To get industry experts talking together about continuous processing of monoclonal antibodies (mAbs), the National Institute for Innovation in Manufacturing Biopharmaceuticals (NIIMBL) in Newark, DE, created the N-mAb case study. As explained by NIIMBL: “The N-mAb case study brought together over 60 industry and government stakeholders from over 20 organizations to develop shared expectations and vocabulary around a control strategy for an integrated continuous bioprocess for a hypothetical monoclonal antibody.”

Instead of a one-off project, NIIMBL developed the N-mAb case study to fuel continuous conversations.

As one example, Gene Schaefer, ScD, a senior fellow at NIIMBL, and his colleagues—including scientists from Amgen, AstraZeneca, CSL Behring, Merck KGaA, and the National Institute of Standards and Technology—reported on “managing deviations from a state of control in real time” when manufacturing mAbs with continuous bioprocessing.

In particular, Schaefer’s team explored a range of deviations, including changes in critical process parameters (CPPs), problems with critical in-process controls (IPCs), analysis of samples, real-time changes in continuous processing, and even disposal of mAbs that fail to meet the desired criteria.

Deviation management

If something goes wrong in the continuous processing of mAbs, some parts of a lot can be impacted, other parts not, and there can be some mixing of the two. So, a bioprocessor needs a deviation-management system that detects the problem and determines its cause and impact. Plus, this system must eliminate any part of a lot that is negatively impacted by a deviation.

To maintain the desired quality of a mAb, the bioprocessor must monitor its production and act on deviations outside of previously set ranges. For example, Schaefer and his colleagues stated: “The severity and duration of the CPP or IPC excursion informs the likelihood of product impact; the root cause identification informs confidence in the ability to restore control; and the proportion of lot impacted informs whether the quality of the lot is acceptable.”

With so many monitoring and decision-making steps, automation is a crucial goal in managing deviations in continuous manufacturing of mAbs.

“Implementation of artificial intelligence or machine learning (AI/ML) to rapidly evaluate the data collected using multivariate models and make automated recommendations to either divert or forward process becomes an appetizing ‘first step’ toward fully automated process control,” Schaefer’s team concluded. “We anticipate an increased role for automation, including AI/ML, in [quality management] systems as technology advances.”

Previous articleDeep Learning without Compromising Confidentiality
Next articleGenetic Disorders Diagnosed More Quickly via Long-Read Sequencing Method