S122
Real-time bioprocess control applications utilizing in situ Raman spectroscopy for process optimization, reduced batch-to-batch variability, improved quality, and advanced process control
Wednesday, July 27, 2016: 11:00 AM
Grand Chenier, 5th Fl (Sheraton New Orleans)
S. Gilliam*, M.E. Cuellar, P. Wiegand, D. Strachan and I. Lewis, Kaiser Optical Systems Inc., Ann Arbor, MI; J. Aon and J. Sun, GlaxoSmithKline, King of Prussa, PA; A. Pitters, H. Lucas and B. Lenain, Kaiser Optical Systems Inc. SARL, Ecully, France
Process Analytical Technologies (PAT) have demonstrated value in bioprocess control, enabling non-invasive and real-time measurements of critical process parameters (CPP) and in applying Quality by Design (QbD) principles. As an
in situ PAT for fermentation or cell culture bioprocesses, a single Raman spectroscopy sensor provides simultaneous measurement of multiple biochemistry and bioreactor CPPs and allows in-process optimizations and corrections in real time. Raman spectra provide rich chemical and physical data, which are used to generate multicomponent qualitative and quantitative predictive models. Unique to Raman is its compatibility with aqueous systems, as water does not interfere with the Raman spectrum. Thus, Raman is well-suited for many upstream and downstream bioprocess applications.
In recent years, Raman has opened up new avenues to bioprocess analytics by utilizing a technology that is robust, scaleable from lab to process, provides in situ knowledge in real-time, and is transferable between cell lines, media feedstocks, and process conditions. During process development, Raman facilitates QbD principles defining the manufacturing design space for optimization of process conditions and product quality. Compared to off-line PAT, in situ Raman reduces costs and overburden associated with sampling and equipment maintenance, eliminates consumables, as well as reduces sterility risk. The wealth of bioprocess information provided by in situ, real-time Raman is harnessed by many companies to deploy efficient, continuous, and hybrid biopharmaceutical manufacturing. We will present examples demonstrating Raman in fermentation and cell culture bioprocesses and provide new developments in data models that account for matrix complexity and facilitate cross-scale model transferability.