Setting Specifications for a Biotech Therapeutic Product in the Quality by Design Paradigm

Manufacturing using meaningful, science-based specifications will ensure that we attain the optimal balance between manufacturing flexibility and product safety.
Jan 01, 2010


Enhanced process and product understanding are the basic tenets of Quality by Design (QbD). A QbD approach for setting specifications would involve harnessing this understanding not only from clinical and nonclinical data available for the product but also from other similar products. Manufacturing using meaningful, science-based specifications will ensure that we attain the optimal balance between manufacturing flexibility and product safety.

The safety of biotech therapeutic products is paramount to their successful commercialization. Concerns that have been frequently cited include adulteration, changes to product quality over its lifecycle, changes to product quality during distribution, complexities of biotech processes, potency, stability, and environmental impact.1 Product specifications have long been regarded as a safeguard with respect to product safety. They have been defined as "a list of tests, references to analytical procedures, and appropriate acceptance criteria which are numerical limits, ranges, or other criteria for the tests described."2 In traditional manufacturing, they have been regarded as the final hurdle that a manufacturing lot must overcome before its release for commercial use. However, in the Quality by Design (QbD) paradigm, they are one part of the overall control strategy that has been designed to ensure product quality and consistency.3,4

This article is the 19th in the "Elements of Biopharmaceutical Production" series and will present an approach for setting specifications for biotech therapeutic products following the QbD principles.


Regulatory procedures that influence specification setting for biotechnology-derived protein products, including the International Conference of Harmonization (ICH) guidelines, have been reviewed in the literature.5,6 In traditional manufacturing, specifications were set based on the small number of large-scale batches that had been manufactured before filing for regulatory approval. Each batch was then tested against those specifications to ensure product safety. In the QbD paradigm, however, prior knowledge can play a major role in setting specifications. This may result in broad specifications for attributes whose relationship to product safety and efficacy is well understood to not be significant through product-specific or platform data and relatively narrow specifications for attributes for which the impact on safety and efficacy is not fully understood or is found to be significant.7,8 Unlike in traditional manufacturing, product specifications under QbD are solely for confirmation of product quality because the process control strategy ensures that the specifications are met.

The concept of a clinical design space can be used to quantify the clinical experience with a product.3 This would be in the form of a multidimensional design space with each critical quality attribute (CQA) serving as a dimension. The size of the clinical design space for a given product will depend on the number of lots put in the clinic, the availability of applicable data from other similar products, and the extent of product heterogeneity that has been introduced during the clinical trials. The clinical design space is expected to be limited in the early phases of clinical development when only a few lots have been introduced into the clinic, but then would grow as the product reaches an advanced stage of product development and more clinical data become available.

The design space concept also can be extended to product quality.3 Similar to the clinical design space, a design space would also be multidimensional, with each CQA serving as a dimension. The product design space will be documented in the regulatory filing in the form of in-process controls and drug substance and drug product specifications that define the acceptable variability in CQAs. The size of the product design space for a given product will depend on a multitude of factors, including:

  • the robustness and capability of the manufacturing process, as determined by performance at small and large scales as well as the design of experiments (DOE) studies performed at small scale9,10
  • the stability of the drug substance and drug product, as determined by the stability studies
  • clinical data from clinical studies that have been performed with the product and other platform products
  • nonclinical data with the product and other platform products (such as data from binding assays, cell based assays, and in vivo assays)11–14
  • the capability of the analytical methods that are being used to test for the specifications8
  • the level of understanding of the impact of the CQAs on safety and efficacy of the product.3,4,7

Table 1. Typical release tests used for monoclonal antibody products.15 Also shown are mock specifications and data for five lots used in the clinical trials.
Table 1 presents a subset of the typical release tests that have been reported in the literature for monoclonal antibody (MAb) products.15 The tests include indicators of quantity (protein concentration), purity (chromatography), identity (electrophoresis, peptide mapping), potency (antigen binding assay), impurities (host cell proteins [HCPs], nucleic acids, endotoxins), and other general properties (pH, volume, appearance). The table also presents proposed specifications for a mock MAb product along with release data from five lots that have been put into the clinic. Per the discussion above, the product design space will be defined through the specifications submitted in the regulatory filing.