Contract Pharma: What are the major problems with biopharma sampling, the way it is traditionally done today…both with the mechanics, and also the statistics, modeling.  How does that affect the ability to control processes?
Clint Pepper: There are three primary issues seen in bioreactor autosampling. The first is loss of bioreactor sterility. Running bioreactors is an expensive and time-consuming process. Even rare sterility loss events have significant impact on a facility’s budget and timeline.

Sampling high cell density cultures is another high profile issue. This problem is related to moving liquids with very high solids content during sampling. Current sampling systems routinely have difficulty drawing samples for the required distance and they frequently become plugged with accumulated cell debris.

The third issue is loss of sample integrity and consistency where the act of autosampling changes the data. The result is a sample that is not representative of what is happening in the bioreactor.

Existing autosampling systems have consistently suffered from one or more of these three failings. In many of the laboratories and pilot plants, these autosampling systems sit idle.

Because there is no reliable autosampling solution, simple process control solutions that have been used in other industries for decades, are still not being implemented in biopharma.  Online tools such as dielectric spectroscopy and Raman have helped, but because they are an indirect measurement of physical processes in the reactor, a physical sample still needs to be taken.

CP: What led you to develop the Modular Automated Sampling Technology (MAST) platform? Please describe the platform and its components and how it works, and how it feeds into automation and control systems.
CP: We did not set out to design and build an autosampler technology. About four years ago, a key client, Pfizer, asked for our help in finding an engineering solution to autosampling problems. To overcome these issues, Bend Research developed multiple novel technologies.

The MAST system is designed to resolve the three key issues seen in current autosampling systems. To reduce the risk of sterility loss in the bioreactor, the sampling module, or Sample Pilot, has been specifically designed to be sanitized and dried after each sample is collected. The sanitization and drying of the Sample Pilot after each use creates a sanitized volume or cavity when the next sample is collected. This greatly reduces the risk of contaminating the bioreactor.

To date, over 2,500 samples have been collected in nine different facilities without causing a loss of sterility.

The MAST system overcomes the issue of sampling high cell density cultures by incorporating a positive displacement pump directly on the sampling module. This allows the sample to be pushed using hydrostatic pressure, which is significantly superior to vacuum draw. To prevent cell debris accumulation, the MAST system flushes and sanitizes all sample contact components of the system after each sample. In addition, the specially designed valve stems and valve body minimize crevices and abrupt line size changes to reduce areas where accumulation can occur.

To address the third issue of sample integrity and consistency, the MAST system maintains the sample in a closed system and minimizes the contact with gases and other fluids. Samples collected manually are very consistent with those collected by the MAST system.

Recently, the MAST system has been expanded to enable sampling from up to eight bioreactors, with the ability to distribute those samples to as many as four destinations. The same technologies successfully employed in the bioreactor sampling modules have been incorporated into these new hardware components. Analytical destinations that have been developed, include the Nova BioProfile FLEX, a Bend Research designed manual sample collection station, and a liquid handling system from Gilson Inc. Integration programs are in development for the Cedex, ViCell, Bio HT and select UPLC and LCMS instruments. A novel automated Cell Removal System is also being developed.

Many of our clients are interested in the automated collection of metabolite data and titer, which they intend to use for process feed control strategies and for monitoring product quality attributes.

CP: So, how long has this been in development and what key technologies drive it?
CP: As I mentioned, we have been working on this technology for more than four years. The focus of this last year has been the development of a multiplex system that can take samples from up to eight sources and send those samples to four destinations. The first system of this kind was successfully deployed for an external client earlier in 2014.

The novel components of this product include: the use of hydrostatic pressure rather than vacuum to move the sample; the novel valve stem and valve body design that limits cell accumulation; and the sanitization process that occurs following every sample.

CP: Who contributed the technology involved (software, control interface, data management, feedback loop?)
CP: Bend Research developed this technology in-house with valuable help from Pfizer, Lilly and others in the biopharma industry. We use Ignition software (Inductive Automation, Sacramento CA) for our operational monitoring, sample scheduling and data historian. The MAST database is Sequel-based (available from multiple sources). The MAST hardware is controlled using an Allen Bradley programmable logic controller (Rockwell Automation, Milwaukee WI). Samples can be collected, processed and retained using a Gilson (Gilson Inc., Middleton, WI) Liquid Handler solution, which has been fully integrated with the MAST system.

A new technology that we have been developing with two other clients is the Cell Removal System. This is an integrated module within MAST that can take an automatically collected sample and remove the cells, providing a sample that can be analyzed using devices such as UPLC and LCMS instruments.

CP: Have you seen increased interest from biopharma companies?
CP: The response by the industry has been tremendous. In the last few years, we have placed MAST prototype systems in eight external client facilities. During a recent MAST technical forum, we were joined by representatives from five major biopharma companies, including Pfizer and Lilly. During the call, each representative provided valuable user input that we will incorporate as we finalize the development of the autosampling system.

Data collected to date shows very tight consistency between samples collected manually and samples collected automatically by the MAST system.

Multiple partners have shown a high level of interest in the Cell Removal System and integrating it with the Sample Collection and Processing capability of the Gilson. This combination of capabilities could save dozens of man-hours each week that are currently dedicated to sample collection and processing.

CP: What are your plans for the future?
CP: The MAST system is part of our overall BioSolutions offering that is aimed at integrating on-line and at-line data collection and real-time data analysis, integration and visualization. This combination of integrated hardware and software has been installed and piloted in a state-of-the-art facility that we recently commissioned at our Bend, Oregon facility. This facility is available for clients to visit, to run non-proprietary cell lines for feasibility demonstrations and to develop and optimize bioprocesses for transfer to their clinical and commercial facilities.

The MAST system is in a pre-commercial, late stage development phase. We are continuing to refine the product based on user requirements and lessons learned from early adopter installations. The goal for this year is to install multiple MAST systems at client facilities that have integrated the lessons we have learned from experiments conducted in Bend’s bioprocess demonstration and optimization facility, as well as input from early adopters. Our long-term goal is to provide an automated aseptic sampling platform that can be used from the bench- scale to cGMP.

Our advancement in MAST and related software and laboratory space is just a portion of the Capsugel Dosage Form Solutions (DFS)/Bend Research investment in biologics. We have also implemented existing and emerging technologies to improve formulations and methods of processing large molecules in order to achieve certain goals, including moving away from cold chain storage, providing alternatives to lyophilization and enabling oral and inhaled protein delivery via particle and dosage form engineering.