Robust platform for inline Raman monitoring and control of perfusion cell culture.

Perfusion cell culture has been gaining increasing popularity for biologics manufacturing due to benefits such as smaller footprint, increased productivity, consistent product quality and manufacturing flexibility, cost savings, and so forth. Process Analytics Technologies tools are highly desirable for effective monitoring and control of long-running perfusion processes. Raman has been widely investigated for monitoring and control of traditional fed batch cell culture process. However, implementation of Raman for perfusion cell culture has been very limited mainly due to challenges with high-cell density and long running times during perfusion which cause extremely high fluorescence interference to Raman spectra and consequently it is exceedingly difficult to develop robust chemometrics models. In this work, a platform based on Raman measurement of permeate has been proposed for effective analysis of perfusion process. It has been demonstrated that this platform can effectively circumvent the fluorescence interference issue while providing rich and timely information about perfusion dynamics to enable efficient process monitoring and robust bioreactor feed control. With the highly consistent spectral data from cell-free sample matrix, development of chemometrics models can be greatly facilitated. Based on this platform, Raman models have been developed for good measurement of several analytes including glucose, lactate, glutamine, glutamate, and permeate titer. Performance of Raman models developed this way has been systematically evaluated and the models have shown good robustness against changes in perfusion scale and variations in permeate flowrate; thus models developed from small lab scale can be directly transferred for implementation in much larger scale of perfusion. With demonstrated robustness, this platform provides a reliable approach for automated glucose feed control in perfusion bioreactors. Glucose model developed from small lab scale has been successfully implemented for automated continuous glucose feed control of perfusion cell culture at much larger scale.

Longitudinal monitoring of cell metabolism in biopharmaceutical production using label-free fluorescence lifetime imaging microscopy.

Chinese hamster ovary (CHO) cells are routinely used in the biopharmaceutical industry for production of therapeutic monoclonal antibodies (mAbs). Although multiple offline and time-consuming measurements of spent media composition and cell viability assays are used to monitor the status of culture in biopharmaceutical manufacturing, the day-to-day changes in the cellular microenvironment need further in-depth characterization. In this study, two-photon fluorescence lifetime imaging microscopy (2P-FLIM) was used as a tool to directly probe into the health of CHO cells from a bioreactor, exploiting the autofluorescence of intracellular nicotinamide adenine dinucleotide phosphate (NAD(P)H), an enzymatic cofactor that determines the redox state of the cells. A custom-built multimodal microscope with two-photon FLIM capability was utilized to monitor changes in NAD(P)H fluorescence for longitudinal characterization of a changing environment during cell culture processes. Three different cell lines were cultured in 0.5 L shake flasks and 3 L bioreactors. The resulting FLIM data revealed differences in the fluorescence lifetime parameters, which were an indicator of alterations in metabolic activity. In addition, a simple principal component analysis (PCA) of these optical parameters was able to identify differences in metabolic progression of two cell lines cultured in bioreactors. Improved understanding of cell health during antibody production processes can result in better streamlining of process development, thereby improving product titer and verification of scale-up. To our knowledge, this is the first study to use FLIM as a label-free measure of cellular metabolism in a biopharmaceutically relevant and clinically important CHO cell line.

Tuning monoclonal antibody galactosylation using Raman spectroscopy-controlled lactic acid feeding.

A key aspect of large-scale production of biotherapeutics is a well-designed and consistently-executed upstream cell culture process. Process analytical technology tools provide enhanced monitoring and control capabilities to support consistent process execution, and also have potential to aid in maintenance of product quality at desired levels. One such tool, Raman spectroscopy, has matured as a useful technique to achieve real-time monitoring and control of key cell culture process attributes. We developed a Raman spectroscopy-based nutrient control strategy to enable dual control of lactate and glucose levels for a fed-batch CHO cell culture process for monoclonal antibody (mAb) production. To achieve this, partial least squares-based chemometric models for real-time prediction of glucose and lactate concentrations were developed and deployed in feedback control loops. In particular, feeding of lactic acid post-metabolic shift was investigated based on previous work that has shown the impact of lactate levels on ammonium as well as mAb product quality. Three feeding strategies were assessed for impact on cell metabolism, productivity, and product quality: bolus-fed glucose, glucose control at 4 g/L, or simultaneous glucose control at 4 g/L and lactate control at 2 g/L. The third feeding strategy resulted in a significant reduction in ammonium levels (68%) while increasing mAb galactosylation levels by approximately 50%. This work demonstrated that when deployed in a cell culture process, Raman spectroscopy is an effective technique for simultaneous control of multiple nutrient feeds, and that lactic acid feeding can have a positive impact on both cell metabolism and mAb product quality.

A Rational Analysis of Uniformity Risk for Agglomerated Drug Substance Using NIR Chemical Imaging.

Early risk detection and quick diagnosis of manufacturing challenges are necessary to support the accelerated development pace of drug product in the current competitive environment. Analytical tools, such as near-infrared (NIR) chemical imaging (CI), can be employed for alerting drug substance uniformity risks in intermediates and the final product of solid dosage forms. In this particular study, the ability to characterize the behavior of agglomerated drug substance throughout process development was enabled by NIR CI to identify uniformity risks with small sample sizes and short turnaround time. Using NIR chemical imaging, the drug substance distribution and cluster size in all intermediates were visualized throughout the drug product process. NIR CI enabled rapid identification of the key unit operations that produced the greatest reduction in cluster size for enhanced distribution of the drug substance. The comil acted as a high shear mixing step to disperse soft lumps prior to roller compaction. Shear forces or pressure during roller compaction was sufficient to break down and disperse the agglomerates further. Ultimately, the process was robust against a range of drug substance input properties such that the uniformity of the final blend was consistently achieved and the agglomerated drug substance had no risks to the drug product process.

In-line ATR-UV and Raman Spectroscopy for Monitoring API Dissolution Process During Liquid-Filled Soft-Gelatin Capsule Manufacturing.

Complete dissolution of the active pharmaceutical ingredient (API) is critical in the manufacturing of liquid-filled soft-gelatin capsules (SGC). Attenuated total reflectance UV spectroscopy (ATR-UV) and Raman spectroscopy have been investigated for in-line monitoring of API dissolution during manufacturing of an SGC product. Calibration models have been developed with both techniques for in-line determination of API potency. Performance of both techniques was evaluated and compared. The ATR-UV methodology was found to be able to monitor the dissolution process and determine the endpoint, but was sensitive to temperature variations. The Raman technique was also capable of effectively monitoring the process and was more robust to the temperature variation and process perturbations by using an excipient peak for internal correction. Different data preprocessing methodologies were explored in an attempt to improve method performance.

Synthetic training sets for the development of discriminant functions for the detection of volatile organic compounds from passive infrared remote sensing data.

A novel synthetic data generation methodology is described for use in the development of pattern recognition classifiers that are employed for the automated detection of volatile organic compounds (VOCs) during infrared remote sensing measurements. The approach used is passive Fourier transform infrared spectrometry implemented in a downward-looking mode on an aircraft platform. A key issue in developing this methodology in practice is the need for example data that can be used to train the classifiers. To replace the time-consuming and costly collection of training data in the field, this work implements a strategy for taking laboratory analyte spectra and superimposing them on background spectra collected from the air. The resulting synthetic spectra can be used to train the classifiers. This methodology is tested by developing classifiers for ethanol and methanol, two prevalent VOCs in wide industrial use. The classifiers are successfully tested with data collected from the aircraft during controlled releases of ethanol and during a methanol release from an industrial facility. For both ethanol and methanol, missed detections in the aircraft data are in the range of 4 to 5%, with false positive detections ranging from 0.1 to 0.3%.

Wavelet analysis used for spectral background removal in the determination of glucose from near-infrared single-beam spectra.

Wavelet analysis is developed as a preprocessing tool for use in removing background information from near-infrared (near-IR) single-beam spectra before the construction of multivariate calibration models. Three data sets collected with three different near-IR spectrometers are investigated that involve the determination of physiological levels of glucose (1-30 mM) in a simulated biological matrix containing alanine, ascorbate, lactate, triacetin, and urea in phosphate buffer. A factorial design is employed to optimize the specific wavelet function used and the level of decomposition applied, in addition to the spectral range and number of latent variables associated with a partial least-squares calibration model. The prediction performance of the computed models is studied with separate data acquired after the collection of the calibration spectra. This evaluation includes one data set collected over a period of more than 6 months. Preprocessing with wavelet analysis is also compared to the calculation of second-derivative spectra. Over the three data sets evaluated, wavelet analysis is observed to produce better-performing calibration models, with improvements in concentration predictions on the order of 30% being realized relative to models based on either second-derivative spectra or spectra preprocessed with simple additive and multiplicative scaling correction. This methodology allows the construction of stable calibrations directly with single-beam spectra, thereby eliminating the need for the collection of a separate background or reference spectrum.

Airborne passive Fourier transform infrared remote sensing of methanol vapor from industrial emissions.

Passive Fourier transform infrared (FT-IR) remote sensing measurements are applied to the detection of methanol vapor plumes released from a chemical manufacturing facility. With the spectrometer mounted in a downward-looking mode on a fixed-wing aircraft, overflights of the facility are made during the methanol release. Signal processing and pattern recognition methods are applied to the acquired data for the purpose of constructing an automated classification algorithm for the methanol detection. The analysis is based on the use of short, digitally filtered segments of the raw interferogram data collected by the spectrometer. The classifiers are trained with data collected on the ground by use of an experimental protocol designed to simulate background conditions observed from the air. Optimization of the digital filtering and interferogram segment parameters leads to successful classifiers based on 100 or 120 interferogram points. The optimal interferogram segment location is found to be 95-points displaced from the centerburst, and the best performing digital filters are centered on the methanol C-O stretching band at 1036 cm(-1) and have a passband full-width at half-maximum of 100 to 160 cm(-1). The best classifiers achieve classification errors of less than 1% and are observed to be resistant to possible interference effects from species such as ethanol and ozone. This work demonstrates the utility of airborne passive FT-IR remote sensing measurements of volatile organic compounds under complex background conditions such as those encountered while monitoring an operating industrial facility.