Higher area deprivation index is associated with poorer local control and overall survival in non‐metastatic nasopharyngeal carcinoma

Neighborhood socioeconomic deprivation impacts outcomes in various cancers. We examined this association in nasopharyngeal carcinoma (NPC) patients using the area deprivation index (ADI).


| INTRODUCTION
Radiotherapy is the backbone of treatment for early-stage nasopharyngeal carcinoma (NPC), with combinations of chemotherapy to radiotherapy conferring a survival advantage in more advanced cases. 1,2In recent years, adoption of intensity-modulated radiotherapy (IMRT), 3 advances in imaging, 4 and trials in locally advanced 5 and metastatic disease 6 have led to significant improvements in locoregional control and overall survival, as well as reduced treatment-related toxicities for patients with NPC.
Despite these overall advancements, disparities in NPC outcomes persist, potentially influenced by factors such as socioeconomic status (SES) and neighborhood disadvantage, 7 which may influence access to advanced technologies and optimal multidisciplinary management.SES has been linked to cancer outcomes in numerous studies, which suggest that lower SES and neighborhood disadvantage are associated with worse survival, increased disease burden, and reduced access to optimal cancer care in breast, 8 colorectal, 9 prostate, 10 and lung cancers. 11he National Institutes of Health (NIH)-funded Neighborhood Atlas is a publicly available user-friendly tool that provides area deprivation index (ADI) data 12 with granular local-level mapping of neighborhoods throughout the United States, ranking neighborhoods on socioeconomic disadvantage.ADI is a composite measure of 17 census variables based on income, education, housing, and household characteristics.The primary objective of this study is to explore the association of neighborhood and SES disadvantage, as measured by the ADI, with outcomes in patients with non-metastatic NPC managed with curative-intent radiotherapy.

| Study design and data collection
This was a single-institution retrospective cohort study comprised of patients with newly diagnosed nonmetastatic NPC and treated with curative-intent radiotherapy from 1980 to 2023.Data were obtained from the institutional cancer registry and electronic medical records.Baseline patient, cancer, and treatment variables were abstracted.Demographic, clinical, treatment, and follow-up data were extracted and anonymized.About 10% of patients received their initial radiotherapy at another center; if available, the details of their treatments were abstracted and included.ADI for each patient was abstracted from the Neighborhood Atlas website in the form of state deciles and national percentiles, available at https://www.neighborhoodatlas.medicine.wisc.edu/.

| Study population and staging criteria
The study population consisted of patients diagnosed with American Joint Committee on Cancer (AJCC) 8th edition T1-T4 N0-N3 M0 Stage I-IVA NPC, who received curative-intent definitive radiotherapy during the study period.Patients who underwent upfront curative-intent surgery were excluded, as were non-epithelial nasopharyngeal malignancies.Patients with incomplete records were also excluded from the analysis.

| Variable definitions
The primary exposure variable was ADI expressed as a state decile score, a validated measure of SES and neighborhood disadvantage.Other variables collected included age, sex, race, Charlson Comorbidity Index (CCI), AJCC stage, treatment type (radiotherapy alone or chemoradiotherapy), treatment dates and details, last follow up, and vital status.

| Statistical analysis
Categorical data were analyzed using chi-square tests, while continuous variables were assessed using Student's t-tests.Characteristics of patients with high and low ADI were first compared with ADI assessed as a binary variable (state decile scores 1-5 vs. 6-10).Univariable and multivariable analyses were conducted using Cox proportional hazards regression models to identify factors associated with overall survival (OS), local control (LC), regional control (RC), and distant control (DC), with risk on outcomes assessed with ADI as a binary variable, as quartiles, and as risk per increase in ADI state decile.The final multivariable model was built using clinically relevant variables and using a stepwise backward selection process, retaining variables with a p-value <0.05 in univariable analyses.Results were reported as hazard ratios (HRs) with 95% confidence intervals (CIs).All statistical analyses were performed using STATA 15, 13 and a pvalue ≤0.05 was considered statistically significant.
The demographic characteristics of patients in the high and low deprivation groups were similar with respect to age and sex (Table 1).ADI data could be obtained for 550 (98.0%) patients.Caucasian patients were more likely to be from higher ADI (state decile scores 6-10) neighborhoods versus lower ADI (scores 1-5) at 36.1% versus 12.5% respectively, while East Asian patients were less likely so at 59.0% versus 83.8% ( p < 0.01).A higher proportion of patients from high deprivation neighborhoods had higher medical comorbidities with CCI score ≥4, (49.4% vs. 38.8%),although the difference was not statistically significant ( p = 0.07).

| DISCUSSION
This single-institution, retrospective cohort study investigated the association of neighborhood ADI and cancer outcomes in patients diagnosed with non-metastatic NPC receiving definitive radiotherapy.Patients from neighborhoods with higher ADI were less likely of East Asian ethnicities and more likely to present with more advanced AJCC stage.After adjusting for known confounders including comorbidities and cancer stage, higher ADI remained associated with worse local control and overall survival in our cohort, suggesting that higher ADI was independently associated with poorer oncologic outcomes.14,15 Our results contribute to the growing body of literature highlighting the importance of addressing disparities in cancer outcomes related to SES and neighborhood disadvantage.
The observed association between ADI and survival in patients with NPC appears to be primarily mediated by local control and may be due to several factors.Socioeconomic deprivation is a continuum and can result in limited access to healthcare, financial resources, and ability to take time off from employment, which can result in delayed diagnoses and contribute to patients presenting with more advanced stage and higher comorbidities, as seen in our cohort. 15,16Socioeconomic deprivation may also be associated with treatment disruptions such as missed radiotherapy fractions, delays in completing radiotherapy, and incomplete radiation treatment and chemotherapy cycles, which ultimately result in poorer outcomes. 17,18Additionally, cultural and linguistic disadvantages prevalent in high socioeconomic deprivation neighborhoods may also contribute to observed disparities. 19ur study has several limitations, including its retrospective design and single-institution dataset.A notable limitation of our work is lack of information on World Health Organization histologic subtyping of NPC in our cohort.Nevertheless, subgroup analyses of East Asian participants showed a sustained associated between ADI and cancer outcomes.These associations were not seen in a subgroup analysis of non-East Asian participants, likely due to small sample size.Future studies should evaluate the association of ADI and cancer outcomes in larger cohorts of non-East Asian participants.ADI is a composite regional statistic and SES heterogeneity may vary within each neighborhood and cannot be applied broadly at the individual level.Nevertheless, ADI is a powerful tool to further our understanding of disparities in cancer outcomes.Finally, neighborhood ADI became publicly available in 2018 and for much of the study period when ADI data was not available, neighborhood SES may have changed, and these changes are not reflected in this study.

F I G U R E 1
Overall survival by area deprivation index (ADI) state decile score in study cohort in patients with low deprivation (ADI 1-5) and high deprivation (ADI 6-10).[Color figure can be viewed at wileyonlinelibrary.com] non-East Asian participants did not show an association between ADI score and LC or OS ( p > 0.05).
T A B L E 2 Results of univariable Cox regression analysis for overall survival in the study cohort.There were 554 subjects with available data for univariable analyses, with 290 death events.Local control and area deprivation index (ADI) state decile scores in study cohort stratified by quartiles.[Color figure can be viewed at wileyonlinelibrary.com] 17,22menting targeted interventions to improve treatment attendance, adherence, and escalation of population-specific supportive resources.17,22FI G U R E 2 T A B L E 3 Results of multivariable Cox regression analysis for overall survival in the study cohort.There were 490 subjects with all available data for multivariable analyses, with 238 death events in the cohort.