Advancing Education, Research, and Quality of Care for the Head and Neck oncology patient.
Program Number: B224
Session Name: Poster Session
Impact of body mass index on overall survival in Veterans with oropharyngeal cancer
Allison Rollins, MD1; M Bourdreaux-Kelly, PhD2; JL McCoy2; K Contrera, MD1; S Sridharan, MD1; S Kim, MD1; ME Spector, MD1; JP Zevallos, MD, MPH1; M Nilsen, PhD, RN1; A Mazul, PhD, MPH1; JH Maxwell, MD, MPH1; 1University of Pittsburgh Medical Center; 2VA Healthcare SystemBackground: Oropharyngeal cancer (OPC), often related to the human-papillomavirus (HPV), represents a distinct cohort of patients with fewer comorbidities and higher socioeconomic status compared to those with HPV-unrelated head and neck cancers (HNC). Although most OPC patients respond well to treatment, a minority of patients succumb to their disease. Unfortunately, Veterans with OPC have worse outcomes compared to their non-Veteran counterparts. Body mass index (BMI) is a cursory yet easily obtainable measure shown in single-institution retrospective studies to be linked to survival in patients with HNC. Therefore, this study aims to investigate the impact of BMI on 3-year overall survival (OS) among Veterans with OPC.
Methods: A retrospective review was conducted of Veterans with OPC diagnosed between 2012 and 2022, with medical record data extracted from the Veteran’s Health Administration’s Corporate Data Warehouse. Veterans were identified via ICD-9 and ICD-10 codes. Data were collected on the nearest pre-index diagnosis measures for age, gender, self-reported race and ethnicity, marital status, BMI, smoking status, alcohol status, Elixhauser Comorbidity Index score, rurality as determined from geocoded patient location, and national area deprivation index (ADI). Three-year OS was determined as death from any cause within three years of diagnosis. Chi-square testing and logistic regression modeling were conducted to determine the associations between BMI and 3-year OS.
Results: A total of 25,957 Veterans with OPC were included in this study. The cohort was overwhelmingly White (n=20,418; 85.3%) and male (n=25,537; 98.2%). Fewer patients lived in rural or highly rural areas compared to urban (36.5% versus 63.5%, respectively), and 70.3% had an ADI of less than 75. Patients had an average Elixhauser score of 6.44. Approximately 69% were current or former smokers and 40.4% were current or former alcohol users within the two years prior to diagnosis. Three-year OS for the entire cohort was 69%.
Most patients were overweight at diagnosis (34.7%), followed by optimal (31.9%), obese (28.38%), and underweight (5%). Using logistic regression modeling, reduced BMI at diagnosis was associated with worse 3-year OS (<0.001; Figure 1). Furthermore, patients who were categorized as underweight at diagnosis, compared to overweight, were nearly six times more likely to die within 3 years (Odds Ratio [OR]=5.9; 95% confidence interval [CI]=5.1-6.8). Other factors associated with worse 3-year OS included male gender (OR [95% CI]=1.9[1.4-2.6]; p<0.001), older age (OR [95% CI]=1.05[1.04-1.05]; p<0.001), current/former smokers and alcohol users (OR [95% CI]=1.3[1.2-1.4] and OR [95% CI]=1.4 [1.3-1.5], respectively; both p<0.001). Other factors included higher Elixhauser score, higher ADI, and being unmarried at the time of diagnosis (all p<0.001). On multivariate analysis, BMI remained significantly associated with survival (p<0.001). Race and ethnicity were not significantly associated with 3-year OS.
Conclusion: This study demonstrates a protective effect of higher BMI on OS in Veterans with OPC. In addition to known clinical factors that impact outcomes among patients with OPC, this study also reveals the impact of marital status and area deprivation on survival. In contrast to findings from non-Veteran studies, race and ethnicity did not affect OS in our cohort.
Figure 1.
