A secondary analysis of FDG spatio-temporal consistency in the randomized phase II PET-boost trial in stage II-III NSCLC.

Abstract

FDG-PET scans have shown spatial consistency in NSCLC patients before and following chemoradiotherapy, implying radioresistance. We hypothesized that patients, who received FDG-PET redistributed dose painting, would demonstrate reduced spatial consistency when compared to registered patients or to escalated dose treatment. Stage II-IIIB, inoperable NSCLC patients were randomized in a phase II trial (NCT01024829) to (chemo)radiotherapy of either homogeneous boosting to the primary tumor, or redistributed inhomogeneous boosting to the GTV subvolume (FDG-SUV > 50% SUV_max). Patients who could not be boosted (≥72 Gy) received 66 Gy in 24 fractions. Spatial consistency of pre-treatment and post-treatment (3 months) FDG-PET scans was measured by various overlap fraction thresholds. 66/82 patients analyzed received randomized treatment in the trial. Thresholds of 50% SUV_max pre-treatment and 70% SUV_max post-treatment yielded a median overlap fraction of 0.63 [interquartile range: 0.15-0.93], with similar results for other thresholds. No significant differences were found among overlap fractions of the treatment groups. A high incidence of FDG-uptake in normal lung (grade-1 pneumonitis: 73%) was found post-treatment. FDG redistributed boosting did not reduce FDG spatial consistency from pre-treatment to post-treatment, which was highly variable among patients. The study found high numbers of patients with lung inflammation after treatment.