Staffing, Training and Validation Are Critical to PAT

March 31, 2005
FDA’s involvement has convinced pharma to embrace PAT, but new recruitment and validation issues have arisen as PAT moves into its next phase.
By Paul Thomas, Managing EditorMany manufacturers are “doing” PAT these days, but few individuals truly have a grasp of what it means and what its potential is in the pharmaceutical industry — for process monitoring, and design and control as well.One who knows PAT inside and out is Bruker Optics’ VP Dan Klevisha. Klevisha has been trying for years to convince pharma firms of the merits of PAT. Only since FDA put its full weight behind analytical technologies like those Bruker offers — such as process monitoring IR, NIR and Raman — was he able to make significant headway. Seventeen of the top 20 pharmaceutical firms are now using Bruker for some sort of PAT application, he says.I sat down with Klevisha at last month’s Pittcon show in Orlando, and picked his brain about PAT and where it’s headed.
Bruker developed its Tandem on-line NIR device for monitoring tablet uniformity at a major pharma firm's request; there was no market for such an instrument before FDA began promoting PAT.

What is different about PAT these days? I asked. Unlike a few years ago, he said, every firm has committed substantial resources and has a defined organizational strategy to approach PAT. “The manufacturing segment of pharma has finally accepted PAT for real,” he said.Some firms have been vocal about their PAT efforts, others not so. While he won’t name names, Klevisha says some of the firms leading the pack in terms of progressive PAT strategies are those we hear least from.What’s the greatest risk firms face in implementing PAT? There are two, Klevisha said. First, “they have to make sure that they have a sufficient staffing commitment to support the deployment of a PAT system. This includes having people qualified to transfer that knowledge to the manufacturing staff.” There’s a very long learning curve for PAT applications, he added, and firms need to commit to thorough training programs for personnel.Finding experts to train the manufacturing ranks is not easy, he said. “Process analysis experts from chemistry, defense, and telecom are in heavy demand,” Klevisha said. “The challenge is that these individuals don’t have experience in a pharmaceutical environment.”The second greatest risk is validation of PAT-enabled processes. “The industry has a good sense of validation and equipment qualification,” he said. “But that has to be rethought for PAT. Traditional validation processes are too time-consuming and expensive.” Costs of PAT equipment, in fact, can be dwarfed by the cost of validating its use, Klevisha said. Understanding the science behind PAT will be key for firms to solve validation issues.An understanding of processes grounded in science is, of course, what FDA is looking for. I asked Klevisha what FDA’s PAT initiative has meant to his and Bruker’s business. Bruker has focused on PAT instrumentation for more than 12 years, he noted. But for most of those years, few pharmaceutical firms expressed even the slightest interest in process analytical technologies and systems. As a result, Bruker had no need to pursue technologies that it had within its grasp.That has changed dramatically since the initiative. Klevisha used the example of the company’s Tandem on-line NIR device for monitoring tablet uniformity. After FDA’s change of course, a major pharmaceutical firm asked Bruker to develop the device. “Without PAT, we would not have developed it,” he said. “There would have been no market.”Where is PAT headed in the next two or three years? "In that time, we’ll have a firmer understanding of successful applications," Klevisha said. Right now, comprehensive PAT strategies are just getting under way and offer no real lessons to speak of. With trial and error will come experience. For Bruker, Klevisha believes, that will translate into better and less expensive tools and solutions, and easier and faster deployment.When payback is realized, investment will increase, he continued. “Firms will demand robust, single-purpose, application-specific devices,” he predicts. One company in another industry has already implemented more than a hundred different Bruker devices, Klevisha noted. Eventually, pharmaceutical firms will have similar needs.