Software translates improved diagnostic information to better target radiation at cancer cells

NewsGuard 100/100 Score

As diagnostic imaging and radiation treatment technologies both become more capable and precise, physicians using them face an increasingly difficult task: how to transfer information from the diagnostic systems to the treatment planning systems.

Researchers at the Georgia Institute of Technology and Memorial Sloan-Kettering Cancer Center are solving that problem with an image translation system designed to bridge the gap between advanced diagnostic techniques such as magnetic resonance spectroscopy (MRS) and the latest techniques for delivering precise doses of radiation to tumors. The system can also compensate for how tumors change and shift between diagnosis and treatment, during treatment, between treatments – and even as patients breathe.

Based on advanced computer modeling, the image translation technique would facilitate a biological optimization planning system that uses information about the location and density of cancer cells to deliver escalated doses of radiation to tumors.

“The benefit to the patient would be in improved local tumor control,” said Eva K. Lee, an associate professor with faculty appointments in the School of Industrial and Systems Engineering at the Georgia Institute of Technology and the Winship Cancer Institute at Emory University School of Medicine. “That means the rate of recurrence should be lower and there will be fewer complications affecting the normal tissue. Patients should also experience fewer side effects from the treatment.”

Information on the work, “Combined Modality Treatment Advances – Incorporating Biological Metabolite Information for Cancer Treatment,” was presented Tuesday, October 5, 2004 at the 46th annual meeting of the American Society for Therapeutic Radiology and Oncology (ASTRO) in Atlanta. Sponsored by the Whitaker Foundation and the National Science Foundation, the work is part of Lee’s long-term collaboration with Marco Zaider, professor of Physics in Radiology at Cornell University Medical College, and attending and head of Brachytherapy Physics at Memorial Sloan-Kettering Cancer Center.

A mathematician who has been working on radiation treatment planning optimization for prostate and other forms of cancer, Lee has used the new system for developing radiation therapy plans for patients undergoing the MRS imaging at Memorial Sloan Kettering. The results so far are promising, though controlled clinical studies must still be done to compare the results with standard techniques that treat tumors as homogenous masses.

By analyzing cell metabolism in tumor areas, magnetic resonance spectroscopy can identify regions of the prostate that have denser populations of tumor cells. Combined with data on healthy structures in the prostate, that information can be used to increase the radiation dosage to areas containing more cancer cells, through the placement of radioactive seeds or application of external beams.

But before that can be done, information must be transferred from the MRS system to the treatment planning systems that radiation oncologists use to determine the radiation dosage applied to the tumor.

“Functional imaging involves looking at how the cancer cells actually proliferate within an organ, and for that technology to really be useful, we have to be able to translate it into the clinical setting,” Lee noted. “But until now, these biological factors couldn’t usually be used in treatment because the imaging modalities are so different. This new system allows us to use the diagnostic imaging information in a practical sense. It allows us to put everything together.”

The new system translates the spatial information about tumor concentrations – measured in “voxels” – from the MRS system to the treatment planning system, morphing that information onto the ultrasound/CT images typically used by planning and delivery systems. The voxels can be translated even when the organ has changed shape due to treatment preparation.

Use of the MRS data to escalate radiation dosage is known as biological optimization, and is part of a trend toward increasing customization of treatment based on specific information about a patient’s cancer.

The ability to adapt radiation treatment to changes in organs is particularly important in the radiation treatment of lung cancer. Though patients can be asked to briefly hold their breath, the effectiveness of treatment can be adversely affected by the motion of tumors as patients breathe during treatment.

The system Lee and Zaider are developing can account for those spatial changes over time, tailoring radiation to provide effective dosages to cancer cells even when the tumors are moving.

“The shape of the tumor as well as the position of the tumor both change because the patient is breathing,” Lee noted. “We have to follow the movement and know where the tumor voxels are that need radiation. We want to make sure that the end result – the average dose received by each voxel as it moves – will satisfy the clinical constraints.”

An automated treatment planning system Lee and Zaider developed earlier for prostate brachytherapy improved local tumor control from 65 percent to 95 percent.

The growing power of computer systems has made such treatment planning systems possible. In systems designed to deliver external beam radiation from as many as five to nine different angles, there can be as many as a million variable and constraints – a huge challenge for people designing the system and writing the software code.

“We are really only catching up with what the radiation delivery system can do,” Lee explained. “The modulation of the radiation dose has become quite good, allowing us to vary the intensity over the tumor.”

Lee is also collaborating with clinical researchers on applications of the system in lung cancer treatment.


The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of News Medical.
Post a new comment
You might also like...
New tool unravels the mystery of DNA-RNA interactions in single cells