USP生物测定章节<1032>与<1034>最终版预计明年4月生效

As early as September 2024 and in January 2025 we reported in our newsletter on the draft USP chapters on bioassays. According to the latest information, the final chapters <1032> and <1034> are due to become official in April next year. Objectives and backgroundUSP chapters <1032> and <1034> address two related steps in relative bioactivity assays: developing a suitable method and statistically analysing its results. Bioassays determine the biological activity of a medicinal product using biological systems and typically compare a test sample with a standard.The background is the natural variability of biological systems. A potency value expressed relative to a standard is therefore generally more meaningful and less variable than an absolute measurement. However, it is useful only if the test sample and the standard are sufficiently comparable biologically. Statistical similarity between the response curves supports this assumption, but does not prove it on its own.Content of the two chaptersUSP <1032> - Design and Development of Biological Assays describes how to plan and develop a bioassay so that it is fit for its intended use. Performance requirements should be derived from the assay&rsquo;s intended use and integrated into the method lifecycle. The chapter discusses how experimental design, randomisation, and consideration of groups or systematic effects can help limit bias. It also covers development, qualification, validation, ongoing monitoring and change management, and includes practical examples of experimental designs.Where possible, the assay should reflect the product&rsquo;s biological mechanism of action. If this is not possible, its relevance may be supported by relationships to other assays or to nonclinical or clinical outcomes. The choice of assay involves balancing biological, statistical and practical considerations.USP <1034> - Analysis of Biological Assays explains how to analyse bioassay data statistically to determine relative potency and uncertainty. It links dose-response models to criteria for system and test-sample suitability, similarity assessments, outlier handling, and the calculation of potency and confidence intervals across individual and repeated runs.Similarity should not be assumed merely because a result is not statistically significant. Equivalence testing with scientifically justified limits is recommended. Outliers should be investigated rather than automatically removed solely on the basis of a statistical test. The chapter also considers group outliers and their impact on the analysis. Comparison with the first draftThe development objectives described in our news items cited above about the first draft are largely reflected in the current revised versions. Particularly notable are the stronger integration into the method lifecycle, the alignment of performance requirements with intended use, and the addition of practical experimental-design examples.The previously announced closer alignment is also evident: <1032> links development, suitability and monitoring with the analysis described in <1034>. At the same time, the USP recommendations are aligned with ICH Q14 and Q2(R2).The points previously raised concerning systematic errors, groups and outliers have also been made more specific. The design should account for sources of shared variation through randomisation, blocking and appropriate modelling of fixed and random effects. Ongoing monitoring should detect trends and drift. Outliers should be investigated rather than removed automatically; in addition to individual observations, <1034> also addresses group outliers and refers to the general principles in <1010>.The rationale for equivalence limits has been clarified: equivalence testing is recommended, and the limits should reflect whether a deviation would materially affect the bias or interpretation of the potency estimate. There is therefore no single, universally applicable limit for all products; limits must be scientifically justified for the specific method and product.Log-transformed relative potency is relevant when combining results and when making changes to or comparisons between methods. Weighted models are not ruled out for handling variance, but transformation of the measurements is highlighted as the preferred approach; <1034> does not treat weighted analyses as the standard approach.The considerations raised in 2025 concerning novel products, such as cell and gene therapies, remain relevant. However, the chapters do not prescribe a single &ldquo;matrix solution&rdquo; for these products. Instead, they emphasise the product-specific mechanism of action, the possible need for multiple or orthogonal test methods, and a robust link to the intended function.ConclusionThe general direction of the revisions, as described on the basis of the draft, remains valid. The current versions have developed these objectives into a more lifecycle-oriented and practically implementable framework. The aim is for <1032> to enable the development of a bioassay that is fit for purpose and protected against systematic effects, while <1034> is intended to ensure a statistically transparent analysis, including equivalence and suitability assessments.In practice, this means that limits and controls must be justified on the basis of product biology, intended use and data. A generic potency measurement alone is not sufficient, particularly for complex biologics.Further information on the chapters is available on the USP-NF webseite following paid registration.