Prognostic value of post-induction chemotherapy (18)F-FDG PET-CT in stage II/III non-small cell lung cancer before (chemo-) radiation

INTRODUCTION: The purpose of our present study was to assess the prognostic impact of FDG PET-CT after induction chemotherapy for patients with inoperable non-small-cell lung cancer (NSCLC). MATERIAL AND METHODS: This retrospective study included 50 patients with inoperable stage II/III NSCLC from January 2012 to July 2015. They were treated for curative intent with induction chemotherapy, followed by concomitant chemoradiation therapy or sequential radiation therapy. FDG PET-CT scans were acquired at initial staging (PET(1)) and after the last cycle of induction therapy (PET(2)). Five parameters were evaluated on both scans: SUVmax, SUVpeak, SUVmean, TLG, MTV, and their respective deltas. The prognostic value of each parameter for overall survival (OS) and progression-free survival (PFS) was evaluated with Cox proportional-hazards regression models. RESULTS: Median follow-up was 19 months. PET(1) parameters, clinical and histopathological data were not predictive of the outcome. TLG(2) and ΔTLG were prognostic factors for OS. TLG(2) was the only prognostic factor for PFS. For OS, log-rank test showed that there was a better prognosis for patients with TLG(2)< 69g (HR = 7.1, 95%CI 2.8–18, p = 0.002) and for patients with ΔTLG< -81% after induction therapy (HR = 3.8, 95%CI 1.5–9.6, p = 0.02). After 2 years, the survival rate was 89% for the patients with low TLG(2) vs 52% for the others. We also evaluated a composite parameter considering both MTV(2) and ΔSUVmax. Patients with MTV(2)> 23cc and ΔSUVmax> -55% had significantly shorter OS than the other patients (HR = 5.7, 95%CI 2.1–15.4, p< 0.01). CONCLUSION: Post-induction FDG PET might be an added value to assess the patients’ prognosis in inoperable stage II/III NSCLC. TLG, ΔTLG as well as the association of MTV and ΔSUVmax seemed to be valuable parameters, more accurate than clinical, pathological or pretherapeutic imaging data.