In January, as the new coronavirus was rapidly spreading around the world, scientists at Moderna teamed up with the National Institutes of Health to pursue a potential vaccine based on an experimental genetic technology involving messenger RNA (mRNA), a molecule in every cell that helps translate DNA into biological functions.
The young Cambridge, Mass.-based biotech firm, which has built a Cloud-based platform designed to speed the discovery and development of mRNA therapeutics for a wide range of diseases, was able to quickly act after Chinese officials released the genetic code of the SARS-CoV-2 virus. Working at a record pace, the company delivered a candidate vaccine to the first human vaccine trials just two months later.
In July, Moderna became one of the first U.S. companies to enter Phase III of a clinical trial for a potential coronavirus vaccine, with the Biomedical Advanced Research and Development Authority (BARDA) pledging $955 million to its research and production.
Moderna’s accelerated vaccine push is the culmination of a decade-long quest to build a fully integrated digital biotech company that takes advantage of the latest Industry 4.0 technologies. Its futuristic manufacturing facility in Massachusetts features seamlessly integrated, Cloud-based IT systems and leverages robotics, the Internet of Things (IoT) and artificial intelligence (AI).
“I believe it is a competitive strategic advantage of a company, having our own manufacturing facility—from raw materials to shipping vials to clinical trials,” CEO Stephane Bancel told analysts during a conference call in May.
Moderna is among the emerging generation of biotech firms embracing innovation and seeking to disrupt the incumbent pharma industry, which is often plagued by legacy technology and siloed data, lengthy product-development cycles and risk aversion.
Meanwhile, the global race to find a vaccine against COVID-19 has placed an intense focus on the biotech space, as scientists are seeking to drastically reduce the average time to market and scale up manufacturing to meet global demand.
“The herculean efforts we are seeing from many stakeholders to develop a vaccine is being supported by increased funding from governments and private investors,” said Billy Sisk, life sciences industry manager at Rockwell Automation. “The average time spent to bring a vaccine from discovery to production is 10 years; to reduce this to somewhere between 18-36 months takes collaboration from many stakeholders and the use of disruptive technologies.”
The new approaches to vaccine manufacturing rely on recent advances in synthetic biology. New platform technologies combine gene synthesis, industrial automation, Cloud computing, bioinformatics and machine learning to accelerate all stages of the vaccine R&D pipeline.
“Synthetic biology technologies are empowering advances in biotech R&D across the entire industry,” said Sean MacLeod, CEO of Seattle-based FenoLogica Biosciences, a startup using high-throughput analysis and machine learning to identify the effects genes have on cell phenotypes.
“Recent advances in protein synthesis and genome editing have created the opportunity to rapidly accelerate biotech innovation, transforming the approach to vaccine research, pharmaceutical discovery and manufacturing,” he added.
DNA- and RNA-based platforms have been reporting the fastest results because they produce vaccines using synthetic processes that require no culture or fermentation.
“Traditional vaccine manufacturing processes, like the egg-based vaccine process, are almost 200 years old and are not vertically scalable,” said Dr. Parviz Shamlou, executive director and head of the Jefferson Institute for Bioprocessing at Thomas Jefferson University. “Advanced vaccines are vertically scalable and have the potential to become the basis for a trillion-dollar vaccine industry as we come out of COVID-19.”
Racing against the clock, scientists are evaluating more than 160 potential vaccines in clinical and pre-clinical trials, according to the World Health Organization.
In July, the U.S. said it would pay the pharmaceutical company Novavax $1.6 billion to expedite the development of 100 million doses while Sanofi and GlaxoSmithKline secured up to $2.1 billion as part of “Operation Warp Speed,” the U.S. government’s push to accelerate a vaccine.
Moderna plans to use BARDA’s funding to scale production, hiring up to 150 new workers, including manufacturing staff, engineers and clinical and regulatory personnel.
While investments in commercial manufacturing usually come in the later stages of the drug-development process, companies are now rushing to scale up and automate in parallel with clinical trials—seeking to shave off months from development to production, said Cynthia Pussinen, VP and general manager of life sciences and specialty chemicals at Honeywell.
In conversations with potential clients as well as in industry news, Pussinen is hearing the same message, she said. “I’m hearing exactly what Honeywell has been saying: Let’s bring this into the phases of development early and start preparing your manufacturing facilities for that scale up that will be needed inevitably.”
Honeywell helps biotech firms prepare manufacturing automation designs, leveraging the power of the Cloud, batch software running in the process controller and remote asset management. Pussinen said she could not name clients.
In the past, the complexity, regulation and cost involved in each of the vaccine development stages—exploratory, pre-clinical, clinical development, regulatory review and manufacturing—have dragged out the process, Rockwell’s Sisk said.
Advances in automating data analysis and improving visualization at each step can help ramp things up.
Once scientists identify optimal compounds, the process moves to clinical trials. Advanced analytics and data visualization of expected human response to the potential vaccines allow researchers to conduct tests at scale even before vaccine candidates are administered to patients. Finally, after regulatory approval, manufacturers can combine AI and sensor-based technologies to fine-tune the supply chain, avoiding disruption and the risk of products being spoiled in distribution, Sisk said.
One emergent trend at drugmakers is the use of digital twins, which work as a live replica of all physical processes, said Niels Thomsen. He is VP and global head of insight practice at the IT services provider Atos, which is collaborating with Siemens on a digital twin solution for pharma manufacturing.
Synced up with IoT sensors installed inside the physical plant, the digital twin generates volumes of complex data, giving an instant view of all details of the operations so that every step can be optimized.
For example, a simulation of the chemical-mixing process allows scientists to test multiple variants to make sure the right mix is always used in live production. It can save costs and speed up manufacturing.
“With a digital twin, you can produce more because the quality-assurance process is done in real time,” Thomsen said. “You’ll have less waste, you’ll use fewer resources.”
Meanwhile, Atos’ high-performance computers, which can count thousands of times faster than standard computers, are used for simulation, building predictive models, analyzing the progress of the disease or developing new treatments.
“These powerful machines are performing very demanding calculations that prove to be essential in today’s race against the clock,” Thomsen said.
AI and machine learning can also help researchers sift through vast digital libraries of previous studies and treatments, analyzing properties of thousands of pharmaceutical compounds to pinpoint a potential candidate.
For instance, the Allen Institute for AI’s Semantic Scholar team partnered with a consortium of tech leaders to launch the COVID-19 Open Research Dataset, which uses natural language processing to organize more than 130,000 articles about the virus for the global research community.
Insilico Medicine CEO Alex Zhavoronkov said the company is leveraging modern AI techniques of generative adversarial networks (GAN) and generative reinforcement learning (RL) to rapidly discover novel molecular targets in a variety of diseases and design molecules with the desired properties.
The Hong Kong-based startup has uncovered novel small-molecule inhibitors for the key COVID-19 protein and started pre-clinical development.
“The key to success in pharma AI is massive integration of the systems used to identify biological targets—systems that help design novel molecules, and systems that personalize the treatments and predict the clinical trials outcomes,” Zhavoronkov said. “We need one big pharma brain which can span the discovery and development cycles that take 10 years or even longer and can integrate clinical data back into target discovery.”
The formula for success that Zhavoronkov proffers rings true to Moderna.
Self-described as the first biotech company “born in the Cloud,” it relies on Amazon Web Services for seamless integration across its business, R&D, manufacturing processes and future commercial efforts.
In addition to the Cloud, the company’s strategy embraces integration, the IoT, automation, data analytics and AI.
“We decided from the beginning to build from the ground up a digital biotech,” Marcello Damiani, Moderna’s chief digital and operational excellence officer, said from his home office in France. “We make sure that the company is data centric and that we can get insights from this database to help us improve quality and efficiency and accelerate our learning.
“Once you have the automation, the Internet of Things and the integration on the Cloud, you have data that’s flowing, and you can start doing sophisticated analytics,” he said. “And, of course, the holy grail of all this: predictive analytics and machine learning.”
Just like a computer operating system that can plug and play interchangeably between different programs and apps, Moderna’s platform is taking advantage of the software-like features of mRNA, creating unique sequences that instruct cells in the body to make specific proteins that trigger an immune response.
The company’s drug-design studio contains a library of existing sequence components and embedded AI sequence optimization algorithms.
Once designed, the custom mRNA sequence is sent to a pre-clinical production team using an ordering app, which pinpoints the ideal properties for both the mRNA and its delivery formulations, and automatically performs several AI sequence quality checks and optimizations.
Next, mRNA production is triggered in a pre-clinical production app that orchestrates every step of the process, from the initial creation of the DNA plasmid template to the final formulated mRNA.
The technology allows the biotech firm to work on a record number of drug candidates simultaneously, said Damiani, who joined the company in 2015.
Since then, the company has gone from producing 40 mRNAs for research per month to more than 1,000.
“It means that we can do 1,000 experiments in parallel,” he said. “This gives you an idea how the digitization and the use of this technology can enhance drastically the efficiency of the scientists.”
Two years ago, Moderna opened its 200,000 square-foot manufacturing site in Massachusetts with the capacity to develop materials for pre-clinical toxicology studies, as well as Phase I and II clinical development programs, and to manufacture, test and run fill/finish operations for its mRNA candidates.
Last year, the site won the Facility of the Future award from the International Society for Pharmaceutical Engineering. The company’s approach to integrating digital technologies into its work flows and processes brings the industry “to a new level in the digital era,” the society said.
Inside, scientists are deploying automation and robotics platforms with movable arms and pipettes that can take instructions and relay data back, using the IoT for inventory replenishment.
The plant’s “workhorse” is the Hamilton liquid handling robot, a general-purpose platform with a reconfigurable deck for different equipment. It allows for dozens of unique methods or operations within the production cycle.
Moderna also uses Zebra label printers, as well as robotic freezers that store tubes of material at various stages of the process, including minus-80 Celsius Hamilton Sam freezers and a minus-20 Celsius Liconic freezer.
“We implemented automation as we matured our processes. If you go to our pre-clinical production manufacturing facility, you will see that it’s fully automated. If you go to our clinical manufacturing, although it’s fully digitized, it’s less automated. We have islands of automation, because we are still evolving our processes,” Damiani said.
In May, Moderna signed a 10-year partnership agreement with Lonza, a contract development and manufacturing organization, with the goal to produce one billion doses annually at sites in the United States and Switzerland. The idea is to replicate Moderna’s digitization strategy at scale, he said.
The company’s Phase III trial, which is expected to enroll 30,000 participants in the U.S., will determine if the vaccine, known as mRNA-1273, will be green-lighted for use in the general public.
The biotech industry has a long-term commitment to advanced biomanufacturing, including digital biomanufacturing, said Shamlou of Thomas Jefferson University. As the industry moves from batch and semi-batch to continuous bioprocessing, process automation and in-line, at-line and on-line measurements of process parameters and their control will play important parts of process design and development.
“The ability to rapidly measure and reliably process large amounts of data in real time and then use the information to make sound process decisions with little to no human interference are at the heart of true digital manufacturing,” he said. “That journey started prior to COVID-19, and will continue to accelerate post COVID-19.”
The pressure on drugmakers like Moderna to deliver rapid solutions will continue to accelerate the industry’s digital transformation for years to come, fueling more robust use of AI and predictive analytics, smarter distribution and blockchain technology, Honeywell’s Pussinen said.
“If we were to look across the industry at various systems that are used in manufacturing, what we see right now is the data contained in each individual system,” she said. “ If we are able to look at it all together, think about the power of using that data.
“A more connected approach to the entire industry is starting and will be growing. It’s a very exciting time to be in the life sciences industry, for sure.”
To listen to this article, visit www.sme.org/vaccine.
Connect With Us