Fernando Muzzio, of Rutgers University’s department of chemical and biochemical engineering, began the debate by suggesting that while most unit operations – with the exception of tablet coating – are well understood and could run as continuous processes, fully integrated control is still not possible.
“The real challenge is not so much the individual unit operations or even the mechanical integration of individual unit operations, it is really in the full informatics, PAT, modelling and control integration. It is in developing the systems that can keep track of everything that is going on…that can control themselves and regulate how the process is operating.”
He also rejected the idea that a ‘plug and play’ – where separate technologies are combined simply by connecting them – is currently possible for continuous processing, again citing the lack of control technologies.
“We are not yet at a place where continuous processing is plug-and-play. We need to still do a lot of work in developing the integrated control systems,” said Muzzio, explaining that – unlike in batch processes – all operations in a continuous process happen at the same time, requiring monitoring and central control technologies that are not yet available.
Muzzio’s argument was supported by fellow panel member Moheb Nasr, former US Food and Drug Administration (FDA) QbD champion and current VP of GSK’s regulatory affairs unit, who said: “There is a need for systems engineering controls to be in place to ensure the success of continuous processing operations in addition to the control strategies that need to developed to assure product quality.”
Nasr also suggested that working on these types of controls at an early enough stage in the development of manufacturing processes is difficult for pharmaceutical firms that are uncertain of the future success of their products.
“Better consideration for continuous manufacturing in the early stages is I think, from the industry perspective, very challenging, especially if we don’t know what will be the success or potential success of our drugs. So investing in a new manufacturing platform with the hope that a drug will make it through a clinical trial can be challenging.”
Work being done
On a more positive note Muzzio did say that considerable research is being done in the field of control strategies for continuous processing by academics and pharmaceutical manufacturers. However, he was somewhat critical of equipment makers for not doing more work in this area.
“One place where I haven’t seen enough real implementation of the fully developed control systems yet is at equipment companies, and I think that is a problem.
“We have commercially available, partially integrated systems for a number of years, but the degree of instrumentation, the amount of modelling that has been implemented, the strength of the PAT meters and the integration of the PAT meters into the overall control system… is not where it needs to be. And the concern I have is that that is not being represented clearly to the people looking into buying the machines.”
Regulatory view
Elaine Morefield, the US FDA’s deputy director for review, put forward the regulatory perspective on continuous processing, explaining that – as with all production processes - conversation between industry and the agency is key to successful implementation.
“The regulatory requirements are the same no matter what kind of process you are using. You have to show your control strategy, so the issues when you do continuous processing – API to finished product – include how are you going to define the stability of that product? How are you going to define what your API actually is?”
Morefield added that: “The FDA is willing to work with you [drug manufacturers] on these issues and discuss controls, the appropriate specifications, how you clarify the definition of your API if it isn’t isolatable…if you want to pursue something like this it would be good if you met with us during development.”
Nasr welcomed the FDA’s willingness to debate the issues and stressed his personal belief that formal regulations governing continuous processing would be unhelpful.
“I hope we do not have regulatory guidance at this time because I really think industry and the technology should lead the approach that will be used…and I would advise the regulators to react to that and facilitate that, rather than with their limited knowledge and experience of continuous manufacturing to develop a guidance that tells people how to do things that are not yet clearly and fully understood.”