Takeda and Janssen join effort to detect scratching

Takeda and Janssen are teaming up with wearable device maker Actigraph to validate a digital measure that quantifies the disruptive nighttime scratching that impacts people with skin conditions like psoriasis and atopic dermatitis. The hope is that a measure of so-called nocturnal scratch might one day be used as an endpoint in clinical trials for new drugs.

Called DECODE, the project will conduct a study enrolling roughly 60 participants with either skin condition who will be filmed with infrared cameras while sleeping with an Actigraph device worn on each wrist. The video recordings will be used as ground truth for the validation of an algorithm that uses data from the wearable to discern overnight scratching. The study will also collect polysomnography recordings to validate sleep metrics specific to the skin conditions. Actigraph's chief scientific officer Christine Guo told me the companies hope to have an interim algorithm by the end of next year and that the project will be complete in the first half of 2025.

Nocturnal scratch has been explored in atopic dermatitis in the past with Pfizer publishing feasibility work, but Takeda's head of sensing and measurement Ariel Dowling said the company was particularly interested in exploring the potential in psoriasis, where Takeda has been advancing a drug candidate, TAK-279. Treatment around psoriasis has primarily focused on clearing up the visible signs of the disease and a digital measures of scratch could offer an opportunity to objectively understand how much it impacts people's sleep and quality of life.

With further development, it's conceivable that the Food and Drug Administration might one day accept nocturnal scratch as an endpoint for clinical trials. A phase 3 trial of TAK-279 is slated to be getting under way this year, and Dowling said it's an open question whether the DECODE work will be done quickly enough to be squeezed in. Graham Heap, a Takeda VP leading the TAK-279 franchise, said that data about scratching could be useful as part of a package showing the value of a drug to payers and providers who must embrace new treatments in a field with plenty of existing options. It's worth noting here that many companies use digital measures in post-approval studies of their drugs.

For its part, Actigraph chose the seemingly obscure metric of scratching because it saw an opportunity to get in on the ground floor on a measurement where there is both unmet need and technical feasibility. In the future, scratch measurement could be expanded to a host of other conditions.

"If you look at the number of studies in digital in dermatology, it's actually extremely low. It's lower than many of the other therapeutic areas we work on," said Guo. "But we believe that this is going to really pick up quite quickly, especially if we can demonstrate the use case and provide a turnkey solution to the industry."

Medical devices

FDA warns of Philips CPAP machine overheating

Yet another issue has cropped up with Philips devices used by people with obstructive sleep apnea. The FDA this week issued a safety communication warning that people who use Philips DreamStation 2 continuous positive airway pressure (CPAP) machines should monitor them for overheating. The agency has received 270 reports of problems with the devices since early August. The agency notes that some of these devices may have been issued as replacements for Philips machines that were subject to a gigantic recall.

Hospitals

Mayo Clinic's plan for $5 billion tech-forward campus

Mayo Clinic announced a $5 billion overhaul of its main campus in Rochester, Minn., STAT's Casey Ross reports. The project, to include five new buildings with 2.4 million square feet of space, will merge Mayo's traditional medical services with its increasing investments in artificial intelligence and digital tools. Even as Mayo touts tech, the buildings are fundamentally aimed at expanding the hospital's capacity. CEO Gianrico Farrugia said the construction, which should be completed by 2030, will alleviate the unsustainable demand the hospital expects.

"The increased age and increased size of the populations will be such that overall we are going to need physical spaces — more than I think we as a country have realized," he told Casey. "In order for us to achieve the digital transformation of health care, I strongly believe that we need to at the same time create a physical transformation."

Artificial intelligence

Amazon and Microsoft tout healthcare and pharma customers for generative AI tech

Two of the world's largest technology companies and their partners have over the last few weeks been aggressively pushing out announcements about progress developing — and selling — their generative AI tools.

This week AWS, Amazon's cloud services arm, held its re:Invent conference in Las Vegas, where Pfizer's chief digital and technology officer, Lidia Fonseca joined AWS CEO Adam Selipsky on stage where she detailed how the company had used AWS to stand up an internal platform allowing employees to access large language models and claimed that its use of generative AI would generate cost savings of $750 million to $1 billion per year. (How do you calculate that?) She said that the company was exploring use cases like identifying oncology targets, and profiling companies for potential acquisition. AWS also announced that it was working with Amgen to experiment with generative AI solutions for research and development, including "clinical trial innovation."

Earlier this month, before Microsoft had to help arbitrate the l eadership crisis at ChatGPT developer OpenAI, the company held its annual Ignite conference for IT professionals and developers where it announced the rollout of Copilot for Microsoft 365. The technology helps companies leverage large language models within Microsoft's suite of productivity software. The announcement touted many of the same partners as Amazon, like Pfizer and Amgen plus other recognizable names like Mayo Clinic. A few days after the announcement, Duke Health sent us a release touting how it would be testing out the technology as part of its ballyhooed AI partnership with Microsoft.

Mental health

Analysis suggests notifications don't improve depression apps

A new meta-analysis of 13 randomized clinical trials evaluating app-based interventions for moderate to severe depression suggests that overall they showed positive effects. The publication in JAMA Network Open also begins to explore how characteristics of the interventions might influence outcomes and calls for further study of factors that could "enhance or attenuate the effectiveness" of the apps.

For example, the researchers found that "shorter interventions were associated with greater treatment efficacy" and surprisingly, that those that interventions that used in-app reminders had a smaller effect size than those that used other methods for nudging users. The researchers say this highlights the importance of avoiding reliance on notifications and finding better ways to promote adherence, like in-person reminders. As Megan Ranney, the dean of Yale School of Public Health, told me in a message after posting the findings on LinkedIn, "people with mental illness — like the rest of us — are suffering from notification overload, and 'more' may not be better!"

More around STAT

Check out more exclusive coverage with a STAT+ subscription

Read premium in-depth biotech, pharma, policy, and life science coverage and analysis with all of our STAT+ articles.

What we're reading

Digital technology for diabetes, New England Journal of Medicine
How artificial intelligence could make pregnancy discrimination in employment more common, STAT