“Looking in the patients’ eyes and having a conversation” has motivated Stanford oncologist Ash Alizadeh, MD, PhD, to improve the way we diagnose, talk about and treat cancer.
Patients go home nervous and the care team is nervous, he pointed out, because you’re fighting a battle together to save a life and the things you’re doing are toxic and expensive.
“It’s really sobering to look at how blunt our tools are for getting a sense for whether you’re making progress as you’re going through the course of your therapy,” said Alizadeh in a recent episode of the Sirius radio show The Future of Everything hosted by Russ Altman, MD, PhD.
A key area of his work aims to more accurately predict a patient’s prognosis. He developed a computer algorithm (the focus of a recent Stanford Medicine magazine article) that searches data for information likely to affect the patient’s long-term outcome — generating a unique personalized estimate of risk, called the continuous individualized risk index (CIRI). The goal is to use CIRI to guide personalized therapy selection.
In the episode, he explained that their integrated approach better forecasts a patient’s prognosis by analyzing the complete medical path of the patient, whereas oncologists typically give more weight to the most recent data.
The researchers validated their predictive model using data gathered over time from patients with three types of cancers: diffuse large B-cell lymphoma (DLBCL), chronic lymphocytic leukemia or early-stage breast cancer.
In the study, they also measured the amount of circulating tumor DNA (ctDNA) in the blood of 132 DLBCL patients, before and during their treatment. Circulating tumor DNA is DNA that was shed from dying tumor cells and released into the bloodstream.
For this small group of DLBCL patients, standard methods to forecast how well a patient will do had a predictive index of 0.6, where a perfectly accurate test would score 1 and a random test like a coin toss would score 0.5. Alizadeh’s CIRI score was 0.8 for the same patients — not perfect but markedly better than the current “crystal ball exercise,” he said in a news release.
In the radio show, he also discussed how this predictive model complements his work to develop new technologies for cancer diagnosis and treatment.
For example, he explained measuring ctDNA levels with a non-invasive liquid biopsy may help detect early-stage cancer, guide treatment selection and monitor treatment response. And if liquid biopsies detect cancers at an early stage, this may allow oncologists to leverage their patients’ immune system to attack their cancer, he said.
“So instead of directly attacking the tumor cells with drugs that kill the cancer cells, you now have drugs that engage the immune system to say, ‘Hey, wake up,’” he said. That means the same drug could work for many cancers.
Alizadeh is developing these new techniques to personalize cancer diagnosis and treatment in hopes of improving the outcomes for his patients, he said.
Photo by Pikrepo
This is a reposting of my Scope story, courtesy of Stanford School of Medicine.