Alternative quantification of global lung glycolysis using volumetric FDG PET/CT parameters in lung cancer patients treated with radiotherapy

Objectives We evaluated the feasibility of quantifying global lung parenchymal glycolysis (GLPG) using an alternative semiautomatic lung segmentation technique in lung cancer patients treated with photon and proton radiation therapy (RT), where they may serve as biomarkers for treatment-related inflammation and radiation-induced pulmonary toxicity

Methods We evaluated 18 consecutive patients on this IRB-approved study who underwent pre and post-treatment FDG-PET/CT scans from a cohort of locally advanced non-small cell lung cancer patients treated with definitive (median 66.6 Gy in 1.8 Gy fractions) photon or proton RT at our institution between 2010 - 2014. Lung volume segmentation was conducted using a thresholding function. Global lung glycolysis (GLG) was calculated by summing FDG uptake in the whole lung. The SUVmean and total lesion glycolysis (TLG=SUVmean*metabolic active volume) of primary tumor were quantified using an iterative contrast oriented thresholding algorithm. GLPG was calculated by subtracting TLG from GLG. Parameters of baseline and follow up PET were compared using two-tailed paired t-tests.

Results Overall, we did not observe differences in GLPG between pre- and post-treatment images in the ipsilateral or contralateral lung. In the subset of 9 patients treated with photon RT, there was a significant increase in GLPG in both ipsilateral (mean difference 1727±1807, -3502 to 3807, p=0.0209) and contralateral (mean difference 1222±1349, 185 to 2259, p=0.0263) lungs. In the subset of 9 patients treated with proton RT, no increase in GLPG was observed in either the ipsilateral or contralateral lung (mean difference 152.7±4755, -3502 to 3807, p=0.90 for ipsilateral GLPG)

Conclusions We identified increases in bilateral lung GLPG after photon-based RT. These findings did not extend to patients receiving proton RT. Future studies will examine whether this discrepancy is related to limitations of our quantification methods or inherent differences in proton and photon RT