“Our challenge was to find a molecule that’s almost exclusively present on activated T cells — not just any T cell — because there are many T cells that just sit around resting,” Gambhir said. By coincidence, the molecule he found was the same one that Levy harnessed in his vaccine, a protein on the surface of activated T cells called OX40.
Boiled down, Levy’s cancer vaccine is a package of two stimulating agents. One coaxes T cells into producing OX40 on their surface; the other binds to OX40 and enables the cell to engage with tumor cells. Together the tag-team agents essentially prod loafing immune cells into high gear.
Once the tracer is injected, it scours the entire body, including the immune system, in search of cancer-killing T cells — but only those laden with OX40. Upon meeting, the tracer binds to OX40 and, when hitched together, the radioactive complex glows under a PET scan, revealing only those T cells that have been successfully activated, ready to ravage the tumor. If the scan comes back with low to no signal in the tumor or tumors, it’s an indication that doctors (in theory, as the vaccine and tracer have only been tested in mice) ought to reevaluate the immunotherapy dosage or change the treatment course altogether.
The power of PET
Gambhir’s lab tested the tracer first in cell cultures. They found that the compound was able to suss out activated T cells about 95 percent of the time. Later in mouse models, they still saw success overall, but it was a bit more subdued. In a group of about 50 mice, the PET tracer performed accurately upward of 90 percent of the time.
“It’s really only now that this tactic is coming into play; the PET scan is usually focused on assessing only the tumor cells,” said Gambhir. “But now, with new imaging agents like this, we’re able to image the immune cells, and that’s really the second half of the equation.”
Gambhir acknowledges that one could simply wait to see physical changes in the tumor volume to determine whether the therapy is working. But that poses a problem. It may take weeks, or even months, to definitively see whether the cancer is responding to the treatment. Say the vaccine doesn’t work. In the time it took to find out, the cancer would have continued to spread, becoming more molecularly heterogeneous and even more difficult to treat the next time around. Knowing sooner gives the patient more time to try other options, hopefully leading to better outcomes.
Levy has moved his vaccine into a phase-1 clinical trial. In the next few months, Gambhir plans to move this new OX40 tracer into that same clinical trial, so that the tracer and therapy can be tested in conjunction.
“We were able to predict what was going to happen in mice several weeks out by looking only 48 hours from the start of the immunotherapy. We could figure out which mouse was going to respond to the immunotherapy and which wasn’t before they actually did or did not respond,” Gambhir said. “And that’s exactly what we’re trying to do. We’re trying to show that this approach can, in humans, allow us to image early and thereby let us evolve the therapy quickly.”
We’re trying to show that this approach can, in humans, allow us to image early and thereby let us evolve the therapy quickly.
Gambhir also is pursuing work to establish the OX40 tracer as a diagnostic for other applications, such as the autoimmune disease multiple sclerosis. “It’s important to remember that this is a really general approach to visualizing activated T cells — this shouldn’t be thought of as specifically for cancer immunotherapy alone,” he said. “That’s just one important application.”
The work is an example of Stanford Medicine’s focus on precision health, the goal of which is to anticipate and prevent disease in the healthy and precisely diagnose and treat disease in the ill.
Other Stanford co-authors of the study are Idit Sagiv-Barfi, PhD, instructor of oncology; Kezheng Wang, MD, PhD, a visiting faculty member in the Gambhir lab; postdoctoral scholar Ophir Vermesh, PhD; Debra Czerwinski, life science research assistant; Emily Johnson, life science research professional; and Michelle James, PhD, assistant professor of radiology and of neurology and neurological sciences.
A researcher at Harbin Medical University also contributed to this work.
Stanford’s Department of Radiology also supported the work.