Prognostic value of metabolic tumor burden on 18F-FDG PET in nonsurgical patients with non-small cell lung cancer

Purpose The objective of this study was to assess the prognostic value of metabolic tumor burden on 2-deoxy-2-[F-18]fluoro-D-glucose (F-18-FDG) positron emission tomography (PET)/CT measured with metabolic tumor volume (MTV) and total lesion glycolysis (TLG), independent of Union Internationale Contra la Cancrum (UICC)/American Joint Committee on Cancer (AJCC) tumor, node, and metastasis (TNM) stage, in comparison with that of standardized uptake value (SUV) in nonsurgical patients with non-small cell lung cancer (NSCLC). Methods This study retrospectively reviewed 169 consecutive nonsurgical patients (78 men, 91 women, median age of 68 years) with newly diagnosed NSCLC who had pretreatment F-18-FDG PET/CT scans. The F-18-FDG PET/CT scans were performed in accordance with National Cancer Institute guidelines. The MTV of whole-body tumor (MTVWB), of primary tumor (MTVT), of nodal metastases (MTVN), and of distant metastases (MTVM); the TLG of whole-body tumor (TLG(WB)), of primary tumor (TLG(T)), of nodal metastases (TLG(N)), and of distant metastases (TLG(M)); the SUVmax of whole-body tumor (SUVmaxWB), of primary tumor (SUVmaxT), of nodal metastases (SUVmaxN), and of distant metastases (SUVmaxM) as well as the SUVmean of whole-body tumor (SUVmeanWB), of primary tumor (SUVmeanT), of nodal metastases (SUVmeanN), and of distant metastases (SUVmeanM) were measured with the PETedge tool on a MIMvista workstation with manual adjustment. The median follow-up among survivors was 35 months from the PET/CT (range 2-82 months). Statistical methods included Kaplan-Meier curves, Cox regression, and C-statistics. Results There were a total of 139 deaths during follow-up. Median overall survival (OS) was 10.9 months [95% confidence interval (CI) 9.0-13.2 months]. The MTV was statistically associated with OS. The hazard ratios (HR) for 1 unit increase of ln(MTVWB), ae(MTVT), ae(MTVN), and ae(MTVM) before/after adjusting for stage were: 1.47/1.43 (p < 0.001/< 0.001), 1.06/1.05 (p < 0.001/< 0.001), 1.11/1.10 (p < 0.001/< 0.001), and 1.04/1.03 (p = 0.007/0.043), respectively. TLG had statistically significant associations with OS with the HRs for 1 unit increase in ln(TLG(WB)), ae(TLG(T)), ae(TLG(N)), and ae(TLG(M)) before/after adjusting for stage being 1.36/1.33 (p < 0.001/< 0.001), 1.02/1.02 (p = 0.001/0.002), 1.05/1.04 (p < 0.001/< 0.001), and 1.02/1.02 (p = 0.003/0.024), respectively. The ln(SUVmaxWB) and ae(SUVmaxN) were statistically associated with OS with the corresponding HRs for a 1 unit increase before/after adjusting for stage being 1.46/1.43 (p = 0.013/0.024) and 1.22/1.16 (p = 0.002/0.040). The ae(SUVmeanN) was statistically associated with OS before and after adjusting for stage with HRs for a 1 unit increase of 1.32 (p < 0.001) and 1.24 (p = 0.015), respectively. The ae(SUVmeanM) and ae(SUVmaxM) were statistically associated with OS before adjusting for stage with HRs for a 1 unit increase of 1.26 (p = 0.017) and 1.18 (p = 0.007), respectively, but not after adjusting for stage (p = 0.127 and 0.056). There was no statistically significant association between OS and ae(SUVmaxT), ln(SUVmeanWB), or ae(SUVmeanT). There was low interobserver variability among three radiologists with intraclass correlation coefficients (ICC) greater than 0.94 for SUVmaxWB, ln(MTVWB), and ln(TLG(WB)). Interobserver variability was higher for SUVmeanWB with an ICC of 0.806. Conclusion Baseline metabolic tumor burdens at the level of whole-body tumor, primary tumor, nodal metastasis, and distant metastasis as measured with MTV and TLG on FDG PET are prognostic measures independent of clinical stage with low inter-observer variability and may be used to further stratify nonsurgical patients with NSCLC. This study also suggests MTV and TLG are better prognostic measures than SUVmax and SUVmean. These results will need to be validated in larger cohorts in a prospective study.