By Paul Thomas, Senior Editor

Think QbD is daunting? Double that for the biotech arena, where the size and complexity of molecules and the difficulty in defining the attributes that determine quality and safety mean that QbD boils down to a science of educated guesswork.

For this reason, biopharmaceutical implementations of QbD have lagged behind those of small molecules. But, says Lynne Krummen, Genentech’s senior director of regulatory affairs, this is finally changing. The bio-QbD conversation is now moving from “one of concept to one of implementation,” she says.

Krummen and Pfizer research fellow Amit Banerjee highlighted a May 19 session at the  BIO conference in Atlanta on approaches to Quality by Design for biopharmaceuticals.

Banerjee began by outlining some of the challenges of applying QbD to biologics:

• The inherent complexity of biologics presents challenges to using the principles of a Design Space.
• The link between a product attribute and clinical performance is not often well understood in biologics.
• It is difficult to identify Quality Attributes that are truly critical to safety and efficacy, and their acceptable boundaries.
• Regulatory agencies tend to focus on process consistency and clinical experience, which are often not as well defined in bio.
• It is hard to precisely measure identity and purity in biologics.
• There is a challenge to define a multidimensional design space and to generate functional relationships between quality attributes and process parameters.
• The agency is constantly challenging companies to provide scale-up/scale-down models.

Banerjee went on to outline how Pfizer’s Right First Time (RFT) approach to manufacturing has facilitated Quality by Design projects for RFT’s emphasis on:

• Risk-based and science-based decision-making
• Process understanding to create a design space
• Multivariate experimentation, predictive modeling and efficient experimental design
• Increased transparency with regulators
• Process robustness (capability)

No Easy Task

Despite the synergies of the two philosophies, Banerjee admitted that applying QbD to Pfizer’s biopharma operations has been tough. He used process scale-up/scale-down as an example. “We’ve seen where we model something and it tells us to use a certain set of equations, but when you do it in real life it turns out different than we expected,” he says. Still, he says, it’s important to carefully identify parameters that are affected by scaling up, to use experimentation, simulation and prior knowledge for greater process understanding, and to fine-tune the effects of scale-up or scale-down as more knowledge is gained.

It’s an ongoing challenge, Banerjee admitted. “Is there a pot of gold?” he asks. “We’ll find out.”

Krummen’s talk illustrated just how far bio-QbD has come. She outlined results and lessons from two industry-regulatory collaborations that “will go a long way towards shaping how QbD projects will be organized and used to support product filings with FDA and other regulatory agencies.”

The first is the FDA Pilot Program for QbD-based submissions of biotech products. A similar pilot for small molecules has paid significant benefits. The second is the continued work of the CMC Working Group of industry experts that was responsible for the development of last year’s “ACE Tablet” case study and is drafting a second, more sophisticated case study for publication later this year.

First, Krummen outlined some of the external challenges that biomanufacturers face in addition to those product- and process-related internal ones. It’s difficult to reach agreement on QbD strategies and practices within one company, she says, much less within one industry, within regulatory agencies, and across global regulatory bodies. The work of ICH (the International Council on Harmonization) is key in addressing these challenges and enabling QbD for biopharma, she says.

There are positive signs, she says, that have shifted bio-QbD towards an implementation phase. There has been unprecedented openness on the part of FDA and EMEA to discuss concepts and implementation options, she says. Multiple agency-industry forums have made real progress.

And even though regulatory guidelines for Quality by Design are technically “optional” for industry, “it is becoming more and more apparent that sponsors will need to show elements of QbD in their applications,” Krummen says. Another positive: The EMEA is revising its variations guidance to include the concept of Design Space, and has adopted a new monoclonal antibody (MAb) guidance that allows the use of platform knowledge to support license changes.

Genentech’s Pilot with FDA

Krummen shared information from Genentech’s participation in FDA’s pilot program for biologics QbD. The company has gotten two proposals — one for a new product, one for an expanded change protocol on an approved drug — accepted into the program, though it will be some time before it files for approval with the Agency, she says.

Regarding the drug for the new biologics license application (BLA), Krummen did not share proprietary details but did say that it is MAb-based and that Genentech is leveraging its history and development of MAbs. It is also in the process of applying risk assessment tools to identify Critical Quality Attributes (CQA’s), identify process parameters, Critical Process Parameters (CPP’s) and Design Space, and leveraging the above to develop an appropriate process control strategy.

For the change protocol, Genentech is proposing an “expanded comparability protocol,” a mechanism that FDA has established (and is under consideration in the E.U.). It involves working closely with regulators in the U.S. and E.U. on Design Space issues. Krummen also stated that the manufacturing facility that will eventually receive the product has previously been inspected by FDA and is in cGMP compliance.

A major challenge: Execution will be critical, Krummen says. “You can’t jeopardize approval timelines, and we’ll have to figure out innovative ways to work with regulators to make sure there will be no surprises at the end of the day.” Genentech will also, ambitiously, submit approval filings simultaneously with FDA, and in the E.U. and other ICH regions.

The “A-MAb” Case Study

Next, Krummen discussed the efforts of the CMC Working Group, comprising representatives from her company, as well as Pfizer, GlaxoSmithKline, Eli Lilly, MedImmune, Amgen, and Abbott.

The group’s current case study is based on a fictitious antibody drug substance and drug product call “A-MAb”, Krummen noted. The case study is being designed to illustrate key principles of QbD beyond those presented in the ACE study, and will be available as a teaching tool for industry and agencies.

Why mAb? In order to exploit the large amount of product and process knowledge that’s already available, she says. The goals of the study are to provide:

• Examples of CQA’s
• Various ways to treat data
• Information on how to categorize risk
• Examples of how a design space can be articulated in the context of a filing a risk-based, lifecycle approach to managing changes to DS/DP

“It’s not a filing ... not intended to be the gold standard,” Krummen says. “It’s structured like a book chapter; it will be cut down and showcase different tools, different approaches and different levels of data.”

The case study is expected to be published late this summer, she says.

Banerjee and Krummen say that these developments are the tip of the iceberg. “In the future, all filings will be ‘QbD filings’ or at least have an element of QbD,” Banerjee says.