Proactive, Personalized, Precise: AI and Collaborative Research is Revolutionizing ALS Care

With no single known cause, no definitive diagnostic test and no cure, ALS remains a tragically confounding neurodegenerative disease. Its progression varies dramatically from patient to patient, leaving clinicians and researchers racing to understand why. 

A collaboration between UCSF School of Pharmacy professors Steven Altschuler, PhD, and Lani Wu, PhD, and Catherine Lomen-Hoerth, MD, PhD, director of the ALS Center at UCSF, is developing a new approach to classifying and treating  the complexities of ALS by applying high-throughput phenotypic profiling, advanced imaging and real-time patient data analytics. 

Treatment options for ALS, also known as Lou Gehrig’s disease, have been limited — and modest. The Food and Drug Administration approved the first ALS drug 20 years ago and added another in 2017, with both only slightly slowing progression rates. A more recent addition shows promise for the 1 percent of patients who have a genetic form of ALS, but ALS’ unpredictability makes both research and treatment difficult. 

That’s where Altschuler and Wu come in, using computational tools like machine learning to help expose patterns and subtypes within ALS that were previously invisible, offering stratification as a potential path to more personalized and proactive care. 

A platform for possibility 

“Now that we have this phenotypic profiling, we can begin the process that started 30 years ago in cancer research, figuring out what’s dysregulated and finding good druggable targets,” said Altschuler, who is a professor in the Department of Pharmaceutical Chemistry. “It’s a flywheel of innovation that just doesn’t exist right now in ALS.” 

Lomen-Hoerth and her team, who see hundreds of ALS patients each year, are piloting a clinical dashboard designed by Altschuler and Wu’s group. 

Catherine Lomen-Hoerth, MD, PhD

Patients and caregivers are equipped with Garmin wearable devices that feed data, including oxygen and heart rates, directly into the dashboard in real time, enabling physicians like Lomen-Hoerth to intervene earlier, adjusting treatments or recommending new therapies in a multidisciplinary clinic that includes physical, occupational, speech and respiratory therapists in addition to doctors, nurses, dieticians and social workers. 

“We typically follow patients every three months or six months, but so much happens in between that we can track now, such as detecting falls. We can ask patients and caregivers to respond to our questions and then use AI to sift through that data,” said Lomen-Hearth, a professor of neurology at the UCSF Weill Institute for Neurosciences. “Even though ALS is a 100 percent fatal disease, there’s so much we can do to help slow decline with mechanical things like a breathing machine or a feeding tube.” 

Collaboration and stratification 

Cindy Chew

Research assistant Sabrina Hammerlindl in the Altschuler and Wu labs. 

Combining this increased access to data with sophisticated imaging of deep phenotypic profiling has been a serendipitous breakthrough for Altschuler, Wu and Lomen-Hearth. Each of them has personal connections to ALS. About two years ago, their social circles overlapped, which led to professional introductions and the realization that they were grappling with the same question: How to start the process of precision therapy without genetic targets.

“It’s all come together at this perfect time,” said Altschuler. “Cathy didn’t know about our molecular work. And we didn't know what it was like to work with a patient population in the disease. Once we started talking, this was just the obvious way to provide the maximum benefit...by finding targets for subgroups, not just general drugs for everybody, because those just don’t work.” 

Challenges and hope 

Altschuler said that while ALS is not "infinitely complex," it's also unlikely that it's only a single disease, which is just one of the challenges of translating their research findings into treatments. Most pharmaceutical companies focus on genetic forms of ALS, but Altschuler and Wu are optimistic that their work will widen that scope.

Cindy Chew

Department of Pharmaceutical Chemistry Assistant Researcher Heinz Hammerlindl, PhD, works with ALS patient skin cells in the Altschuler and Wu labs. 

Additional challenges come from the fact that animal models are not useful for studying neurodegeneration, and from the urgent need for an accelerated workaround to the typical stem cell-to-neuron conversion process — because the typical life expectancy after an ALS diagnosis is only two to five years.

Altschuler said the biggest surprise so far from their research has been showing that skin cells from a motor neuron disease carry a signature of the disease, a discovery they hope will eventually lead to enrolling patients in clinical trials. 

Lomen-Hearth said the ALS community is also hopeful that treatments found to be effective for one degenerative disease may work well for others. 

“There is overlap between ALS and frontotemporal dementia,” she said. “And this idea of being able to track patients at home with a wearable and their caregiver, and being able to download that data from looking at a dashboard, that’s something that would be universally applicable to all diseases.” 

Philanthropy and government support 

The UCSF team’s efforts to transform ALS care rely on two distinct yet complementary sources of support: philanthropy and federal funding. The foundational research — everything from developing the deep phenotyping platform to building predictive dashboards and analyzing patient-derived cells — is funded almost entirely by philanthropy, Altschuler said. 

Cindy Chew

Department of Pharmaceutical Chemistry  Data Scientist Karl Kumbier, PhD works with Altschuler and Wu. 

“Federal programs can help with clinical implementation, but the innovation, the discovery, the engine behind this work, that’s donor-funded. And this would be the first time anybody whose ALS is not genetically defined would have a sliver of hope for a personalized therapy.” 

Meanwhile, support from federal Early Access Programs, also known as Expanded Access Programs, has been vital to translating these discoveries into clinical opportunities. These programs enable patients with serious or life-threatening conditions to access investigational drugs or devices outside of clinical trials, when no comparable or satisfactory alternatives exist. By funding the compassionate use of promising therapies even before full FDA approval, they provide a critical bridge between bench research and bedside care. 

Together, these funding sources are helping to drive a future in which patients receive more proactive, personalized care, guided not only by symptoms but by the complex biology unfolding at the cellular level. 

More  

Altschuler and Wu develop new cell-screening approach to speed drug discovery