Biologically Active Volume of Disease (BaVD) Predicts for Survival in Metastatic Non-Small Cell Lung Cancer.

Abstract

PURPOSE/OBJECTIVE(S)\nMultiple randomized trials incorporating locally ablative therapy in oligometastatic non-small cell lung cancer (NSCLC) have used number of metastatic sites to identify which patient populations benefit from aggressive local therapy. In this study, we measure the amount of fluorodeoxyglucose (FDG) activity on positron emission tomography/computerized tomography (PET/CT) scans to quantify the biologically active volume of disease (BaVD) at diagnosis. We hypothesize that BaVD is a significant predictor of overall survival (OS) for NSCLC.\n\n\nMATERIALS/METHODS\nWe reviewed consecutive charts of metastatic NSCLC patients treated at our institution. All patients included had a PET/CT scan prior to starting any therapy. An oncology imaging informatics system was utilized to auto-contour individual PET/CT scans. A threshold SUV of 3 was selected a priori to auto-contour all FDG avid areas on each patient s pre-treatment PET/CT. SUV activity was measured in cubic centimeters (cc). Normal physiologic activity in expected organs including brain, heart, liver, spleen, genitourinary track, and gastrointestinal track was removed through auto-contouring and adjusted manually if needed. All plans were reviewed by a radiation oncologist to confirm FDG avid areas contoured only included areas of gross disease. Kaplan-Meier analysis evaluated overall survival (OS) based on volume of metastatic disease at diagnosis. Multivariate Cox regression (MVA) was performed accounting for age, histology, gender, race, smoking status, molecular mutations, presence of brain metastasis at diagnosis, number of metastatic sites (1-3, 4-5, > 5 sites), and type of systemic treatment received.\n\n\nRESULTS\nA total of 43 patients were included. Median volume of disease using an SUV cutoff of ≥3 was 317 cc (range, 23-2476 cc). Median survival was worse for patients with ≥ 317 cc of disease compared to those < 317 cc, 17 mo vs 22 mo (P\u202f=\u202f0.010). Volume of disease was found to be significant under univariate analysis (UVA) both when analyzed as a continuous (hazard ratio [HR] 1.01, 95% confidence interval [CI] 1.00-1.02, P < 0.001) and categorical variable with a cutoff of 317 cc (HR 3.27, 95% CI, 1.26-8.48, P\u202f=\u202f0.013). On MVA, volume of disease remained a significant predictor of OS (HR 5.19, 95% CI 1.17-22.99, P\u202f=\u202f0.030). Additional variables associated with OS included older age (HR 6.26, 95% CI 1.65-23.76, P\u202f=\u202f0.007) and receipt of chemotherapy/immunotherapy (HR 20.51, 95% CI 1.98-212.09, P\u202f=\u202f0.011) vs targeted therapy. While number of metastatic sites was significant under UVA (HR 8.45, 95% CI 1.10-64.72, P\u202f=\u202f0.040), it was not significant under MVA (HR 5.23, 95% CI 0.56-49.91, P\u202f=\u202f0.164) CONCLUSION: To our knowledge, this is one of the first studies to demonstrate the significance of BaVD as an important predictive factor of OS in metastatic NSCLC. Our results also suggest BaVD may be a stronger correlate to OS compared to number of metastatic sites. Further work is needed to better understand the impact BaVD on OS.