Post-radiation Dental Disease Amongst Head and Neck Cancer Patients

Plan of investigation

Clinical hypothesis The primary hypothesis of the study is that increased mean radiation dose to the teeth and to the 'spared' parotid gland is associated with an increase in the mean number of carious teeth and in the proportion with periodontal disease 2 years post-radiotherapy.

Study design: Prospective cohort study.

• Pre-radiotherapy assessment at baseline and patients rendered dentally fit.
• Post-radiotherapy follow-up assessments at 6 months, 12 months, and 24 months.

Pre-radiotherapy assessment:

In line with current practice, the Multidisciplinary Head and Neck Cancer Team (Royal Victoria Hospital, Belfast) will continue to refer head and neck cancer patients for pre-radiotherapy dental assessment to the School of Dentistry, Belfast. Referred patients will be assessed in the existing 'Head and Neck Clinic' in the School of Dentistry. All potentially eligible patients will be informed of the research study and invited to participate. Each recruited patient will be assessed by a trained and calibrated dentist (examiner). Examiners will undergo pre-study training and calibration to ensure consistent clinical measurements. Inter-examiner consistency will be calculated using the Kappa statistic. A structured protocol will be followed. Radiographs may be requested provided specific criteria have been fulfilled. The same examiner(s) will interpret and record data from all radiographs.

The following measurements will be collected:

• Dental caries: The presence or absence of dental caries on each tooth will be determined using the World Health Organization's criteria and coding system and ICDAS-II, and recorded in a modified version of the Oral Health Assessment Form for Adults, 2013. Radiographic evidence of secondary caries or caries into dentine will also be recorded.
• Periodontal disease: A six-point periodontal charting will be performed for all patients. Clinical attachment loss and probing pocket depths will be measured using a Williams Periodontal Probe and recorded for six sites on each tooth. The 'Centers for Disease Control and Prevention (CDC)', in partnership with the 'American Academy of Periodontology (AAP)' workgroup definition of periodontitis will be used to identify those with active periodontal disease: 2 or more interproximal sites with attachment loss ≥3 mm, and 2 or more interproximal sites with pocket depths ≥4 mm (not on same tooth) or one site with pocket depth ≥5 mm.
• Stimulated salivary flow rate using paraffin pellets.
• Ruler measurement of mouth opening.
• Oral hygiene practice and dietary assessment using the World Health Organization's Oral Health Questionnaire for Adults.
• Subjective assessment of quality of life and experience of xerostomia using questionnaire booklets containing the European Organization for Research and Treatment of Cancer (EORTC) QLQC30 quality-of-life instrument, the associated head and neck specific module HN35, the Oral Health Impact Profile, and the modified Xerostomia Questionnaire.

All patients will receive standardised oral hygiene and dietary advice. A dental treatment plan will be formulated for each patient to ensure adequate dental fitness prior to radiotherapy. Dental treatment required pre-radiotherapy (e.g. restorations and extractions) will be co-ordinated by the School of Dentistry in line with current practice.

Post-radiotherapy assessments:

Participants will be followed-up at 6 months, 12 months, and 24 months post-radiotherapy.

At each follow-up visit, patients will undergo dental assessment in the School of Dentistry, Belfast. The following measurements will be collected (using the same methods and examiners outlined above): charting of dental caries, 6-point clinical attachment loss and probing pocket depth charting, stimulated salivary flow rate, mouth opening, oral hygiene practice, dietary assessment, quality of life, and xerostomia. A structured protocol will be followed. Radiographs may be requested provided specific criteria have been fulfilled. At each visit, patients will receive standardised oral hygiene and dietary advice. Consultants will oversee treatment plans for the restoration of dental health. Required treatment will be co-ordinated by the School of Dentistry in line with current practice.

Radiation dose and tumour location determination Patients will receive radiotherapy as prescribed by their Consultant Clinical Oncologist. The Clinical Oncologist will also plan individual tissue exposures in line with current practice. Using additional research software, doses to the teeth and to the 'spared' parotid gland will be calculated after the teeth and parotid glands have been contoured on each patient's radiotherapy planning CT scan. The Clinical Oncologist will contour the parotid glands in line with current practice. A dentist will contour the teeth on patient-anonymised CT scans. Tumour location will be determined from the initial referral letter sent by the Head and Neck Cancer Multidisciplinary Team.

Sample size Simplified sample size calculations were conducted based upon an independent samples t-test and correlation coefficient. Assuming the standard deviation of the number of carious teeth in post-radiotherapy head and neck cancer patients is 4.0 (based upon the results of a 10-year School of Dentistry audit) and comparing patients receiving over 20 Gray to the 'spared' parotid gland with those receiving under 20 Gray (based upon a 80%/20% distribution seen in a Cancer Centre radiation dose audit), the investigators would require a total sample of 150 patients with available data to have over 80% power to detect a difference of 3.0 in the mean number of carious teeth between the under 20 Gy and over 20 Gy groups as statistically significant at the 5% level. Furthermore, a sample size of 150 patients would allow 80% power to detect, as significant at the 5% level, a correlation coefficient of 0.22 for the association between radiation dose to the 'spared' parotid gland (Gy) and the number of carious teeth. Assuming 30% dropout the investigators would therefore need to recruit 215 patients.

Statistical analysis Primary outcome measures to be analysed include: (1) the number of carious teeth and (2) the presence of periodontal disease.

1. Initial comparisons between the radiation dose to the 'spared' parotid gland and the number of carious teeth will be performed using independent t-tests (categorising radiation dose as <20Gy and >20Gy) and ANOVA (categorising radiation dose as <20Gy, 20-30Gy, 30-40Gy, and >40Gy) as appropriate. Adjusted analysis will be performed using multiple linear regression, with the number of carious teeth as the outcome variable, radiation dose to the 'spared' parotid gland as an explanatory variable (categorised as above) along with potential confounding variables: dental radiation dose, oral hygiene practice, diet, and mouth opening. Without categorising radiation dose, a separate analysis will be conducted using linear regression with the number of carious teeth as the outcome variable and radiation dose to the 'spared' parotid gland (Gy) as the explanatory variable to give the increase in the number of carious teeth per 10 unit increase in Gray. If the number of carious teeth is not normally distributed transformations will be used prior to analysis or equivalent non-parametric techniques will be used.
2. Initial comparisons of the association between radiation dose to the 'spared' parotid gland (categorised as above) and the presence of periodontal disease (Yes/No) will be performed using chi-square tests. Adjusted analysis will be performed using logistic regression with presence of periodontal disease as the outcome variable, radiation dose to the 'spared' parotid gland as the explanatory variable and dental radiation dose, oral hygiene practice, diet, smoking, medical history, and mouth opening as potential confounders. Similar analyses will be undertaken to determine the effects of dental radiation dose and tumour location on the number of carious teeth and the presence of periodontal disease in post-radiotherapy head and neck cancer patients.

Secondary outcomes to be analysed include: quality of life, salivary flow rate, diet, oral hygiene practice, mouth opening, xerostomia (all change from baseline); tooth loss, and denture wear. Dichotomous outcomes will be analysed using chi-square tests and logistic regression as above. Continuous outcomes will be analysed as above, apart from variables with baseline measurements, for which ANCOVA will be used to compare outcome measurements adjusting for baseline. Adjusted analyses will use multiple linear regression models as above but with baseline measurements as an additional explanatory variable.

A micro-costings study will also be undertaken to evaluate patient and healthcare costs in relation to the diagnosis and treatment of pre- and post-radiotherapy dental disease.

• Patient costs to be evaluated include: cost of travel to dental appointments, time off work to attend for dental treatment, cost of dental treatment incurred pre and post-radiotherapy (general or private practice), out-of-pocket expenses (e.g. pain-relief medications).
• Healthcare costs to be evaluated include: clinical and administration staff costs for pre- and post-radiotherapy dental assessments and treatment, equipment, general overheads, remuneration of primary care dental practitioners.

Cost savings due to improved treatment planning of patients will also be hypothesised.