To Explore a Representative Hypoxic Parameter to Predict the Treatment Response and Prognosis Obtained by [ 18 F]FMISO-PET in Patients with Non-small Cell Lung Cancer

Purpose To explore a representative hypoxic parameter to predict the treatment response and prognosis for [ 18 F]fluoromisonidazole ([ 18 F]FMISO) positron emission tomography (PET)/X-ray computed tomography (CT) in patients with non-small cell lung cancer (NSCLC). Procedures Twenty-nine patients with NSCLC underwent FMISO-PET scans before chemoradiotherapy (CRT). The maximum standard uptake values (SUVmax) in the tumor, normal lung, aortic arch, and vertical ridge muscle were measured, and the tumor-to-lung (T/L) ratios, tumor-to-blood (T/B) ratios, ands tumor-to-muscle (T/M) ratios were calculated and analyzed. Fractional hypoxic volume (FHV) was expressed as percentage of hypoxic volume. Results SUVmax, T/L ratio, T/B ratio, and FHV were all significantly different between the responders and the non-responders (SUVmax, 2.07 ± 0.53 vs. 2.61 ± 0.69, P = 0.026; T/L ratio, 3.16 ± 0.85 vs. 4.09 ± 1.46, P = 0.047; T/B ratio, 1.27 ± 0.20 vs. 1.48 ± 0.32, P = 0.042; 38.92 ± 18.47 vs. 52.91 ± 11.29 %, P = 0.020). However, the T/M ratio was not significantly different between the two populations (1.46 ± 0.31 vs. 1.67 ± 0.33, P = 0.098). The correlation ratio between hypoxic parameters and treatment responses ranged from high to low as FHV ( r = 0.412); SUVmax ( r = 0.400); T/L ratio ( r = 0.379), P < 0.05; and T/B ratio ( r = 0.355), P = 0.059. According to the area under curve (AUC) to predict response, the hypoxic parameters were arranged as FHV (AUC = 0.748), SUVmax (AUC = 0.731), T/L ratio (AUC = 0.719), and T/B ratio (AUC = 0.705 ) . Binary logistic regression analyses showed that FHV was the only independent predictor for treatment response with the P value of 0.038. In the progression-free survival (PFS) prediction, both FHV and SUVmax reached statistical significance by Kaplan–Meier plots (FHV, 46.99 %, P = 0.010; SUVmax, 1.99, P = 0.046) while only FHV was the independent prognostic factor in multivariate analysis by Cox proportional hazard model ( P = 0.037). Conclusion FHV may be a representative hypoxic parameter to predict the CRT response and PFS in patients with NSCLC.