Bucking the Trend

Not surprisingly, industries operating in a free market tend to emulate and implement successful practices from divergent industries. This tendency is particularly true for technology-based industries such as biopharmaceuticals and software development—two industries that are customer-focused and rapidly changing. Although the term "platform" dates back hundreds of years, the concept of developing a platform or framework off of which subsequent products can be launched originated in software development. Current industry leaders in software development have implemented this concept with great success. Over the past decade or so, biopharmaceutical companies have applied platform processes to various aspects of their operations. The concept has gained considerable traction within the field of bioprocess development as an increasing number of groups have implemented it and begun touting its benefits in quick succession. However, as with any idea, there are benefits as well as disadvantages that need to be assessed prior to and concurrent with implementation.

TOUTING THE PLATFORM

The central argument in favor of a platform process, particularly in the context of bioprocess development, is that it expedites project timelines. By leveraging similarities between new molecules, a platform process enables resources to be deployed efficiently, reducing costs while also shrinking timelines. Everything from cell-culture expansion steps to filter sizing can be grouped together to form a comprehensive, thaw-to-formulation platform process. Particularly in the context of preclinical and early-phase clinical production needs, platform processes allow for rapid development of molecules. Even for late-phase clinical production needs, a platform process provides a starting point for subsequent development, readying a process for regulatory approval and eventual commercialization.

More specifically, platform processes also allow bioprocess development groups to focus on those aspects of the process that require greater evaluation or development such as viral clearance and bioreactor harvest criteria. For instance, viral clearance must be demonstrated under the conditions employed during purification steps to ensure the safety of the end product. Orthogonal steps capable of removing viral contaminants by alternate modes of action must be employed, with the performance of each step quantified. Because each molecule is different, with different physiochemical properties such as its isoelectric point (pI), and purified under slightly different conditions (i.e., buffers, pH, and conductivity), viral clearance must be demonstrated routinely. Similarly, specific cell-culture conditions can degrade the desired product based on culture duration (i.e., temperature and pH), cell viability, and the onset of apoptosis.

Determining how these and other factors affect product quality is vital to successful transfer and implementation of a GMP process. It is crucial to understand these operations because of their potential to affect release of the final product. Therefore, both operations need to be explored in detail. With a platform process in-hand, many noncritical operations can be ignored and run based on the platform, focusing instead on critical operations such as the two described above. This ability to prioritize effectively while maintaining a high level of process understanding is another benefit of platform technology.

Another realized benefit of employing platform processes is that it simplifies the complexity of the interactions taking place between groups inside and more importantly outside of the process development group. Groups such as quality assurance, quality control, logistics/supply chain, and manufacturing can all benefit from process development's use of platform processes. For instance, vendors for everything from growth media to chromatography resins can be assessed once (audited and certified by quality assurance) and relied upon routinely because multiple programs will use the same raw materials and consumables. Similarly, specifications defined by quality control and material numbers defined by logistics can be defined once and be used for multiple programs. Implementation of a platform process also espouses confidence in manufacturing groups because preparing and executing batch records are simplified. In other words, all of these groups external to process development benefit from standardized, well-defined workflows that have been previously tested and shown to be effective; a hallmark of a true platform process. In fact, the better the platform, the more these desirable characteristics can be achieved.

PITFALLS, REMORSE, AND THE PATH FORWARD

Just as with any other beneficial procedure or method, there are complications and problems that result from over-reliance of the singular platform approach. For instance, proponents of platform processes often argue technical knowledge can be acquired and developed by continually improving the platform. This notion is a fallacy as often platform improvement is discouraged due to the effect these changes would have on external groups already operating in the platform framework. Furthermore, implementing subtle improvements become less and less likely as these potential changes require greater justification. Recognizing that innovation arises from a series of small, seemingly inconsequential steps, the inertia around modifying platform processes can lead to stagnation and inefficiencies over time.

In addition, with a singular platform process in place, molecules with "platform friendly" physical and/or chemical properties may be favored over other molecules. That choice might not appear damaging, but the link between a molecule's manufacturability and its efficacy is weak or nonexistent. So, funneling molecule after molecule through development based primarily on its ability to conform seems ineffective and out-of-touch, thereby questioning the validity of the system that spawned this result. Such a system is unable to self-correct as the bias for conformity is difficult to eliminate. A troubling extension of this bias is that certain molecules are effectively killed if they do not fit the platform or require significant development. This action is appalling because it represents a true disservice to patients and the very business of biotechnology. To terminate a program with potential life-saving properties because it does not fit the cookie-cutter mold is beyond egregious. When development groups refuse to work outside of the platform environment, their failure is everyone's failure.

With the profound long-term effect platform processes and the mindset that goes hand-in-hand with its uninterrupted implementation can have on bioprocess development, a different approach is needed. An approach is needed that leverages the technical knowledge and overall skill set not only to develop new processes, but also troubleshoot existing processes. Such expertise requires continued exposure to an ever-expanding assortment of unit operations, techniques, and methodologies. Think of it as a scientist's repertoire or toolbox that is continually updated and employed. Every method and every unit operation has to be mastered such that it can be reworked and reconfigured as needed.

Every set of conditions tested is another data point, and as data points are collected into a comprehensive database, decisions can be made for new processes. It is important to note that not all aspects of each unit operation or method have to be identified upfront before implementation. Such characterization comes over time, as certain parameters are locked down and others are varied to understand the design space and identify where the process can potentially fail. Instead, the focus can be on functionality and utility—what purpose might a particular step play and how. For example, defining the expectations for a capture step (i.e., purity, aggregation, and profile of impurities) can instantly exclude or highlight certain types of resin based on their performance. Applying this very notion systematically across all unit operations can identify weak spots that merit greater scrutiny while also providing several possible avenues to explore concurrently.

A key responsibility for development groups that can often be overlooked is troubleshooting; assisting manufacturing and technical services groups in addressing unexpected issues. Troubleshooting by definition requires a systematic approach to identify the source of a particular failure or potential problem. Underlying that logical, structured approach is technical knowledge, without which troubleshooting cannot be accomplished. A modular platform approach would in effect, expose development groups to a whole assortment of operations and methods, thus enhancing their overall technical knowledge and ability to troubleshoot. For example, exploring multiple ways in which cell lines can be transfected and screened can lead to a better understanding of how particular cell lines behave and what triggers might disrupt productivity or product quality. Similarly, experience with multiple formulation buffers can lead to the development and implementation of screening tools aimed at delivering the right formulation for a specific mode of administration, concentration, and stability. Working in a modular framework essentially builds a layered, powerful database that can be employed to rapidly develop processes without forgoing actual development, as the singular platform approach does.

CONCLUSIONS

The benefits of developing and instituting platform processes are undeniable because the versatility of the platform drives cost-effectiveness and overall efficiencies, especially for routine operations. However, once established, a platform approach can be difficult to modulate as incremental changes are heavily scrutinized particularly by key stakeholders outside of development. As a result, a singular platform technology can lead to stagnation if not routinely challenged and refined. An alternative approach, one that has a longer timeframe in mind, is to develop expertise around many distinct unit operations that can be assembled into different combinations. By pairing an understanding of the capabilities and limitations of each unit operation from seed train operations to ultrafiltration and chromatography, with platform evaluation techniques, new processes can be quickly constructed. By constructing a framework that is more modular in form than true platform technologies, a whole host of molecules can be produced, purified, and formulated rapidly without sacrificing efficiencies and process robustness. Monoclonal antibodies may be prevalent today, but that may not be true several years from now as technologies advance and the industry inches closer to personalized medicine demarked by the use of many different types of therapeutic molecules. Being ready for any molecule that comes out of research is not only prudent, but fundamentally a core responsibility shared by all development groups.

ACKNOWLEDGMENTS

The author would like to Luca Di Noto and Derek Adams for support and feedback during preparation of the manuscript.

PRATIK JALURIA, PHD, is an associate director in upstream development at Alexion Pharmaceuticals, Cheshire, CT, jaluriap@alxn.com.