To examine the impact of comorbidity on overall survival (OS) in a population‐based study of patients with head and neck cancer who were treated between 2009 and 2011. Data of 1094 patients with primary head and neck carcinomas without distant metastasis from the Thuringian cancer registries were evaluated concerning the influence of patient's characteristics and comorbidity on OS. Data on comorbidity prior to head and neck cancer diagnosis was adapted to the Charlson Comorbidity (CCI), age‐adjusted CCI (ACCI), head and neck CCI (HNCCI), simplified comorbidity score (SCS), and to the Adult Comorbidity Evaluation–27 (ACE‐27). Most patients were male (80%; median age: 60 years; 50% stage IV tumors). Smoking, alcohol abuse, and anemia were registered for 38%, 33%, and 23% of the patients, respectively. Predominant therapy was surgery + radiochemotherapy (30%), surgery (29%), and surgery + radiotherapy (21%). Mean CCI, ACCI, HNCCI, SCS and ACE‐27 were 1.0 ± 1.5, 2.6 ± 2.1, 0.6 ± 0.8, 4.4 ± 4.2, and 0.9 ± 0.9, respectively. Median follow‐up was 25.7 months. Multivariable analyses showed that higher age, higher UICC stage, no therapy, including surgery or radiotherapy, alcohol abuse, and anemia, higher comorbidity were independent risk factors for worse OS (all P < 0.05). According to the discriminatory power analysis none of the five comorbidity scores was superior to the other scores to prognosticate OS. This population‐based study showed that comorbidity is frequent in German patients with head and neck cancer and is an important risk factor for poor OS. Comorbidity should be routinely assessed and taken into account in prospective clinical trials.

CSS and OS did not improve during the study period. Incidence of parotid cancer increased significantly in elderly patients. It seems that grading is next to patient's age the better predictor of OS than TNM stage.

No abstract

No abstract

Background This population-based study investigated the influence of different lymph node (LN) classifications on overall survival (OS) in head and neck cancer (HNC). Methods 401 patients (median age: 57 years; 47% stage IV) of the Thuringian cancer registries with diagnosis of a primary HNC receiving a neck dissection (ND) in 2009 and 2010 were included. OS was assessed in relation to total number of LN removed, number of positive LN, LN ratio, and log odds of positive LN (LODDS). Results Mean number of LODDS was 0–0.96 ± 0.57. When limiting the multivariate analysis to TNM stage, only the UICC staging (stage IV: HR 9.218; 95% CI 2.721–31.224; p < 0.001) and LODDS >  – 1.0 (HR 2.120; 95% CI 1.129–3.982; p = 0.019) were independently associated with lower OS. Conclusion LODDS was an independent and superior predictor for OS in HNC in a population-based setting with representative real-life data.