Search

Regulatory authorities in biopharmaceutical industry emphasize process design by process understanding but applicable tools that are easy to implement are still missing. Soft sensors are a promising tool for the implementation of the Quality by Design (QbD) approach and Process Analytical Technology (PAT). In particular, the correlation between viable cell counting and oxygen consumption was investigated, but problems remained: Either the process had to be modified for excluding CO 2 in pH control, or complex k L a models had to be set up for specific processes. In this work, a non-invasive soft sensor for simplified on-line cell counting based on dynamic oxygen uptake rate was developed with no need of special equipment. The dynamic oxygen uptake rates were determined by automated and periodic interruptions of gas supply in DASGIP® bioreactor systems, realized by a programmed Visual Basic script in the DASware® control software. With off-line cell counting, the two parameters were correlated based on linear regression and led to a robust model with a correlation coefficient of 0.92. Avoidance of oxygen starvation was achieved by gas flow reactivation at a certain minimum dissolved oxygen concentration. The soft sensor model was established in the exponential growth phase of a Chinese Hamster Ovary fed-batch process. Control studies showed no impact on cell growth by the discontinuous gas supply. This soft sensor is the first to be presented that does not require any specialized additional equipment as the methodology relies solely on the direct measurement of oxygen consumed by the cells in the bioreactor. • Non-invasive soft sensor method for simplified on-line cell counting. • Implemented in exponential growth phase of CHO cells. • Dynamic oxygen uptake rate as only predictor resulted in a correlation coefficient of 0.92. • No negative influences on cell growth as oxygen starvation was prevented. • 5-k cross validation showed RMSE of 0.7 × 106 cells/mL. [ABSTRACT FROM AUTHOR]

The control of ergot alkaloids in biotechnological processes is important in the context of obtaining new strain producers and studying the mechanisms of the biosynthesis, accumulation and secretion of alkaloids and the manufacturing of alkaloids. In pharmaceuticals, it is important to analyze the purity of raw materials, especially those capable of racemization, quality control of dosage forms and bulk drugs, stability during storage, etc. This review describes the methods used for qualitative and quantitative chemical analysis of ergot alkaloids in tablets and pharmaceutic forms, liquid cultural media and mycelia from submerged cultures of ergot and other organisms producing ergoalkaloid, sclerotias of industrial Claviceps spp. parasitic strains. We reviewed analytical approaches for the determination of ergopeptines (including their dihydro- and bromine derivatives) and semisynthetic ergot-derived medicines such as cabergoline, necergoline and pergolide, including precursors for their synthesis. Over the last few decades, strategies and approaches for the analysis of ergoalkaloids for medical use have changed, but the general principles and objectives have remained the same as before. These changes are related to the development of new genetically improved strains producing ergoalkaloids and the development of technologies for the online control of biotechnological processes and pharmaceuticalmanufacturing ("process analytical technologies," PAT). Overall, the industry is moving toward "smart manufacturing." The development of approaches to production cost estimation and product quality management, manufacturing management, increasing profitability and reducing the negative impact on personnel and the environment are integral components of sustainable development. Analytical approaches for the analysis of ergot alkaloids in pharmaceutical rawmaterials should have high enough specificity for the separation of dihydro derivatives, enantiomers and R-S epimers of alkaloids, but low values of the quantitative detection limit are less frequently needed. In terms of methodology, detection methods based on mass spectrometry have become more developed and widespread, but NMR analysis remains in demand becauseof its high accuracy and specificity. Both rapid methods and liquid chromatography remain indemand in routine practice, with rapidanalysis evolving toward higher accuracy owing to improved analytical performance and new equipment. New composite electrochemical sensors (including disposable sensors) have demonstrated potential for real-time process control. [ABSTRACT FROM AUTHOR]

AbstractPlant-based food consumption has increased substantially owing to its positive effects on human and global health. However, ensuring the quality and safety of plant-based foods remains a challenge. Diagnostic ultrasonic technology is widely used for rapid and nondestructive determination owing to its ability to penetrate optically opaque materials, strong directivity, rapid detection capabilities, low equipment costs, and ease of operation. This review provides a comprehensive understanding of diagnostic ultrasonic technology by summarizing the principles of food characterization, factors that influence detection accuracy and methods to mitigate their impact, composition of ultrasonic machine systems, and application of diagnostic ultrasound for monitoring plant-based foods. The detection principle of ultrasonic technology is based on empirical equations that establish a relationship between the ultrasonic and physicochemical indicators of food. To improve the detection accuracy, a compensation mechanism for the temperature and pressure should be established, measurement distances should be set in the far-field region, and liquid samples should be degassed. Furthermore, the sample platform design and the choice of detection mode depend on the nature of the food. Combining ultrasonic technology with machine learning techniques presents promising prospects for real-time process monitoring in the food and beverage industries. [ABSTRACT FROM AUTHOR]

Media preparation parameters contribute significantly to media quality, cell culture performance, productivity, and product quality. Establishing proper media preparation procedures is critical for ensuring a robust CHO cell culture process. Process analytical technology (PAT) enables unique ways to quantify assessments and improve media quality. Here, cell culture media were prepared under a wide range of temperatures (40–80°C) and pH (7.6–10.0). Media quality profiles were compared using three real‐time PATs: Fourier‐transform infrared (FTIR) spectroscopy, Raman spectroscopy, and excitation‐emission matrix (EEM) spectroscopy. FTIR and Raman spectroscopies identified shifts in media quality under high preparation temperature (80°C) and at differing preparation pH which negatively impacted monoclonal antibody (mAb) production. In fed‐batch processes for production of three different mAbs, viable cell density (VCD) and cell viability were mostly unaffected under all media preparation temperatures, while titer and cell specific productivity of mAb decreased when cultured in basal and feed media prepared at 80°C. High feed preparation pH alone was tolerated but cell growth and productivity profiles deviated from the control condition. Further, charge variants (main, acidic, basic species) and glycosylation (G0F, afucosylation, and high mannose) were examined. Statistically significant differences were observed for one or more of these quality attributes with any shifts in media preparation. In this study, we demonstrated strong associations between media preparation conditions and cell growth, productivity, and product quality. The rapid evaluation of media by PAT implementation enabled more comprehensive understanding of different parameters on media quality and consequential effects on CHO cell culture. [ABSTRACT FROM AUTHOR]

This study addresses the critical issue of cardiovascular diseases (CVDs) as the leading cause of death globally, emphasizing the importance of stent delivery catheter manufacturing. Traditional manufacturing processes, reliant on destructive end-of-batch sampling, present significant financial and quality challenges. This research addresses this challenge by proposing a novel approach: a closed-loop cyber-physical production system (CPPS) employing non-destructive process analytical technology (PAT). Through a mixed-method approach combining a comprehensive literature review and the development of a CPPS prototype, the study demonstrates the potential for real-time quality control, reduced production costs, and increased manufacturing efficiency. Initial findings showcase the system's effectiveness in streamlining production, enhancing stability, and minimizing defects, translating to substantial financial savings and improved product quality. This work extends the author's previous research by comparing the validated system's performance to that of pre-implementation manual workflows and inspections, highlighting tangible and intangible improvements brought by the new system. This paves the way for advanced control strategies to revolutionize medical device manufacturing. Furthermore, the study proposes a generalized CPPS framework applicable across diverse regulated environments, ensuring optimal processing conditions and adherence to stringent regulatory standards. The research concludes with the successful demonstration of innovative approaches and technologies, leading to improved product quality, patient safety, and operational efficiency in the medical device industry. [ABSTRACT FROM AUTHOR]

Secreted metabolites are an important class of bio‐process analytical technology (PAT) targets that can correlate to cell conditions. However, current strategies for measuring metabolites are limited to discrete measurements, resulting in limited understanding and ability for feedback control strategies. Herein, a continuous metabolite monitoring strategy is demonstrated using a single‐use metabolite absorbing resonant transducer (SMART) to correlate with cell growth. Polyacrylate is shown to absorb secreted metabolites from living cells containing hydroxyl and alkenyl groups such as terpenoids, that act as a plasticizer. Upon softening, the polyacrylate irreversibly conformed into engineered voids above a resonant sensor, changing the local permittivity which is interrogated, contact‐free, with a vector network analyzer. Compared to sensing using the intrinsic permittivity of cells, the SMART approach yields a 20‐fold improvement in sensitivity. Tracking growth of many cell types such as Chinese hamster ovary, HEK293, K562, HeLa, and E. coli cells as well as perturbations in cell proliferation during drug screening assays are demonstrated. The sensor is benchmarked to show continuous measurement over six days, ability to track different growth conditions, selectivity to transducing active cell growth metabolites against other components found in the media, and feasibility to scale out for high throughput campaigns. [ABSTRACT FROM AUTHOR]

Real‐time selective protein quantification is an integral component of operating continuous chromatography processes. Partial least squares models fit with spectroscopic UV‐Vis absorbance data have demonstrated the ability to selectively quantify proteins. With standard continuous chromatography equipment that is only capable of measuring absorbance at a few user‐defined wavelengths, the problem of selecting appropriate wavelengths that maximize the measurement capability of the instrument remains unaddressed. Therefore, we propose a method for selecting wavelengths for continuous chromatography equipment. We illustrate our method using sets of protein mixtures composed of bovine serum albumin and lysozyme. The first step is to refine the raw wavelength set with a statistical t‐test and an absorbance magnitude test. Then, the wavelengths within the refined spectroscopic range are ranked. Three existing techniques are evaluated – sequential forward search, variable importance to projection scores, and the least absolute shrinkage and selection operator. The best technique (in this case, sequential forward search) determines a subset of three wavelengths for further evaluation on the BioSMB PD. We use an exhaustive approach to determine the final wavelength set. We show that soft sensor models trained from the method's wavelength selections can quantify the two proteins more accurately than from the wavelength set of 230, 260 and 280 nm, by a factor of four. The method is shown to determine appropriate wavelengths for different path lengths and protein concentration ranges. Overall, we provide a tool that alleviates the analytical bottleneck for practitioners seeking to develop advanced monitoring and control methods on standard equipment. Application of a wavelength selection method that consists of a spectroscopic wavelength refinement step, followed by a wavelength ranking step and then a final wavelength verification step using in‐line absorbance data. The soft sensor partial least square models trained from the method's wavelength selections can selectively quantify proteins more accurately than from a set of 230, 260, and 280 nm wavelengths. The proposed method is shown to generate appropriate wavelengths for data sets of different path length and concentration ranges. [ABSTRACT FROM AUTHOR]

In the era of Biopharma 4.0, process digitalization fundamentally requires accurate and timely monitoring of critical process parameters (CPPs) and quality attributes. Bioreactor systems are equipped with a variety of sensors to ensure process robustness and product quality. However, during the biphasic production of viral vectors or replication‐competent viruses for gene and cell therapies and vaccination, current monitoring techniques relying on a single working sensor can be affected by the physiological state change of the cells due to infection/transduction/transfection step required to initiate production. To address this limitation, a multisensor (MS) monitoring system, which includes dual‐wavelength fluorescence spectroscopy, dielectric signals, and a set of CPPs, such as oxygen uptake rate and pH control outputs, was employed to monitor the upstream process of adenovirus production in HEK293 cells in bioreactor. This system successfully identified characteristic responses to infection by comparing variations in these signals, and the correlation between signals and target critical variables was analyzed mechanistically and statistically. The predictive performance of several target CPPs using different multivariate data analysis (MVDA) methods on data from a single sensor/source or fused from multiple sensors were compared. An MS regression model can accurately predict viable cell density with a relative root mean squared error (rRMSE) as low as 8.3% regardless of the changes occurring over the infection phase. This is a significant improvement over the 12% rRMSE achieved with models based on a single source. The MS models also provide the best predictions for glucose, glutamine, lactate, and ammonium. These results demonstrate the potential of using MVDA on MS systems as a real‐time monitoring approach for biphasic bioproduction processes. Yet, models based solely on the multiplicity and timing of infection outperformed both single‐sensor and MS models, emphasizing the need for a deeper mechanistic understanding in virus production prediction. [ABSTRACT FROM AUTHOR]

Coating thickness is a critical quality attribute of many coated tablets. Functional coatings ensure correct drug release kinetics or protection from light, while non-functional coatings are generally applied for cosmetic reasons. Traditionally, coating thickness is assessed indirectly via offline methods, such as weight gain or diameter growth. In the past decade, several methods, including optical coherence tomography (OCT) and Raman spectroscopy, have emerged to perform in-line measurements of various subclasses of coating formulations. However, there are some obstacles. For example, when using OCT, a major challenge is scattering pigments, such as titanium dioxide and iron oxide, which make the interface between the coating and the tablet core difficult to detect. This work explores novel OCT image evaluation techniques using unsupervised machine learning to compute image metrics. Certain image metrics of highly scattering coatings are correlated with the tablet thickness, and hence indirectly with the coating thickness. The method was demonstrated using a titanium dioxide rich coating formulation. The results are expected to be applicable to other scattering coatings and will significantly broaden the applicability of OCT to at-line and in-line coating thickness measurements of a much larger class of coating formulations. [Display omitted] • Optical coherence tomography for indirect coating thickness measurement of highly scattering coatings. • Machine learning for image analysis to extract and quantify light reflection. • Non-destructive coating thickness measurement through light reflection properties. [ABSTRACT FROM AUTHOR]

Spray freeze-drying has emerged as a promising alternative to conventional vial freeze-drying. In this study, we have developed a non-contact monitoring technique for the freeze-drying of microparticles utilizing microwave resonance spectroscopy. Spectral changes, indicative of the degree of drying, were successfully captured during the freeze-drying process of particulate samples ranging in size from approximately 0.6–3.5 mm. The observed spectral patterns demonstrated significant dependence on both the mean particle size and the formulation type (mannitol or sucrose). The partial least squares method was employed to extract data series strongly correlated with the drying progress. Multiple regression analysis was then utilized to derive regression equations, yielding values representing the drying progress based on the intensity values at selected frequencies within the spectra. The resulting regression equations accurately replicated the experimentally estimated drying kinetics. Notably, a robust regression equation was obtained, demonstrating applicability to various formulations and particle sizes, with coefficient of determination values ranging from 0.95 to 0.99. Furthermore, it was suggested that a correlation between the obtained spectra and the change in moisture content during secondary drying. Microwave resonance spectroscopy proves to be a versatile technique for monitoring freeze-drying processes, offering insights that can enhance the efficiency and adaptability of this critical pharmaceutical manufacturing step. [ABSTRACT FROM AUTHOR]

Several key technologies for advancing biopharmaceutical manufacturing depend on the successful implementation of process analytical technologies that can monitor multiple product quality attributes in a continuous in‐line setting. Raman spectroscopy is an emerging technology in the biopharma industry that promises to fit this strategic need, yet its application is not widespread due to limited success for predicting a meaningful number of quality attributes. In this study, we addressed this very problem by demonstrating new capabilities for preprocessing Raman spectra using a series of Butterworth filters. The resulting increase in the number of spectral features is paired with a machine learning algorithm and laboratory automation hardware to drive the automated collection and training of a calibration model that allows for the prediction of 16 different product quality attributes in an in‐line mode. The demonstrated ability to generate these Raman‐based models for in‐process product quality monitoring is the breakthrough to increase process understanding by delivering product quality data in a continuous manner. The implementation of this multiattribute in‐line technology will create new workflows within process development, characterization, validation, and control. [ABSTRACT FROM AUTHOR]

Suspensions of protein antigens adsorbed to aluminum‐salt adjuvants are used in many vaccines and require mixing during vial filling operations to prevent sedimentation. However, the mixing of vaccine formulations may generate undesirable particles that are difficult to detect against the background of suspended adjuvant particles. We simulated the mixing of a suspension containing a protein antigen adsorbed to an aluminum‐salt adjuvant using a recirculating peristaltic pump and used flow imaging microscopy to record images of particles within the pumped suspensions. Supervised convolutional neural networks (CNNs) were used to analyze the images and create "fingerprints" of particle morphology distributions, allowing detection of new particles generated during pumping. These results were compared to those obtained from an unsupervised machine learning algorithm relying on variational autoencoders (VAEs) that were also used to detect new particles generated during pumping. Analyses of images conducted by applying both supervised CNNs and VAEs found that rates of generation of new particles were higher in aluminum‐salt adjuvant suspensions containing protein antigen than placebo suspensions containing only adjuvant. Finally, front‐face fluorescence measurements of the vaccine suspensions indicated changes in solvent exposure of tryptophan residues in the protein that occurred concomitantly with new particle generation during pumping. [ABSTRACT FROM AUTHOR]

The ability for nuclear magnetic resonance (NMR) spectroscopy to provide quantitative, structurally rich information makes this spectroscopic technique an attractive reaction monitoring tool. The practicality of NMR for this type of analysis has only increased in the recent years with the influx of commercially available benchtop NMR instruments and compatible flow systems. In this study, we aim to compare 19F NMR reaction profiles acquired under both on‐line continuous‐flow and stopped‐flow sampling methods, with modern benchtop NMR instrumentation, and two reaction systems: a homogeneous imination reaction and a biphasic activation of a carboxylic acid to acyl fluoride. Reaction trends with higher data density can be acquired with on‐line continuous‐flow analyses, and this work highlights that representative reaction trends can be acquired without any correction when monitoring resonances with a shorter spin–lattice relaxation time (T1), and with the used flow conditions. On‐line stopped‐flow analyses resulted in representative reaction trends in all cases, including the monitoring of resonances with a long T1, without the need of any correction factors. The benefit of easier data analysis, however, comes with the cost of time, as the fresh reaction solution must be flowed into the NMR system, halted, and time must be provided for spins to become polarized in the instrument's external magnetic field prior to spectral measurement. Results for one of the reactions were additionally compared with the use of a high‐field NMR. [ABSTRACT FROM AUTHOR]

The application of compact NMR instruments to hot flowing samples or exothermically reacting mixtures is limited by the temperature sensitivity of permanent magnets. Typically, such temperature effects directly influence the achievable magnetic field homogeneity and hence measurement quality. The internal‐temperature control loop of the magnet and instruments is not designed for such temperature compensation. Passive insulation is restricted by the small dimensions within the magnet borehole. Here, we present a design approach for active heat shielding with the aim of variable temperature control of NMR samples for benchtop NMR instruments using a compressed airstream which is variable in flow and temperature. Based on the system identification and surface temperature measurements through thermography, a model predictive control was set up to minimise any disturbance effect on the permanent magnet from the probe or sample temperature. This methodology will facilitate the application of variable‐temperature shielding and, therefore, extend the application of compact NMR instruments to flowing sample temperatures that differ from the magnet temperature. [ABSTRACT FROM AUTHOR]

Background: Charge related heterogeneities of monoclonal antibody (mAb) based therapeutic products are increasingly being considered as a critical quality attribute (CQA). They are typically estimated using analytical cation exchange chromatography (CEX), which is time consuming and not suitable for real time control. Raman spectroscopy coupled with artificial intelligence (AI) tools offers an opportunity for real time monitoring and control of charge variants. Objective: We present a process analytical technology (PAT) tool for on-line and real-time charge variant determination during process scale CEX based on Raman spectroscopy employing machine learning techniques. Method: Raman spectra are collected from a reference library of samples with distribution of acidic, main, and basic species from 0–100% in a mAb concentration range of 0–20 g/L generated from process-scale CEX. The performance of different machine learning techniques for spectral processing is compared for predicting different charge variant species. Result: A convolutional neural network (CNN) based model was successfully calibrated for quantification of acidic species, main species, basic species, and total protein concentration with R2 values of 0.94, 0.99, 0.96 and 0.99, respectively, and the Root Mean Squared Error (RMSE) of 0.1846, 0.1627, and 0.1029 g/L, respectively, and 0.2483 g/L for the total protein concentration. Conclusion: We demonstrate that Raman spectroscopy combined with AI-ML frameworks can deliver rapid and accurate determination of product related impurities. This approach can be used for real time CEX pooling decisions in mAb production processes, thus enabling consistent charge variant profiles to be achieved. [ABSTRACT FROM AUTHOR]

This review provides an overview of the latest advancements and applications in multi‐dimensional liquid chromatography coupled with mass spectrometry (mD‐LC‐MS), covering aspects such as inter‐laboratory studies, digestion strategy, trapping column, and multi‐level analysis. The shift from an offline to an online workflow reduces sample processing artifacts, analytical variability, analysis time, and the labor required for data acquisition. Over the past few years, this technique has demonstrated sufficient maturity for application across a diverse range of complex products. Moreover, there is potential for this strategy to evolve into an integrated process analytical technology tool for the real‐time monitoring of monoclonal antibody quality. This review also identifies emerging trends, including its application to new modalities, the possibility of evaluating biological activity within the mD‐LC set‐up, and the consideration of multi‐dimensional capillary electrophoresis as an alternative to mD‐LC. As mD‐LC‐MS continues to evolve and integrate emerging trends, it holds the potential to shape the next generation of analytical tools, offering exciting possibilities for enhanced characterization and monitoring of complex biopharmaceutical products. [ABSTRACT FROM AUTHOR]

Abstract Secreted metabolites are an important class of bio‐process analytical technology (PAT) targets that can correlate to cell conditions. However, current strategies for measuring metabolites are limited to discrete measurements, resulting in limited understanding and ability for feedback control strategies. Herein, a continuous metabolite monitoring strategy is demonstrated using a single‐use metabolite absorbing resonant transducer (SMART) to correlate with cell growth. Polyacrylate is shown to absorb secreted metabolites from living cells containing hydroxyl and alkenyl groups such as terpenoids, that act as a plasticizer. Upon softening, the polyacrylate irreversibly conformed into engineered voids above a resonant sensor, changing the local permittivity which is interrogated, contact‐free, with a vector network analyzer. Compared to sensing using the intrinsic permittivity of cells, the SMART approach yields a 20‐fold improvement in sensitivity. Tracking growth of many cell types such as Chinese hamster ovary, HEK293, K562, HeLa, and E. coli cells as well as perturbations in cell proliferation during drug screening assays are demonstrated. The sensor is benchmarked to show continuous measurement over six days, ability to track different growth conditions, selectivity to transducing active cell growth metabolites against other components found in the media, and feasibility to scale out for high throughput campaigns.

Imaging and image-based process analytical technologies (PAT) have revolutionized the design, development, and operation of crystallization processes, providing greater process understanding through the characterization of particle size, shape and crystallization mechanisms in real-time. The performance of corresponding PAT models, including machine learning/artificial intelligence (ML/AI)-based approaches, is highly reliant on the data quality used for training or validation. However, acquiring high quality data is often time consuming and a major roadblock in developing image analysis models for crystallization processes.To address the lack of diverse, high-quality, and publicly available particle image datasets, this paper presents an initiative to create an open-access crystallization-related image database: OpenCrystalData (OCD, at www.kaggle.com/opencrystaldata/datasets). The datasets consist of images from different crystallization systems with different particle sizes and shapes captured under various conditions. The initial release consists of four different datasets, addressing the estimation of particle size distribution using in-situ images for different categories of particles and detection of anomalous particles for process monitoring purposes. Images are collected using various instruments, followed by case-specific processing steps, such as ground-truth labeling and particle size characterization using offline microscopy. Datasets are released on the online collaborative platform Kaggle, along with specific guidelines for each dataset. These datasets are aimed to serve as a resource for researchers to enable learning, experimentation, development, and evaluation and comparison of different analytical approaches and algorithms. Another goal of this initiative is to encourage researchers to contribute new datasets focusing on various systems and problem statements. Ultimately, OpenCrystalData is intended to facilitate and inspire new developments in imaging-based PAT for crystallization processes, encouraging a shift from time-consuming offline analysis towards comprehensive real-time process insights that drive product quality.

Abstract Background Monitoring and control of both growth media and microbial biomass is extremely important for the development of economical bioprocesses. Unfortunately, process monitoring is still dependent on a limited number of standard parameters (pH, temperature, gasses etc.), while the critical process parameters, such as biomass, product and substrate concentrations, are rarely assessable in-line. Bioprocess optimization and monitoring will greatly benefit from advanced spectroscopy-based sensors that enable real-time monitoring and control. Here, Fourier transform (FT) Raman spectroscopy measurement via flow cell in a recirculatory loop, in combination with predictive data modeling, was assessed as a fast, low-cost, and highly sensitive process analytical technology (PAT) system for online monitoring of critical process parameters. To show the general applicability of the method, submerged fermentation was monitored using two different oleaginous and carotenogenic microorganisms grown on two different carbon substrates: glucose fermentation by yeast Rhodotorula toruloides and glycerol fermentation by marine thraustochytrid Schizochytrium sp. Additionally, the online FT-Raman spectroscopy approach was compared with two at-line spectroscopic methods, namely FT-Raman and FT-infrared spectroscopies in high throughput screening (HTS) setups. Results The system can provide real-time concentration data on carbon substrate (glucose and glycerol) utilization, and production of biomass, carotenoid pigments, and lipids (triglycerides and free fatty acids). Robust multivariate regression models were developed and showed high level of correlation between the online FT-Raman spectral data and reference measurements, with coefficients of determination (R2) in the 0.94–0.99 and 0.89–0.99 range for all concentration parameters of Rhodotorula and Schizochytrium fermentation, respectively. The online FT-Raman spectroscopy approach was superior to the at-line methods since the obtained information was more comprehensive, timely and provided more precise concentration profiles. Conclusions The FT-Raman spectroscopy system with a flow measurement cell in a recirculatory loop, in combination with prediction models, can simultaneously provide real-time concentration data on carbon substrate utilization, and production of biomass, carotenoid pigments, and lipids. This data enables monitoring of dynamic behaviour of oleaginous and carotenogenic microorganisms, and thus can provide critical process parameters for process optimization and control. Overall, this study demonstrated the feasibility of using FT-Raman spectroscopy for online monitoring of fermentation processes.

Development and optimization of inclusion body (IB) refolding processes is often based on empirical strategies. Thus, generation of in-depth process understanding is crucial to shift towards more knowledge-driven approaches. Within this contribution the goal was to establish a systematic approach for the reliable quantification of process dynamics in protein refolding processes. Established offline- analytical tools like HPLC-based methods or photometric assays were analyzed regarding their specifications and consequent suitability to quantify protein states. An in-silico study was conducted to define requirements for state quantification by analyzing the influences of error, discrete sampling intervals and the pace of the reaction on signal quality. Refolding and aggregation reaction rates were calculated by finite difference approximation with errors propagated from the corresponding folding state. Application of the defined principles on two real Lactate dehydrogenase fed-batch refolding processes resulted in two individual sampling strategies. At reaction rates of up to 0.58 g h−1 description of the ongoing dynamics could be done accurately with high sampling intervals during the first few hours of processing as shown by comparison to a noise free simulation. The proposed method has the potential to be transferred to other proteins and to facilitate the development of model-based monitoring & control strategies. [Display omitted] • Systematic selection of analytical tools for protein refolding states. • Analysis of influencing factors on signal quality of protein refolding and aggregation rates. • Quantitative analysis of folding dynamics in fed-batch refolding processes. • Sampling strategy for maximization of information via offline data generation. [ABSTRACT FROM AUTHOR]

Sr(OH)2 is an indispensable strontium compound extensively harnessed in sugar refining, strontium lubricating wax formulation, and polymer plastic stabilization. Sr(OH)2·8H2O is the prevalent hydrate form of Sr(OH)2. Deprived of moisture via vacuum drying, Sr(OH)2 can be procured from Sr(OH)2·8H2O. Sr(OH)2·8H2O particles with larger sizes exhibit impressive attributes such as facile solid–liquid divergence, elevated product purity, expedient drying, and resilience to agglomeration, which have garnered significant interest. Given the superior quality of the product and the dependability of the process, process analytical technology (PAT) has been extensively employed in the pharmaceutical sector, rendering it feasible to employ PAT to fabricate large-particle Sr(OH)2·8H2O crystals. This study utilizes industrial SrCO3 to prepare high-purity Sr(OH)2·8H2O with a purity of over 99.5%. The growth process of single crystals was observed using a hot-stage microscope, and the growth process of large-particle Sr(OH)2·8H2O was optimized and regulated online using PAT. The optimal process conditions were optimized, and large-particle Sr(OH)2·8H2O crystals were obtained by adding crystal seeds. On this basis, we proposed a seed control mechanism for Sr(OH)2·8H2O. [ABSTRACT FROM AUTHOR]

Hydrodistillation is traditionally a green technology for the manufacturing of natural products that are volatile. As well as acknowledged process intensification methods such as microwave support for energy efficiency to move towards climate neutral operation, digital twins combined with process analytical technology for advanced process control enables reliable operation of an optimal operation point regarding lowest cost of goods, as well as lowest global warming potential equivalent. A novel process control enabled by digital twin technology has shown to reduce the ecological footprint of the extraction by up to 46.5%, while reducing the cost of extraction by 22.4%. Additionally, skilled operator time is reduced, and the sustainable plant material is utilized most efficiently. The approach is ready to apply, but broad industrialization seems to be held back by unclear business cases and lack of comprehension of decision makers. This is in drastic contrast to the political demand for climate neutrality goals and the cost pressure by worldwide completion. [ABSTRACT FROM AUTHOR]

Digital twins have emerged as a powerful concept for real-time monitoring and analysis, facilitating Quality by Design integration into biopharmaceutical manufacturing. Traditionally, lyophilization processes are developed through trial-and-error, incorporating high security margins and inflexible process set points. Digital twins enable the integration of adaptable operating conditions and implementation of automation through Advanced Process Control (APC) with Process Analytical Technology (PAT) and validated physicochemical models that rely on heat and mass transfer principles, allowing us to overcome the challenges imposed by the lyophilization process. In this study, a digital twin for freeze-drying processes is developed and experimentally validated. Using the digital twin, primary drying conditions were optimized for controlled nucleation and annealing methods by carrying out a few laboratory tests beforehand. By incorporating PAT and modeling, the digital twin accurately predicts the product's temperature and drying endpoint, showing smaller errors than the experiments. The digital twin significantly increases productivity by up to 300% while reducing the costs by 74% and the Global Warming Potential by 64%. [ABSTRACT FROM AUTHOR]

Background: Monitoring and control of both growth media and microbial biomass is extremely important for the development of economical bioprocesses. Unfortunately, process monitoring is still dependent on a limited number of standard parameters (pH, temperature, gasses etc.), while the critical process parameters, such as biomass, product and substrate concentrations, are rarely assessable in-line. Bioprocess optimization and monitoring will greatly benefit from advanced spectroscopy-based sensors that enable real-time monitoring and control. Here, Fourier transform (FT) Raman spectroscopy measurement via flow cell in a recirculatory loop, in combination with predictive data modeling, was assessed as a fast, low-cost, and highly sensitive process analytical technology (PAT) system for online monitoring of critical process parameters. To show the general applicability of the method, submerged fermentation was monitored using two different oleaginous and carotenogenic microorganisms grown on two different carbon substrates: glucose fermentation by yeast Rhodotorula toruloides and glycerol fermentation by marine thraustochytrid Schizochytrium sp. Additionally, the online FT-Raman spectroscopy approach was compared with two at-line spectroscopic methods, namely FT-Raman and FT-infrared spectroscopies in high throughput screening (HTS) setups. Results: The system can provide real-time concentration data on carbon substrate (glucose and glycerol) utilization, and production of biomass, carotenoid pigments, and lipids (triglycerides and free fatty acids). Robust multivariate regression models were developed and showed high level of correlation between the online FT-Raman spectral data and reference measurements, with coefficients of determination (R2) in the 0.94–0.99 and 0.89–0.99 range for all concentration parameters of Rhodotorula and Schizochytrium fermentation, respectively. The online FT-Raman spectroscopy approach was superior to the at-line methods since the obtained information was more comprehensive, timely and provided more precise concentration profiles. Conclusions: The FT-Raman spectroscopy system with a flow measurement cell in a recirculatory loop, in combination with prediction models, can simultaneously provide real-time concentration data on carbon substrate utilization, and production of biomass, carotenoid pigments, and lipids. This data enables monitoring of dynamic behaviour of oleaginous and carotenogenic microorganisms, and thus can provide critical process parameters for process optimization and control. Overall, this study demonstrated the feasibility of using FT-Raman spectroscopy for online monitoring of fermentation processes. [ABSTRACT FROM AUTHOR]

Degumming is an oil refinement process in which the naturally occurring phospholipids in crude vegetable oils are removed. Enzymatic degumming results in higher oil yield and more cost-efficient processing compared to traditional degumming processes using only water or acid. Phospholipase C hydrolyses phospholipids into diglycerides and phosphate groups during degumming. The diglyceride content can therefore be considered a good indicator of the state of the enzymatic reaction. This study investigates the use of near-infrared (NIR) spectroscopy and chemometrics to monitor the degumming process by quantifying diglycerides in soybean oil in both off-line and on-line settings. Fifteen enzymatic degumming lab scale batches originating from a definitive screening design (with varying water, acid, and enzyme dosages) were investigated with the aim to develop a NIR spectroscopy prediction method. By applying tailored preprocessing and variable selection methods, the diglyceride content can be predicted with a root mean square error of prediction of 0.06% (w/w) for the off-line set-up and 0.07% (w/w) for the on-line set-up. The results show that the diglyceride content is a good indicator of the enzyme performance and that NIR spectroscopy is a suitable analytical technique for robust real-time diglyceride quantification. Graphical abstract This is a visual representation of the abstract. [ABSTRACT FROM AUTHOR]

Purpose: This study evaluates the use of the closed feed frame as a material sparing approach to develop near-infrared (NIR) spectroscopic calibration models for monitoring blend uniformity. The effect of shear induced by recirculation on NIR spectra was also studied. Methods: Calibration models were developed using NIR spectra obtained in the closed feed frame for two cases. For case 2, blends that flowed through the open feed frame were predicted with the model. The shear effect of the feed frame on the blends was assessed through the characterization of powder properties before and after recirculation. Results: The physical characterization of the blends confirmed that the powder properties were not altered after recirculation within the closed feed frame. Both calibration models provided highly accurate predictions of the test sets with low bias (0.03% w/w and -0.06% w/w) and relative standard error of prediction (1.9% and 3.7%), respectively. The predictive performance of the calibration models was not affected by the shear effect. Conclusion: Recirculation within the closed feed frame did not change the physical properties of the blends studied. The prediction of blends flowing through the open feed frame was possible with a calibration model developed in the closed feed frame. The closed feed frame could reduce the materials needed to develop calibration models by more than 90%. [ABSTRACT FROM AUTHOR]

Raman spectroscopy is a useful tool for polymorphic form-monitoring during the crystallization process. However, its application to solute concentration estimation in two-phase systems like crystallization is rare, as the Raman signal is influenced by various changing factors in the crystallization process. The development of a robust calibration model that covers all variations is complex and represents a major challenge for the implementation of Raman spectroscopy for in-line monitoring and control of the solution crystallization process. This paper describes the development of a Raman-based calibration model for estimating the solute concentration of the active pharmaceutical ingredient ceritinib. Several different calibration approaches were tested, which included both temperature and spectra of clear solutions and slurries/suspensions. It was found that the concentration of the ceritinib solution could not be accurately predicted when suspended crystals were present. To overcome this challenge, the approach was enhanced by including additional variables related to crystal size and solid concentration obtained via in-line process microscopy (chord-length distribution percentiles D10, D50 and D90) and turbidity. Partial least squares regression (PLSR) and artificial neural network (ANN) models were developed and compared based on root mean square error (RMSE). ANN models estimated the solute concentration with high accuracy, with the prediction error not exceeding 1% of the nominal solute concentration. [ABSTRACT FROM AUTHOR]

Chemometric modeling for spectral data is considered a key technology in biopharmaceutical processing to realize real-time process control and release testing. Machine learning (ML) models have been shown to increase the accuracy of various spectral regression and classification tasks, remove challenging preprocessing steps for spectral data, and promise to improve the transferability of models when compared to commonly applied, linear methods. The training and optimization of ML models require large data sets which are not available in the context of biopharmaceutical processing. Generative methods to extend data sets with realistic in silico samples, so-called data augmentation, may provide the means to alleviate this challenge. In this study, we develop and implement a novel data augmentation method for generating in silico spectral data based on local estimation of pure component profiles for training convolutional neural network (CNN) models using four data sets. We simultaneously tune hyperparameters associated with data augmentation and the neural network architecture using Bayesian optimization. Finally, we compare the optimized CNN models with partial least-squares regression models (PLS) in terms of accuracy, robustness, and interpretability. The proposed data augmentation method is shown to produce highly realistic spectral data by adapting the estimates of the pure component profiles to the sampled concentration regimes. Augmenting CNNs with the in silico spectral data is shown to improve the prediction accuracy for the quantification of monoclonal antibody (mAb) size variants by up to 50% in comparison to single-response PLS models. Bayesian structure optimization suggests that multiple convolutional blocks are beneficial for model accuracy and enable transfer across different data sets. Model-agnostic feature importance methods and synthetic noise perturbation are used to directly compare the optimized CNNs with PLS models. This enables the identification of wavelength regions critical for model performance and suggests increased robustness against Gaussian white noise and wavelength shifts of the CNNs compared to the PLS models.

Fully automated at-line terahertz time-domain spectroscopy in transmission mode is used to measure tablet porosity for thousands of immediate release tablets. The measurements are rapid and non-destructive. Both laboratory prepared tablets and commercial samples are studied. Multiple measurements on individual tablets quantify the random errors in the terahertz results. These show that the measurements of refractive index are precise, with the standard deviation on a single tablet being about 0.002, with variation between measurements being due to small errors in thickness measurement and from the resolution of the instrument. Six batches of 1000 tablets each were directly compressed using a rotary press. The tabletting turret speed (10 and 30 rpm) and compaction pressure (50, 100 and 200 MPa) were varied between the batches. As expected, the tablets compacted at the highest pressure have far lower porosity than those compacted at the lowest pressure. The turret rotation speed also has a significant effect on porosity. This variation in process parameters resulted in batches of tablets with an average porosity between 5.5 and 26.5%. Within each batch, there is a distribution of porosity values, the standard deviation of which is in the range 1.1 to 1.9%. Destructive measurements of disintegration time were performed in order to develop a predictive model correlating disintegration time and tablet porosity. Testing of the model suggested it was reasonable though there may be some small systematic errors in disintegration time measurement. The terahertz measurements further showed that there are changes in tablet properties after storage for nine months in ambient conditions.

ABSTRACTThe implementation of process analytical technologies is positioned to play a critical role in advancing biopharmaceutical manufacturing by simultaneously resolving clinical, regulatory, and cost challenges. Raman spectroscopy is emerging as a key technology enabling in-line product quality monitoring, but laborious calibration and computational modeling efforts limit the widespread application of this promising technology. In this study, we demonstrate new capabilities for measuring product aggregation and fragmentation in real-time during a bioprocess intended for clinical manufacturing by applying hardware automation and machine learning data analysis methods. We reduced the effort needed to calibrate and validate multiple critical quality attribute models by integrating existing workflows into one robotic system. The increased data throughput resulting from this system allowed us to train calibration models that demonstrate accurate product quality measurements every 38 s. In-process analytics enable advanced process understanding in the short-term and will lead ultimately to controlled bioprocesses that can both safeguard and take necessary actions that guarantee consistent product quality.

Process analytical technology (PAT) is increasingly applied in the crystallization process for continuous monitoring of some of the key process parameters and product quality features. Very important process variables for cooling crystallization are the temperature and concentration of the mother liquor. Continuous measurement of concentration is made possible by advanced in situ spectroscopic instruments. Attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR), used in this work belongs to such methods. The calibration model which estimates the concentration of the solution in real time can be developed using machine-learning methods. The aim of this work was to develop and analyse partial least squares regression (PLSR) and neural network models for modelling the dependence of the concentration of the active ingredients, xylometazoline hydrochloride in n-butanol, on temperature and spectral data obtained by measurements with an ATR-FTIR spectrometer. In this work, pre-processing of the collected data was performed with MSC technique (multiplicative scatter correction), Min-Max and Z-score normalization; the number of neurons in the first and second hidden layers, the number of hidden layers, the type of learning algorithm applied (ADAM, NADAM, RMSprop), and the influence of the type of transfer function (ReLU, sigmoid, tanh) on the quality of the developed neural networks were analysed. Considering values of coefficient of determination and mean square error, developed models gave very good results on all four datasets. The neural network model gave coefficients of determination in the range of values from 0.9979 to 0.9989, and the mean square error from 0.0020 to 0.0011. With the PLSR model, coefficients of determination from 0.9990 to 0.9995, and mean square errors from 0.0009 to 0.0005, were obtained. Obtained results showed that the pre-processing of the data and the addition of a second hidden layer of the neural network in this study did not have a major impact on the final results. This type of monitoring and control of the process would lead to more efficient production with a lower probability of error, enabling the pharmaceutical industry to bring products to market faster. [ABSTRACT FROM AUTHOR]

Current manufacturing and development processes for therapeutic monoclonal antibodies demand increasing volumes of analytical testing for both real‐time process controls and high‐throughput process development. The feasibility of using Raman spectroscopy as an in‐line product quality measuring tool has been recently demonstrated and promises to relieve this analytical bottleneck. Here, we resolve time‐consuming calibration process that requires fractionation and preparative experiments covering variations of product quality attributes (PQAs) by engineering an automation system capable of collecting Raman spectra on the order of hundreds of calibration points from two to three stock seed solutions differing in protein concentration and aggregate level using controlled mixing. We used this automated system to calibrate multi‐PQA models that accurately measured product concentration and aggregation every 9.3 s using an in‐line flow‐cell. We demonstrate the application of a nonlinear calibration model for monitoring product quality in real‐time during a biopharmaceutical purification process intended for clinical and commercial manufacturing. These results demonstrate potential feasibility to implement quality monitoring during GGMP manufacturing as well as to increase chemistry, manufacturing, and controls understanding during process development, ultimately leading to more robust and controlled manufacturing processes. [ABSTRACT FROM AUTHOR]

In recent years a substantial effort has been devoted to the development of allogeneic cell-based immunotherapies with the aim of reducing costs and improving the accessibility of such treatments. NK cells have the potential to target a wide range of cancer cells without causing graft-versus-host disease (GvHD), making them an attractive option as off-the-shelf therapies for cancer treatment. Translating from laboratory production of NK cells to biomanufacturing technology for off-the-shelf products requires bioreactor scale-up and process optimization which will likely involve integrating biological understanding of NK cells with modern sensors for in-line monitoring of key process variables that impact NK cell proliferation and function. The development of NK biomanufacturing can benefit from the tremendous amount of knowledge accumulated from the biomanufacturing of therapeutic proteins. A quality by design (QbD) approach can be applied in NK cell biomanufacturing even in its early stages of process development. Identifying and characterizing critical quality attributes (CQA) and critical process parameters (CPPs) will help in developing a robust and efficient manufacturing process. Cell therapy offers the potential for curative treatment of cancers. Although T cells have been the predominantly used cell type, natural killer (NK) cells have attracted great attention owing to their ability to kill cancer cells and because they are naturally suitable for allogeneic applications. Upon stimulation by cytokines or activation by a target cell, NK cells proliferate and expand their population. These cytotoxic NK cells can be cryopreserved and used as an off-the-shelf medicine. The production process for NK cells thus differs from that of autologous cell therapies. We briefly outline key biological features of NK cells, review the manufacturing technologies for protein biologics, and discuss their adaptation for developing robust NK cell biomanufacturing processes. [ABSTRACT FROM AUTHOR]

[Display omitted] Transmission Raman spectroscopy (TRS) is a process analytical technology tool for nondestructive analysis of drug content in tablets. Although wet granulation is the most used tablet manufacturing method, most TRS studies have focused on tablets manufactured via direct compression. The effects of upstream process parameter variations, such as granulation, on the prediction performance of TRS quantitative models are unknown. We evaluated the effects of process parameter variations during granulation on the prediction performance of the TRS quantitative model. Tablets with a drug concentration of 1%w/w were used. We developed PLS calibration models for the drug concentration range of 70–130% label claims. Subsequently, we predicted the drug content of the tablets with different granulation parameters. The results of our study demonstrate that the variation in the predicted recovery due to the variation in granulation parameters was practically acceptable. The calibration model showed a good prediction performance for tablets manufactured at different granulation scales and thicknesses. Therefore, we conclude that TRS quantitative models are robust to variations in upstream processes, such as granulation and downstream variations in tableting parameters. These results suggest that TRS is a versatile non-destructive quantitative analysis method that can be applied in tablet manufacturing. [ABSTRACT FROM AUTHOR]

Manufacturers of nanoparticle-based products rely on detailed information about critical process parameters, such as particle size and size distributions, concentration, and material composition, which directly reflect the quality of the final product. These process parameters are often obtained using offline characterization techniques that cannot provide the temporal resolution to detect dynamic changes in particle ensembles during a production process. To overcome this deficiency, we have recently introduced Optofluidic Force Induction (of2i) for optical real-time counting with single particle sensitivity and high throughput. In this paper, we apply of2i to highly polydisperse and multi modal particle systems, where we also monitor evolutionary processes over large time scales. For oil-in-water emulsions we detect in real time the transition between high-pressure homogenization states. For silicon carbide nanoparticles, we exploit the dynamic of2i measurement capabilities to introduce a novel process feedback parameter based on the dissociation of particle agglomerates. Our results demonstrate that of2i provides a versatile workbench for process feedback in a wide range of applications. [ABSTRACT FROM AUTHOR]

In-process monitoring and control of biomanufacturing workflows remains a significant challenge in the development, production, and application of cell therapies. New process analytical technologies must be developed to identify and control the critical process parameters that govern ex vivo cell growth and differentiation to ensure consistent and predictable safety, efficacy, and potency of clinical products. This study demonstrates a new platform for at-line intracellular analysis of T-cells. Untargeted mass spectrometry analyses via the platform are correlated to conventional methods of T-cell assessment. Spectral markers and metabolic pathways correlated with T-cell activation and differentiation are detected at early time points via rapid, label-free metabolic measurements from a minimal number of cells as enabled by the platform. This is achieved while reducing the analytical time and resources as compared to conventional methods of T-cell assessment. In addition to opportunities for fundamental insight into the dynamics of T-cell processes, this work highlights the potential of in-process monitoring and dynamic feedback control strategies via metabolic modulation to drive T-cell activation, proliferation, and differentiation throughout biomanufacturing. [ABSTRACT FROM AUTHOR]

Process analytical technology (PAT) tools such as Raman Spectroscopy have become established tools for real time measurement of CHO cell bioreactor process variables and are aligned with the QbD approach to manufacturing. These tools can have a significant impact on process development if adopted early, creating an end‐to‐end PAT/QbD focused process. This study assessed the impact of Raman based feedback control on early and late phase development bioreactors by using a Raman based PLS model and PAT management system to control glucose in two CHO cell line bioreactor processes. The impact was then compared to bioreactor processes which used manual bolus fed methods for glucose feed delivery. Process improvements were observed in terms of overall bioreactor health, product output and product quality. Raman controlled batches for Cell Line 1 showed a reduction in glycation of 43.4% and 57.9%, respectively. Cell Line 2 batches with Raman based feedback control showed an improved growth profile with higher VCD and viability and a resulting 25% increase in overall product titer with an improved glycation profile. The results presented here demonstrate that Raman spectroscopy can be used in both early and late‐stage process development and design for consistent and controlled glucose feed delivery. [ABSTRACT FROM AUTHOR]

Process analytical technology (PAT) in late-stage drug product development is typically used for real-time process monitoring, in-process control, and real-time release testing. In early research and development (R&D), PAT usage is limited as the manufacturing scale is relatively small with frequent changes and only a few batches are produced on an annual basis. However, process understanding is critical at early R&D in order to identify process and formulation boundaries, so PAT applications could be particularly useful in early-stage R&D. For oral solid dosage form, conventional HPLC-based content uniformity (CU) methods with sampling of 3 tablets per stratified sampling location in early R&D are typically not sufficient to identify these manufacturing process boundaries and temporal profile. Here, we report a screening CU method based on a multivariate model using transmission Raman spectroscopy (TRS) data on a phase-appropriate calibration set of only 16 tablets. This initial model was used for multiple pre-GMP development batches to provide critical information about blend uniformity and content uniformity (CU). In this work, the precision of the TRS method was evaluated; multiple spectral preprocessing approaches were compared regarding their effects on measurement precision as well as their ability to mitigate the photo bleaching effects during precision experiments. Overall, the TRS-based CU method was much faster than a traditional HPLC-based method allowing a much larger number of tablets to be screened. This larger number of analyzed tablets enabled the processes boundaries and temporal changes in CU to be identified while providing proper statistical assurance on product quality. [ABSTRACT FROM AUTHOR]

The automation of traditional Chinese medicine (TCM) pharmaceuticals has driven the development of process analysis from offline to online. Most of common online process analytical technologies are based on spectroscopy, making the identification and quantification of specific ingredients still a challenge. Herein, we developed a quality control (QC) system for monitoring TCM pharmaceuticals based on paper spray ionization miniature mass spectrometry (mini-MS). It enabled real-time online qualitative and quantitative detection of target ingredients in herbal extracts using mini-MS without chromatographic separation for the first time. Dynamic changes of alkaloids in Aconiti Lateralis Radix Praeparata (Fuzi) during decoction were used as examples, and the scientific principle of Fuzi compatibility was also investigated. Finally, the system was verified to work stably at the hourly level for pilot-scale extraction. This mini-MS based online analytical system is expected to be further developed for QC applications in a wider range of pharmaceutical processes.

Bioreactor parameters can have significant effects on the quantity and quality of biotherapeutics. Monoclonal antibody products have one particularly important critical quality attribute being the distribution of product glycoforms. N‐linked glycosylation affects the therapeutic properties of the antibody including effector function, immunogenicity, stability, and clearance rate. Our past work revealed that feeding different amino acids to bioreactors altered the productivity and glycan profiles. To facilitate real‐time analysis of bioreactor parameters and the glycosylation of antibody products, we developed an on‐line system to pull cell‐free samples directly from the bioreactors, chemically process them, and deliver them to a chromatography‐mass spectroscopy system for rapid identification and quantification. We were able to successfully monitor amino acid concentration on‐line within multiple reactors, evaluate glycans off‐line, and extract four principal components to assess the amino acid concentration and glycosylation profile relationship. We found that about a third of the variability in the glycosylation data can be predicted from the amino acid concentration. Additionally, we determined that the third and fourth principal component accounts for 72% of our model's predictive power, with the third component indicated to be positively correlated with latent metabolic processes related to galactosylation. Here we present our work on rapid online spent media amino acid analysis and use the determined trends to collate with glycan time progression, further elucidating the correlation between bioreactor parameters such as amino acid nutrient profiles, and product quality. We believe such approaches may be useful for maximizing efficiency and reducing production costs for biotherapeutics. [ABSTRACT FROM AUTHOR]

In the production of antibody‐drug conjugates (ADCs), the conjugation reaction is a central step defining the final product composition and, hence, directly affecting product safety and efficacy. To enable real‐time monitoring, spectroscopic sensors in combination with multivariate regression models have gained popularity in recent years. The extended Kalman filter (EKF) can be used as so‐called soft‐sensor to fuse sensor predictions with long‐horizon forecasts by process models. This enables the dynamic update of the current state and provides increased robustness against experimental noise or model errors. Due to the uncertainty associated with sensor and process models in biopharmaceutical applications, the deployment of such soft‐sensors is challenging. In this study, we demonstrate the combination of an uncertainty‐aware sensor model with a kinetic reaction model using an EKF to monitor a site‐directed ADC conjugation reaction. As the sensor model, a Gaussian process regression model is presented to realize a time‐variant determination of the sensor uncertainty. The EKF fuses the time‐discrete predictions of the amount of conjugated drug from the sensor model with the time‐continuous predictions from the kinetic model. While the ADC species are not distinguishable by on‐line recorded UV/Vis spectra, the developed soft‐sensor is able to dynamically update all relevant reaction species. It could be shown that the use of time‐variant process and sensor noise computation approaches improved the performance of the EKF and achieved a reduction of the prediction error of up to 23% compared with the kinetic model. The developed framework proved to enhance robustness against noisy sensor measurements or wrong model initialization and was successfully transferred from batch to fed‐batch mode. In future, this framework could be implemented for model‐based process control and be adopted for other ADC conjugation reaction types. [ABSTRACT FROM AUTHOR]

Fostered by novel analytical techniques, digitalization, and automation, modern bioprocess development provides large amounts of heterogeneous experimental data, containing valuable process information. In this context, data-driven methods like machine learning (ML) approaches have great potential to rationally explore large design spaces while exploiting experimental facilities most efficiently. Herein we demonstrate how ML methods have been applied so far in bioprocess development, especially in strain engineering and selection, bioprocess optimization, scale-up, monitoring, and control of bioprocesses. For each topic, we will highlight successful application cases, current challenges, and point out domains that can potentially benefit from technology transfer and further progress in the field of ML. Bioprocess development requires identification of robust design spaces for specific bioproducts and involves efficient strain selection, bioprocess optimization, scale-up, and optimal control strategies for robust industrial production. Beyond multivariate data analysis, deep learning, reinforcement learning, and other novel ML techniques start to complement and replace traditional data analysis approaches to accelerate screening, optimization, and control procedures. Transfer learning is emerging as a means to leverage the potential of historic data to guide novel production processes. No single algorithmic solution will be suitable for all aspects of bioprocess development. Instead, a flexible combination of various techniques is required to enhance the whole development pipeline. Fast impact is expected in autonomous strain selection and the optimization of bioprocess parameters. The application of ML for scale-up has a high impact but needs further development. [ABSTRACT FROM AUTHOR]

Viable cell concentration (VCC) is an essential parameter that is required to support the efficient cultivation of mammalian cells. Although commonly determined using at‐line or off‐line analytics, in‐line capacitance measurements represent a suitable alternative method for the determination of VCC. In addition, these latter efforts are complimentary with the Food and Drug Administration's initiative for process analytical technologies (PATs). However, current applications for online determination of the VCC often rely on single frequency measurements and corresponding linear regression models. It has been reported that this may be insufficient for application at all stages of a mammalian cell culture processes due to changes in multiple cell parameters over time. Alternatively, dielectric spectroscopy, measuring capacitance at multiple frequencies, in combination with multivariate mathematical models, has proven to be more robust. However, this has only been applied for retrospective data analysis. Here, we present the implementation of an O‐PLS model for the online processing of multifrequency capacitance signals and the on‐the‐fly integration of the models' VCC results into a supervisory control and data acquisition (SCADA) system commonly used for cultivation observation and control. This system was evaluated using a Chinese hamster ovary (CHO) cell perfusion process. [ABSTRACT FROM AUTHOR]