PET Scans May Predict Response in NSCLC Treatment

ANN ARBOR, Mich., July 19 -- It may be possible to determine whether non-small cell lung cancer is responding to radiation therapy before the course of treatment is over, a small study evaluating the predictive value of PET imaging suggests.

Changes in peak [¹⁸F]fluorodeoxyglucose (FDG) activity in tumors on PET images obtained during radiation therapy correlated significantly with post-treatment outcomes (P<0.001), investigators reported in the July 20 issue of the Journal of Clinical Oncology.

FDG metabolic activity in the lung did not change significantly from pre- to post-treatment PET scans, said Feng-Ming Spring Kong, M.D., Ph.D., of the University of Michigan, and colleagues.

In addition to using PET scans to predict long-term outcomes, the correlation between PET scans during and after treatment suggests FDG-PET imaging might be used to alter planned treatment, they added.

Efforts to individualize treatment for non-small cell lung cancer continues to be handicapped by an inability to determine whether a patient has responded to a given therapy until after completion of treatment, the authors noted. Several studies have shown that changes in tumor uptake of FDG correlated with pathologic response and overall survival. Post-treatment normalization of the standard uptake value appears to be a marker of complete tumor response and an indicator of favorable prognosis.

Conventionally, FDG-PET scans to assess treatment response are performed at least six to eight weeks after completion of radiation therapy and two weeks after chemotherapy. However, such an approach has limited applicability in clinical practice because it affords no opportunity to alter first-line treatment.

The Michigan investigators hypothesized that tumor FDG activity decreases during radiation therapy, that no confounding FDG activity occurs during treatment in the irradiated lung, and that changes in tumor FDG activity during radiation therapy correlates with the ultimate response.

To test their hypotheses, the group performed FDG-PET imaging in 15 patients undergoing radiation therapy for stage I-III non-small cell lung cancer. Scans were performed two weeks before the start of therapy, after delivery of about a 45-Gy dose of radiation (of a planned 60-Gy total dose), and three to four months after completion of fractionated radiation therapy.

Each of the three scans was evaluated for tumor and lung response by a nuclear medicine specialist and a radiation oncology. Response was determined on the basis of peak FDG activity, which is a reproducible measurement commonly used in clinical practice, the investigators stated. The primary endpoint of the study was tumor response at three to four months after completion of therapy.

On the basis of changes in tumor FDG activity, the investigators determined that 11 patients had partial metabolic responses, two had complete metabolic responses, and two had stable metabolic disease (including one non-FDG-avid tumor). Peak FDG activity decreased in all 14 of the FDG-avid primary tumors but not in the non-FDG-avid tumor.

The mean normalized SUV decreased from 5.2 before radiation therapy to 2.5 during treatment and to 1.7 post-treatment. The change in mean normalized standard uptake value differed significantly for all time-point comparisons (P=0.002 to P<0.001).

Peak FDG activity in the irradiated lung did not change significantly between pre-treatment and during treatment. Post-treatment scans revealed a small but statistically significant increase in FDG activity compared with scans during treatment (P=0.01). The findings disproved "the traditional belief that radiation causes inflammatory changes in normal lung tissue during RT," the authors asserted.

The PET imaging was performed with a PET-CT scanner. Changes in tumor FDG activity during radiation therapy had a significant association with post-treatment CT assessment of tumor response (P=0.03).

Although encouraged by the results, the investigators noted the study's limitations: small sample size, heterogeneous study population, and non-uniform treatment regimens.