AAPM: Treat tumors as spatially different, not homogenous masses
PHILADELPHIA—While most cancer therapies deliver a uniform amount of radiation to the tumor as a whole, cancer masses are not homogenous. Researchers at the University of Wisconsin, Madison, discovered that most of the head and neck tumors in their study contained three statistically different subpopulations with distinct profiles, and they presented the study July 19 at the annual meeting of the American Association of Physicists in Medicine (AAPM).

Stephen Bowen, PhD candidate, and colleagues in the department of medical physics evaluated the PET/CT scans of 13 patients with head and neck cancer and measured three different characteristics: metabolism, cell proliferation and oxygen deprivation, or hypoxia. Previous studies had shown that these three factors can vary within a tumor, and each is known to effect how a tumor reacts to treatment, Bowen said.

Investigators used three different PET tracers as surrogates for the different characteristics: 18F-FDG for metabolism, 18F-FLT for cell proliferation and 61-CuATSM for hypoxia. Computer algorithms classified the regions based on these three parameters.

Bowen and colleagues partitioned seven tumors: one into two subpopulations and six into three subpopulations. For the three partition cases, the subpopulations were found to be spatially separated into peripheral, central and intermediate regions.

FDG uptake was 55 percent lower for the peripheral subpopulations and 19 percent lower for central subpopulations relative to the overall mean. On the other hand, FLT uptake was 51 percent lower for the peripheral subpopulations, but 52 percent higher for the central subpopulations. The researchers did not observe any significant difference in CuATSM uptake among subpopulations.

"We demonstrated that tumors can be partitioned into multiple subpopulations with distinct characteristics using multi-phenotypic molecular imaging," Bowen said. "This subpopulation-based study may lead to better understanding in designing personalized cancer therapy."

For example, it may be possible to develop future therapies that up the dose given to radiation-resistant cells and drop the dose given to radiation-sensitive cells, said senior author Robert Jeraj, PhD, from the department of medical physics at the University of Wisconsin School of Medicine and Public Health in Madison, Wis.

"There are some regions that are overtreated in a tumor and some that are undertreated. The idea of 'dose painting' is to treat each region properly," Jeraj said.

One potential target for increasing the dose, said Jeraj, is the 20 percent of tumor cells that show high hypoxia, a low metabolic rate and low proliferation. Candidates for a lower dose include the 30 percent of cells that show high proliferation, but low hypoxia and intermediate metabolism.

New tools must be developed, however, to measure how one region changes size in relationship to the tumor as a whole, he said.