Conference PaperPDF Available

Distributions of OSA Therapy Outcomes Based on a Trial Oral Appliance and/or Supine Avoidance



The study characterizes treatment outcome responder phenotypes using oral appliance therapy and/or supine avoidance.
2019 Clinical Research Award recipient of the American Academy of Dental Sleep Medicine, San Antonio, Tx
Distributions of OSA Therapy Outcomes Based on a Trial Oral Appliance and/or Supine Avoidance
Daniel Levendowski MBA1, Dominic Munafo MD2, Samuel J Clark MD2, Bretton Hevener RPSGT2,
Dorian Cruz Arista CDA2, Todd Morgan DMD3
1Advanced Brain Monitoring, Carlsbad, CA, 2Sleep Data, San Diego, CA, 3Sleep Alliance, San Diego, CA
Introduction: A previous cross-over study design reported that the combination of oral appliance (OA) and position
therapies was more effective than either therapy used individually.1 This study assessed the distributions of potential
outcomes from oral appliance (OA) therapy and/or supine avoidance using a two-night home sleep apnea test (HSAT)
and a Trial OA.
Methods: One-hundred and fifty-two adult patients with a diagnostic AHI >5, 54+12.8 years, and 40% female
were included in this retrospective analysis. All patients were fitted with a Trial OA (Apnea Guard®, Advanced
Brain Monitoring, Carlsbad, CA) to assess the likelihood of response to OA therapy. The Trial OA was advanced
to 70% of the distance from neutral bite to maximum protrusion, with females delivered either 5.5 or 6.5 mm
vertical dimension of occlusion (VDO) and males fitted with a 6.5 or 8 mm VDO depending on tongue scalloping.2
A two-night HSAT (NightOne, Philips Healthcare, Monroeville PA) was performed at Baseline on night one and with the
Trial OA on night two. Response to the Trial OA required an AHI reduction >30% for those with a pre-treatment AHI
<20, and >40% reduction in those with a pre-treatment AHI >20. Positional OSA (POSA) was characterized when the
overall AHI/non-supine AHI > 1.4 and >20 minutes in both supine and non-supine positions.3 Patients who slept <20
minutes non-supine were characterized as supine dominant. Two-tailed Chi-squared and t-test analyses were used to
identify group differences.
Results: Sixty-six percent of the patients were identified as responders with the Trial OA (100/152). Prior to OA
treatment, the prevalence of POSA (when it could be determined) was 70% of those who subsequently responded to
the Trial OA and 48% in non-OA responders (p<0.02). The odds of a patient with POSA being an OA responder was 2.5
(p<0.02, 95% CI range 1.2-5.4).
Patients were then stratified into one of four phenotypes based on response to OA and/or having POSA, resulting in
the following distributions by group:
OA Responder with: Proportion Non- OA Responder with: Proportion
1. No residual POSA
3. Positional OSA
2. Residual POSA
4. Non-positional OSA
Forty-one percent of the patients responded to the Trial OA with no residual POSA (Type 1)(62/152); 20% achieved an
efficacy equivalent to CPAP (i.e., AHI <5 events/h) and 32% of the cohort had an AHI < 10 events/h (Fig. 1).
Twenty-five percent of those who exhibited a response to the Trial OA either had residual POSA or were Supine
Dominant (Type 2). Based on the non-supine AHI, if the Trial OA was combined with supine avoidance, 16% of the
cohort could achieve an additional 60+21% reduction in overall AHI (Fig. 2.a.), with 14% having an non-supine AHI <5
events/h. OA responders who were supine dominant and would likely benefit from supine avoidance occurred in 9%
of the cases (Fig. 2.b.).
Fig. 1
Apnea Guard
2019 Clinical Research Award recipient of the American Academy of Dental Sleep Medicine, San Antonio, Tx
In the patients who did not respond to the Trial OA (Type 3), the Baseline HSAT identified POSA in 11% of the cases
(Fig. 3.a.). If supine avoidance was provided as an alternative to a Custom OA in this group, a mean reduction in
overall AHI of 66 + 20% could be achieved; eight patients would have an AHI < 5. Four OA non-responders were non-
POSA at baseline but converted to POSA after introduction of OA therapy (Fig. 3.b.). The combination of OA plus
supine avoidance in this sub-group could have reduced the overall AHI by 65 + 24%.
Only 20% of the patients would likely not be effectively treated with either an OA and/or supine avoidance (Type 4)
(Fig. 4). Patients who did not respond to the Trial OA and were non-POSA (Type 4) had significantly greater body fat as
compared to those who responded to OA with no residual POSA (Type 1)( p < 0.003, 28.7+5.1 vs. 32.9+8.0) (Fig. 5).
Conclusions: The combination of oral appliance and/or supine avoidance therapies could effectively treat 80% of the
patients (all but Type 4) with 39% potentially achieving an outcome equivalent to CPAP (i.e., AHI <5). Based on the
Trial OA, approximately 66% of patients would benefit from OA therapy (Types 1 & 2) and 39% could benefit from
supine avoidance (Types 2 and 3). Patients with POSA are more likely to respond to OA therapy, a finding consistent
with a previous report.4
Fig. 2.b.
Fig. 2.a.
Fig. 3.a.
Fig. 4
Fig. 3.b.
Fig. 5
1. A randomized, controlled trial of positional therapy versus oral appliance therapy for position-dependent sleep apnea. Benoist L, de Ruiter M, de Lange et al. Sleep
Med 2017; 34: 109-117
2. Initial evaluation of a titration appliance for temporary treatment of obstructive sleep apnea. Levendowski D, Morgan T, Westbrook P, J Sleep Disord Ther 2012; 1(1)
3. A systematic comparison of factors that could impact treatment recommendations for patients with Positional Obstructive Sleep Apnea (POSA). Levendowski D,
Oksenberg A, Vicini C et al. Sleep Med 2018; 50: 145-151
4. Patients with positional versus nonpositional obstructive sleep apnea: a retrospective study of risk factors associated with apnea-hypopnea severity. CHung J,
Enciso R, Levendowski D, et al. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2010; 110: 605-610.
ResearchGate has not been able to resolve any citations for this publication.
Full-text available
Abstract Objective/Background Systematically compare four criteria for Positional Obstructive Sleep Apnea (POSA) based on AASM 2007 and 2012 hypopnea scoring definitions. Patients/Methods 142 records acquired by in-home polysomnography (Sleep Profiler PSG2TM) were retrospectively analyzed using AHI based on the American Academy Sleep Medicine 2007 and 2012 criteria (AHI2007 and AHI2012). Positional obstructive sleep apnea (POSA) was characterized using four criteria: Amsterdam Positional OSA Classification (APOC), supine AHI twice the non-supine AHI (Cartwright), Cartwright plus the non-supine AHI <5 (Mador), and the overall AHI severity at east 1.4 times the non-supine severity (Overall/NS-AHI). Results Correlations between the Cartwright and Overall/NS-AHI criteria increased with the inclusion of a more relaxed definition of hypopneas (AHI2007=0.79 and AHI2012=0.86, P<0.00001). The prevalence of POSA based on the Cartwright and Overall/NS-AHI criteria was approximately 60% in those with at least mild OSA by AHI2007and AHI2012. A 16% reduction in POSA prevalence for AHI2012 vs. AHI2007 was attributed to the increased incident of mild OSA. For identificatio of those expected to have 25% or 35% reductions in SDB severity with positional therapy, Cartwright and Overall/NS-AHI exhibited the strongest sensitivity and Overall/NS-AHI and Mador the best specificity. Conclusions The four criteria used to identify POSA have similarities and differences. While there were similarities between the Cartwright and Overall/NS-AHI criteria in the detection of POSA prevalence across both scoring criteria, the Overall/NS-AHI provided the most consistent detection of those most likely to demonstrate important reductions in sleep disordered breathing severity if supine sleep is avoided.
Full-text available
Abstract Background: Custom oral appliances that adjustably advance the mandible provide superior outcomes when treating patients with moderate or severe sleep apnea. Custom appliances, however, are expensive, must be fitted by a dentist, and the likelihood of successful outcomes are difficult to predict. An inexpensive trial appliance, if proven efficacious, might be used to predict custom appliance outcomes or to provide temporary therapeutic benefit. Objective: The aim of this initial study was to assess the treatment efficacy of a novel titration oral appliance with that of an optimized custom appliance. Methods: Seventeen patients, treated with a custom oral appliance for at least one year, successfully completed a three-night home sleep test. The baseline obstructive sleep apnea severity was established on Night 1 with seven patients exhibiting severe, six moderate and four mild apnea/hypopnea indexes. Patients were randomly assigned to wear their custom appliance or the titration appliance on Nights 2 and 3. Results: Significant reductions in the mean overall and supine apnea indexes (p < 0.05), and the overall (p < 0.01) and supine (p < 0.05) apnea/hypopnea indexes were observed for both the titration and custom appliances. The proportion of patients who exhibited at least a 50% reduction in the overall apnea index and supine apnea/hypopnea were similar for the titration and custom appliance (~60%). The custom appliance reduced the overall apnea/hypopnea index by 50% in a greater proportion of the patients compared to the titration appliance (77% vs. 53%). The titration appliance significantly reduced the degree of hypoxic exposure across sleep disordered breathing events overall (p <0.05) and supine (p < 0.01). Patients found their custom appliance was more comfortable than the titration appliance, but preferred the titration appliance to no therapy. Conclusion: The titration appliance may be useful in assessing oral appliance treatment efficacy. When set to 70% of maximum protrusion, the titration appliance may provide immediate, temporary therapeutic benefit.
Full-text available
The aim of this study was to investigate the differences in and risk factors for positional and nonpositional obstructive sleep apnea (OSA). One hundred twenty-three nonpositional (supine apnea-hypopnea index [AHI] < 2 times the lateral AHI), 218 positional (supine AHI ≥ 2 times the lateral AHI), and 109 age-, gender-, and BMI-matched patients with positional OSA performed 2 nights of sleep study. Gender, age, BMI, and percentage of time in supine position, and percentage of time snoring louder than 40 dB were evaluated as risk factors. Both unmatched positional and matched positional patients had less severe overall AHI values, higher mean SpO(2), lower percentage time SpO(2) less than 90%, and lower percentage of time snoring when compared with the nonpositional group. Overall AHI scores were associated with increasing age and percentage of time snoring for positional and nonpositional groups. However, BMIs were associated with the overall AHI only in the nonpositional group. The influence of position on OSA severity may contribute to the choice and prognosis of treatment and may represent 2 distinct groups with probable anatomic differences.
Objective To compare the effectiveness of positional therapy (PT) with the sleep position trainer (SPT) to oral appliance therapy (OAT) in patients with mild-to-moderate positional obstructive sleep apnea (POSA). Methods Multicenter, prospective, randomized, controlled trial. Patients with mild-to-moderate POSA (apnea-hypopnea index (AHI) ≥5 ≤30/hour sleep) were randomized for PT or OAT. Polysomnography was repeated after 3 months. Efficacy, adherence, mean disease alleviation (MDA), quality of life, dropouts and adverse events were evaluated. Results A total of 177 patients were screened for the study; 99 underwent randomization and 81 completed the study. Intention-to-treat (ITT) analysis of median [IQR] AHI showed a reduction in the PT group from 13.0 [9.7-18.5] to 7.0 [3.8-12.8], p<0.001 and in the OAT group from 11.7 [9.0-16.2] to 9.1 [4.9-11.7], p<0.001. Mean adherence (≥4 hours/night, ≥5 days/week) was 89.3 ± 22.4% for SPT versus 81.3 ± 30.0% in OAT patients, p=0.208. Conclusions Oral appliance therapy and positional therapy were equally effective in reducing the median AHI in patients with mild-to-moderate POSA. The results of this study have important implications for future OSA treatment guidelines and daily clinical practice.