ABSTRACT
PURPOSE: The tumour molecular profile predicts the activity of epidermal growth factor receptor (EGFR) inhibitors in non-small-cell lung cancer (NSCLC). However, tissue availability and tumour heterogeneity limit its assessment. We evaluated whether [(18)F]FDG PET might help predict KRAS and EFGR mutation status in NSCLC. METHODS: Between January 2005 and October 2011, 340 NSCLC patients were tested for KRAS and EGFR mutation status. We identified patients with stage III and IV disease who had undergone [(18)F]FDG PET/CT scanning for initial staging. SUVpeak, SUVmax and SUVmean of the single hottest tumour lesions were calculated, and their association with KRAS and EGFR mutation status was assessed. A receiver operator characteristic (ROC) curve analysis and a multivariate analysis (including SUVmean, gender, age and AJCC stage) were performed to identify the potential value of [(18)F]FDG PET/CT for predicting KRAS mutation. RESULTS: From 102 patients staged using [(18)F]FDG PET/CT, 28 (27%) had KRAS mutation (KRAS+), 22 (22%) had EGFR mutation (EGFR+) and 52 (51%) had wild-type KRAS and EGFR profiles (WT). KRAS+ patients showed significantly higher [(18)F]FDG uptake than EGFR+ and WT patients (SUVmean 9.5, 5.7 and 6.6, respectively; p < 0.001). No significant differences were observed in [(18)F]FDG uptake between EGFR+ patients and WT patients. ROC curve analysis for KRAS mutation status discrimination yielded an area under the curve of 0.740 for SUVmean (p < 0.001). The multivariate analysis showed a sensitivity and specificity of 78.6% and 62.2%, respectively, and the AUC was 0.773. CONCLUSION: NSCLC patients with tumours harbouring KRAS mutations showed significantly higher [(18)F]FDG uptake than WT patients, as assessed in terms of SUVpeak, SUVmax and SUVmean. A multivariate model based on age, gender, AJCC stage and SUVmean might be used as a predictive marker of KRAS mutation status in patients with stage III or IV NSCLC.
