A79-04 Differences Between Mean Disease Alleviation of Cpap and Mad Treatment Are Associated With Non-Invasively Determined Site of Collapse Probability

Rationale Continuous positive airway pressure (CPAP) and mandibular advancement devices (MAD) are commonly used treatments for obstructive sleep apnea (OSA). Mean Disease Alleviation (MDA) is used to quantify overall clinical effectiveness of OSA treatments by multiplying adherence and efficacy. While studies have shown similar overall effectiveness of CPAP and MAD, little is understood about what drives differences in effectiveness of both treatments within patients. The aim of this study was to assess associations between non-invasive site-of-collapse probability, derived from a recently developed and validated polysomnography-based method, and differences in MDA between CPAP and MAD. Methods 60 patients (median[Q1;Q3] age:52.0[45.0;60.0]years, body mass index (BMI):28.1[26.0;30.4] kg/m², 85% male, AHI:24.9[18.6;31.8] events/h) of the FLOSAT trial, in which participants underwent 3 months of MAD followed by 3 months of CPAP therapy, were included in this secondary analysis. Non-invasive site of collapse probabilities were calculated at baseline using the validated algorithm. MDA was calculated as objective adherence (% total sleep time) x efficacy (% AHI reduction) at 3 months. Logistic regression assessed associations between site of collapse probability and the odds of having higher MDA with MAD versus CPAP. Analyses were adjusted for baseline AHI, age, sex and BMI. Secondary analyses were performed to separately assess associations with adherence and efficacy. Results 26/60 patients showed higher MDA with MAD vs CPAP. 40 patients showed higher MAD adherence; 6 patients showed higher MAD efficacy. Multivariable logistic regression analysis showed increased odds for higher MDA with MAD versus CPAP in patients with higher probability of complete concentric collapse at the level of the palate (CCCp, p = 0.034 OR[95%CI]: 2.0[1.1;3.9] per 1 SD (0.36)). However, secondary analyses showed an opposite effect for efficacy: patients with lower probability of CCCp showed a 3-fold higher odds for greater efficacy with MAD (2.8[1.2;6.2], p = 0.014). Furthermore, a younger age was associated with increased odds for higher efficacy with MAD (p = 0.034, 2.0[1.1;4.00] per 1SD (11.1)). There was a tendency for better adherence in CCCp with MAD versus CPAP, but this did not reach statistical significance (p = 0.10, OR[95%CI]: 1.8[0.89;3.45]) Conclusions Higher probability of CCCp was associated with greater overall effectiveness (MDA) of MAD compared to CPAP. However, CCCp probability was inversely associated with efficacy, suggesting higher MDA in patients with CCCp may have been driven by better adherence, despite not reaching significance. These findings underscore the complexity of predicting treatment response, as physiological traits may influence both efficacy and adherence, ultimately determining overall clinical effectiveness.