The feasibility study was meant to see whether personalized cancer genetic sequencing and drug sensitivity recommendations could be done in a “clinically reasonable amount of time.”
The future of personalized cancer treatments will be enabled by technology to understand the genetic variants of a person’s disease, as well as tools to determine the best treatment pathways for the individual.
A new feasibility study from Stanford Medical Center, along with startups Notable Labs and Tempus, sketched out a possible model of that approach.
The study – which was focused on MDS cancer patients who had unsuccessfully undergone the standard of care for their disease – was an attempt to create personalized recommendations about what other potential treatment options may work.
Researchers at Stanford took blood samples from patients and sent them to Chicago-based Tempus for genomic sequencing of the person’s cancer. Another sample was sent to Notable, which used its AI-based automated laboratory system to test hundreds of potential drug treatments against the disease and provided treatment recommendations based on that analysis.
Stanford clinicians then used this data (alongside the clinical features of the patient) to create a report with individualized recommendations that could inform future treatment.
While Notable has been working with Stanford on other clinical validation research, this particular study had a time limit of 30 days for the required analysis.
Key to providing answers that can be used in living patients is ensuring that analysis is done in a “clinically reasonable amount of time,” said Peter Greenberg, a Stanford hematologist and director of the Stanford MDS Center.
“If we want to use this type of test on a much larger scale, the turnaround time is vitally important to get treatments recommendations to patients in a way that’s actionable,” Notable CEO and founder Matt de Silva said.
Foster City, California-based Notable has raised more than $20 million from investors and recently announced the launch of its first drug development program focused on targeting pediatric leukemia.
The feasibility study involved a small 20 person patient group. According to Notable, early clinical data indicates average positive and negative predictive values of 84 percent.
Researchers from Notable, Stanford and Tempus are working to publish a more detailed paper on the study later this year and will present further data at the European Hematology Association annual meeting in June.
“Once we have this technology, the next step is to do prospective clinical trials to determine whether the drugs that were recommended by the analysis are actually useful in vivo,” Greenberg said.
De Silva said the company’s research collaboration with Stanford has revealed important insights into how to design future clinical trials around patients with high sensitivity to the drug being tested.
He referenced one previous study undertaken to predict potential clinical response to a range of MDS treatments that was able to segment patients who would and would not likely benefit from the standard of care for the disease.
“We need to do additional trials in the second or third line settings, but the data we’re generating hints that there’s certainly a role in the front line setting, which is holy grail of what we’re trying to do,” De Silva said.
“It means that rather than giving all patients the same drugs, and only using this approach after all other options have failed, they’ll get the right drug the first time and for much longer.”