Pfizer “aggressively” pursuing cost saving real-time release

Pfizer is “aggressively” moving toward real-time release (RTR) and believes return on investment can be realised in a year.

Adopting RTR eliminates the need to do conventional end-of-line testing by using process analytical technologies (PAT) to gather quality data during manufacturing. With technology in place and regulators showing support the onus is now on pharmaceutical companies to make the shift.

We see no hurdles to moving forward [with RTR]”, Gerry Migliaccio, senior vice president, network performance at Pfizer Global Supply, told a US Food and Drug Administration (FDA) committee meeting.

It is now up to companies to satisfy themselves that they have the level of process control to adopt RTR, Migliaccio said, and Pfizer is approaching this point. “We’re certainly moving as aggressively as we can to [adopt RTR]”, Migliaccio said.

Switching to RTR requires upfront investments, in PAT and training for example, but payback can be quick. “We’ve seen examples where the return on investment is as short as 12 months”, Migliaccio said, although depending on product volumes and other factors it can take longer.

Supporters of RTR claim it will generate savings by cutting quality control laboratory costs and lowering throughput time, which, in turn, reduces inventory.

Beyond RTR

For Migliaccio, intelligence based manufacturing, also known as adaptive processes, is the “ultimate”. In intelligence based manufacturing, the impact of input variability is reduced through changes to controls.

For example, intelligence based manufacturing would help manage the “fair amount of variability in excipients” pharma faces. The goal is models that allow manufacturers to say “if my particle size is between ‘a’ and ‘b’, I use these conditions, if it’s between ‘b’ and ‘c’, I use these conditions”.

Quality-by-Design (QbD) will help reach this point. “The more effective we are in understanding what’s critical to quality, developing robust design spaces [and] predictive models, [the sooner] we’ll get to this adaptable process”, Migliaccio said.