Whole Body Imaging Sheds Light on the Success of Lung Cancer Therapy
In a first study on mice and patients, researchers show that a new PET-imaging agent can help cancer doctors predict the success of lung cancer therapy.
Non-small cellular lung cancer (NSCLC) is a debilitating form of lung cancer. It killed about 160,000 Americans in 2013 and accounts for about 14 percent of new cancer cases reported each year in the U.S. Doctors label it a chemotherapy rebel since it does not often yield to drugs.
One of the challenges in treating NSCLC is predicting patients’ response to therapy, both before beginning it and during early treatment. Current monitoring practice relies on blood tests and biopsies to look for genetic mutations that might signal whether the therapy has worked, but the approach has shortcomings.
“If we have to wait too long to see if the tumor shrinks, by that time it may already have evolved further and be much more difficult to treat with a second round of therapy if it didn’t respond,” said Sanjiv Sam Gambhir, a radiologist at the Stanford University School of Medicine.
So Gambhir and his colleagues set out to find new ways doctors can predict a treatment’s success within days of beginning it.
His team already knew that the class of drugs known as tyrosine kinase inhibitors (TKI) help to treat some patients with NSCLC. The drugs target mutations in the genes that code for the epidermal growth factor receptor, or EGFR, in the tumor.
Gambhir’s team created a copycat of a TKI drug, called F-MPG, with the radioactive isotope fluorine-18 added as a tracer element. Tests on cultured cells showed that the modified drug can bind to EGFRs in NSCLC cells, so the researchers put it in mice that carried NSCLC cells under their skin. Then the animals were scanned using positron emission tomography.
“If that mutation [in EGFR] exists, this imaging agent locks in on [the mutations] and accumulates much more on those areas than in areas where the mutation doesn’t exist,” said Gambhir. This means the tracer can easily identify cancerous cells throughout the body likely to respond to TKI treatment.
Next, Gambhir and his team tried the same approach on human patients. Collaborating with researchers from the Molecular Imaging Research Center of Harbin Medical University in China, they imaged 75 patients with NSCLC.
Since the researchers were not allowed to use the PET scan to medically classify patients and decide their course of treatment at this stage of their research, all the patients received TKI drugs as a treatment. Over the course of more than a year, the researchers checked back on the status of the patients.
“The PET scan did very well in predicting who’s going to respond to [TKI] therapy,” Gambhir said. This is important because patients who are not likely to respond to TKI drugs should be given other lung cancer drugs, he said.
The PET images could also predict which patients who responded to the therapy were unlikely to have their cancer come back within the study window.
The results are published in Science Translational Medicine.
Geoff Higgins, a Cancer Research U.K. clinician scientist at the University of Oxford in England, said the results sound promising, but he wonders about the sensitivity of 18F-MPG to detect rare EGFR mutations, and whether the approach can meaningfully influence clinical management yet. “Since genetic analysis on tumor or plasma samples can be routinely and cheaply undertaken, the sensitivity/specificity of the scan will need to be extremely good to enter routine practice,” he said.
It’s true that 18F-MPG lit up in a few patients in places that had no EGFR mutation. But if the specificity is improved and supplemented with results from larger multisite clinical trials, Gambhir believes in about four years, 18F-MPG may be regular in cancer clinics and “let us better manage patients by customizing therapies instead of throwing a drug at everyone.”