Monitoring Mammalian Cell Cultivations for Monoclonal Antibody Production Using Near-Infrared Spectroscopy
Near-infrared (NIR) spectroscopy as a process monitoring and process supervision technique is reviewed in the context of biomanufacturing.
An industrial pilot-plant mammalian cell cultivation process has been chosen to illustrate the use of on-line in-situ NIR monitoring by means of an immersion transflectance NIR probe.
NIR calibration development must be performed carefully and should incorporate a number of steps to obtain a properly validated model which exhibits long-term robustness and is independent of process scale. A description of such good modelling practises is given. In general, NIR can be as accurate as the reference methods employed and at least as precise provided that sufficient spectral selectivity and sensitivity exists.
NIR can also be used as a direct technique for very fast process monitoring and process supervision, thus enabling one to follow the trajectory of a process. This alternative to the indirect use of NIR through laborious calibration development with direct reference methods has been little explored. Since NIR is sensitive to both chemical and physical properties, the analysis of whole samples enables relevant process information to be captured and thus generates better process state estimates than by simply looking at defined process parameters one at a time.
KeywordsProcess Analytical Technology Biomanufacturing Process Spectro scopy NIR mammalian cells cultivation
Symbols and Abbreviations
High performance liquid chromatography
Process analytical technologies
Principal component analysis
Correlation coefficient for cross-validation predictions
Correlation coefficient for external validation predictions
Root mean square error of cross-validation
Root mean square error of prediction
Standard error of laboratory
Standard normal variate
Variable importance plot
Dr Licinia O. Rodrigues (4TUNE Engineering Ltd) for discussions on the material in the paper. Mrs Miriam Ahlert (Roche Diagnostics GmbH, Germany) for support in the analytical work.
Numerical method for feature selection based on the mechanisms of biologic evolution
Matrix describing how the original variables relate to the new principal components of a PCA
Factorisation method typically used for regression when large numbers of collinear variables are present
Data factorisation methods that creates new orthogonal variables called principal components as linear combination of the original variables capturing the most possible variance in the original data
Projection of an observation (sample) in the principal component space of a PCA
- 2.Menezes JC, Ferreira AP, Rodrigues LO, Brás LP, Alves TP (2009) Chemometrics role within the PAT context: examples from primary pharmaceutical manufacturing. In: Steve B, Romà T, Beata W (eds) Comprehensive chemometrics, chap 13.11, 4 vols. Elsevier, AmsterdamGoogle Scholar
- 6.FDA (2004) Guidance for industry PAT - a framework for innovative pharmaceutical development, manufacturing, and quality assurance, September 2004Google Scholar
- 7.ICH (2005) Guideline Q8 - pharmaceutical development, November 2005Google Scholar
- 8.Kourti T (2004) Process analytical technology and multivariate statistical process control. Wellness Index of process and product (Part 1). J Process Anal Technol 1(1):13–19Google Scholar
- 9.Kourti T (2005) Process analytical technology and multivariate statistical process control. Wellness Index of process and product (Part 2). J Process Anal Technol 2(1):24–28Google Scholar
- 10.Kourti T (2006) Process analytical technology and multivariate statistical process control. Wellness Index of process and product (Part 3). J Process Anal Technol 3(1):18–24Google Scholar
- 11.Montague G (1997) Monitoring and control of fermentors. Institution of Chemical Engineers, Rugby, UKGoogle Scholar
- 13.Williams P, Norris K (eds) (2001) Near-infrared technology in the agricultural and food industries. American Association of Cereal Chemists, St Paul, MN, USAGoogle Scholar
- 15.Ciurczak EW, Drennen JK (2002) Pharmaceutical and medical applications of near-infrared spectroscopy. Marcel Dekker, New YorkGoogle Scholar
- 19.Siesler HW, Ozaki Y, Kawata S, Heise HM (eds) (2002) Near-infrared spectroscopy: principles, instruments, applications. Wiley, Weinheim, GermanyGoogle Scholar
- 20.Ozaki Y, Amari T (2000) Near-infrared spectroscopy in chemical process analysis. In: Chalmers JM (ed) Spectroscopy in process analysis. Sheffield Academic, Sheffield, UK, pp 53–95Google Scholar
- 21.Raghavachari R (ed) (2001) Near-infrared applications in biotechnology. Marcel Dekker, New York, USAGoogle Scholar
- 26.González-Vara A, Vaccari G, Dosi E, Trilli A, Rossi M, Matteuzzi D (2000) Enhanced production of L-(+)- lactic acid in chemostat by Lactobacillus casei DSM 20011 using ion-exchange resins and cross-flow filtration in a fully automated pilot plant controlled via NIR. Biotechnol Bioeng 67:47–156Google Scholar
- 33.Macaloney G, Hall JW, Rollins MJ, Draper I, Anderson KB, Preston J, Thompson BG, McNeil B (1997) The utility and performance of near-infra red spectroscopy in simultaneous monitoring of multiple components in a high cell density recombinant Escherichia coli production process. Bioprocess Biosyst Eng 17:157–167Google Scholar
- 39.Wülfert F (2004) Temperature-robust multivariate calibration, Phd Thesis (English 135 pp), University of Amsterdam, The Netherlands, available online http://dare.uva.nl/record/161626 (accessed June 13, 2008)Google Scholar
- 40.Næs T, Isaksson T, Fearn T, Davies T (2002) A user-friendly guide to multivariate calibration and classification. NIR, ChichesterGoogle Scholar
- 41.Otto M (1999) Chemometrics: statistics and computer application in analytical chemistry. Wiley, ChichesterGoogle Scholar