Neighborhood deprivation and clinical outcomes among head and neck cancer patients

Abstract

The unique effects of neighborhood-level economic deprivation on survival, recurrence, and second primary malignancy development were examined using adjusted Cox proportional hazards regression models among 1151 incident squamous cell carcinomas of the head and neck patients. Cancer site was examined as a potential moderator. Main analyses yielded null results; however, interaction analyses indicated poorer overall survival [HR=1.59 (1.00–2.53)] and greater second primary malignancy development [HR=2.99 (1.46–6.11)] among oropharyngeal cancer patients from highly deprived neighborhoods relative to less deprived neighborhoods. Results suggest a dual focus on individual and neighborhood risk factors could help improve clinical outcomes among oropharyngeal cancer patients.

Introduction

Squamous cell carcinomas of the head and neck (SCCHN) are cancers of the epithelial cells in the head and neck (excluding the eyes, ears, brain, thyroid, and esophagus), and are the main histological subtype of head and neck cancers. One of the strongest behavioral risk factors for SCCHN incidence is tobacco use, followed to a lesser extent by excessive alcohol use (Blot et al., 1988, Diaz et al., 2003, Lewin et al., 1998). Other factors related to SCCHN incidence include male sex, older age, African American race, and oncogenic human papillomavirus (HPV) status. Oncogenic HPV status is particularly relevant to oropharyngeal cancer (i.e., cancers that include the middle part of the throat including the soft palate, the base of tongue, the tonsils, and the back and side throat wall), with HPV-positive patients being younger, less likely to have a significant tobacco and alcohol use history, and more likely to have a better prognosis than HPV-negative patients (Chaturvedi et al., 2008). In addition to these demographic and clinical factors, research suggests that individuals of lower socioeconomic status (SES) have higher rates of SCCHN incidence, even when controlling for tobacco and alcohol use status (Johnson et al., 2008).

In total, SCCHN account for between 3% and 5% of incident cancer cases in the US [cf. (NCI, 2011)], and approximately 6% of cancers worldwide (Lefebvre, 2005). Standard treatment options for SCCHN include surgery and/or radiotherapy for early stage or localized disease, with chemotherapy as an adjuvant treatment in advanced stages of disease (ACS, 2011). Despite relatively low incidence rates, treatment costs associated with SCCHN are significant, with recent estimates topping $3.1 billion dollars annually in the US alone [cf. (NCI, 2011)]. Survival rates for SCHHN vary based on disease stage at clinical presentation, and disease stage at presentation does not appear to be related to patients' SES (Johnson et al., 2010). Stages I and II disease represent localized cancers, Stages III, IVa, and IVb represent locally advanced disease with metastasis to local lymph nodes, and Stage IVc represents distant metastasis. Five-year survival rates for patients with SCCHN are about 91% for Stage I disease, 77% for Stage II, 61% for Stage III, 32% for Stage IVa, 25% for Stage IVb, and <4% for Stage IVc. [cf. (Lefebvre, 2005)]. Although 5-year survival rates for SCCHN are relatively high in most cases, patients with SCCHN frequently suffer mortality from comorbid illnesses and relatively high rates of second primary malignancy development [4%, cf. (Lefebvre, 2005)]. In addition, functional morbidities resulting from treatment that affects speech, swallowing, and facial esthetics may result in social and occupational impairments and a potentially lower quality of life (Bonanno et al., 2010, Buckwalter et al., 2007, Givens et al., 2009, Goldstein et al., 2007, Mowry et al., 2010, Terrell et al., 1998, Teymoortash et al., 2010, Yao et al., 2007). Thus, the fiscal, social, and clinical consequences of SCCHN are significant, warranting the need to comprehensively understand the factors that affect treatment outcome among these patients.

So far, studies have cited a number of sociodemographic and clinical factors that affect SCCHN clinical outcomes (i.e., prognosis and survival), including age, sex, race/ethnicity, tumor stage, tobacco use, comorbidity (Piccirillo et al., 2004), HPV status, treatment modality (Chaturvedi et al., 2008, Fakhry et al., 2008, Molina et al., 2008, Pytynia et al., 2004, Schrank et al., 2011), and cancer site (Chaturvedi et al., 2008, Schrank et al., 2011). A number of studies have also focused on the significant relations of SES and SCCHN outcomes, with lower SES predicting poorer clinical outcomes [e.g., (Arbes et al., 1999, Chu et al., 2011; Molina et al., 2008)]. However, these studies have relied on neighborhood-level SES as a proxy for patient-level SES, largely because individual-level SES information is often unavailable within population-based survival databases (e.g., cancer registries). As a result, the true influence of neighborhood SES on SCCHN outcomes, over and above the influence of individual-level SES, remains unstudied. Clarifying whether neighborhood-level SES plays a unique role in SCCHN outcomes is important to improving interventions, as significant associations would suggest that a dual focus on individual and neighborhood factors would be desirable for improving outcomes (Winkleby et al., 2006). Also, the clinical landscape of head and neck cancers has been changing in recent years, with rising rates of oropharyngeal cancers and declining rates of non-oropharyngeal cancers (Sturgis and Ang, 2011). A better understanding of contextual influences on SCCHN by cancer site is needed, as site is associated with prognosis and survival outcomes (Chaturvedi et al., 2008, Schrank et al., 2011).

The purpose of the current study was to examine the unique effect of neighborhood-level SES on survival, recurrence, and the development of second primary malignancies among patients from Texas and Louisiana who were assessed and treated for incident SCCHN at a single, large multidisciplinary cancer center in the southwest. To our knowledge, this is the first study to examine the associations between neighborhood-level SES and SCCHN outcomes while controlling for individual-level income as well as other sociodemographic and clinical variables. In addition to examining relations of neighborhood-level SES with clinical outcomes among all SCCHN patients, the potential for moderation by cancer site was also examined.