Researchers at NYU Grossman School of Medicine and its Perlmutter Cancer Center have released data from a new blood test that that has the potential to identify which patients will benefit from immunotherapies targeting cancer. The team, whose results appear today in the journal Clinical Cancer Research, developed a panel of autoantibodies that if present in a patient’s blood prior to I/O treatment has the potential to accurately predict whether a patient’s cancer would recur, or if they would experience autoimmune side effects as a result of the therapy.
“Our results show that the new research test, by predicting whether a patient will respond to a treatment or experience side effects, has the potential to help physicians make more precise treatment recommendations,” said first author Paul Johannet, MD. Who was a postdoctoral fellow in the lab of senior study author Iman Osman, MD, the Rudolf L. Baer professor of dermatology and member of Perlmutter Cancer Center. “With further validation, this composite panel might help patients to better balance the chances of treatment success against severe side effects.”
The study examined patients who received adjuvant immunotherapy, where the aim is to keep cancer from returning after prior treatment. The researchers theorized that some patients may have higher levels of autoantibodies prior to treatment, but not enough to be identified as an autoimmune disease. This “hidden” susceptibility, the researchers theorized, could be triggered by checkpoint inhibitor drugs and, in turn, cause higher levels of immune response-related side-effects.
To test this hypothesis, the investigator obtained blood samples from more than 950 patients enrolled in one of two Phase III randomized controlled clinical trials of adjuvant checkpoint inhibitors in patients with advanced melanoma. The tumors for these patients had been surgically removed, and the blood samples were collected prior to the patients receiving adjuvant therapy. Using a chip-based panel with more than 20,000 proteins, the team collected data on which proteins were present in each patient’s blood prior to treatment.
Using statistical modeling, co-senior author Judy Zhong, PhD, and colleagues developed a score-based prediction system for each treatment. Patients who showed a high autoantibody recurrence score had their cancer return earlier than those with a low score. In addition, patients who had a high level of autoantibodies prior to treatment were more likely to experience adverse effects from the drugs.
“That we identified 283 autoantibody signals shows that the biological phenomena underlying recurrence and toxicity are complex, and cannot be driven one or two biomarkers,” said Osman.
With the testing panels and modeling completed in melanoma, the NYU team now plans to test the predictive value of their autoantibody signatures in patients with other cancer types that have checkpoint inhibitors as approved treatments.