The Evolution of Quality by Design (QbD) for Biologics

In 2008, the U.S. Food and Drug Administration (FDA) launched a quality by design (QbD) pilot program for biotechnology products (1, 2), seeking volunteers from pharmaceutical companies to participate in a pilot program involving the submission of quality (chemistry, manufacturing, and controls) information for biotechnology products. Since this time, a number of guidance documents on the subject have been published, including the Chemistry, Manufacturing and Controls (CMC) Biotech Working Group A-Mab (monoclonal antibody) (3) and the CMC Vaccines Working Group A-VAX (vaccine) (4) case studies (available from the California Separation Science Society/International Society for Pharmaceutical Engineering (CASSS/ISPE) and the Parenteral Drug Association (PDA) websites, respectively) that provide practical examples from industry consortiums illustrating the application of QbD for a representative A-Mab and vaccine. It has been reported that at least two products submitted by Genentech and Roche as part of the FDA biotechnology pilot program have since been approved (5).

But the application of QbD for a biotechnology product presents a number of potential compliance challenges, as scaled-down laboratory or pilot-scale studies supporting a QbD-based manufacturing process are often undertaken within research and development laboratories (6). Design space studies are typically undertaken throughout the stages of product development, and when combined with risk assessments, three-dimensional design space models and a perceived sprinkle of statistical assumptions may present a challenge for both regulatory assessors and good manufacturing practice (GMP) inspectors because an assessment of process robustness ultimately requires an evaluation of small-scale development and characterisation data. In addition, the subject matter experts responsible for the small-scale studies and characterisation reports that support the proven acceptable ranges for unit operations may not be based at the site used for commercial manufacture. Process ranges are also not typically varied significantly when demonstrating manufacturing consistency, with process validation batches undertaken within an established normal operating range often at a midpoint or set point (7), and therefore the robustness of the design space and manufacturing process is not routinely challenged at commercial scale. It is assumed that operating within a proven acceptable range is considered to have no product quality impact and does not require regulatory approval if supported by design space studies (8).

But a joint FDA/European Medicines Agency (EMEA) question and answer (Q&A) document published on October 24, 2013 (9) provides a reflection of the EMEA's and FDA's position on the subject of design space verification. It appears that QbD is evolving—industry will now be expected to undertake design space verification studies at commercial scale over the lifecycle of the product and process, particularly when an applicant cannot demonstrate that the design space is scale-independent. The publication also indicates that movement from one area to another within a design space (for example to establish a new normal operating range) that is not anticipated to require regulatory pre-approval may pose higher or unknown risks due to potential scale-up effects and/or model assumptions. Although the Q&A document was not specifically written for biotechnology products, the document states that the principles laid down for chemical products are also applicable to biological products. But design space verification should not require the verification of entire regions of the design space, or verification of the edges of failure for unit operations. Instead the large-scale approach should be guided by risk assessment, focusing on parameter criticality, robustness, and the potential effect of scale-dependent parameters on the product. It is now European authorities' expectation that a protocol for design space verification be submitted as part of the quality section (3.2.R) of the common technical document, which typically includes details of validation studies that may not have been completed but are intended to be conducted (10). At the time of submission, a proposed design space that is not verified at commercial scale should now be accompanied by an appropriate verification protocol. The FDA Guidance for Industry (7) refers to continued process verification as an ongoing program established after performance qualification to collect and analyse data to ensure the process remains in a state of control. With the appropriate planning it may be possible to design a comprehensive continuous process verification program to include the evaluation of parameter ranges and other studies at process scale to support manufacturing robustness, such as evaluating the impact of cumulative worst-case processing conditions for critical steps.

The regulatory expectation to undertake design space verification at commercial scale will also serve another purpose—the transfer of ownership for critical or scale-dependent process ranges from small-scale development laboratories to commercial-scale manufacturing, un-burdening development groups from the continued ownership of the supporting data. The FDA's expectation is that such plans for design space verification are available at the manufacturing site. Design space verification during the lifecycle of the product will also ensure that potential process robustness issues observed will be captured as part of a site quality system. Manufacturing deviations, non-conformances, investigations, out-of-specification results, and subsequent corrective and preventative actions that occur when process ranges are evaluated within a GMP production facility will be documented and may be reviewed by product, by batch, or included as part of a site quality metric, providing production-scale information to assess whether the design space established in the laboratory is indeed representative of the process operated at commercial scale. The Q&A also indicates that if verification data proves that the extent of movement within the design space is considered high risk (critical quality attributes are met but are close to the edge of failure identified at laboratory- or pilot-scale), then process validation in the new area of the design space should be considered.

It would appear that QbD for biologics is likely to evolve further over time as regulators begin to evaluate data submitted as part of QbD pilot programs. QbD is considered by many as an effective method for developing new products and processes, enabling effective technology transfer and the optimization and improvement of existing processes. The application of QbD also provides enhanced process and product knowledge that may improve commercial success rates and to support platform approaches for process development. Some biotechnology companies have been using QbD principles to support platform processes for many years reducing both cost and time, but for other manufacturers QbD is currently perceived to add less value, with studies associated with additional development time and upfront investment.

• © PDA, Inc. 2014