反应和氧化反应；
（5）Intentional chemical modifications, such as PEGylation sites and characteristics 有意的化学修饰，如聚乙二醇化结合位点及其特征。
Sponsors should conduct extensive structural characterization of both the proposed product and the reference product in multiple representative lots to understand the lot-to-lot variability of both products in the manufacturing processes. Lots used for the analyses should support the biosimilarity of both the clinical material used in the clinical study(ies) intended to support a demonstration of biosimilarity, and the to-be-marketed proposed product, to the reference product. Characterization of lots manufactured during process development for the proposed product may also be useful. Sponsors should justify the selection of the representative lots, including the number of lots.
In addition, FDA recommends that sponsors analyze the finished dosage form of multiple lots of the proposed product and the reference product, assessing excipients and any formulation effect on purity, product- and process-related impurities, and stability. Differences in formulation between the proposed product and the reference product are among the factors that may affect the extent and nature of subsequent animal or clinical testing. A sponsor considering manufacturing changes after completing the initial analytical similarity assessment or after completing clinical testing intended to support a 351(k) application should perform an additional analytical similarity assessment with lots manufactured by the new process and the reference product and establish comparability of the proposed product manufactured by the old and new manufacturing processes. The nature and extent of the changes may determine the extent of the analytical similarity and comparability studies and any necessary additional studies.
If the reference product or the proposed product cannot be adequately characterized with state-of-the-art technology, the application for the proposed product may not be appropriate for submission under section 351(k) of the PHS Act; and the sponsor should consult FDA for guidance on the appropriate submission pathway.
如果利用最先进的技术仍不能全面描述类似药和参比制剂的特征，则PHS Act351(k) 可能不适用于该产品的申请，申请人应咨询FDA，请FDA指导PHS Act 351(k)是否适用于这种蛋白产品的申请。
The pharmacologic activity of protein products should be evaluated by in vitro and/or in vivo functional assays. In vitro assays may include, but are not limited to, biological assays, binding assays, and enzyme kinetics. In vivo assays may include the use of animal models of disease (e.g., models that exhibit a disease state or symptom) to evaluate functional effects on pharmacodynamic markers or efficacy measures. A functional evaluation comparing a proposed product to the reference product using these types of assays is also an important part of the foundation that supports a demonstration of biosimilarity and may be used to scientifically justify a selective and targeted approach to animal and/or clinical testing.
Sponsors can use functional assays to provide additional evidence that the biologic activity and potency of the proposed product are highly similar to those of the reference product and/or to support a conclusion that there are no clinically meaningful differences between the proposed product and the reference product. Such assays also may be used to provide additional evidence that the MOA of the two products is the same to the extent the MOA of the reference product is known. Functional assays can be used to provide additional data to support results from structural analyses, investigate the consequences of observed structural differences, and explore structure-activity relationships. These assays are expected to be comparative so they can provide evidence of similarity or reveal differences in the performance of the proposed product compared to the reference product, especially differences resulting from variations in structure that cannot be detected using current analytical methods. FDA also recommends that sponsors discuss limitations of the assays they used when interpreting results in their submissions to FDA. Such discussions would be useful for the evaluation of analytical data and may guide whether additional analytical testing would be necessary to support a demonstration of biosimilarity.
Functional assays can also provide information that complements the animal and clinical data in assessing the potential clinical effects of minor differences in structure between the proposed product and the reference product. For example, cell-based bioactivity assays may be used to detect the potential for inducing cytokine release syndrome in vivo. The available information about these assays, including sensitivity, specificity, and extent of validation, can affect the amount and type of additional animal or clinical data that may be needed to establish biosimilarity. As is the case for the structural evaluation, sponsors should justify the selection of the representative lots, including the number of lots.
The PHS Act also requires that a 351(k) application include information demonstrating biosimilarity based on data derived from animal studies (including the assessment of toxicity), unless FDA determines that such studies are not necessary in a 351(k) application. Results from animal studies may be used to support the safety evaluation of the proposed product and more generally to support the demonstration of biosimilarity between the proposed product and the reference product.
1. Animal Toxicity Studies动物毒理学研究
As a scientific matter, animal toxicity data are considered useful when, based on the results of extensive structural and functional characterization, uncertainties remain about the safety of the proposed product that need to be addressed before initiation of clinical studies in humans (assuming results from animal studies can meaningfully address the remaining uncertainties).
The scope and extent of any animal toxicity studies will depend on information about the reference product, information about the proposed product, and the extent of known similarities or differences between the two. As described further in section IX, FDA encourages sponsors to initiate early discussions with the Agency with regard to their biosimilar development plans, including identifying appropriate scientific justifications for not conducting an animal toxicity study or for the scope and extent of such a study.
If comparative structural and functional data using the proposed product provide strong support for analytical similarity to a reference product, then limited animal toxicity data may be sufficient to support initial clinical use of the proposed product. Such a study may be non-sacrificial and include endpoints that measure in-life parameters, PD, and PK (with an assessment of immunogenicity).
If the structural and functional data are limited in scope or there are concerns about the proposed product quality, a general toxicology study may be needed that includes full animal pathology, histopathology, PD, PK, and immunogenicity assessments. When animal toxicology studies are conducted, it will be useful to perform a comparative study with the proposed product and the reference product (i.e., comparative bridging toxicology studies). The selection of dose, regimen, duration, and test species for these studies should provide a meaningful toxicological comparison between the two products. It is important to understand the limitations of such animal studies (e.g., small sample size, intra-species variations) when interpreting results comparing the proposed product and the reference product. For a detailed discussion on the design of animal toxicology studies relevant to biological products, see the ICH guidance for industry S6(R1) Preclinical Safety Evaluation of Biotechnology-Derived Pharmaceuticals (ICH S6(R1)).
Safety data derived from animal toxicity studies generally are not expected if clinical data (e.g., from studies or marketing experience outside the United States) using the proposed product are available (with the same proposed route of administration and formulation) that provide sufficient evidence for its safe use, unless animal toxicity studies are otherwise needed to address a specific product quality concern.
Animal toxicity studies are generally not useful if there is no animal species that can provide pharmacologically relevant data for the product (i.e., no species in which the biologic activity of the product mimics the human response). For a detailed discussion about demonstrating species relevance, see the criteria described in ICH S6(R1). However, there may be some instances when animal data from a pharmacologically nonresponsive species (including rodents) may be useful to support clinical studies with a proposed product that has not been previously tested in human subjects, for example, comparative PK and systemic tolerability studies. If animal toxicity studies are not warranted based on an acceptable scientific justification, additional comparative in vitro testing (using human cells or tissues when appropriate) is encouraged. Data derived using human cells can provide important comparative information between the proposed product and the reference product regarding potential clinical effects (section VII.B), particularly in situations where there are no animal species available for safety testing.
In general, nonclinical safety pharmacology, reproductive and developmental toxicity, and carcinogenicity studies are not warranted when the proposed product and the reference product have been demonstrated to be highly similar through extensive structural and functional characterization and animal toxicity studies (if such studies were conducted).
2. Inclusion of Animal PK and PD Measures动物PK/PD研究
Under certain circumstances, a single-dose study in animals comparing the proposed product and the reference product using PK and PD measures may contribute to the totality of evidence that supports a demonstration of biosimilarity. Specifically, sponsors can use results from animal studies to support the degree of similarity based on the PK and PD profiles of the proposed product and the reference product. PK and PD measures also can be incorporated into a single animal toxicity study, where appropriate. Animal PK and PD assessment will not negate the need for human PK and PD studies.
3. Interpreting Animal Immunogenicity Results动物免疫学研究
Animal immunogenicity assessments are conducted to assist in the interpretation of the animal study results and generally do not predict potential immune responses to protein products in humans. However, when differences in manufacturing (e.g., impurities or excipients) between the proposed product and the reference product may result in differences in immunogenicity, measurement of anti-therapeutic protein antibody responses in animals may provide useful information. Additionally, differences observed in animal immunogenicity assessments may reflect potential structural or functional differences between the two products not captured by other analytical methods.
The sponsor of a proposed product must include in its submission to FDA information demonstrating that “there are no clinically meaningful differences between the biological product and the reference product in terms of the safety, purity, and potency of the product.”
The nature and scope of the clinical study or studies will depend on the nature and extent of residual uncertainty about biosimilarity after conducting structural and functional characterization and, where relevant, animal studies. The frequency and severity of safety risks and other safety and effectiveness considerations (e.g., poor relationship between pharmacologic effects and effectiveness) for the reference product may also affect the design of the clinical program. The scope of the clinical program and the type of clinical studies (i.e., comparative human PK, PD, clinical immunogenicity, or clinical safety and effectiveness) should be scientifically justified by the sponsor.
As a scientific matter, FDA expects a sponsor to conduct comparative human PK and PD studies (if there is a relevant PD measure(s)) and a clinical immunogenicity assessment. In certain cases, the results of these studies may provide adequate clinical data to support a conclusion that there are no clinically meaningful differences between the proposed biosimilar product and the reference product. However, if residual uncertainty about biosimilarity remains after conducting these studies, an additional comparative clinical study or studies would be needed to further evaluate whether there are clinically meaningful differences between the two products.
1.Human Pharmacology Data人体药理学数据
Human PK and PD profiles of a protein product often cannot be adequately predicted from functional assays and/or animal studies alone. Therefore, human PK and PD studies comparing a proposed product to the reference product generally are fundamental components in supporting a demonstration of biosimilarity. Both PK and PD studies (where there is a relevant PD measure(s)) generally will be expected to establish biosimilarity, unless a sponsor can scientifically justify that such a study is not needed. Even if relevant PD measures are not available, sensitive PD endpoints may be assessed if such assessment may help reduce residual uncertainty about biosimilarity.
通常，仅通过功能测定和/或动物研究不能充分评估某种蛋白制品的人体PK和PD。因此，类似药和参比制剂的人体PK 和PD 比较数据是支持生物相似性的基本组成，也是必须的，除非申请人科学论证其不必要性。即使没有PD研究方法，也可以通过评估PD敏感终点来帮助减少生物相似性的不确定性。
Sponsors should provide a scientific justification for the selection of the human PK and PD study population (e.g., patients versus healthy subjects) and parameters, taking into consideration the relevance and sensitivity of such population and parameters, the population and parameters studied for the licensure for the reference product, as well as the current knowledge of the intra-subject and inter-subject variability of human PK and PD for the reference product. For example, comparative human PK and PD studies should use a population, dose(s), and route of administration that are adequately sensitive to allow for the detection of differences in PK and PD profiles. FDA recommends that, to the extent possible, the sponsor select PD measures that (1) are relevant to clinical outcomes (e.g., on mechanistic path of MOA or disease process related to effectiveness or safety); (2) are measurable for a sufficient period of time after dosing to ascertain the full PD response and with appropriate precision; and (3) have the sensitivity to detect clinically meaningful differences between the proposed product and the reference product. Use of multiple PD measures that assess different domains of activities may also be of value.
申请人要说明PK 和PD 研究中试验人群（如患者和健康人的比例）和参数选择的科学合理性，要综合考虑这些人群和参数之间的关联度和灵敏度、服用参比制剂的人群和参数研究以及参比制剂在受试者体内或受试者之间PK 和PD的相关变化。例如，人类PK/PD比较性研究采用的人群、剂量和给药途径应具有足够灵敏度，以检测PK/PD的差异性。FDA建议，PD参数应选择：（1）和临床结果相关，例如与MOA作用机制或疾病发展过程的安全性或有效性相关；（2）给药后的一段时间内可测性，确保全PD反应的精确测定；（3）足够灵敏，可以检测到两药之间的差异。使用多个PD指标来评估不同反应活动也可能可取的。
When there are established dose-response or systemic exposure-response relationships (response may be PD measures or clinical endpoints), it is important to select, whenever possible, a dose(s) for study on the steep part of the dose-response curve for the proposed product. Studying doses that are on the plateau of the dose-response curve is unlikely to detect clinically meaningful differences between the two products. Sponsors should predefine and justify the criteria for PK and PD parameters for studies included in the application to demonstrate biosimilarity
A human PK study that demonstrates similar exposure (e.g., serum concentration over time) for the proposed product and the reference product may provide support for a demonstration of biosimilarity. A human PK study may be particularly useful when the exposure correlates with clinical safety and effectiveness. A human PD study that demonstrates a similar effect on a relevant PD measure(s) related to effectiveness or specific safety concerns (except for immunogenicity, which is evaluated separately) represents even stronger support for a biosimilarity determination.
In certain cases, establishing a similar clinical PK, PD, and immunogenicity profile may provide sufficient clinical data to support a conclusion that there are no clinically meaningful differences between the two products. PK and PD parameters are generally more sensitive than clinical efficacy endpoints in assessing the similarity of two products. For example, an effect on thyroid stimulating hormone (TSH) levels would provide a more sensitive comparison of two thyroxine products than an effect on clinical symptoms of euthyroidism.
In cases where there is a meaningful correlation between PK and PD results and clinical effectiveness, convincing PK and PD results may make a comparative efficacy study unnecessary. For example, similar dose-response curves of the proposed product and the reference product on a relevant PD measure, combined with a similar human PK profile and clinical immunogenicity profile, could provide sufficient evidence to support a conclusion of no clinically meaningful differences. Even if there is still residual uncertainty about biosimilarity based on PK and PD results, establishing a similar human PK and PD profile may provide a scientific basis for a selective and targeted approach to subsequent clinical testing.
For PD studies using products with a short half-life (e.g., shorter than 5 days), a rapid PD response, and a low incidence of immunogenicity, a crossover design is appropriate, when feasible. For products with a longer half-life (e.g., more than 5 days), a parallel design will usually be needed. Sponsors should provide a scientific justification for the selection of study dose (e.g., one dose or multiple doses) and route of administration.
FDA recommends that sponsors consider the duration of time it takes for a PD measure to change and the possibility of nonlinear PK. FDA also encourages consideration of the role of modeling and simulation in designing comparative human PK and PD studies.
2.Clinical Immunogenicity Assessment临床免疫学评价
The goal of the clinical immunogenicity assessment is to evaluate potential differences between the proposed product and the reference product in the incidence and severity of human immune responses. Immune responses may affect both the safety and effectiveness of the product by, for example, altering PK, inducing anaphylaxis, or promoting development of neutralizing antibodies that neutralize the product as well as its endogenous protein counterpart. Thus, establishing that there are no clinically meaningful differences in immune response between a proposed product and the reference product is a key element in the demonstration of biosimilarity. Structural, functional, and animal data are generally not adequate to predict immunogenicity in humans. Therefore, at least one clinical study that includes a comparison of the immunogenicity of the proposed product to that of the reference product will be expected. FDA encourages that, where feasible, sponsors collect immunogenicity data in any clinical study, including human PK or PD studies.
The extent and timing of the clinical immunogenicity assessment will vary depending on a range of factors, including the extent of analytical similarity between the proposed product and the reference product, and the incidence and clinical consequences of immune responses for the reference product. For example, if the clinical consequence is severe (e.g., when the reference product is a therapeutic counterpart of an endogenous protein with a critical, nonredundant biological function or is known to provoke anaphylaxis), a more extensive immunogenicity assessment will likely be needed to support a demonstration of biosimilarity. If the immune response to the reference product is rare, a premarketing evaluation to assess apparent differences in immune responses between the two products may be adequate to support biosimilarity. In addition, in some cases certain safety risks may need to be evaluated through postmarketing surveillance or studies.
The overall immunogenicity assessment should consider the nature of the immune response (e.g., anaphylaxis, neutralizing antibody), the clinical relevance and severity of consequences (e.g., loss of efficacy of life-saving therapeutic and other adverse effects), the incidence of immune responses, and the population being studied. FDA recommends use of a comparative parallel design (i.e., a head-to-head study) in treatment-naïve patients as the most sensitive design for a premarketing study to assess potential differences in the risk of immunogenicity. However, depending on the clinical experience of the reference and proposed products (taking into consideration the conditions of use and patient population), a sponsor may need to evaluate a subset of patients to provide a substantive descriptive assessment of whether a single cross-over from the reference product to the proposed biosimilar would result in a major risk in terms of hypersensitivity, immunogenicity, or other reactions. The design of any study to assess immunogenicity and acceptable differences in the incidence and other parameters of immune response should be discussed with FDA before initiating the study. Differences in immune responses between a proposed product and the reference product in the absence of observed clinical sequelae may be of concern and may warrant further evaluation (e.g., extended period of follow-up evaluation).
The study population used to compare immunogenicity should be justified by the sponsor and agreed to by the Agency. If a sponsor is seeking to extrapolate immunogenicity findings for one condition of use to other conditions of use, the sponsor should consider using a study population and treatment regimen that are adequately sensitive for predicting a difference in immune responses between the proposed product and the reference product across the conditions of use. Usually, this will be the population and regimen for the reference product for which development of immune responses with adverse outcomes is most likely to occur (e.g., patients on background immunosuppressants would be less likely to develop immune responses than patients who are not immunosuppressed).
The selection of clinical immunogenicity endpoints or PD measures associated with immune responses to therapeutic protein products (e.g., antibody formation and cytokine levels) should take into consideration the immunogenicity issues that have emerged during the use of the reference product. Sponsors should prospectively define the clinical immune response criteria (e.g., definitions of significant clinical events such as anaphylaxis), using established criteria where available, for each type of potential immune response and should obtain agreement from FDA on these criteria before initiating the study.
The duration of follow-up evaluation should be determined based on (1) the time course for the generation of immune responses (such as the development of neutralizing antibodies, cell-mediated immune responses) and expected clinical sequelae (informed by experience with the reference product), (2) the time course of disappearance of the immune responses and clinical sequelae following cessation of therapy, and (3) the length of administration of the product. For example, for chronically administered agents, the follow-up period is recommended to be 1 year unless a shorter duration can be scientifically justified based on the totality of the evidence to support biosimilarity.
As a scientific matter, a sponsor should evaluate the following antibody parameters in the clinical immunogenicity assessment:
（1）Titer, specificity, relevant isotype distribution, time course of development, persistence, disappearance, impact on PK, and association with clinical sequelae效价（滴定度）、特异性、相关同型的分布、发生进程、持续时间、消失进程、对PK的影响以及和临床后遗症的联系。
（2）Neutralization of product activity: neutralizing capacity to all relevant functions (e.g., uptake and catalytic activity, neutralization for replacement enzyme therapeutics) 产物活性的中和：对所有相关功能的中和能力（例如，摄取和催化活性，替代酶治疗剂的中和）。
The sponsor should develop assays capable of sensitively detecting immune responses, even in the presence of the circulating drug product (proposed product and reference product). The proposed product and the reference product should be assessed in the same assay with the same patient sera whenever possible. FDA recommends that immunogenicity assays be developed and validated early in development, and the validation should consider both the proposed product and the reference product. Sponsors should consult with FDA on the sufficiency of assays before initiating any clinical immunogenicity assessment.
D. Clinical Studies – General Considerations临床研究——一般考虑
3. Comparative Clinical Studies临床可比性研究
As a scientific matter, a comparative clinical study will be necessary to support a demonstration of biosimilarity if there is residual uncertainty about whether there are clinically meaningful differences between the proposed product and the reference product based on structural and functional characterization, animal testing, human PK and PD data, and clinical immunogenicity assessment. A sponsor should provide a scientific justification if it believes that a comparative clinical study is not necessary.
The following are examples of factors that may influence the type and extent of the comparative clinical study data needed:
a. The nature and complexity of the reference product, the extensiveness of structural and functional characterization, and the findings and limitations of comparative structural, functional, and nonclinical testing, including the extent of observed differences. 参比制剂的性质和复杂性，结构和功能特性，结构功能比较分析的结果和限制条件，非临床研究（包括两药差异）。
b. The extent to which differences in structure, function, and nonclinical pharmacology and toxicology predict differences in clinical outcomes, in conjunction with the degree of understanding of the MOA of the reference product and disease pathology. 结构、功能、非临床药理学和毒理学方面的差异能够预测临床结果差别的程度，以及对参照药作用机制和疾病药理的理解程度。
c．The extent to which human PK or PD is known to predict clinical outcomes (e.g., PD measures known to be relevant to effectiveness or safety).已知人体PK/PD 数据能够预测临床结果的程度（比如，PD结果和临床有效性相关）。
d．The extent of clinical experience with the reference product and its therapeutic class, including the safety and risk-benefit profile (e.g., whether there is a low potential for off-target adverse events), and appropriate endpoints and biomarkers for safety and effectiveness (e.g., availability of established, sensitive clinical endpoints).参比制剂临床使用历史和治疗类别，包括安全性和风险效益比（例如无效案例的概率是否够低），以及是否有合适的安全性和有效性终点和生物标志物（如，已经建立的、敏感的临床终点）。
e．The extent of any other clinical experience with the proposed product (e.g., if the proposed product has been marketed outside the United States).试验药物的临床使用历史（例如，该产品是否已在美国以外地区销售）
A sponsor should provide a scientific justification for how it intends to use these factors to determine what type(s) of clinical study(ies) are needed and the design of any necessary study(ies). For example, if a comparative clinical study is needed, a sponsor should explain how these factors were considered in determining the design of such a study, including the endpoint(s), population, similarity margin, and statistical analyses.
Additionally, specific safety or effectiveness concerns regarding the reference product and its class (including history of manufacturing- or source-related adverse events) may warrant more comparative clinical data. Alternatively, if there is information regarding other biological products that could support a biosimilarity determination (with marketing histories that demonstrate no apparent differences in clinical safety and effectiveness profiles), such information may be an additional factor supporting a selective and targeted approach to the clinical program.
A sponsor should use endpoints that can assess clinically meaningful differences between the proposed product and the reference product in a comparative clinical study. The endpoints may be different from those used as primary endpoints in the reference product’s clinical studies if they are scientifically supported. As discussed in section VII.D.1, certain endpoints (such as PD measures) are more sensitive than clinical endpoints and, therefore, may enable more precise comparisons of relevant therapeutic effects. There may be situations when the assessment of multiple PD measures in a comparative clinical study will enhance the sensitivity of the study. The adequacy of the endpoints depends on the extent to which PD measures correlate with clinical outcome, the extent of structural and functional data support for biosimilarity, the understanding of MOA, and the nature or seriousness of outcome affected.
The choice of study population should allow for an assessment of clinically meaningful differences between the proposed product and the reference product. Often the study population will have characteristics consistent with those of the population studied for the licensure of the reference product for the same indication. However, there are cases where a study population could be different from that in the clinical studies that supported the licensure of the reference product. For example, if a genetic predictor of response was developed following licensure of the reference product, it may be possible to use patients with the response marker as the study population.
Sample Size and Duration of Study样本量和试验周期：
The sample size for and duration of the comparative clinical study should be adequate to allow for the detection of clinically meaningful differences between the two products. As discussed in section VII.D.1, certain endpoints, such as PD measures, may be more sensitive than clinical endpoints and facilitate the conduct of a smaller study of limited duration. In such cases where the size and duration of the comparative clinical study may not be adequate for the detection of relevant safety signals, a separate assessment of safety and immunogenicity may be needed.
Study Design and Analyses研究设计和分析：
A comparative clinical study for a biosimilar development program should be designed to investigate whether there are clinically meaningful differences between the proposed product and the reference product. The design should take into consideration the nature and extent of residual uncertainty that remains about biosimilarity based on data generated from comparative structural and functional characterization, animal testing, human PK and PD studies, and clinical immunogenicity assessment.
Generally, FDA expects a clinical study or studies designed to establish statistical evidence that the proposed product is neither inferior to the reference product by more than a specified margin nor superior to the reference product by more than a (possibly different) specified margin. Typically, an equivalence design with symmetric inferiority and superiority margins would be used. Symmetric margins would be reasonable when, for example, there are dose-related toxicities.
In some cases, it would be appropriate to use an asymmetric interval with a larger upper bound to rule out superiority than lower bound to rule out inferiority. An asymmetric interval could be reasonable, for example, if the dose used in the clinical study is near the plateau of the dose-response curve and there is little likelihood of dose-related effects (e.g., toxicity). In most cases, use of an asymmetric interval would generally allow for a smaller sample size than would be needed with symmetric margins. However, if there is a demonstration of clear superiority, then further consideration should be given as to whether the proposed product can be considered biosimilar to the reference product.
In some cases, depending on the study population and endpoint(s), ruling out only inferiority may be adequate to establish that there are no clinically meaningful differences between the proposed product and the reference product. For example, if it is well established that doses of a reference product pharmacodynamically saturate the target at the clinical dose level and it would be unethical to use lower than clinically approved doses, a non-inferiority (NI) design may be sufficient.
A sponsor should provide adequate scientific justification for the choice of study design, study population, study endpoint(s), estimated effect size for the reference product, and margin(s) (how much difference to rule out). Sponsors should discuss their study proposal(s) and overall clinical development plan with FDA before initiating the comparative clinical study(ies).
4. Extrapolation of Clinical Data Across Indications临床数据在不同适应症之间的外推
If the proposed product meets the statutory requirements for licensure as a biosimilar product under section 351(k) of the PHS Act based on, among other things, data derived from a clinical study or studies sufficient to demonstrate safety, purity, and potency in an appropriate condition of use, the applicant may seek licensure of the proposed product for one or more additional conditions of use for which the reference product is licensed. However, the applicant would need to provide sufficient scientific justification for extrapolating clinical data to support a determination of biosimilarity for each condition of use for which licensure is sought.
在PHS Act 351（K）项下，如果类似药的安全性、纯度和效力等研究结果表明其和参比制剂在一适应症下具有生物相似性，那么类似药申请用于与参比制剂获批的其他适应症的许可。但是，申请人需要提供科学合理的理由，说明该临床数据能够支持其他适应症的生物相似性。
Such scientific justification for extrapolation should address, for example, the following issues for the tested and extrapolated conditions of use:
The MOA(s) in each condition of use for which licensure is sought; this may include 每种适应症的作用机制，包括：
--The target/receptor(s) for each relevant activity/function of the product相关活性或功效的作用靶点或受体；
--The binding, dose/concentration response, and pattern of molecular signaling upon engagement of target/receptor(s) 作用靶点或受体的结合、剂量/浓度反应和分子信号模式；
--The relationships between product structure and target/receptor interactions 药物结构和靶点/受体相互作用的关系；
--The location and expression of the target/receptor(s) 靶点/受体的位置和表达。
The PK and bio-distribution of the product in different patient populations (Relevant PD measures may also provide important information on the MOA.) 不同患者中药物的PK和分布参数（相关PD测量值可能也会提供作用机制的重要信息）。
The immunogenicity of the product in different patient populations 不同患者群体中产品的免疫原性。
Differences in expected toxicities in each condition of use and patient population (including whether expected toxicities are related to the pharmacological activity of the product or to off-target activities) 每种适应症和患者群体中毒理学的差异，包括药理活性和非药理活性相关的毒性。
Any other factor that may affect the safety or efficacy of the product in each condition of use and patient population for which licensure is sought 其他任何可能影响所申请适应症的安全性或有效性的因素。
Differences between conditions of use with respect to the factors described above do not necessarily preclude extrapolation. A scientific justification should address these differences in the context of the totality of the evidence supporting a demonstration of biosimilarity.
In choosing which condition of use to study that would permit subsequent extrapolation of clinical data to other conditions of use, FDA recommends that a sponsor consider choosing a condition of use that would be adequately sensitive to detect clinically meaningful differences between the two products.
The sponsor of a proposed product may obtain licensure only for a condition of use that has been previously licensed for the reference product. If a reference product has a condition of use that was licensed under section 506(c) of the FD&C Act and 21 CFR part 601, subpart E (accelerated approval), and the reference product’s clinical benefit in this condition of use has not yet been verified in postmarketing studies, the proposed product sponsor should consider studying another condition of use for which the reference product is licensed to avoid potential complications in the event that postmarketing studies fail to verify the clinical benefit of the reference product for the condition of use.
申请人获取的许可只能是参比制剂已获批的适应症。若参比制剂的某适应症根据FD＆C Act第506（c）节和21 CFR第601部分E子部分（加速审批）获得许可，但参比制剂该适用证的临床优势尚未在上市后研究中被验证，那么申请人应考虑研究参比制剂获得许可的另一适用证，以避免在参比制剂该适应症上市后监测中未能验证其临床优势或有潜在的并发症。
Robust postmarketing safety monitoring is an important component in ensuring the safety and effectiveness of biological products, including biosimilar therapeutic protein products.
Postmarketing safety monitoring should first take into consideration any particular safety or effectiveness concerns associated with the use of the reference product and its class, the proposed product in its development and clinical use (if marketed outside the United States), the specific condition of use and patient population, and patient exposure in the biosimilar development program. Postmarketing safety monitoring for a proposed product should also have adequate mechanisms in place to differentiate between the adverse events associated with the proposed product and those associated with the reference product, including the identification of adverse events associated with the proposed product that have not been previously associated with the reference product. Rare, but potentially serious, safety risks (e.g., immunogenicity) may not be detected during preapproval clinical testing because the size of the population exposed likely will not be large enough to assess rare events. In particular cases, such risks may need to be evaluated through postmarketing surveillance or studies. In addition, as with any other biological product, FDA may take any appropriate action to ensure the safety and effectiveness of a proposed product, including, for example, requiring a postmarketing study or clinical trial to evaluate certain safety risks.
Because some aspects of postmarketing safety monitoring are product-specific, FDA encourages sponsors to consult with appropriate FDA divisions to discuss the sponsor’s proposed approach to postmarketing safety monitoring.
Many product-specific factors can influence the components of a product development program intended to establish that a proposed product is biosimilar to a reference product. Therefore, FDA will ordinarily provide feedback on a case-by-case basis on the components of a development program for a proposed product. In addition, it may not be possible to identify in advance all the necessary components of a development program; and the assessment of one element (e.g., structural analyses) at one step can influence decisions about the type and amount of subsequent data for the next step. For these reasons, FDA recommends that sponsors use a stepwise approach to establish the totality of the evidence that supports a demonstration of biosimilarity.
FDA also advises sponsors intending to develop biosimilar products to meet with FDA to present their product development plans and establish a schedule of milestones that will serve as landmarks for future discussions with the Agency. FDA anticipates that early discussions with FDA about product development plans and about the approaches to providing adequate scientific justifications will facilitate biosimilar development.
FDA 还建议申请者先向FDA 说明其研究方案，并建立关键咨询讨论的时间表。FDA 认为，提前沟通讨论研发计划及资料递交方法有助于类似药的研究与审批。
As used in this guidance, the following terms are defined below 本指南中使用的术语如下：
Biological product生物制品：“a virus, therapeutic serum, toxin, antitoxin, vaccine, blood, blood component or derivative, allergenic product, protein (except any chemically synthesized polypeptide), or analogous product, or arsphenamine or derivative of arsphenamine (or any other trivalent organic arsenic compound), applicable to the prevention, treatment, or cure of a disease or condition of human beings.” 用以预防，治疗，诊断疾病或人体机能的病毒、治疗性血清、毒素、抗毒素、疫苗、血液、血液成分或衍生物，引起过敏的产品、蛋白质（化学合成多肽除外），或者类似产品，或胂凡纳明及其衍生物（或其他三价有机砷化物）。
Biosimilar or biosimilarity生物相似或生物相似性：“the biological product is highly similar to the reference product notwithstanding minor differences in clinically inactive components,” and that “there are no clinically meaningful differences between the biological product and the reference product in terms of the safety, purity, and potency of the product.” 尽管临床非活性成分存在微小差异，一种生物制品和参比制剂高度相似，且两种产品在安全性、纯度和效力方面无临床有意义差别。
Chemically synthesized polypeptide化学合成多肽：any alpha amino acid polymer that (a) is made entirely by chemical synthesis and (b) is less than 100 amino acids in size. 完全由化学合成的，氨基酸少于100个的α-氨基酸聚合物。
Product产品：when used without modifiers in this guidance, is intended to refer to the intermediates, drug substance, and/or drug product, as appropriate. The use of the term product is consistent with the use of the term in ICH Q5E. 本指南中的“Product（不带修饰语时）”可以指中间体、原料药、和/或药品。和ICH Q5E中的定义一致。
Protein蛋白质：any alpha amino acid polymer with a specific defined sequence that is greater than 40 amino acids in size. 由超过40 个氨基酸按照特定序列构成的α-氨基酸聚合物。
Reference product参比制剂：the single biological product licensed under section 351(a) of the PHS Act against which a biological product is evaluated in a 351(k) application. 按照PHS Act351(a) 路径获批的生物制品，其他生物制品通过351(k)路径申请时以其作为对比对象。