Is Tech Transfer Broken? Can the Principles of QbD Fix It?

While scant data exists to suggest how efficient, or inefficient, technology transfer is within drug manufacturers' development efforts, it's pretty clear that most companies hardly have tech transfer down pat, and still others just don't get it at all. That's a recurring theme in a podcast posted on PharmaManufacturing.com yesterday, in which tech transfer experts Stephen Perry, Russ Somma, Emil Ciurczak, and Paul McKenzie offer harsh criticisms of the tech transfer projects they've witnessed first hand. And thankfully they offer advice as well.

Sometimes the problems are simple--development teams just fail to write critical information down in their notebooks, "something they should have learned in college!" Ciurczak says. But the experts also make clear that practices associated with Quality by Design--namely, proactive risk assessment and data management--can eliminate problems in the early stages of development and facilitate later transfer. Perry notes how developers typically send data on successful process studies to their partners, but not data from failed studies that can be just as enlightening.

Perry is the author of Pharmaceutical Manufacturing's January cover story on doing tech transfer the right way, and he suggests that one thing that often dooms tech transfer projects (internal or external) is the inability of the technology sender to share necessary information with its partner:

There is no such thing as “too much” or “too detailed” information. Such dedicated tools as assay summaries, detailed process descriptions, and bills of materials should be created and used . Other information from the sender, such as regulatory submissions, development reports, raw data, validation reports, etc., can be extremely useful. This work would have to be done anyway as part of the future process validation and to support regulatory submissions.

One of Somma's mantras is simplifying tech transfer, which seems to run counter to the idea of providing as much detailed information as possible. He, as well as Oracle's Arvindh Balakrishnan, explain how this paradoxical challenge can be overcome in a podcast which is now the lead feature on our PharmaQbD.com site.

Finally, Centocor's McKenzie is a proponent of recipe-based development as a means of ensuring consistency throughout scaleup and as handoff occurs between partners. Here is a summary of McKenzie's philosophy from last fall's Bioprocess International conference, as well as a blast from the past: Pharmaceutical Manufacturing's September 2007 cover story featuring McKenzie in his days with BMS.

The success or failure of tech transfer is hard to measure, but that doesn't mean that it shouldn't be a key focus of initiatives to accelerate and lessen the costs of drug development.

Visualizing the Golden Batch with S88

ISA-S88 is a standard for batch control that many proponents of Quality by Design are adopting. S88 is a design philosophy for processes and recipes (see here for Wikipedia’s pretty good summary) that is becoming more and more relevant in the pharmaceutical industry.

One S88 believer is Centocor’s Paul McKenzie, formerly of Bristol-Myers Squibb. McKenzie, who spoke at the Bioprocess International Conference last fall, explains the value of S88 as follows (as related by Agnes Shanley, who was in attendance):

Each process will have its own raw materials, quantities, process parameters, recipe procedures, and equipment requirements. . . . With a portable recipe, equipment is separated out as its own class, and you assign an equipment class to each area, a good approach for when a process is changing, McKenzie said.

The approach allows users to create a series of recipes. Everyone collects data in the same way, he explained, so that tech transfer becomes ownership of that recipe. As it moves along, you have a master recipe and then a control recipe specific to each actual unit you run. . . .

This in turn facilitates a QbD approach.

“I don’t want people innovating on what they call something in a MS word doc,” McKenzie said, but in terms of parameters, using a common global language to describe process and product.” With this approach, one recipe can be used across the world and global facilities, he said. The method makes it easy to link to regulatory filings and to visualize the “golden batch.”

For the full summary of McKenzie's talk, see "Paul McKenzie's Recipe for Success with QbD".

QbD Presentations on Parade: DynoChem's User Meeting

DynoChem has made available presentations from its user group this past May. Registration is required to access the content, but among the materials available are:

Roles of Mechanistic and Empirical Modeling/DOE in Achieving Quality by Design, by Paul Stonestreet of GSK

Practical Insight on New Model Development: Filtration and Centrifugation, by Rich Ballenger of Abbott

Process Modeling Based Approach Towards Quality by Design for an API Synthetic Step, by Shawn Brueggemeier, Emily Reiff, Olav Lyngberg, Lindsay Hobson, and Jose Tabora, BMS

Practical Aspects of Distillation Modeling in DynoChem, by Carolyn Cummings of Amgen

An Example from GSK's Design for Manufacture Initiative: Use of Dynochem in Conjunction with Lab and Pilot Scale Data to Advance Process Understanding, by Dharmesh Bhanushali of GSK

How Process Safety and Environmental Lab Can Guide Process Development, by Viviane Massonneau of Merck

Lean and Green: The Value of API Process Design, by David am Ende, Pfizer

--Paul Thomas