A Biopharmaceutical Industry Perspective on Single-Use Sensors for Biological Process Applications

, , , , , , , , ,
BioPharm International, BioPharm International-04-01-2015, Volume 28, Issue 4

This article presents first-hand perspectives from industry users to suppliers of single-use sensors.

 

Single-use equipment (also referred to as disposables) is providing the biopharmaceutical industry with more flexibility and enabling the faster development and marketing of products (1, 2, 3). At the same time, single-use equipment needs suitable single-use sensors that meet requirements similar to traditional durable equipment operations (4, 5). Industry practitioners would like “plug-and-play” single-use equipment. While improvements to single-use sensors have been evolving, experiences have shown that there are still more opportunities for improvement to achieve the requirements necessary for successful implementation in the manufacturing setting. This article presents first-hand perspectives from industry users to suppliers of single-use sensors. It provides insight regarding important industry requirements that will enable more successful implementation of single-use technology. 

The information presented herein evolved from a Pharmaceutical Process Analytics Roundtable (PPAR) meeting. PPAR is a forum for practitioners from within the biopharmaceutical industry who meet each year to discuss common areas of interest and to exchange experiences related to process analytical technology (PAT). One outcome of the 2013 meeting was a common interest among the PPAR participants in communicating the perspectives of the industry with regard to its need for successful single-use sensor applications. Several of the meeting participants had experienced challenges when implementing single-use sensors, and although these sensors continue to evolve with improvements, there are still challenges to using them. Improvements are necessary to meet user needs and process requirements. And, while there is progressive adoption of the technology for use in bio-product development, it lags in the manufacturing sector, partially due to the need for more robust cost-effective single-use sensors.  

In this article, “single-use” terminology is used to match common industry nomenclature such as SUBs (single-use bioreactor or bag) and SUMs (single-use mixers), but the term “disposable” could be used interchangeably. The main focus of the authors is on the sensor that is in direct contact with the product and is integral to the single-use “bag” portion of the equipment, such as a fluorescent dissolved-oxygen (DO) sensor. The main requirement is a compatible sensor that fully enables the cost-effective operation of single-use equipment. No matter what form a sensor may take, the expectations discussed in this article remain the same, and all forms are categorized as single-use sensors. The sensor can take the form of a durable reusable device, for example, that is enabled by a membrane or is capable of sensing the measured attribute through the bag while maintaining sterility within the bag. Similarly, it can take some other form, as long as it achieves the key objectives expected during its use. Furthermore, these expectations exist for single-use sensor applications in both development and manufacturing settings that can be adopted for large- and small-molecule platforms.  

The overall objective of this article is to encourage the suppliers of sensors to help industry practitioners reduce implementation time, which will increase the likelihood of success in using their products. This success is expected to result in a greater number of installed single-use sensors in a variety of applications. 

Survey of applications

Participants of the PPAR shared their experiences with single-use sensors without mentioning specific suppliers or brand names. The most commonly described applications were those that employed single-use sensors for pH and dissolved oxygen (DO). Consequently, they were the most often discussed applications that presented measurement challenges. 

Generally, single-use pH sensors that are integrated with single-use equipment are based on optical principles. Single-use pH sensors can be applied in almost all biological processes from initial seed build, to bioreactors, to downstream isolation and purification of the drug substance.  Likewise, single-use pH sensors are applicable across the lifecycle of a product: from lab-scale development, to clinical pilot scale and, finally, to full commercial-scale manufacturing. Among the PPAR participants, single-use sensors are generally only used in development and clinical pilot-scale operations. There are concerns that the single-use pH sensors are not robust enough for manufacturing operations. Some of the challenges encountered with using single-use pH sensors include the following:

• Lot-to-lot variability in the materials used in sensor construction causing variations in sensor performance

• A need for sensor equilibration time after wetting

• Unknown strength of optical signal drifting with time in various applications

• A need for truly calibration-free sensors with self-diagnosis to understand what could possibly cause an off measurement.  

Single-use DO sensors are most commonly used in upstream operations. They are normally more robust than pH sensors and do not need frequent adjustments. During PPAR discussions, however, participants agreed that there is a need for more guidance regarding the impact of anti-foams on DO sensor performance. 

There are challenges that are common to both pH and DO sensors: 

• Lack of a system for tracking the number of light pulses (This results in sensor drift and it would be valuable to incorporate a correction factor)

• Difficulties recalibrating the transmitter when sensor drift occurs

• Connection problems between the optical fiber and the sensor (This results in signal attenuation due to light scatter and a lack of robustness)

• Inadequate/insufficient interface with existing data and standard industry control systems

• Need for different sensor and transmitter configurations based on scale

• Lack of data and documentation showing the effects for various sterilization methods (e.g., steam sterilization, gamma irradiation, ethanol, ethylene oxide, peroxide)

• Impact of air during gamma irradiation causing leachable or extractable compounds that compromise the product  

• Sensor chemistry effects due to the glue used and procedures for affixing to the single-use equipment

• Redundant pH and DO capability should be incorporated in SUBs.

As users of these technologies, we agreed that it is imperative for the manufacturers and suppliers of single-use sensors to be aware of our concerns and realize the challenges related to sensor use. Beyond that, suppliers should address these concerns, and all challenges should be explored and mitigated before their products are marketed for use in the industry. Expert guidance, dependable information, and real recommendations for success should accompany the products into the field.

Ease of use

Ease of use is important when scientists and engineers first evaluate new equipment, but it becomes essential when technicians routinely use the equipment in a pilot plant or a manufacturing setting. Lab environments are flexible, and scientists can more readily adapt to stand-alone equipment and custom user interfaces. Systems operated by technicians (in product development or manufacturing), however, need to be more robust and easily operated. Therefore, ease of implementation is also important, and that is why standardized plug-and-play devices have tremendous benefits, as they are easily adaptable in traditional systems.

When evaluating a new single-use sensor or transitioning to full deployment in a pilot plant or manufacturing setting, technician training and standard operating procedure/batch record revisions to support the new system become more difficult when the setup and user interface differ considerably from legacy systems already in use. It can also be challenging to deploy a dedicated computer and software to run a stand-alone system in some settings. One key aspect that affects robustness is the suitability for routine use in industrial environments. Optical single-use sensors requiring fiber optic connections, for example, have on some occasions been negatively affected by routine activity that led to the cable or connector being incidentally pulled or bumped.

User interface with single-use sensors should mimic that of standard instruments. A standard transmitter (or a reasonably comparable equivalent), for example, should be used for connectivity and for sensor setup. Furthermore, as some sensors will be used in pilot plants or manufacturing settings, sufficient operational robustness in an industrial environment is required, and hardware specifications, such as being able to mount systems in standard rails, must be considered.

Integration

Easy integration of a new single-use sensor with the practitioner’s existing industrial control systems and/or process data historians will enable trouble-free implementation and ultimately result in success of use. Experience suggests that suppliers of single-use sensors may not be fully aware of, or do not always choose to follow, the industrial standards used by traditional sensors. While new single-use sensors can be used as stand-alone equipment in many lab applications, incorporating industrial standards into the design of the sensor would allow for its seamless integration with existing legacy systems used in manufacturing operations. Integration with industrial systems would also facilitate data transfer to online databases and process data historians.

Suppliers should provide single-use sensors that follow industrial communication protocols (e.g., ModBus, ProfiBus, etc.) and communicate with standard industrial equipment (e.g., DeltaV, PLCs) using OPC standard. It is recommended that suppliers collaborate with the companies that develop integration systems so that such features can be incorporated and be available as a standard option on sensors. 

Overall, “off-the-shelf” systems, which provide the benefits of reducing implementation time and increasing the probability of success using the system, are preferred.

GMP implementation

It is important to demonstrate that sensors do not significantly leach or extract any deleterious foreign chemicals to the process under normal and extreme process conditions of pH, temperature, and time. Also, when transitioning from a traditional sensor to a new single-use sensor (e.g., pH) in a manufacturing setting, a comparability assessment may be required to establish that the measurement capability of the new single-use sensor is acceptable when compared with the legacy sensor.

Single-use sensors must be fabricated and assembled in facilities that meet good manufacturing practice (GMP) quality standards if they are to be used in GMP production. Besides Class VI certification, Code of Federal Regulations, Title 21 Part 11 and other ANSI/AAMI/ISO guidelines (6) and certifications, suppliers should ideally generate extractable data under a model matrix of test conditions and solutions to demonstrate that the extractable from any new single-use sensor contacting materials is acceptable. Reliable information should be provided regarding the process conditions under which the sensor should not be used. Absence of such information and supporting data makes the process clearance and potential drug substance safety impact assessment by the practitioner more difficult to complete, and it may result in an assessment that is not as detailed and quantitative as that which is required for GMP use.

Formal documentation of the single-use sensor’s measurement uncertainty should also be included with the sensor, along with information about the expected precision and accuracy and robustness of the sensor measurement under typical use. 

Additional perspectives

The field of sensors continues evolving with the contribution of specialized fields such as optics, chemistry, electronics, and computers. The biopharmaceutical industry needs more novel sensors for more efficient development and operation of processes and to ensure product quality. Hence, a desired list for new sensor development is useful to guide suppliers to meet current and future needs. Areas were identified that are impeding the implementation of single-use sensors into GMP (and are, therefore, opportunities for improvement that could provide suppliers with a competitive advantage to enable a more rapid industry acceptance).  Collectively, these areas include connectivity, standard interfaces, accuracy and reproducibility, mechanical durability, pre-calibration with in-situ recalibration, and options for sterilization. 

Although single-use pH and DO sensors are already on the market, there are two aspects that industry practitioners would like to have commercially available:

• Single-use pH sensors with wider range measurements and long (years) shelf life. The currently available single-use pH sensors have a narrow range of measurement, which limits its applicability to upstream operations. A constant supply of some sensors that are critical to the operation of processes must be maintained; therefore, a two- or three-year shelf life of a sensor will provide flexibility in managing the inventory for multiple development projects and commercial processes.

• Harmonization that enables fast implementation. Harmonization among sensors can be implemented on three different levels:

  • Mechanical-the connection, adaptation, or mechanical interface of a sensor in the equipment to be monitored should be standard so that sensors from different suppliers can be installed in the same single-use equipment. 

  • Electrical-wired or wireless interface of sensors should follow industry standards. The interchangeability of a wired or wireless set up provides more flexibility to the industry.

  • Software-use of industrial communication protocols is recommended to communicate with standard industrial equipment.

Other single-use sensors that are commercially available, such as pressure, flow, and cell density, are becoming more common in research and development. Some future improvements in these other single-use sensors that would benefit the biopharmaceutical industry include the following:

• Precalibrated sensors and self-diagnostics. The perception that the single-use sensor is more convenient and more efficient to use than standard re-usable devices makes it more attractive. Sensors that are pre-calibrated and ready to use “out of the box” will provide a beneficial time savings to development and manufacturing. In the same way, self-diagnostics can efficiently provide assurance that the sensor is working properly or alert the user when it malfunctions, without the need for a highly trained technician to repeatedly check sensor performance.

• Preinstalled in single-use equipment (i.e., SUBs, SUMs). Some single-use sensors are purchased and received directly from suppliers and are subsequently installed in-house, compromising process sterility. It is recommended that suppliers of the single-use equipment and those who provide the sensors collaborate to deliver a complete single-use system with trusted integrity.

Finally, to match traditional analytical capabilities, the following is a list of single-use sensors that could make a significant difference in the biopharmaceutical industry:

• Single-use capacitance, optical density, and spectroscopy-based sensor show a great deal of promise in bioprocessing applications. Simple univariate sensors have become inexpensive commodities in the automotive industry. The bridge from there to the biopharmaceutical industry may be a simple change in the form factor. Even small spectrometers, such as mid-infrared, near-infrared, and Raman, are being discussed in these same terms, with the promise of low cost while still providing high selectivity (7-10).

• Single-use “actuators” that enable sampling and delivery from single-use equipment to other bench-top equipment. The sampling devices that are currently available are bulky or are aftermarket add-ons. Offline sampling using single-use equipment is still needed and a streamlined and inexpensive sampling delivery system is desired.

• Single-use refractive index sensors offer quick measurement of some media composition as it changes over time in a process. If such a single-use sensor can be designed and developed, it would offer an alternative to offline sampling of some media components. Alternatively, it can also be used as a single-use temperature sensor.

• Single-use protein concentration sensors (from low to high concentration) would contribute to efficiency in upstream and downstream operations within the industry. Meanwhile, single-use on-line viability, foam detector, and CO2 sensors can positively impact the efficiency of SUBs.

Conclusion

While there have been favorable successes using single-use sensors in single-use equipment, a number of challenges have nonetheless been experienced by several PPAR participants. The authors hope that suppliers of single-use sensors will use this information to address the various challenges that they encountered while using this valuable technology. Hopefully, future sensors for measuring additional attributes added to the single-use sensor analytics portfolio will be as good as or better than traditionally available analytics. Ultimately, the supplier who addresses these challenges will have a competitive advantage when selling to the biopharmaceutical industry. 

Disclaimers

The views expressed in this paper are the personal views of the contributing authors and do not necessarily reflect the official position of their respective organizations. 

References

1. R. Eibl and D. Eibl, Eds., Single-Use Technology in Biopharmaceutical Manufacture (Wiley, 2011).

2. W. Ding, and J. Martin, BioProcess International  7(5), 46-51 (2009).

3. T. Sandle, and M.R. Saghee, M.R., Journal of Commercial Biotechnology 17(4), 319-329 (2011).

4. H. Weichert, J. Lueders, and M. Becker et al., BioProcess International 12(5), 20-24 (2014).

5. I. Bauer, I. Poggendorf, and S. Spichiger et al., BioProcess International 10(8), 56-61 (2012).

6. G. Calafiore, A. Koshelev, and S. Dhuey et al., Light: Science & Applications 3, 203-210 (2014).

7. Y. Park and S.H. Choi, Journal of Nanophotonics 7(1), (2013).

8. A. Nitkowski, K.J. Preston, and N. Sherwood-Droz et al., “Sensing Systems using Chip-based Spectrometers,” Proceedings of SPIE 9083, (2014), p. 908332.

9. J. Malinen, A. Rissanen, and H. Saari et al., Proceedings of SPIE, 9101, (2014), p. 91010C.

10. W. Ding, G. Madsen, and E. Mahajan et al., Pharmaceutical Engineering, 34(6), 74-853, (2014).

About the Authors

Bradley H. Diehl is manager, Process Analytical Sciences Group, Pfizer.

Mark A. LaPack is research advisor, Small Molecule Design & Development, Eli Lilly and Company.

Tony Y. Wang is senior engineer, Digital Integration and Predictive Technologies, Amgen.

Robert E. Kottmeier is senior manager, BioProcess R&D Manufacturing, Pfizer.

Stacey M. Kaneshiro is associate engineering advisor, Bioproduct Research & Development, Eli Lilly and Company.

Michael C. Brandenstein is scientist, Amgen.

Yongchun Zhang is principal engineer, Bayer Business and Technology Services.

Yuk Chun Chiu is group head PAT Biologics at Bayer Technology Services, Bayer HealthCare.

Seongkyu Yoon is associate professor and director of Biomanufacturing Center University of Massachusetts.

Victor M. Saucedo is senior engineer, Process Development Engineering, Genentech, saucedo.victor@gene.com.

Article DetailsBioPharm International
Vol. 28, Issue 4
Pages: 28-31, 53
Citation: When referring to this article, please cite it as B. Diehl et al., "A Biopharmaceutical Industry Perspective on Single-Use Sensors for Biological Process Applications," BioPharm International  28 (4) 2015.