Depressive Symptoms, Diurnal Salivary Cortisol, and the Role of Inflammation in Treatment Outcomes in Head and Neck Cancer

Presentation: C064
Topic: Pharynx / Larynx Cancer
Type: Poster
Date: Thursday, April 19, 2018
Session: 9:00 AM - 7:00 PM
Authors: Cynthia Duck, JD1, Courtney Brinkman2, Christina Albert, BA3, Whitney Rebholz, PhD3, Paula M Chilton, PhD4, Thomas C Mithcell, PhD5, Sandra E Sephton, PhD2, Jeffrey Bumpous, MD3, Elizabeth Cash, PhD6
Institution(s): 1University of Louisville School of Medicine, 2University of Louisville Dept. of Psychological and Brain Sciences, 3University of Louisville Department of Otolaryngology-Head and Neck Surgery & Communicative Disorders, 4Christine M. Kleinert Institute for Hand & Microsurgery, 5Institute for Cellular Therapeutics; Department of Microbiology & Immunology, 6University of Louisville Department of Otolaryngology-Head and Neck Surgery & Communicative Disorders; Department of Psychological & Brain Sciences

Objective. Head and neck cancer (HNC) patients experience depressive symptoms early in the cancer trajectory at a higher rate than patients with other cancers. We have previously found these to be predictive of poorer overall HNC survival. To examine potential mediators of these effects, we piloted the collection of circadian, endocrine, and immune data among a sample of HNC patients. We hypothesized that depressive symptoms would be related to circadian disruption and systemic inflammation, as well as short (acute toxicities, tumor response) and longer-term treatment outcomes (progression-free survival; PFS).

Methods. Recently diagnosed HNC patients presenting to a Multidisciplinary Clinic (N=53) reported on depressive symptoms (Patient Health Questionnaire; PHQ-9), provided saliva samples for diurnal cortisol assessment twice daily over six days, and had serum drawn for multiplex 13-bead array immune assay, prior to the initiation of treatment. Records review two years post treatment documented frequency of toxicities within six months of treatment end (mucositis, dermatitis, xerostomia, dysphagia, dehydration, fibrosis, pain, fatigue, and esophagitis), tumor response to treatment, and PFS. Hierarchical and logistic regression, and Cox proportional hazard models tested hypotheses. Spearman correlations were used to determine which inflammatory analytes were associated with depression and cortisol and should be included in hypothesis tests.

Results. Patients were mostly (60%) male, averaging 58 years of age, primarily presenting with late (43% stage IV) cancers. Oral (30%), oropharyngeal (28%), laryngeal (15%), hypopharyngeal (8%), and other cancers of the head and neck region (18%) were included. More than half of the sample received radiation in combination with another treatment modality. Median follow-up was 294 days.

Depressive symptoms were not significantly related to cortisol or immune measures, or longitudinal treatment and PFS outcomes. Higher mean bedtime cortisol levels were marginally associated with poorer tumor response to treatment (HR=5.6; 95% CI=0.974-32.198; p=.054), and significantly associated with interferon-gamma (r-square=.406, p=.017) and interleukin-1beta (r-square=.385, p=.025), though gender explained a significant amount of model variance. Higher mean bedtime cortisol was predictive of poorer two-year PFS (HR=8.511; 95% CI=1.076-67.308; p=.042), and this association was robust to the influence of gender.

Conclusions. Contrary to expectation, depressive symptoms were generally mild in this sample of HNC patients and were not significantly related to cortisol, inflammatory analytes, or outcomes within two years of treatment. Gender differences accounted for associations between salivary cortisol and acute treatment outcomes. These findings are consistent with recommendations from RTOG to consider the impact of gender on treatment outcomes. Higher mean bedtime cortisol levels were related to poorer two-year PFS, a finding that was independent of the effects of gender. This relationship highlights biomarker pathways that may influence cancer progression. Consideration of depression and salivary cortisol is warranted in a larger sample with longer-term follow-up that may be better poised to replicate our previously observed relationship between pre-treatment depression and overall HNC survival. [Supported by a UofL Multidisciplinary Research Intramural Grant]