Effects of Living in Food Priority Areas on Treatment and Outcomes in Patients with Stage III Nonsmall Cell Lung Cancer

Oncological outcomes and treatment-related toxicities have been closely correlated to nutrition status before, during and after definitive therapies for multiple cancer sites, including nonsmall cell lung cancer (NSCLC). In some inner cities, up to a quarter of Americans reside in areas with poor access to nutritious food i.e. food priority areas (FPAs). While the effects of living in FPAs have been characterized in comorbid conditions such as diabetes, hypertension and cardiovascular disease, there is limited data on effects of living in FPAs on cancer care. We hypothesized that residence in FPAs influences treatment choice, compliance and outcomes. An institutional, retrospective analysis was performed on 422 consecutive patients with stage III NSCLC definitively treated between January 2000 and December 2016. Chi-square tests were used to compare categorical variables stratified by residence in FPAs. The Kaplan-Meier analysis and the Cox proportional hazard models were used to analyze overall survival (OS) and freedom-from-recurrence (FFR) in patients that did and did not reside in FPAs (as determined by zip code) at the time of cancer diagnosis. Cox regression with forward model selection was used for the multivariate analysis. Thirty-one percent of our stage III NSCLC patients (N = 130) resided in FPAs. Residence in FPAs significantly correlated with being black, single, having a low median income, and a poor performance status (all p<0.001). Though there was no significant difference in pre- or post-chemoradiation (CRT)-BMI (pre: 27.0± 5.8 vs.26.7 ± 6.5, p = 0.2, post: 25.8± 5.3 vs 27.6 ± 20.1, p = 0.34)when stratified by FPA status, patients that resided in FPAs had significantly lower pre-treatment mean albumin levels (3.8 ± 0.6 vs. 3.4 ± 0.8, p<0.001). Patients receiving concurrent CRT (94% vs. 90%), an RT dose ≥60 Gy (88% vs. 81%) or consolidation chemotherapy (70% vs 67%,) were indistinguishable between patients living outside or within FPAs, respectively. However, patients from FPAs were less likely to undergo trimodality therapy after neoadjuvant CRT (36% vs 12%, p<0.001), which translated to higher local and distant recurrence (15% vs. 6%, p = 0.023). However, this did not translate to statistical significance in OS (25m vs. 23m) or FFR (19m vs. 15m) for patients that did not and did reside in FPAs, respectively. In this largest reported institutional analysis examining impact of FPA in patients with stage III NSCLC, we demonstrate the socioeconomic chasm that exists within the population. Residing in an FPA influenced patient demographics, receipt of aggressive therapy (i.e. surgery) and correlated with distinctive patterns of failure. To overcome limitations of retrospective analysis, future efforts are underway to prospectively characterize the nutritional needs of our cancer patients.