Predictors of Depression and Anxiety in Head and Neck Cancer Patients

Presentation: AHNS-062
Topic: Other
Type: Oral
Date: Thursday, May 2, 2019
Session: 10:15 AM - 11:00 AM Scientific Session 8 - Outcomes & Miscellaneous
Authors: Ashok R Jethwa, Katrina Hueniken, Catriona M Douglas, Geoffrey Liu, Andrew Bayley, Shao Hui Huang, Aaron Hansen, David P Goldstein, Madeline Li, John R de Almeida
Institution(s): Princess Margaret Cancer Center

Background: Depression affects up to 40% of head and neck cancer (HNC) patients. Additionally, patients with oral cavity/pharynx and larynx cancer have the third and fourth highest rates of suicide among cancer sites, respectively. Unfortunately, HNC patients with depression have been shown to have decreased survival and decreased functional outcomes. However, few studies have investigated risk factors or predictors of depression in HNC patients. We aim to identify patient factors, such as out of pocket costs or occupational loss, or disease factors, such as tumor site, stage, or treatment modality which may predict development of depression.

Methods: Patients with a new diagnosis of HNC (lip, oral cavity, pharynx, larynx, unknown primary) treated at the Princess Margaret Cancer Center/University Health Network between October 22, 2012 and December 31, 2017 were included in the study. Patients prospectively completed the Edmonton Symptom Assessment Scale (ESAS) at routine clinic visits between 10 and 14 months post-diagnosis and baseline clinical, treatment and pathologic data was collected. We excluded those patients who had recurrence less than one month prior to completing questionnaires. Patients were screened for anxiety and depression using scores of 4 or greater using the ESAS, a validated screening tool for psychological distress among cancer patients. Associations between patients who screened positive for anxiety/depression and predictor variables such as age, tumor site, stage, ECOG performance status, treatment modality, smoking and alcohol history were analyzed using univariable and multivariable logistic regression.

Results: Four hundred ninety-five patients were identified and included in our study. The mean age of the study group was 59.6 (SD=10.1), with a majority of male patients (79%). Sixty-seven percent and 66% of patients reported a smoking and drinking history, respectively. Of all patients, 78% had stage III-IV disease at diagnosis. Primary tumor sites included pharynx (57%), larynx (15%), lip/oral cavity (23%), and unknown primary (5%). Most patients had high performance status ECOG 0 (65%), ECOG 1 (33%), and ECOG 2+ (3%). Treatment modalities included chemoradiotherapy (43%), surgery alone (32%), surgery followed by adjuvant (chemo)radiotherapy (16%), and radiation alone (8%). At one year post diagnosis, 17% and 19% of patients screened positive for depression and anxiety, respectively. A younger age at diagnosis and disease progression were associated with a positive ESAS screening score for anxiety on multivariable analysis (OR=0.969, p=0.012; OR 2.662, p=0.021, respectively). ECOG status (ECOG 1 vs 0 and ECOG 2+ vs 0), and disease progression were associated with a positive ESAS screening score for depression on multivariable analysis (OR=1.917, p=0.016; OR=7.570, p=0.002; OR=3.014, p=0.011, respectively). Disease site, treatment modality, and smoking/alcohol history showed no significant association with depression/anxiety. The median lost household income was $25,000 and out of pocket cost (mid-treatment) was $498. Neither of these variables was associated with anxiety or depression on univariable analysis.

Conclusion: Younger age at diagnosis, poorer ECOG performance status, and disease progression are associated with higher rates of anxiety/depression in HNC patients after treatment while tumor stage, disease site, and treatment modality did not show a significant association.