(SciePro/Shutterstock)
The cloud is a natural place for startups with large-scale computing and data storage needs but an uncertain future. For two startups aiming to stop cancer, including Lyell Immunopharma and Hurone AI, the public cloud powered by Amazon Web Services is an ideal place to foster promising but unproven new technologies.
Lyell Immunopharma was founded in 2018 with the goal of developing processes for manufacturing customized cell therapies to treat cancer. This type of cell therapy has had some success in blood cancers, and Lyell is one of the first startups to see if it can also help treat solid tumors.
This San Francisco-based biotechnology has developed a promising method to reactivate individual patients’ T cells and turn them into cancer killers. As part of the Phase 1 clinical trial, Lyell will build her 72,000-square-foot manufacturing facility in Bothell, Washington, where she can test her proprietary cell-reprogramming technology.
Manufacturing custom cell therapies is challenging enough, but meeting the regulatory requirements of the US Food and Drug Administration is even more challenging. In order to not only develop high-quality products that fight cancer, but also to ensure they meet FDA safety standards, Lyell needs to collect and process large amounts of data, said Lyell’s CEO, Elizabeth Homans said.
“There are a lot of handovers from different stakeholders,” Homans said. Data Nami“Manufacturing plants have a lot of activities going on. Applying genetic and epigenetic reprogramming techniques, complying with regulatory specifications, quality control, etc., all take up an enormous amount of time.”
This is where AWS comes into play. While the new factory was under construction during the COVID-19 lockdown, Lyell worked with his AWS to collect and process data from every step of the manufacturing process, Homans said. says.
Lyell Immunotherapy develops customized cancer treatments by training patients’ own cells (Image credit: Lyell Immunotheraphy)
“AWS gave us visibility data at every step of that process,” Homans said in an interview at the re:Invent conference two weeks ago. “What we have been able to do with them is take advantage of their deep knowledge of the Internet of Things to get all the data from electronic platforms like bioreactors, all the equipment, all the equipment in the manufacturing facility. Getting data about what was happening to those cells at every step.”
All data from the manufacturing line is collected digitally (Bothell LyFE Manufacturing Center is the first cGMP certified plan to be paperless) and uploaded in real time to Lyell’s data cloud running on AWS. Being able to keep all the data from the manufacturing plant in the same place that keeps the data from the research and clinical development organization is critical to Lyell’s success, Homans said.
“While this means that we have this huge library of data that helps us understand what is going on in a particular batch or in a particular patient’s cells, we can do correlation analysis and say, ‘In the study, How does it compare to what you see?” How does this compare to the engineering run you did when you submitted your IDN? [investigational new drug application]? ‘”She said. ‘Understanding the biology of our cells is of paramount importance because it will help us develop better treatments.'”
Also, storing all data in a data lake in a standard format eliminates the need for multiple steps to perform analysis. At the biotechnology company Homans previously worked for, researchers struggled to get answers to their questions because of the extensive data manipulation required.
“If you wanted global data, it was sometimes difficult to get that data because it wasn’t being passed back and forth. We could actually combine data from all the systems, but it was very cumbersome and clumsy and took a long time,” she said. It’s about putting the sheets together…and then you have to mix them up.”
AWS not only helped set up the data architecture, but also provided guidance on how to structure the facility to accommodate rapid data collection. Lyell attended his AWS workshop called “The Art of Possibility” and encouraged attendees to “throw away everything they thought they knew about building a manufacturing facility,” he said. Homans says.
The beauty of cell therapy from the perspective of biotechnology (and its investors, publicly traded companies) is that we can quickly see if the treatment is working. If any of Lyell’s two Phase 1 trials are successful, the company hopes to be ready to move forward quickly, even to the potentially all-important Phase 2.
“It’s important to be prepared and not to mess around with the system and try to get it right,” she said. “Because if you’re successful, you really don’t want to lose time. At that point, patients are literally waiting and you need to get to market as soon as possible.”
Better cancer outcomes for Africans
Another startup trying to put cancer in its place is Hurone AI, a Seattle, Washington-based company that helps oncologists in sub-Saharan Africa better detect and treat cancer. We build AI-assisted applications that help
Kingsley Ndoh, founder and chief strategist at Hurone AI, says there is one oncologist for every 3,000 cancer patients in sub-Saharan Africa, compared to about 150 cancer patients in the United States. There is one oncologist for you.
“One of the main challenges in cancer treatment is the inability of patients to cope with the symptoms and side effects of treatment,” Ndoh said. Data Nami“What Huron AI is doing at this stage is providing a platform that can do that remotely and streamline the work of doctors.”
Kingsley Ndoh, Founder and Chief Strategist at Hurone AI, holds a PhD in Medicine from the University of Joss, an MPH in Global Health from the University of Washington, and is a faculty member in the Department of Global Health at the same university.
Much of the oncology field data is collected from around the world. But most of that data comes from whites of European descent, who suffer from different types of cancer at different rates than blacks of African descent, Ndoh said.
For example, a study conducted in Nigeria in collaboration with Memorial Sloan Kettering Cancer Center in New York found that the median age at diagnosis of colorectal cancer in Nigeria is 49 years. Chadwick, a black American who starred in four of his comic book movies as Black His Panther, said his boss was diagnosed with colorectal cancer at the age of 39 and died in 2020 at the age of 43. pointed out that
“Medical experts in Nigeria have always recommended screening at age 50, primarily because they used US and European data,” Ndoh told re:Invent. “It will be important for us to do this kind of AI and data science work in Africa because doing so will provide us with invaluable data that can be modeled and personalized, such as prevention and confirmation of some data points. Because you can get [from someone], as well as family history, encourage me to say, “Okay, I need to do this cancer screening right now.” ”
The lack of good clinical data from blacks also appears during treatment planning. For example, the drug Atroxene has been shown to be an effective treatment for breast cancer in American women. But black women haven’t achieved the same level of success as white women, West Her Virginia study shows.
“What they found was that African-American women in that study had worse side effects than their white counterparts and were therefore less compliant with the medication, which was not demonstrated in clinical trials. “I mean, these are the kinds of things to do. [work] Many of these people share genetics, so research in Africa could help us understand here in the United States. ”
Hurone AI hopes to help fill gaps in oncology data (Image credit: Hurone AI)
Huron AI is currently participating in a randomized controlled trial to determine whether an AI-powered approach to precision oncology is superior to standard of care. Rather than collecting individual genes for testing, the company uses genetic data collected at the village level in Rwanda to inform research.
Based on the results of that trial, which Ndoh says is due soon, the company will provide more funding with the ultimate goal of providing AI assistance to oncologists in Africa. decide whether to
For Huron AI, running data science workloads on AWS has been very helpful. With just five employees at the company, spread all over the world, from Washington and California to Israel and Rwanda, setting up and running an on-premises system was impractical.
By participating in the AWS Health Equity Initiative, Hurone AI has access to the computing resources it needs. As part of an initiative launched in 2021, AWS has pledged to donate $40 million over three years to organizations that promote health equity.
“At this stage, we can save $3,000 to $10,000 a month,” Ndoh says. “This is basically how much we would spend on cloud costs in Rwanda. So AWS, through cloud credits, gave us the headroom to run and scale quickly.”
Related products:
Inside AWS’ New Supply Chain Planning Service
AI ‘fingerprinting’ combats hard-to-detect cancers
Why AI has yet to deliver on its healthcare promises
