Evaluation of FLT-PET and DWI-MRI in Patients With NSCLC Treated With a Platinum-based Doublet as Preoperative Chemo

This is a prospective, multicenter, single-arm imaging trial. Patients with NSCLC will undergo 18F-FLT-PET/CT and DWI-MRI scans on three separate occasions: at baseline, at 14 days (maximum +/- 1 days deviation is acceptable) after first administration of chemotherapy and finally after up to 4 cycles of chemotherapy, and then followed by surgery. The quantification of 18F-FLT SUV and ADC will be measured to assess tumor characteristics and response to therapy. Patients will be first registered into the EORTC system after signing the informed consent form. The site will have to complete all the study related procedures within 6 weeks prior to treatment and all eligibility criteria should be met before the patient can be enrolled into the study. Both 18F-FLT-PET/CT and DWI-MRI will be assessed independently in this trial. Therefore the overall type I error of 0.1 will be split by two in order to test independently each imaging biomarker with a one-sided type I error of 0.05. In order to demonstrate with 95% confidence interval (one sided) that the correlation between the imaging biomarker change and the pathological response is \> 0.5 (H0: rho ≤ 0.5) with 90% power if the true correlation is 0.8 (H1: rho \> 0.5), 31 lesions are needed. As, in this population, patients only have the primary tumor, 31 eligible and evaluable patients will be needed for each primary endpoint. If all included patients have both 18F-FLT-PET/CT and DWI-MRI then 31 eligible and evaluable patients will be enough. If some patients have only one type of scans (18F-FLT-PET/CT or DWI-MRI), the sample size would need to be adapted to have 31 patients with each type of scans. In addition, the total number of patients to be registered may be increased to 40 patients for each primary endpoint to take into account some screening failure. For the primary analysis of correlation between relative change in imaging biomarkers (18F-FLT-SUV or ADC) and pathological quantification, the correlation coefficient will be reported using a one-sided 95% confidence interval, and tested as a one-sided comparison to the null hypothesis (H0: ρ ≤ 0.5). All secondary objectives to correlate preoperative imaging biomarkers and IHC biological markers or tumor volume will use the same analysis as cited above with 99% confidence intervals. All measures will be analyzed in a random effect ANOVA model allowing for within center-variability. Quality assurance is planned for the control of data consistency, on-site monitoring, audits, and quality assurance in pathological response assessment and in imaging.