At-Line NIR Spectroscopy As a Simple and Effective PAT Monitoring Technique In Mab Cultivations During Process Development and Manufacturing

Thursday, October 20, 2011: 2:15 PM
Conrad A (Hilton Minneapolis)
Christian Hakemeyer1, Silke Werz1, Jose Cardoso Menezes2, Francisca Folque3, Gledson Jose3 and Ulrike Strauss1, (1)Pharmaceutical Biotech Production & Development, Roche Diagnostics GmbH, Penzberg, Germany, (2)Chemical & Biological Engineering Department, I.S.T., Lisbon, Portugal, (3)Technology Services & Products, 4Tune Engineering Ltd, Lisbon, Portugal

The experience of using at-line NIR transmittance spectroscopy on supernatant samples from a Mab cultivation process spanning several scales, clones and development years, from 2.5 l to 10,000 l is described.  The collected spectra after preprocessing were used to (1) do process state estimation and (2) to obtain several culture parameters.  Multivariate process trajectories were computed from NIR spectra acquired at-line and were used to: enhance process understanding across different scales up to industrial scale, assess batch-to-batch variability and scale-up issues, and examine the relative importance (criticality) of different sources of variability.  The proposed approach demonstrates that NIR monitoring carried out at-line on samples already taken for reference methods analysis, (1) is very effective in replacing in-process monitoring of some critical process parameters by reference methods, (2) enables batch trajectories as defined by multiple culture parameters to be computed, (3) can be used to do guided sampling (i.e., near harvest or to investigate a deviation from the nominal trajectory by established bioanalytical methods) and (4) has potential use in implementing process corrections or even control strategies.  The proposed approach is applicable throughout process development as well as to existing large-scale validated CHO bioprocesses, doesn’t require more complex and expensive set-ups (viz., in-situ, on-line probes and process equipment) and is capable of supporting continuous improvement efforts on legacy and new but non-QbD Mab manufacturing processes.

Extended Abstract: File Not Uploaded