Technology transfer is what keeps pharmaceutical innovation alive. Yet, at many companies, this strategically important function is still handled in a very random way. At worst, projects are abandoned outright; at best, weeks, months, and millions of dollars in development costs are lost.
We read of universities and companies failing to transfer technology at the very start of the value chain. But later, efforts can also falter when moving a process from R&D to scaleup, or from an operating company to a contract manufacturing partner.
Tech transfer is very complex, but the problems that impede it often aren’t. First, there can be a failure to plan, to designate “project managers” to function as central communications points.
Companies can also fail to distinguish clearly between process development functions and tech transfer. In such cases, the tech transfer team may be stuck sorting out issues that should have been resolved earlier.
In other cases, some members of the development team may hold on to vital information, either thinking it’s not important, or planning to provide it “after everything is all wrapped up.” In this case, the development partner (internal or external) has to guess at information that should have been provided.
There may also be a tendency to move too quickly to process issues, instead of dealing with analytical assay transfers, one of the major reasons for tech transfer failures in biopharmaceutical manufacturing.
No Such Thing as Too Much Information
With tech transfer, there is no such thing as too much relevant information, says Stephen Perry, CEO of Kymanox, a company that specializes in facilitating tech transfer from operating companies to CMO’s. Even if the information is rather raw, he says, it may prove useful to your tech transfer partner.
Mr. Perry recalled a situation involving a biopharm company that had transferred analytical assays to a CMO in another part of the world. The company did not provide all the information, and sent data, piecemeal (and some of it out of date) to different people.
As a result, the CMO spent months trying to qualify the assays but couldn’t.
Eventually, a face-to-face visit was needed, and partners learned that one assay required a very specific reagent made by only one manufacturer in the world (who, it turned out, had stopped selling it). By the time the meeting took place, six months had been lost.
There may also be a tendency to “edit” the information sent to tech transfer partners, Perry says. Assay summaries and BOM’s (including compendial status) are all important, he says. Another key is to transfer analytical assays before the process, and to include sample storage and stability data.
Trust, But Verify
Small-scale verification should be confirmed by CMO’s at the bench scale before scaleup and large production. This is particularly important for supporting time, temperature and pH studies, Perry says. All too often, he says, scaleup is done prematurely, using less-than-robust models.
Much of what is easy to do in a lab will be just the opposite in a manufacturing setting. Consider mixing and pH adjustment. Don’t skip the dull work (for instance, assessing hold times) because it’s critical for the other side.
And, Perry advises, never skip pre-GMP engineering runs, even if you are using very similar equipment and conditions. “These are dress rehearsals for GMP. They allow you to ensure that batch records and SOPs are correct, and they facilitate operator training,” Perry says. Depending on the level of risk entailed, he recommends between one and six runs.
Some of the problems with pharma tech transfer may be traced back to information and documentation. Most tech transfer projects today are handled in a very document-centric way, explained Paul McKenzie, VP of Centocor, in a recent PharmaManufacturing.com webcast. As a result, analytical assays lag significantly behind the processes they are meant to support.
Everyone involved in the project, across functions, shares the same information in documents, but each reflects the person’s individual function and preferences. McKenzie advocates a move to standards, and defining data in terms of recipes, so that each group is dealing with a single view of the world.
That vision may still be a long way off from today’s pharma tech transfer reality. In the meantime, a simple focus on transferring adequate information, in the best form and at the right time, will go a long way to improving results.