Quantitative PET imaging could help monitor cancer treatment
Caren B. Les
It was expected that, in the US in 2007, approximately 63,000 people would be diagnosed with non-Hodgkin’s lymphoma, a cancer of the lymphatic system.
A recent study compared visual and quantitative analyses of PET scans for assessing the prognoses of patients with diffuse large B-cell lymphoma, a common and aggressive form of non-Hodgkin’s lymphoma. The ability to predict tumor response in the early stages of chemotherapy could result in modification of the treatment. Patients demonstrating an initial positive response to therapy, for example, could be given less-intense treatment.
These PET scans of a 25-year-old woman with non-Hodgkin’s lymphoma were captured as part of a study in which visual and quantitative scan interpretations were compared for their effectiveness in evaluating prognoses. Captured prior to chemotherapy, the scan on the left (anterior view) shows bulky mediastinal uptake. Captured after two cycles of chemotherapy, the scan on the right (left anterior oblique view) was interpreted as positive, visually — two residual foci were diagnosed, as indicated by the arrows. However, quantitatively, the reduction in maximum standardized uptake value per body weight was measured at 80.7 percent, indicating good early response. The patient remained in full remission after 36 months of follow-up. Reprinted with permission of The Journal of Nuclear Medicine.
Researchers at the Henri Mondor Hospital in Créteil, France, conducted the study to assess the value of early 18F-fluorodeoxyglucose PET by using the standardized uptake value of the radiotracer’s metabolic rate compared with visual analysis to differentiate the patients’ responses to chemotherapy. According to the lead scientist, professor Michel Meignan, the quantitative analysis of pre- and post-therapeutic tumor images seemed to be superior to the visual analysis, but both analyses still would be necessary.
The standardized uptake value describes the ratio between the local increase of radioactivity resulting from tumor aggressiveness and the homogenous normal tissue activity. The higher the value, the more active the tumor. This quantitative index reflects the reduction of metabolic activity following first-line chemotherapy.
Ninety-two patients who were newly diagnosed with the disease underwent PET prior to and following two cycles of chemotherapy, at midtherapy. The scans were assessed both visually and by the quantitative parameters.
The standardized uptake value tumor-to-normal ratios and their changes over time were compared with visual analysis for predicting event-free and overall survival of the patient. Receiver-operating-characteristic analysis, a curve study, was used to determine the value that allows the best separation of the categories of the patients who relapse and those who do not. According to Meignan, a 65.7 percent reduction of the tumor activity obtained at midtherapy was the best value for discriminating between patients who survived without presenting a new pathological event from those who presented new occurrence of the disease.
The researchers determined, using quantitative analysis, that the accuracy of early PET to predict event-free survival was 76.1 percent. In comparison, visual analysis produced an accuracy of 65.2 percent. In the study, 14 patients considered positive on visual analysis could have been reclassified as good responders, based on quantitative analysis.
Meignan said that the group intends to investigate the identification of a subset of patients according to the distribution and the evolution of the uptake and the relationship of the tumor uptake with the abnormalities observed with the CT. He added that refining and automating the quantitative analysis, as well as introducing new PET radiotracers, would be helpful. Combining the visual analysis with the quantitative study would still be required, he emphasized.
The Journal of Nuclear Medicine, October 2007, pp. 1626-1632.
MORE FROM PHOTONICS MEDIA