Feasibility and Predictive Accuracy of an In-Home Computer Controlled Mandibular Positioner in Identifying Favorable Candidates for Oral Appliance Therapy

I. BACKGROUND AND SIGNIFICANCE Obstructive sleep apnea (OSA) is a common disease that carries significant risks for cardiovascular disease, mortality, and economic costs. Almost thirty years ago, initial population studies found the prevalence of OSA to be five to nine per cent of the adult population. Excess body weight is a risk factor for the development of OSA, and the recent rise in prevalence of obesity has led to revised estimates of OSA prevalence, now at seventeen per cent of the adult population. OSA is poorly recognized clinically; currently 85% of apneics remain undiagnosed and untreated. OSA derives fundamentally from structural abnormalities of the pharynx that cause pharyngeal narrowing or closure during sleep and produce recurrent apneas and hypopneas. During wakefulness compensatory neuromuscular reflexes protect the pharynx from collapse. These reflexes are lost during sleep, leaving the collapsible human pharynx susceptible to narrowing or closure. Nasal continuous positive airway pressure (CPAP), comprised of an air generator and nose mask, is the standard therapy for OSA. CPAP delivers positive pressure to the pharyngeal lumen, thereby dilating it and eliminating obstruction. While this therapy is highly efficacious, it is cumbersome and its effectiveness is compromised by a relatively low adherence rate. Adherence depends on the methods used to initiate therapy and on the severity of OSA, being higher in subjects with more severe hypoxemia and excessive daytime sleepiness. In current practice, CPAP adherence rate appears to approximate fifty per cent. The only currently available alternative to CPAP is oral appliance (OA) therapy. Oral appliances maintain patency of the airway during sleep by stabilizing and protruding the mandible and/or the tongue. The most commonly used type of OA is a custom-made mandibular repositioner (MR) which protrudes the mandible. Mandibular protrusion in paralyzed subjects dilates both the velopharynx and the oropharynx. In clinical practice, a specialist dentist fabricates a custom-fitted appliance covering upper and lower teeth. The appliance is then empirically adjusted to progressively protrude the mandible until a therapeutic end-point is reached. OA therapy is better accepted by the OSA subject than nasal CPAP, and self-reported adherence rates are high. Unfortunately, OA therapy is not uniformly effective in OSA. Reported effectiveness ranges from 50 to 70 percent, and our recent study found a 58% percent success rate. In short, two therapies are currently available for treating OSA, and each has its own shortcomings: 1) CPAP is efficacious but not uniformly tolerated by subjects; and 2) OAT is not highly efficacious but is better tolerated than CPAP. II. CURRENT STATUS OF MEDICAL PRACTICE Because of the uncertainties regarding use of OA therapy for treating OSA, current practices in North America focus primarily on the use of nasal CPAP. Virtually all subjects found to have OSA receive a trial of nasal CPAP. If they prove non-adherent with this therapy, they may then be offered OA therapy. The American Academy of Sleep Medicine recommends OA therapy as a CPAP alternative in subjects with OSA of mild to moderate severity. However, lacking a valid test for clinically selecting subjects with OSA who will have a favorable response to OA therapy, reimbursement for OA therapy is usually provided only for apneics who fail CPAP. The effectiveness of OA therapy could be improved by screening OSA subjects and prospectively identifying those suitable for this therapy. Unfortunately, at the present time, a method for selecting favorable candidates for OA therapy is lacking. Moreover, even if a subject were known to be a favorable candidate, the dentist currently has no way to determine the target therapeutic protrusive position. Studies of the passive pharynx indicate that the response of the pharynx to mandibular protrusion is 'dose dependent'; incremental mandibular protrusion produces corresponding pharyngeal enlargement. However, clinical experience shows that excessive mandibular protrusion is undesirable, producing side effects, such as, pain and tooth movement that lead to discontinuation of therapy. In some cases, over-protrusion can worsen OSA. Current practice is for the dentist to progressively protrude the mandible until a symptomatic response occurs, and then reassess the subject to determine if OSA has resolved. Thus, prospective identification of suitable candidates and of a target protrusive position would greatly facilitate treatment of OSA with OA. Owing largely to its association with excess body weight, obstructive sleep apnea (OSA) has become arguably the most prevalent chronic non-communicable disease in industrialized societies, assuming that obesity is not classified a disease. In addition to impairing quality of life, OSA conveys an increased risk of cardiovascular disease and vehicular accidents. Finally, the majority of apneics remain undiagnosed and untreated. Thus, OSA has all the hallmarks of a major public health problem, affecting all age groups and increasing in epidemic proportions. Unfortunately, current medical practices seem ill suited to dealing with the challenges posed by OSA. Perhaps the most unsettling aspect of present medical practices relates to the cumbersome, uninviting and expensive methods currently used to diagnose and treat the disease. Common clinical experience indicates that individuals often resist undergoing the standard diagnostic test, an overnight in-hospital polysomnogram, because of the extensive attachments to the body surface and because, if OSA is present, continuous positive airway pressure (CPAP) will be administered during the latter half of the night. In other words, as public awareness of sleep apnea has increased so also has the negative impression of current testing and treatment. Any effective solution to the public health challenges posed by sleep apnea will have to consider the need for less cumbersome and off-putting clinical methods. In other words, the current approach to OSA, i.e., overnight polysomnogram plus CPAP, simply will not deal with the problem. While simpler, less expensive home monitoring appears to be gaining acceptance of sleep physicians, CPAP continues to be virtually the only therapy offered to most patients, despite the low adherence rated documented in numerous studies. The dominant alternative, at the present time, is mandibular repositioner (MR) therapy, using a custom fitted dental appliance that protrudes the mandible during sleep. Despite overwhelming patient preference for OA over CPAP, only 5% of patients diagnosed with OSA receive this therapy. The likely reason for the physician bias in favor CPAP therapy is that it is benign and highly efficacious. OA therapy, while similarly benign, is efficacious in only 50-70% of unselected cases. Thus, the clinician prefers prescribing the more reliable therapy, having less concern about the low adherence rate. To summarize, current practice of diagnosis and treatment of OSA in North America emphasizes expensive in-hospital testing and initiation of a therapy which is difficult for many patients to accept and use regularly. III. NEW TECHNOLOGIES The proposed research evaluates new technology that could contribute to ameliorating the difficulties outlined above by allowing accurate selection of patients for OA therapy using a test that is performed in the home. This home technology is, in turn, derived from a device that Zephyr Sleep Technologies has invented, developed and now have brought to market, called remotely controlled mandibular positioner (RCMP). In order to have a more convenient and less expensive method for selecting favorable candidates for OA therapy, Zephyr Sleep Technologies has developed a companion technology for use in the home. Like the RCMP, the patient sleeps with the motorized mandibular positioner in place. However, the position of the mandibular protruder is controlled by a computer, rather than a sleep technician, making it a computer controlled mandibular positioner (CCMP). As in the RCMP study, respiratory airflow and oxygen saturation are recorded, and the CCMP accurately detects apneas and hypopneas in real time. However, bioelectric signals necessary for sleep staging are not recorded. The CCMP protrudes the mandible when apneas or hypopneas are detected. In addition, it conducts proactive testing, moving the mandible and evaluating the effect on airflow. Using both types of information, the controller decides whether or not to move the mandible, in which direction, and by how much. Finally, Zephyr Sleep Technologies has developed an analytical algorithm that examines the response of the moving apnea hypopnea index (AHI, the sum of apneas and hypopneas per unit time) to mandibular protrusion and makes a prediction of therapeutic success with OA. The CCMP uses the same motorized mandibular protruder used for the MATRx product. The motor is enclosed in a housing which is attached to the upper and lower appliances. The movement of the linear actuator exerts a force on the lower jaw to either protrude or retrude it. A series of dental trays for upper and lower teeth have been fabricated. This allows a close fitting of the tray to the subject's teeth so that a minimum amount of tray material occupies the inner surface of the teeth. A highly flexible, thin wire connects the motor in the mandibular protruder to a controller on a bedside table which, in turn, is connected to a laptop computer. As well, an input signal box on the bedside table is connected to two naris catheters that are positioned in the patient's nose and a commercially available finger oximeter. The input signal box is connected to the laptop and these two signals (oxygen saturation and airflow) are used by the computer algorithms to decide how to move the mandible. The MATRx controller receives commands from the computer that cause it to protrude or retrude the mandible in small, defined steps. The controller also has a wired output connection to the input box of computer, continuously providing the computer with information regarding the position of the mandible. The limits of excursion by the actuator are set by the technician in the software, according to the full retrusion and full protrusion values noted by the dentist. Thus, the computer can move the mandible only between predefined limits during the CCMP study. The signal box also receives input signals from a microphone and a position sensor mounted on the mandibular positioner as well as an input from a force transducer inside the positioner. IV. STUDY PROCEDURES Participants will receive a two night in-home cardio-pulmonary evaluation during sleep using the Snore Sat Recorder. The results will provide baseline data. Prior to the in-home CCMP study, the subject will be assessed by the dental co-investigator to ensure suitability for the trial and use of the device. They will then be fitted with an upper and a lower dental tray by the dental co-investigator. The dentist fills each tray with a silicone impression material and applies them to the upper and lower teeth. The relative position of upper and lower trays is documented by the numerical reading from a linear scale imprinted on the sliding strut of the lower tray. The subject moves the mandible to the rest position, the most retruded position, and the most protruded position. The scale is read by the dentist at each position. The data collected is provided to the research coordinator and will be input into the device to control the range of protrusion tested by the device. The clinical coordinator will visit the subject in the home during the day of the first night test and set up the computer equipment and mandibular positioner. To attach the subject's dental trays, the upper and lower trays are clipped to the actuator housing with the actuator piston fully retracted. The maximum allowable protrusive position as determined by the dentist will be entered in the computer to prevent excessive protrusion of the mandible. The computer controls movement of the linear actuator in protrusive or retrusive directions in increments of 0.2 mm through the keyboard on the computer. The position of the actuator is connected as an input to the computer. Just prior to starting the study, the subject will lie supine in bed and insert the dental trays into the mouth with the trays in the rest position. The clinical coordinator will also provide the subject with full instruction on how to use the device during the study. The subject will be instructed to properly insert the trays, connected to the mandibular positioner, in the mouth. The subject will also learn to apply the finger oximeter probe, and the naris cannula. The subject will be taught how to start the CCMP test, pause it and terminate it at the end of the study. At bedtime, the subject will apply the naris cannula and the probe, insert the tray in the mouth and initiate the CCMP study by the appropriate command on the laptop. The subject then falls asleep, and as apneas and/or hypopneas appear, the computer will protrude the mandible from rest position in 0.2 mm steps. Once all respiratory disturbances are eliminated (absence of apneas, hypopneas), proactive testing is performed in which the mandible is protruded or retruded 0.6 mm. If the latter causes a decrease in peak inspiratory airflow or if the former causes an increase in this variable, the mandible is protruded 0.6 mm, and so on. The following morning, the subject will awaken and remove the oral appliance, terminate the CCMP program and transmit the night's data to the server by a mouse click on a send button. The subject will complete a Device Satisfaction Questionnaire each morning. The data from the first night titration will be retrieved and automatically analyzed to judge the efficaciousness of the mandibular protrusion in eliminating pharyngeal obstruction and to identify the effective target protrusive position that will be required for successful therapy. If the predicted success criterion is not met, the subject is tentatively predicted to be a therapeutic failure, and a target of 90% of full protrusion is selected. After the analysis has been completed, the clinical coordinator will return to the subject's home and set the target position for the second night CCMP study. During the home visit prior to the second night study, the clinical coordinator will discuss with the subject his/her experience during the first night, and will carefully inquire regarding any discomfort or problems the subject may have experienced. The coordinator sets the CCMP for the Refinement and Verification program and enters the appropriate target distance. At bedtime the subject prepares for the study as on the first night. The CCMP algorithm moves the mandible to the target position and holds it there for the entire night, if the moving AHI (i.e. AHI for the time after the target is reached, initially measured over at least 2 hr) remains below 10 hr-1, ±0.6 mm. Otherwise the computer will protrude the mandible in 0.2 mm steps and repeat the test at a higher protrusive level. The following morning, the subject will awaken and remove the oral appliance, terminate the CCMP program and transmit the night's data to the Zephyr server by a mouse click on a send button. The subject will complete the Device Satisfaction Questionnaire. The computer calculates an AHI for the full night. The results from the first and second night titration studies will be provided to a sleep physician who will review the results and refine the prediction to arrive at a final prediction of therapeutic outcome and target protrusive setting. The target protrusive position determined from the analysis of the combined titration studies will then be sent to the dentist. For patients predicted to be a success, the minimum effective position will be provided. For patients predicted to be a failure, the scale reading at 70 percent of maximum protrusion will serve as a sham target value. The dentist will be blinded to whether the CCMP test results were "positive" or "negative". The subject will be referred to the dental co-investigator, and fitted with a custom oral appliance at no cost while blinded to the results of the CCMP studies. Follow-up portable cardio-pulmonary monitoring will be performed on two consecutive nights with the oral appliance at target protrusive distance. If the AHI is greater than 10 hr-1, the subject will be referred to the dental co-investigator for further mandibular protrusion. This adjustment will be performed in a minimum period of 3 weeks. The subject will be retested and the therapy judged successful according to the criteria outlined above. If the AHI is greater than 10hr-1, once maximum clinical protrusion of the mandible has been achieved, the OA therapy will be considered a failure.