It’s time for pharmacometricians and statisticians to leverage their expertise and break down drug development barriers.

By Paul Thomas, Senior Editor

What happens when everybody (or at least the relevant people) in a pharmaceutical organization has learned how to do modeling and simulation? This is the premise that Ted Grasela, PhD, President and CEO of model-based drug development specialists Cognigen Corp., is currently operating under. Pharmacometricians and development statisticians have honed their craft to the degree that, Grasela says, they can step forward and make their work a focal point of efforts to speed and improve the development process.

Grasela gave a talk at the AAPS 2009 conference in Los Angeles entitled “Engineering the Pharmacometrics Enterprise: Science in Support of Science.” While he and his Cognigen colleagues once advised primarily PK/PD scientists in early development, they are now working more with late-stage groups as manufacturers better integrate their development efforts.  

But there is much work to be done. “My thesis is that integration is an essential missing ingredient in most drug companies,” he noted. “People are innovating and imaginating all over the place . . . but often these islands of innovation that never get linked with one another.”

Major impediments to productivity in drug development that Grasela sees include empirical decision-making, walls and barriers between functional areas, as well as the failure to effectively integrate knowledge over the life of a product. “One scientist used to do it all,” Grasela said. “It’s much more now about integrated project teams” but these teams do not always function as they could.

Grasela holds a fairly straightforward philosophy for further team integration.

Strategy is the first focus. Grasela encourages an “as-is vs. to-be” gap analysis based upon the target product profile to determine whether or not development teams are really making progress or not.

Process is next. “There are usually unmet process, informatic and systematic needs of the knowledge integration process,” he said. “The knowledge integration workflow process must be defined, provisioned, and governed.” For example, Grasela promotes a systematic methodology for defining inputs, outputs, and deliverables and ensuring that “the right people are responsible for the right steps.”

Lean Six Sigma tools such as 5S can also be effective integrators. Studies have shown that “40% of effort in an average data assembly project is rework.” Those tools that improve communications between scientists and IT professionals are particularly effective, he noted.

Methodology development is Grasela’s third emphasis. Pharmacometric models are only going to become more complex, he noted, so identifying the challenges of the data assembly effort is critical. “What does the data warehouse look like that allows data to be accessed readily and without confusion and ambiguity?” is one question that must be front and center. “Just the simple act of performing a cross-study assessment of data can be very informative,” he added.

Finally, training is essential to sustained integration efforts. Integrated Project Teams must encompass three critical skills, and therefore training must support these three areas:
•    The ability to optimize the data analysis and knowledge integration process
•    The ability to design development program strategies that take advantage of modeling and simulation
•    The ability to impact the entire R&D life cycle.

This leads back to Grasela’s original point—that modeling and simulation experts need to explore how to influence their organization on a broader spectrum than they have. Senior management in particular can be taken to task for poor decisionmaking in the past, and there needs to be a more concerted effort by project teams to focus management’s attention on modeling data, particularly of the predictive variety.

And development teams must be efficient and act quickly. “Management will never wait!” said Grasela. “The only way that I have found to get around this is to figure out how to change the process so that we get access to data on time, it comes to us in the appropriate format, and we have the ability to do the modeling in a time frame that delivers it when all of the important people are sitting at the table.”