WASHINGTON, DC - Metastatic cancer patients often don't have the time to determine whether the treatment they're receiving is beneficial: A new tumor scoring system may help, according to a presentation on Thursday at the Society for Imaging Informatics in Medicine (SIIM) meeting.
A retrospective study found that a new quantitative visual-based scoring system of CT images to assess tumor changes in patients with metastatic melanoma was a better early predictor of progression-free survival than the well-established Response Evaluation Criteria in Solid Tumors (RECIST) standard. Although validation of the new scoring system is still in progress, it has the potential to replace the RECIST standard.
Little work has been done to validate imaging as a way to measure a patient's clinical response to new therapies that are being developed to treat advanced cancer, according to Dr. Ronald Gottlieb, clinical professor of radiology at the Arizona Cancer Center of the University of Arizona. Most work has measured tumor lesion size on CT exams using the RECIST standard.
RECIST measurements are currently used for all patients in clinical trials. Only a single measurement per lesion is taken, and summed measurements can easily be entered into data collection forms.
However, the standard is not very accurate for estimating tumor volume of lesions with complex shapes, and it's outdated because it does not address various combinations of imaging features of tumors responding to a variety of cancer treatments. RECIST is cumbersome for radiologists to use in daily practice, and reader reproducibility and interobserver agreement of measurements are also issues.
Gottlieb and colleagues developed a simplified quantitative visual scoring system to code CT imaging findings on radiology reports. A score of 1 was assigned if a tumor or lesion increased in size, a score of 0 represents no change, and a score of -1 represents a decrease in size. If new tumors appeared, the number of organs involved was numerically noted. All numbers were summed, and changes were noted compared to prior studies.
"The system is analogous to BI-RADS classification," Gottlieb said. "It doesn't require expensive hardware or software, is easy to remember, and is easy to reproduce."
To test the project, the research team retrospectively evaluated 65 consecutive patients with advanced metastatic melanoma who were receiving chemotherapy at the Arizona Cancer Center from 2004 through 2009. The majority of patients were men (72%), with a median age of 65. All patients received a CT exam at baseline, plus at least one or more subsequent exams.
The median time of progression-free survival was 3.5 months, and overall survival was 7.2 months. Patients were followed for up to 25 months; 55 of the 65 patients died.
"We chose metastatic melanoma because the gold standard we were assessing the scoring system against was progression-free survival," Gottlieb said. "Sadly, patients with this disease do not live long. However, this made them ideal subjects for an initial proof-of-concept study."
The visual-based scoring system was more accurate than RECIST projections to predict patients responding to treatment and those whose conditions remained stable. Scores were comparable for patients who were declining. The visual-based system was also more sensitive to changes as they occurred, and it showed a high level of reproducibility.
"The next step is to see how well the scoring system performs with respect to other types of cancer," Gottlieb said. "If it continues to perform well, we then need to find a way to extend its use into daily work."
"We also hope that the scoring system could be used for patients who are not enrolled in clinical trials," he added. "RECIST is only used for patients enrolled in clinical trials, which in the U.S. represents only about 5% of all cancer patients. If scoring is considered to be good for these cancer patients, why not also include the other 95% of patients receiving some type of treatment?"
