跳到正文
制药经理人· GxP知识岛·· 4 天前精选AI 评分77

EMA更新《药品质量:问答(第2部分)》并新增13个问答

EMA发布《药品质量:问答(第2部分)》20261118

AI 导读

EMA于2026年11月18日更新《药品质量:问答(第2部分)》,新增13个问答,重点为仅通过体外比较质量属性证明治疗等效性时的比较方法与文件要求。适用对象为化学药品,全身作用的口服制剂不在范围内,参比批次须来自欧盟,非欧盟批次数据不得用于该工作。

推荐理由

问答逐条说明了体外比较质量属性证明治疗等效性时的批次选择、统计方法与质量标准制定要点,可对照定位申报模块3的相关章节。

正文

2026年11月18日,欧洲药品管理局(EMA)更新其《药品质量:问答(第2部分)》,新增13个问答。如下:

图 EMA官网《药品质量:问答(第2部分)》新增章节原文(2026年11月18日)

Comparison of quality attributes to demonstrate therapeutic equivalence (New Nov 2026)

通过质量属性比较证明治疗等效性(2026年11月新增)

引言与范围

In July 2021 the ‘Reflection paper on statistical methodology for the comparative assessment of quality attributes in drug development (EMA/CHMP/138502/2017)’ was adopted by the CHMP and published. The reflection paper addresses questions related to comparison objectives, sampling strategies, sources of variability, and options (or limitations) for statistical inference. It also provides more detailed guidance for how to perform the comparison task using empirical sample data and is intended to establish a common language, thereby improving common understanding.

2021年7月,CHMP通过并发布《药品研发中质量属性比较评估的统计学方法反思文件》(EMA/CHMP/138502/2017)。该反思文件讨论了比较目的、抽样策略、变异来源以及统计推断的选项(或局限)等相关问题,也就如何利用经验样本数据开展比较任务给出了更详细的指导,旨在建立共同语言,从而改善共识理解。

However, the reflection paper does not provide clear guidance on how comparison of quality attributes should be conducted and described when used to demonstrate therapeutic equivalence by in-vitro comparison only. This Question and Answer document provides such guidance.

然而,该反思文件并未就「仅以体外比较证明治疗等效性时,质量属性比较应如何开展和描述」给出明确指导。本问答文件即提供这方面的指导。

Therapeutic equivalence (TE) means that the efficacy and safety profile of the test- and reference products is sufficiently comparable so that a clinically relevant difference between products can be reliably excluded. TE can be demonstrated by clinical endpoint studies or bioequivalence studies but can sometimes also be adequately substantiated just by the comparison of the pharmaceutical form, route of administration, composition and critical quality attributes (CQAs).

治疗等效性(TE)指受试制剂与参比制剂的疗效和安全性特征具有充分可比性,以致可以可靠地排除两者之间存在具有临床意义的差异。TE可通过临床终点研究或生物等效性研究证明,但有时仅比较剂型、给药途径、处方组成和关键质量属性(CQA),也足以获得充分支持。

范围

This Q&A is applicable for the comparison of quality attributes to demonstrate TE, of chemical products, by in-vitro comparison only and setting finished product quality specifications in such context; concepts defined herein can also be used for comparisons of quality attributes for other objectives, where justified. Oral products for systemic action are excluded, as for these drug products, sufficient guidance is already available. This Q&A does not replace existing guidance; it is intended to act as a resource to identify additional considerations for a wide variety of pharmaceutical dosage forms, some of which may already be the subject of an EMA guideline. For clarity, existing EMA guidance takes precedence.

本问答适用于化学药品仅通过体外比较证明TE时的质量属性比较,以及在此情形下制定成品质量标准;文中定义的概念在有合理依据时,也可用于其他目的的质量属性比较。全身作用的口服制剂不在范围内,因为这类药品已有充分指导可用。本问答不替代现有指导原则;其定位是作为一项资源,用于识别各类剂型需额外考虑的事项,其中部分剂型可能已有EMA指导原则覆盖。明确而言,现有EMA指导原则优先。

1. 比较质量属性以支持治疗等效性的框架是什么

Define and justify those critical quality attributes that may have an impact on the efficacy and safety of the drug product and consequently need to be compared.

界定并论证可能影响制剂疗效和安全性、因而需要比较的关键质量属性。

Justify those attributes not considered critical (and therefore do not need to be compared).

对未被判定为关键的属性(因而无需比较)给出理由。

Define and justify the similarity condition for each CQA (refer to Q3 below and EMA/CHMP/138502/2017 section 4.2 for further details).

为每个CQA界定并论证相似性判定条件(详见下文第3问及EMA/CHMP/138502/2017第4.2节)。

Define the marketed reference product and characterise the reference product by testing all CQAs for a justified number of batches that together are representative of the variability observed in the reference product. Typically, this includes batches of varying ages, to also understand the stability behaviour of CQAs over the shelf life.

确定在欧盟上市的参比制剂,并对经论证数量的批次检测全部CQA,以完成参比制剂表征;这些批次合起来应能代表参比制剂观察到的变异。通常应包括不同存放时间的批次,以便同时了解CQA在货架期内的稳定性表现。

Consider the observed variability in the reference product for each CQA, including over the shelf life and storage condition of the reference product.

考虑参比制剂每个CQA的实测变异,包括货架期内以及参比制剂储存条件下的变异。

Define and justify the statistical method of comparison and associated acceptance criteria for each CQA based on the attribute, the similarity condition, and observed variability in the reference product.

根据属性本身、相似性判定条件以及参比制剂的实测变异,为每个CQA确定并论证统计比较方法和相应的可接受标准。

Justify the selected number of test- and reference batches to include in the comparison.

论证纳入比较的受试制剂与参比制剂批次数量。

Incorporate points 6 and 7 in a Quality Attributes Data Comparison Protocol together with the sampling and testing plan and the validated analytical methods (descriptions of the methods or a reference to the description of the method in Module 3). The protocol should be internally approved prior to use.

将第6、7项纳入《质量属性数据比较方案》,连同取样与检验计划、经验证的分析方法(方法描述,或引用模块3中的方法描述)。方案应在使用前经内部批准。

Test the reference- and test batches for the CQAs. The studies should preferably be conducted by side-by-side testing of products in the same analytical campaign.

对参比批次和受试批次检测CQA。研究最好在同一分析批内,以受试与参比样品并列检测的方式进行。

Evaluate the results, draw conclusions and summarise in an Comparison Report.

评价结果、得出结论,并在《比较报告》中汇总。

2. 其他规格能否豁免体外检测

No. If there are more than one strength, the in-vitro comparison with the reference product should be conducted for each strength. Where more than two strengths are proposed, a bracketing approach may be applied if adequately justified. Bracketing cannot be applied over different qualitative compositions.

不可以。若有多个规格,应针对每个规格与参比制剂开展体外比较。拟申请规格超过两个时,如依据充分,可采用括号法。定性组成不同的规格之间不得采用括号法。

3. 关键质量属性(CQA)应考虑哪些相似性判定条件

The similarity condition should be chosen based on the (potential) relationship between quality attributes and the efficacy and safety of the product. Two examples:

相似性判定条件应基于质量属性与产品疗效、安全性之间(潜在的)关系来选择。示例如下:

Similarity related to maximum allowed differences between test and reference. Similarity in mean values can be used for quantitative CQAs and is the preferred similarity condition for rate- and strength-related CQAs (e.g. assay, in vitro release test, in vitro permeation test). Next to a maximum allowed difference in means, the difference in variability (deviation) may also be part of the similarity condition of some CQAs.

与受试和参比之间最大允许差值相关的相似性。对定量CQA可采用均值相似性,这也是速率相关和强度相关CQA(如含量、体外释放试验、体外渗透试验)的首选相似性判定条件。除均值的最大允许差值外,某些CQA的相似性条件还可包含变异(偏差)的差异。

Following the principle that the entirety of reference product batches on the market represents acceptable quality, similarity based on the overlap of the test distribution with a reference range, defined based on the underlying distribution for the reference product. This may be more suitable for physico-chemical CQAs (e.g. viscosity, density, pH, surface tension) where the mean with a fixed percentage as acceptance criteria is not always scientifically meaningful.

遵循「市场上全部参比制剂批次代表可接受质量」的原则,以受试制剂分布与参比范围的交叠判定相似;参比范围依据参比制剂的潜在分布确定。对理化类CQA(如黏度、密度、pH、表面张力),这种做法可能更合适——对这类属性,以均值加固定百分比作为可接受标准未必总有科学意义。

4. 比较所用的受试批次和参比批次有什么要求

Test batches should be representative of the intended, validated commercial process, and incorporate known contributors to variability (e.g., different batches of active substance, excipients, and devices, different manufacturing sites, where applicable). Test batches should be full scale, or pilot scale if adequately justified that scale has no impact on the CQAs. Use of sub-batches is not allowed.

受试批次应能代表拟定的、经验证的商业生产工艺,并涵盖已知的变异来源(如活性物质、辅料和装置的不同批次;适用时的不同生产场地)。受试批次应为商业生产规模;如能充分论证规模对CQA无影响,可为中试规模。不得使用亚批。

Information on the countries of origin for the reference batches should be provided, and the selection of batches should be justified. Data of non-EU batches should not be used for this exercise. If trends in CQAs are observed over storage of the product, test- and reference batches of different ages, preferably covering the whole shelf-life, should be included. Reference product batches tested as part of the product characterization should also be used (re-analysed) for the comparison exercise, unless adequately justified (e.g. out of shelf-life).

应提供参比批次的原产国信息,并对批次选择给出理由。非欧盟批次的数据不得用于本项工作。若观察到CQA在贮存过程中出现趋势性变化,应纳入不同存放时间的受试批和参比批,最好覆盖整个货架期。产品表征阶段检测过的参比批次也应(重新检测)用于比较,除有充分理由(如已过货架期)外。

5. 比较工作应使用多少批次、每批多少样品

A sufficient quantity of reference and test batches, at least three but – typically 6 -10 for the reference product - should be tested to capture the full range of variability and to ensure robust statistical assessment. Where there is high variability of the CQA, analysis of a larger number of batches and/or more units per batch should be considered. For similarity based on the range of a specific attribute in the reference product, testing even larger numbers of batches is advised.

应有足够数量的参比批次和受试批次接受检测——至少3批,参比制剂通常为6–10批——以覆盖全部变异范围并保证统计评估稳健。CQA变异较大时,应考虑增加批次数或每批检测单元数。若相似性判定基于参比制剂某一特定属性的实测范围,建议检测更多批次。

The sample size and number of batches of test- and reference product required for the statistical comparison of quality attributes should be based on the reference data and the selected method of comparison. Factors to consider for sample size calculation include variability in the reference product, expected difference between test- and reference product, power analysis, and variability of the analytical procedure(s) used.

质量属性统计比较所需的样本量,以及受试、参比批次数,应根据参比数据和选定的比较方法确定。样本量计算应考虑的因素包括:参比制剂的变异、受试与参比之间的预期差异、把握度分析,以及所用分析方法的变异。

6. 何时允许合并批次结果

The variability between batches within the populations (test or reference) should be thoroughly investigated. If no statistical between-batch variability is observed, results of batches of the same population may be pooled. However, it is difficult to rule out between-batch variability especially in small datasets. Therefore, it is safer to use a statistical method for the comparison (see Q7) that does assume between-batch variability and pooling of data of batches is discouraged.

应充分考察同一总体(受试或参比)内部的批间变异。若未观察到统计学意义上的批间变异,同总体各批结果可以合并。但批间变异在小样本数据中尤其难以排除。因此,更稳妥的做法是选用假定存在批间变异的统计比较方法(见第7问),并不鼓励合并批次数据。

7. 应使用何种相似性方法和可接受标准

The comparison should be based on sound statistical methodologies that are suitable for the in-vitro data to be collected. Therefore, detailed justification of the statistical method and acceptance criteria used should be provided and will be assessed case-by-case. If prior knowledge is available, that can and should be used. Failure to demonstrate a difference should not be interpreted as proof of similarity. The method should be suitable to demonstrate similarity.

比较应基于适合所采集体外数据的、可靠的统计学方法。因此,应对所用统计方法和可接受标准给出详细论证,并由审评方逐案评估。若已有先验知识,可以且应当加以利用。未能证明存在差异,不应被解读为相似性的证明。方法应适合证明相似性。

The following information should be presented, at a minimum:

至少应呈现以下信息:

the specific statistical methodology to be applied, along with a comprehensive justification for the use of that methodology, including any limitations.

拟采用的具体统计方法,以及使用该方法的完整论证,包括其局限。

the circumstances under which the specific statistical methodology will be used.

采用该统计方法的具体情形。

how the statistical method will be used, including why the generated data is suitable for evaluation using the chosen statistical methodology.

统计方法如何使用,包括为何所生成的数据适合用所选统计方法评价。

a discussion on risk of false positive conclusion and false negative conclusion

关于假阳性结论与假阴性结论风险的讨论。

any literature references.

相关文献。

In this context it is remarked that the acceptance range might be widened depending on the reference product variability, which overall includes the sample and batch variability and the variability of the analytical method applied (as per validation outcome, see Q9) and that the variability of the analytical method applied is expected to be state-of-the–art.

在此需说明,可接受范围可随参比制剂的变异而放宽。该变异总体上包括样品变异、批次变异以及所用分析方法的变异(依验证结果,见第9问);所用分析方法的变异应处于当前最先进水平。

Three known approaches (A, B, and C) are listed, but do not exclude other approaches:

文件列出三种已知路径(A、B、C),但不排除其他路径:

Similarity related to maximum allowed differences between test and reference

路径A:与受试和参比之间最大允许差值相关的相似性。

Comparison may be based on comparison of arithmetic means or comparison of geometric means.

可基于算术均值比较,也可基于几何均值比较。

Similarity related to other approaches

路径B:基于其他路径的相似性。

For other approaches, such as methods and criteria based on a range observed in the reference product, a scientific advice from a competent authority may be considered. In the specific case that the range in the CQAs observed in the test product is fully within the range observed in the reference product, and that range observed in the reference product is applied as in-process control or shelf-life specification of the test product, where relevant to control these CQAs in the test product, similarity might be adequately addressed for that CQA.

对于其他路径,例如以参比制剂实测范围为基础的方法和标准,可考虑向监管机构申请科学建议。有一种具体情况:若受试制剂CQA的实测范围完全落在参比制剂的实测范围之内,且参比制剂的实测范围被用作受试制剂的过程控制标准或货架期标准(在需要以此控制受试制剂该CQA时),则该CQA的相似性可视为已获充分说明。

Similarity for specific types of products

路径C:特定类型产品的相似性。

Similarity methods and the associated acceptance criteria may have already been established in case of specific dosage forms or product types: in such a case reference is made to the published guidance documents.

对于特定剂型或产品类型,相似性方法及相应可接受标准可能已经确立;此时应参照已发布的指导原则文件。

8. 放行标准和货架期标准应如何制定

Where therapeutic equivalence is based solely on a quality comparison against the reference product, and the quality of the test product should therefore be similar to that of the reference product, the specifications for the test product should also be based on the ranges observed in the reference product tested with the validated analytical methods (usually those found in 3.2.P.5.2). Differently from the comparison exercise (where sample results can be used for assessment of similarity), for the purpose of setting specification, it is the range of batch results (of the reference product) to be used. Obvious outliers should be disregarded. The specifications of the test product should normally not be outside the ranges observed for the reference product. Additional testing of more reference batches may be required to meet this expectation.

当治疗等效性完全建立在与参比制剂的质量比较之上、因而受试制剂的质量应与参比制剂相似时,受试制剂的标准也应基于参比制剂用经验证分析方法实测所得的范围(通常即3.2.P.5.2中的范围)。与比较环节不同——比较时可用样品结果评价相似性——制定标准时应使用参比制剂批次结果的范围。明显离群值应忽略。受试制剂的标准通常不应超出参比制剂实测范围。为满足这一要求,可能需要补测更多参比批次。

If the range of the reference product is much wider than the range of the test product, this does not by default allows a very wide specification range for the test product. If the variability in the test product is less, from batch-to-batch quality control point of view, the specification of the test product should be tightened based on the process capability of the test product.

若参比制剂的范围远宽于受试制剂,这并不当然允许受试制剂采用很宽的标准范围。若受试制剂的变异更小,从批间质量控制的角度,应依据受试制剂的工艺能力收紧其标准。

Specifications for impurities in the test product may be assessed on their own merits, unless these impurities have a clear relation with efficacy and safety of the product, e.g. free iron in iron carbohydrate colloids may impact therapeutic equivalence.

受试制剂的杂质标准可单独评估其合理性,除非这些杂质与产品疗效、安全性有明确关联——例如碳水化合物铁胶体中的游离铁可能影响治疗等效性。

9. 分析检验方法应描述和验证到什么程度

For methods that are used for in-process controls and drug product release- and shelf-life testing, reference can be made to the corresponding Module 3 sections. Where other methods are used, these should be described and suitably validated. The extent of validation should be justified. . Where relevant, the discriminative power of the analytical methods to differentiate between acceptable- and not acceptable products should be substantiated.

用于过程控制和制剂放行、货架期检验的方法,可引用模块3的相应章节。使用其他方法时,应对这些方法进行描述并适当验证,验证程度应给出理由。适用时,应论证分析方法区分可接受产品与不可接受产品的区分能力。

10. 批次生产和批次检验设施有什么法定要求

Sites used for testing comparability and the release of the test batches used in the comparison should be defined. Comparison testing should be conducted by sites that are approved for GMP release testing, or it should be declared that the performing laboratory is qualified to undertake the studies and that an effective quality system is in place.

应明确用于可比性检验的场地,以及比较所用受试批次的放行场地。比较检验应由已获准开展GMP放行检验的场地实施;否则应声明实施实验室具备开展该研究的资质,且已建立有效的质量体系。

Test batches should be manufactured under GMP.

受试批次应在GMP条件下生产。

11. 结果应如何呈现

A Comparison Report should be provided; sample data should be provided in an annex to this report. Additionally, the report should include the following:

应提交《比较报告》,样品数据作为该报告的附件。此外,报告还应包括以下内容:

Introduction and Scope

引言与范围。

Statistical methodology (including software version) and associated (acceptance) criteria

统计方法(含软件版本)及相应的(可接受)标准。

Summary of the data, including any calculations, diagrams, figures, plots and other relevant methods to present (summary) data.

数据汇总,包括任何计算、图表、图形、绘图及其他呈现(汇总)数据的相关方法。

Discussion and conclusion for each CQA compared, and overall conclusion on equivalence from a quality perspective.

每个所比较CQA的讨论与结论,以及从质量角度对等效性的总体结论。

12. 如何比较不可定量的CQA

A quantitative comparison should be provided for all CQAs. If this is not possible, a justified approach to support similarity is required, e.g. photographs of microstructures or morphological forms.

所有CQA都应提供定量比较。确实无法定量时,需给出有依据的支持相似性的方法,例如显微结构或形态学照片。

13. 该项工作应在申报资料的哪一节描述

All documentation on the comparison exercise, including the Quality Attributes Data Comparison Protocol and Comparison Report, should be included in Module 3 (3.2.P.2.3). The summary with conclusion together with reference to the documentation in Module 3 should be included in Module 5.

比较工作的全部文件,包括《质量属性数据比较协议》和《比较报告》,应纳入模块3(3.2.P.2.3)。带有结论的摘要连同对模块3文件的引用,应纳入模块5。

来源:制药经理人 · mp.weixin.qq.com