Explaining the effects of socio-economic deprivation on survival in a national prospective cohort study of 1909 patients with head and neck cancers.

BACKGROUND: Socio-economic differences in survival from head and neck cancers are among the largest of any malignancies. Population-based data have been unable to explain these differences. AIMS: To describe survival from head and neck cancers in a large cohort of patients for whom a range of socio-economic, demographic, behavioural and casemix data was available. METHODS: Prospective cohort study using data from the Scottish Head and Neck Audit on all patients diagnosed with a head and neck cancer in Scotland between 1st September 1999 and 31st August 2001 linked to General Register Office for Scotland death records to 30th June 2006. Cox proportional hazards models were produced to describe adjusted hazards of death according to socio-economic circumstances, using validated area-based DEPCAT scores. RESULTS: Data on 1909 patients were analysed. 71.0% were male and mean age was 64.3 (SD 12.2) years. Overall 5-year survival was 45.6% (95% CI: 43.4-47.8%). In order of strength of association in univariate regression, World Health Organisation Performance Status, disease stage, patient age, tumour site, smoking status, alcohol use, tumour differentiation, and deprivation were significant predictors of all-cause mortality but after multiple adjustment, deprivation was no longer an independent predictor of survival. CONCLUSIONS: Socio-economic differentials in survival from head and neck cancers are determined by a mixture of risk factors, some of which may be amenable to targeted earlier detection methods and lifestyle interventions. However, further research is needed to understand the impacts of performance status in more deprived patients.

Components of socioeconomic risk associated with head and neck cancer: a population-based case-control study in Scotland.

The complex associations between socioeconomic circumstances and risk for head and neck cancer are under-explored. We investigated components of social class and their relative influence on the risk of head and neck cancers by studying 103 patients (age range 24-80 years) who had been diagnosed with cancer of the head and neck between April 2002 and December 2004, and 91 controls who were randomly selected from general practitioners' lists. Information about occupation, education, smoking, and alcohol consumption was collected at personal interview. Socioeconomic circumstances were measured at an individual level (education, occupational social class, unemployment), and by area-based measures of deprivation. Odds ratios (OR) and 95% confidence intervals (CI) were computed using unconditional logistic regression and multivariate analyses. People living in the most deprived areas (OR=4.66, 95% CI 1.79-12.18); and those who were unemployed (OR=2.27, 95% CI 1.21-4.26) had a significantly higher risk of cancer than those with high levels of educational attainment (OR=0.17, 95% CI 0.05-0.58). Significance was lost for all measures of social class when adjustments were made for smoking and consumption of alcohol. Smoking was the only significant risk factor (OR=15.53, 95% CI 5.36-44.99) in the multivariate analysis. A high risk of head and neck cancer was consistently associated with poor socioeconomic circumstances, and there were strong links for specific components however smoking dominated the overall profile of risk. We propose a framework for future socioeconomic analyses.

Sentinel node in head and neck cancer: use of size criterion to upstage the no neck in head and neck squamous cell carcinoma.

BACKGROUND: Anatomical imaging tools demonstrate poor sensitivity in head and neck squamous cell carcinoma (HNSCC) patients with clinically node-negative necks (cN0). This study evaluates nodal size as a staging criterion for detection of cervical metastases, utilizing sentinel node biopsy (SNB) and additional pathology (step-serial sectioning, SSS; and immunohistochemistry, IHC). METHODS: Sixty-five patients with clinically N0 disease underwent SNB, with a mean of 2.4 nodes excised per patient. Nodes were fixed in formalin, bisected, and measured in 3 axes before hematoxylin-eosin staining. Negative nodes were subjected to SSS and IHC. SNB-positive patients underwent modified radical neck dissection. RESULTS: Maximum diameter was larger in levels II and III (13.1 and 13.2 mm) when compared with level I (10.5 mm; p = .004, p = .018), while minimum diameter was constant. Positive nodes were larger than negative nodes (p = .007), but nodes found positive by SSS/IHC were not significantly larger than negative nodes for either measurement (p = .433). Sensitivity and specificity were poor for all measurements. CONCLUSIONS: Nodal size is an inaccurate predictor of nodal metastases and should not be regarded as an accurate means of staging the clinically N0 neck.

Chewing and swallowing after surgical treatment for oral cancer: functional evaluation in 196 selected cases.

One hundred ninety-six patients treated for oral cancer between 1992 and 1999 self-scored their speech, chewing, and swallowing using a new self-questionnaire (Functional Intraoral Glasgow Scale) developed at Canniesburn Hospital, Glasgow, to assess the functional efficiency of patients treated for intraoral cancer. The patients were distributed into 12 homogeneous groups, according to the site and size of surgical resection, carefully mapped out on standard diagrams of the oral cavity. The functional outcome for chewing and swallowing was correlated to the site and size of resected tissue, to the reconstruction modality, and to radiotherapy and compared with the speech quality. The general trend is very similar for both chewing and swallowing; the smaller the resections, the better the functional outcome. Chewing was mostly affected by resections of the floor of the mouth, whereas swallowing was mostly affected by demolition of the base of the tongue and of the retromolar trigone. Speech showed a better postoperative recovery than chewing and swallowing. The reconstruction modality did not influence the eventual outcome for either function. Radiotherapy in combination with surgery is a negative functional prognostic factor. A correlation between site and size of excision and functional outcome is presented using color multiple-view diagrams for immediate appreciation to identify positive and negative prognostic factors.

Objective assessment of speech after surgical treatment for oral cancer: experience from 196 selected cases.

In 1992, a personal computer-based workstation for speech-digitized analysis was developed in conjunction with Canniesburn Hospital and Edinburgh University to measure all dispersion in speech after surgery for oral cancer. The voices of 196 patients with tumor of the oral cavity were recorded preoperatively and postoperatively. Surgical resection was carefully mapped out on standard diagrams of the oral cavity. Patients' recordings were assessed for conversational understandability by two referees. Patients also self-scored their speech using the Functional Intraoral Glasgow Scale self-questionnaire. Many patients had similar if not identical resections; therefore, 12 homogeneous groups were identified. Functional outcome for speech was correlated with the site and size of resected tissue and with the reconstruction modalities. The original association of an objective, computer-based tool and two subjective assessment tools proved to be the most suitable investigation method for speech. The general pattern was for consistently better speech quality with smaller excisions. The reconstruction modalities did not seem to influence the overall speech quality, as it was related mainly to the extent of surgical demolition. The authors present a detailed correlation between site and size of excision and functional outcome using color multiple-view diagrams for immediate appreciation. Positive and negative prognostic factors were identified in surgery for oral cancer.