Tolerance and Benefits of Mandibular Advanced Device for Snoring and Sleep Apnea in Oropharyngeal Cancer.

The department of stomatology of the CHUM (Centre Hospitalier Université de Montreal) mandate is to ensure oral management of patients with oropharyngeal cancers (OPC), including papillomavirus (HPV), treated with radiotherapy (RTH), surgery and chemotherapy. At CHUM clinic, several patients who have had RTH treatments (with or without surgery or/and chemotherapy) complain of snoring and respiratory arrest during sleep. HPV-associated OPC, squamous cell carcinoma, has had a net increasing incidence for more than 20 years in North America (4.6/100,000) and male mortality has increased by 3.1% per year since 2001. In Quebec, the incidence is almost 26 cases per million. Worldwide, it is ranked to be the 13 most common cancer in male aged 25-39 years; the male/female ratio is 2.5 (female have more HPV cervix related cancer) - Global oral health status report: towards universal health coverage for oral health by 2030- (Geneva: World Health Organization; 2022).

Snoring occurs in nearly 40% of the general population, increases with age and can be one of the telltale signs of sleep apnea (brief and repetitive breathing stops at risk of mortality if left untreated). Snoring can be associated with poor sleep quality and produces a loud disturbing noise for the sleeping partner. Obstructive sleep apnea (OSA) is reported by 2-4% of the general population and by almost 30% of the elderly mainly related to obesity and heart disorders. The risk factors associated with these conditions are lost of airway muscle tone during sleep, retrognathia, deep/narrow palate, obesity and age. Fatigue and drowsiness are markers of OSA. Following RTH, there may be oropharyngeal tissue fibrosis that may contribute to snoring and sleep apnea.

Sleep disordered breathing are known in individuals with OPC. Our first results of the ongoing survey seem to indicate that severe snoring, causing drowsiness or fatigue, would be present in 47% of patients with OPC. A recently published literature review, from our team, also found a snoring self-reports prevalence from 33% to 82%. A polygraphic sleep recording study found that 65% of individuals with head and neck cancer (including OPC) have snoring, a value that increases to 92% when OSA is concomitant. On the other hand, OSA seem to be present in over 70% of OPC patients, before and after treatment. However, causality remains to be demonstrated; these sleep disorders are often pre-existing and can worsen as a result of RTH.

In adults, snoring is mainly controlled by devices correcting the dorsal position of sleep as well as with mandibular advanced device (MAD). Regarding apnea, positive pressure vessels (PPC or CPAP) are the most effective. The MAD is indicated for mild to moderate OSA or for low CPAP tolerance. To date, following RTH, there are no studies conducted under strict conditions. Some 'suggestions' in favour of standard treatments, PPC or MAD, were made. In addition, no studies have estimated the nasal discomfort and dryness associated with these treatments in the OPC population. Intuitively, nylon MAD, thin and with little contact on the mucous membranes, could be indicated in the presence of OPC. To date, only one study, based on 2 cases, suggests the use of MAD in the presence of OPC. There is therefore a gap of evidences in the literature as to whether patients treated for OPC presenting snoring and OSA would be able to tolerate and benefit from MAD.

Objectives: Conduct a pilot study, with a mixed quantitative and qualitative design, to reduce snoring or sleep apnea using an MAD, and improve the quality of life and sleep.

Hypotheses: Individuals who have received RTH as part of a treatment for OPC and who have snoring or OSA, can 1) tolerate and 2) benefit from treatment with MAD.

Experimental method and approach.

• Study design: Pilot trial, mixed and sequential design, randomized with active MAD vs. non-active MAD (control) and single-blind condition analyses for the participant. The D-SAD nylon MAD is marketed by Panthera Dental (Quebec) and provided free of charge as part of this project. The study includes a total of 8 visits to CHUM: 5 for clinical appointments and 3 nights of home sleep data collection. The 3 nights of sleep recordings will be spaced 4-5 weeks apart for adaptation to MAD. Night 1 will be for basal level, and nights 2 and 3 for recordings with active MAD or control. This will be done under the supervision of sleep lab director of CHUM, collaborator for over 25 years and residents. -- - Population: Adult patients (+18 years, female or male; ratio of 1/4 toward men; female is a rare phenotype of OSA pre-menopause) who received RTH in the Department of Radiation Oncology, with snoring-respiratory complaints during sleep, and follow-up in Stomatology at CHUM for OPC. Exclusions: Pregnant or breastfeeding woman; edentation preventing retention of the MAD; taking opioid-type medications; history of major depression, uncontrolled hypertension, stroke; under treatment with CPAP for sleep apnea; regular and intense orofacial pain; under pharmacological treatment for insomnia; excessive alcohol or drug use (e.g., cannabis, opioid). Recruitment: With the help of CHUM radiation oncology and dental staff (dentist and dental hygienist, resident). An informed consent will be completed.
• Sample size: It is expected that 24 participants (risk of 2-4 dropouts based on our experience) will be recuited. A 30% reduction in snoring and apnea/h is expected; this based on publications and an estimate of sample size. Indeed, a statistician, estimated that for a 30% decrease in snoring/night time (base at 8.9%/night, SD 3.4%) 15 participants are required for an effect size (ES) of 0.79; for apnea-hypopnea index/h (base at 11.5, SD 5)19 participants are required for a ES of 0.7. For subjective reports of snoring intensity (9 participants are required; ES of 1.11), however, an article reported significant results with 18 subjects. [20] A final sample of 20-22 subjects is targeted. These estimates are comparable to those in our previous orthotic studies of bruxism and OSA populations done by our group members.
• Types of data and collection: After completing the consent, the participant will be asked to complete demographic, sleep and basic medical information. It will be built from validated questionnaires estimating quality of life (SF-12), sleep (PSQI), sleepiness (Epworth), risk of apnea (Stop Bang), fatigue (Chalder), and orofacial pain. The data will be entered via WEB (REDCap) and saved on the CHUM/CITADELLE server. Following the 3 home sleep recordings, paticipants will assess on questionnaire their perception of change in sleep quality based on the validated scales used in our previous studies on apnea and MAD. In addition, the comfort of MAD (tolerance) and the perception of the spouse will also be estimated as in our previous studies. The quantitative home sleep data, namely duration and snoring indexes, apnea-hypopnea, instability and sleep duration, will be calculated, as already published, by the measurements obtained (breathing, oxygen saturation, sleep position, pulse, snoring) with the ambulatory recording instruments (Alice TM, Philips, USA) from CHUM sleep laboratory and analyzed by qualified technical staff. Qualitative data on MAD treatment expectations and oral health experiences related to POC (chewing, pain) will be estimated in a subgroup of participants (n=15), before and after MAD treatment. This will be done by a semi-structured interview.
• Statistics and design for analysis: Like sleep analyses, the results will be analyzed blindly with the following tests: 1) parametric tests such as: t-test for inter-group comparisons, linear mixed models and linear mixed models for repeated measurements, followed by contrasts for post-hoc comparisons where appropriate using the Tukey method (SPSS software); 2) nonparametric tests such as chi-square, Mann-Whitney and non-parametric ANOVA for repeated measurements followed by contrasts for post-hoc comparisons where appropriate using the Tukey method (for category data and/or when the data distribution does not correspond to a normal distribution); 3) correlation tests, such as Pearson's or Spearman's correlation coefficient, if applicable. A biostatistician will oversee the analyses. Qualitative data will be analyzed to extract recurring themes from transcripts of verbatims using software, N*VIVO. The collection of themes will be done until saturation. The software allows data coding and categorization for final analysis .