Ben Locwin, Contributing Editor04.20.17
Compared with just two years ago, biomanufacturing continues to advance and evolve. For example, according to BioPlan Associates’ 2015 annual report, more than 90% of facilities were using single use/disposable technologies, and planning and future modeling was designed to balance out the various costs and risks associated with disposable vs. stainless steel technologies for the long term.
Is technology the end-all?
Currently, more than 9 out of 10 contract manufacturing organizations (CMOs) are using disposables, and they are continuing to see benefits, though they are also reporting some dissatisfaction with the offerings and future trajectory of the disposables market.
Oftentimes, technology is erroneously considered a proxy for the hard work and activities in the industry. It’s easy to point to something like wearable tech, social media platforms, single-use disposables and hold that as a mental heuristic for the cause of pharma industry improvements. But as the saying goes, “The map is not the territory,” and just because process improvements are happening concurrently and coincidentally with the use of disposables and other technology, doesn’t mean that those are the key drivers of that improvement.
Most organizations continue to report that they are making improvements to their processes and practices internally and methodologically. Lean and Six Sigma process improvement practices continue to be adopted, refined, and pushed through organizations, which leads to greater scrutiny of all business practices, and opportunities to change how work is approached. But because internal practices and methodologies aren’t measurable industry-wide in the same way that total units sold of single-use technologies are, these work practice evolutions tend to be minimized and under-reported, when in reality are the unsung heroes of industry improvement.
Increasing confidence in analytical testing continues in the industry, and the thinking tends to go that better analytics and higher-precision analytical results leads to better performance, which isn’t necessarily true. In the last two years, however, there has been extraordinarily more focus and work on biosimilars in the industry. The current analytical requirements for biosimilars are generally not stringent enough, and according to many industry experts, perhaps not similar enough to the trial rigor required of the innovator drug to make biosimilars an area of continued large-scale refinement.
Biotherapeutic developers and CDMOs are tending to see better results than CMOs by refinements in how cell cultures are optimized, and how developability is approached in ways that can harness advancements in risk assessments to increase the likelihood of performance of candidate molecules.
Technology in pharma and biomanufacturing
Gartner’s Symposium in Quarter 4, 2016 identified a number of technological platform advances that will in some form permeate many business models, and pharmaceutical manufacturing is no different.
AI and machine learning. Artificial intelligence (AI) and machine learning involve various levels of deep learning, neural networks, natural language processing (NLP—not to be confused with neurolinguistics programming), and other non-algorithmic-based rule systems. I have used a few of these systems being piloted for clinical trials to help refine designs of trials and ‘learn’ from real-time trial data.
Intelligent apps. Clearly, there are benefits to apps, which remind patients to, for example, take their medications, or adhere to some other healthcare treatment protocol. But apps that can ‘learn’ continually from ever-changing input data can provide context and suggestions to users when used for more complex applications, such as in the clinical trials oversight as mentioned above.
Virtual and augmented reality for pharma. Using virtual reality (VR), or augmented reality, for medical training is something that is just beginning to occur in a compelling way. I have used some of these technologies firsthand, and their ability to impact how training and understanding of complex things occurs, is monumental. The major drawback here is that it’s a ‘worst practice’ to simply take content that would have been taught in a lecture or online and force it into three dimensions of VR. Just as there is a ‘right’ modality for certain training applications, VR and AR are particularly suited for complex visuals where 3-D manipulation can help deeper understanding.
Digital twin. If this is the first place and time you’re hearing of this term and technology, I can guarantee that it won’t be the last! A digital twin is a dynamic model of a physical system that uses data collected by sensors to interpret a sense of its state, develop responses to environmental changes, and improve operations. Top-level digital twins have input from three major sources: metadata of classification and composition; data about current status (temperature, spatial location, time series, mass, velocity); and onboard analytics to develop responses (algorithms and adaptive learning rules for various input levels). Preventive maintenance and instrument/equipment status can be predicted and ‘operationanability maxima’ can be achieved (i.e., as in a design-of-experiments view, there are certain operating ‘maxima’ that give the best possible outputs given a series of inputs).
Conversational systems. Imagine optical character recognition (OCR) on steroids. OCR is a machine system, which can ‘read’ text, say in printed form, and use it to fill out fields and record data. Conversational systems use ambient meshes to ‘listen’ to discussions and dialogs, and then build machine learning from it. Being able to access cloud or stored data can also allow the conversational system to build sorts-of ‘mind maps’ with hyperlinked data and programs/apps that would link together complex chains of reasoning.
Mesh app, service architecture, and adaptive security. Unless you’re an expert in IT and security technology, you may have just had your eyes glazed over. But there are some very innovative features here. A mesh app would take many back-end services of a company or organization and link them together—desktop, telephone, smart devices, internal personnel measurement systems, biomanufacturing distributed control systems (DCS), and any other services. It would all be scalable and allow a continuous experience for each user in one place. Of course, the latter part of this section’s title refers to the fact that these deep linkages present huge challenges for digital security. Continuous monitoring of user behavior and system behavior has been deemed a requirement of good security, but also poses Big Brother privacy concerns for most users. This is the antecedent to good platforms and security measures in the Internet of Things (IoT).
Conclusion
So the major themes above can be grouped into ‘intelligence everywhere,’ ‘intertwining of the digital and the real,’ and the ‘mesh of platforms.’
Of course, there’s a tenet in pharma biomanufacturing that the industry lags other high-tech industries by about 20 years. Oftentimes, the reason most cited is that pharma manufacturing is shackled because of regulatory compliance requirements (FDA, EMA, MHRA, etc.), but many of the other competitively advanced industries—aerospace, food and beverage, automotive—are similarly regulated and in some cases to a higher degree.
What is clear is that we will continue to see an industry in flux, and how we approach biomanufacturing as an industry will depend on many factors, not the least of which includes a new presidential cabinet in the U.S., pressures to reduce drug prices, drug pricing negotiation rules applied between pharma and payers, and a new commissioner at FDA.
References
Ben Locwin
Contributing Editor
Ben Locwin, PhD, MBA, MS, is a contributing editor to Contract Pharma and the president of Healthcare Science Advisors. He is an expert contact for the American Association of Pharmaceutical Scientists (AAPS), a committee member of the American Statistical Association (ASA), and has been featured by the CDC, the Associated Press, The Wall Street Journal, Forbes, and other media outlets. Follow him at @BenLocwin.
Is technology the end-all?
Currently, more than 9 out of 10 contract manufacturing organizations (CMOs) are using disposables, and they are continuing to see benefits, though they are also reporting some dissatisfaction with the offerings and future trajectory of the disposables market.
Oftentimes, technology is erroneously considered a proxy for the hard work and activities in the industry. It’s easy to point to something like wearable tech, social media platforms, single-use disposables and hold that as a mental heuristic for the cause of pharma industry improvements. But as the saying goes, “The map is not the territory,” and just because process improvements are happening concurrently and coincidentally with the use of disposables and other technology, doesn’t mean that those are the key drivers of that improvement.
Most organizations continue to report that they are making improvements to their processes and practices internally and methodologically. Lean and Six Sigma process improvement practices continue to be adopted, refined, and pushed through organizations, which leads to greater scrutiny of all business practices, and opportunities to change how work is approached. But because internal practices and methodologies aren’t measurable industry-wide in the same way that total units sold of single-use technologies are, these work practice evolutions tend to be minimized and under-reported, when in reality are the unsung heroes of industry improvement.
Increasing confidence in analytical testing continues in the industry, and the thinking tends to go that better analytics and higher-precision analytical results leads to better performance, which isn’t necessarily true. In the last two years, however, there has been extraordinarily more focus and work on biosimilars in the industry. The current analytical requirements for biosimilars are generally not stringent enough, and according to many industry experts, perhaps not similar enough to the trial rigor required of the innovator drug to make biosimilars an area of continued large-scale refinement.
Biotherapeutic developers and CDMOs are tending to see better results than CMOs by refinements in how cell cultures are optimized, and how developability is approached in ways that can harness advancements in risk assessments to increase the likelihood of performance of candidate molecules.
Technology in pharma and biomanufacturing
Gartner’s Symposium in Quarter 4, 2016 identified a number of technological platform advances that will in some form permeate many business models, and pharmaceutical manufacturing is no different.
AI and machine learning. Artificial intelligence (AI) and machine learning involve various levels of deep learning, neural networks, natural language processing (NLP—not to be confused with neurolinguistics programming), and other non-algorithmic-based rule systems. I have used a few of these systems being piloted for clinical trials to help refine designs of trials and ‘learn’ from real-time trial data.
Intelligent apps. Clearly, there are benefits to apps, which remind patients to, for example, take their medications, or adhere to some other healthcare treatment protocol. But apps that can ‘learn’ continually from ever-changing input data can provide context and suggestions to users when used for more complex applications, such as in the clinical trials oversight as mentioned above.
Virtual and augmented reality for pharma. Using virtual reality (VR), or augmented reality, for medical training is something that is just beginning to occur in a compelling way. I have used some of these technologies firsthand, and their ability to impact how training and understanding of complex things occurs, is monumental. The major drawback here is that it’s a ‘worst practice’ to simply take content that would have been taught in a lecture or online and force it into three dimensions of VR. Just as there is a ‘right’ modality for certain training applications, VR and AR are particularly suited for complex visuals where 3-D manipulation can help deeper understanding.
Digital twin. If this is the first place and time you’re hearing of this term and technology, I can guarantee that it won’t be the last! A digital twin is a dynamic model of a physical system that uses data collected by sensors to interpret a sense of its state, develop responses to environmental changes, and improve operations. Top-level digital twins have input from three major sources: metadata of classification and composition; data about current status (temperature, spatial location, time series, mass, velocity); and onboard analytics to develop responses (algorithms and adaptive learning rules for various input levels). Preventive maintenance and instrument/equipment status can be predicted and ‘operationanability maxima’ can be achieved (i.e., as in a design-of-experiments view, there are certain operating ‘maxima’ that give the best possible outputs given a series of inputs).
Conversational systems. Imagine optical character recognition (OCR) on steroids. OCR is a machine system, which can ‘read’ text, say in printed form, and use it to fill out fields and record data. Conversational systems use ambient meshes to ‘listen’ to discussions and dialogs, and then build machine learning from it. Being able to access cloud or stored data can also allow the conversational system to build sorts-of ‘mind maps’ with hyperlinked data and programs/apps that would link together complex chains of reasoning.
Mesh app, service architecture, and adaptive security. Unless you’re an expert in IT and security technology, you may have just had your eyes glazed over. But there are some very innovative features here. A mesh app would take many back-end services of a company or organization and link them together—desktop, telephone, smart devices, internal personnel measurement systems, biomanufacturing distributed control systems (DCS), and any other services. It would all be scalable and allow a continuous experience for each user in one place. Of course, the latter part of this section’s title refers to the fact that these deep linkages present huge challenges for digital security. Continuous monitoring of user behavior and system behavior has been deemed a requirement of good security, but also poses Big Brother privacy concerns for most users. This is the antecedent to good platforms and security measures in the Internet of Things (IoT).
Conclusion
So the major themes above can be grouped into ‘intelligence everywhere,’ ‘intertwining of the digital and the real,’ and the ‘mesh of platforms.’
Of course, there’s a tenet in pharma biomanufacturing that the industry lags other high-tech industries by about 20 years. Oftentimes, the reason most cited is that pharma manufacturing is shackled because of regulatory compliance requirements (FDA, EMA, MHRA, etc.), but many of the other competitively advanced industries—aerospace, food and beverage, automotive—are similarly regulated and in some cases to a higher degree.
What is clear is that we will continue to see an industry in flux, and how we approach biomanufacturing as an industry will depend on many factors, not the least of which includes a new presidential cabinet in the U.S., pressures to reduce drug prices, drug pricing negotiation rules applied between pharma and payers, and a new commissioner at FDA.
References
- BioPlan Associates. (2015). Twelfth annual report and survey of biopharmaceutical manufacturing capacity and production.
- Zurdo, J., Arnell, A., Obrezanova, O., Smith, N., Gallagher, T., de la Cuesta, R., Locwin, B. (2015). Developability assessment workflows to de-risk biopharmaceutical development. in Kumar, S., & Kumar Singh, S. (eds.). Developability of biotherapeutics: Computational approaches. CRC Press.
Ben Locwin
Contributing Editor
Ben Locwin, PhD, MBA, MS, is a contributing editor to Contract Pharma and the president of Healthcare Science Advisors. He is an expert contact for the American Association of Pharmaceutical Scientists (AAPS), a committee member of the American Statistical Association (ASA), and has been featured by the CDC, the Associated Press, The Wall Street Journal, Forbes, and other media outlets. Follow him at @BenLocwin.
