Accelerating Process Development with Automated Aseptic Sampling
Challenges of Bioprocess Sampling
Bioprocess development consists of iterative, preferably short, cycles aimed at generating data to optimize workflow steps. Scaling up data collection can accelerate optimization but demands continuous sample collection and analysis throughout experiments.
When using process analytical technology (PAT) to ensure quality within the biopharmaceutical workflow, frequent sample collection for analysis is crucial for identification and control of critical process parameters (CPP) and critical quality attributes (CQA) within a specified design space.
Frequent manual sampling presents challenges in terms of time, labor, and sterility risks, often constrained by staffing schedules. A low sampling frequency results in slow data generation which significantly delays critical process decisions, sometimes by weeks. This delay hinders comprehensive control and prevents corrective action and process optimization, particularly in cell culture, forcing the operator to rely on a step-by-step control of parameters.
Advantages of Automated Sampling for Your Bioprocess
Automated aseptic sampling and delivery of materials to on-line sample preparation and analytical instruments improve the throughput and time coverage of PAT. When implemented correctly, automated sampling reduces experiment run time and allows staff to focus on more value-generating tasks such as data analysis and implementing optimization steps. Increased data density and immediacy is also possible, which condenses development cycle times and informed decisions via rapid data feedback loops.
This page describes the evaluation of the MAST® Autosampling Solution as part of an automated PAT system implemented by Takeda Pharmaceuticals. The intention was to reduce the duration of bioprocess development cycles with rapid, data-based decisions, overcoming the constraints of a lengthy sample-to-data timeline. Three MAST® Autosampling Solutions were implemented for daily bioreactor sampling and on-line data acquisition to characterize cell culture performance and product quality.
The study results confirm how automated sampling facilitates resource efficiency and accelerates turnaround times while generating data that were comparable to those obtained using manual methods. By utilizing the MAST® Autosampling Solution, Takeda was able to achieve new efficiencies in process development including:
- 10x reduction in Full Time Equivalent (FTE) time
- 10x reduction in experiment run time
- 10x faster turnaround times for analytical results
- 60% reduction in the number of analysts needed
Autosampling Technology: Maintaining Sterility of your Sample Source
The MAST® Autosampling Solution collects, directs, and reliably transfers samples to analytical devices automatically and aseptically:
- MAST® Sample Pilots draw samples into an aseptic zone using a patented valve system to ensure sterility of the source vessel and avoid the contamination risk presented by other sampling systems such as automated devices, syringes, and probes.
- Positive displacement pumps push samples of varying viscosities over long distances rather than pulling samples with a weaker vacuum force.
- Upon delivery of the sample, a sanitation and drying step clears the lines for the next draw and prevents clogging.
- Every sample is recorded with metadata for complete traceability as to the origin, analytical purpose, and operator.
Each component of the modular and scalable solution was designed for continuous, long-term operation. To date, more than 60 MAST® Autosampling Solutions installed at customer sites have collected more than 150,000 samples with no recorded instances of sterility issues caused by the system. A detailed investigation comparing automated and manual sampling confirmed method reliability and maintenance of sterility when utilizing MAST® Autosampling Solution.
Case Study:
Experimental Design
In this study, eight bench-scale bioreactors were equipped with MAST® Sample Pilots and Controllers that delivered 1–2 samples per day to bioprocessing analyzers and an on-line ultra-high performance liquid chromatography (UHPLC) following cell removal using the MAST® Cell Removal System and a Verity® GX-271 Liquid Handler (Gilson Inc., Middleton, WI, USA). Figure 1 shows an example of a workflow setup.
Samples were collected and analyzed from the following bioprocessing streams:
- Purification and titer concentration determination via Protein A-based chromatography
- Aggregate analysis via size-exclusion chromatography (SEC)
- Charge variant analysis via cation exchange chromatography (CEX)
- A multi-attribute method (MAM) measuring 15 parameters via liquid chromatography-mass spectroscopy (LC-MS)
A key performance metric was the more efficient use of personnel compared to manual methods for monitoring. As such, data acquisition via manual sampling and off-line analytics were compared to automated sampling and on-line analysis with a focus on resources and time needed to complete experiment runs. Resources, time, cell culture performance, and product quality attributes were evaluated over the course of 14 days.
Figure 1.Example workflow in a laboratory using the MAST® Autosampling Solution for automated on-line sampling and analysis to support rapid process development.
Improving Resource Efficiency and Multiparametric Insight with Autosampling
Results of the study demonstrated that manual sampling and testing required five full-time analysts while the same experimental runs were performed by two analysts supervising the automated sampling and on-line purification and testing (Table 1). By eliminating manual sampling schedules and tedious testing activities, team members could instead focus on optimizing process development and refining methodologies. Takeda realized a 60% reduction in the number of analysts required per testing panel, a result of streamlining experimental preparation and data analysis from 48 hours to 4 hours.
Automated sample collection, processing, and delivery removed manual sampling bottlenecks, enabling concurrent analyses and multiparametric insights. Each sample provided a comprehensive set of parameter values, allowing for a more thorough characterization of each individual timepoint. As such, the experiments provided a deeper understanding of bioreactor conditions, streamlining development cycles.
Accelerating Process Development with Data Immediacy and Near Real-time Analyses
Table 2 provides a comparison of the time required for manual sampling and off-line testing with automated sampling and on-line testing. Turnaround time to results was 10 days for the manual process compared to 0.16 days for the automated process. This immediacy of data and near real-time analysis creates the opportunity to significantly accelerate process development.
Comparability of Data from Automated and Manual Sampling Processes
As shown in Figure 2, titer concentrations over the 14-day evaluation period were similar for manual sampling with off-line analysis and automated sampling with on-line analysis. The MAST® Autosampling Solution was operated independently of staff schedules as it does not require manual intervention. The same figure also shows that coupling the MAST® Autosampling Solution to bioreactors had no negative impact on cell titer. In addition, no contamination events were observed across 25 evaluations using the automated sampling process.
Figure 2.Bioreactor titer sampled manually or with the MAST® Autosampling Solution
CEX, SEC, and MAM by LC-MS results were comparable for off-line and on-line methods (Figure 3). Both manual and automated methods successfully differentiated product quality of process runs under different bioreactor conditions (standard and elevated run temperatures). The differences in percent high-molecular weight species between off-line and on-line methods were attributable to scale differences between the purification methods.
Figure 3.Automated sampling with on-line analysis generated CEX (A, B and C), SEC (D and E) and MAM (F and G) datasets compared to results from manual sampling and off-line analysis.
Increasing Data Resolution with Automated Sampling
Bioprocesses often run for several days and critical events such as peak cell growth can occur at any time. Managing a manual sampling regime in these situations can be limited by staff availability. Similarly, data collection during time-course studies may be interrupted when staffing is challenging such as overnight and on weekends.
Automated sample collection overcomes this limitation as data from an experimental run can be collected more consistently and more frequently throughout the course of the study. This improves overall process understanding by enabling the capture of progressive variation with greater granularity. For example, more frequent analysis of bioreactor contents can reveal details about the trigger, course, and impact of detrimental or beneficial changes to culture conditions. The additional information allows identification of more impactful manipulations that improve efficiency and better-targeted interventions to correct deviations.
Galactosylation of protein biologics may strongly influence their function and is impacted by bioreactor conditions; changes can occur rather abruptly, and manual sampling is often insufficient to accurately reconstruct response dynamics. Figure 4 shows the changes in percent product galactosylation measured in samples drawn from a perfusion bioreactor subjected to a step-change in galactose concentration from 1 to 10 mM. Samples drawn manually or via the MAST® Autosampling Solution were analyzed by UHPLC to measure response in percent galactosylation.
Daily manual sampling showed an increase in galactosylation that appeared to occur evenly over 2 or more days (Figure 4A). With more frequent automated sampling (4x/day) using the MAST® Automated Sampling Solution, it was possible to pinpoint that most of the response occurred 12–24 hours after the boost in galactose followed by a slowdown (Figure 4B). This level of understanding of when and how process parameters trigger changes is essential to define critical product quality attributes.
Figure 4.Comparison of manual (4A) and automated high-frequency sampling (4B) to characterize increased galactosylation in response to step changes in galactose.
Conclusion
Use of PAT is essential to optimize bioprocesses and enable more rapid and well-informed, data-driven decisions. Slow, labor-intensive manual sampling is insufficient to leverage the full potential of PAT. Manual sampling also increases the possibility of handling errors and consumes valuable time and resources that are better invested in translating empirical insights into actions.
Results of the study described above underscore the accuracy and comparability of data generated with PAT automated through the MAST® Autosampling Solution. This system provides many benefits including:
- Robust, near real-time and on-line results to accelerate process development
- Elimination of contamination risk due to manual sampling
- Elimination of handling variation
- Reduction in development cycles
- Augmentation of data generation capabilities
- Programmed or on-demand sampling at any time
- Assurance that an experimental run is recorded in its entirety, at predefined sampling points
Use of the MAST® Autosampling Solution guarantees a uniform and consistent stream of data, providing a clear and uninterrupted view of the bioprocess. In addition to the dramatic reduction in turnaround time for results and the reduced sampling burden placed on staff, automated sampling can be rapidly implemented in an existing bioprocess.
To continue reading please sign in or create an account.
Don't Have An Account?