Primary aim: The MOSAIC trial aims to assess the impact of a mandibular advancement device (MAD) on Apnea-Hypopnea Index (AHI) in Asian patients with Heart Failure with reduced Ejection Fraction (HFrEF) and obstructive sleep apnea (OSA). The investigators hypothesize that the AHI was 60% lower after 3-month treatment with MAD than with sham MAD.

Secondary aims: The investigators also aim to determine i. the interaction between ethnicity (Chinese, Malay, Indians) and the effects of MAD in lowering AHI; ii. the effect of MAD on cardiac remodeling (LVEDVI assessed by cardiac magnetic resonance imaging [CMR]); iii. the characteristic craniofacial skeletal anatomy (using coned beam computed tomography [CT]) associated with OSA in Asian patients with HFrEF; iv. the association between self-reported adherence to MAD and cardiac remodeling; v. the effects of MAD on biomarkers of HF (N-terminal pro-B-type natriuretic peptide [NT-proBNP],high sensitivity cardiac troponin T [hs cTnT], high-sensitivity C-reactive protein [hs-CRP], and ST2);

Rationale: OSA is associated with incident HF. The investigators will study Asian patients because a body of evidence suggests mechanisms for OSA differ between Asians and Caucasians. While obesity is the major contributing factor in Caucasians, craniofacial skeletal anatomy (short mandible, maxilla, and cranial base and a large mandibular volume) plays an important role in the development of OSA among Asians. Using cone beam CT, it has been shown that Asians have shorter mandibular, maxillary, and cranial base lengths and a greater mandibular volume compared with Caucasians. Using a MAD to adjust maxillary-mandibular juxta-positioning to maintain a patent airway may be an ethnic-specific approach to treat OSA in Asians.

The MOSAIC is a randomized, double-blind, placebo-controlled, cross-over trial to be conducted at the National University Health System. It is a collaboration between the National University Heart Centre Singapore, the Faculty of Dentistry, and the University Medicine Cluster. Apart from evaluating the effect of MAD on AHI- a conventional measure of OSA severity, the investigators will also explore the effect of MAD on cardiac remodeling. Adverse remodeling equates to progressive worsening of cardiac function, and slowing or reversing such remodeling is a goal of HF therapy. Ventricular measures including LV end diastolic volume index (LVEDVI) have proven to be meaningful surrogates reflecting the beneficial effect of therapeutic agents such as angiotensin-converting enzyme inhibitors and beta-adrenergic blocking agents on the remodeling process and on "hard" morbid and mortal end points.

Screening. Patients (n=200) will undergo polysomnography and coned beam CT if the following criteria are met: (i) Asians (Chinese, Malay or Indians) living in Singapore (ii) symptomatic HF (NYHA classes II-III) due to ischemic cardiomyopathy, (iii) LVEF <45% (as determined by echocardiography) (iv) clinical stability for ≥1 month, (v) optimal medical therapy including a diuretic, an angiotensin-converting enzyme inhibitor or an angiotensin receptor blocker, and a beta-blocker. The doses of these background medications should be stable for ≥1 month. Exclusion criteria:-(i) known OSA on treatment, (ii) acute coronary syndrome within the preceding 3 months, (iii) severe valvular heart disease, (iv) sustained ventricular arrhythmia, (v) stroke with residual neurological deficits, (vi) contraindication to CMR such as a pacemaker or cardioverter-defibrillator implant.

Overnight Polysomnography will be performed at the National University Health System (NUHS) Cardiosleep Research Laboratory following a standard laboratory protocol with a computerized recording system. All sleep studies will be scored according to the American Academy of Sleep Medicine guidelines, by an experienced sleep physician. Apnea is defined as a complete cessation in airflow lasting ≥10 s and hypopnea as a reduction in airflow or thoraco-abdominal movement by ≥50% for ≥10 s. Apneas will be classified as (i) obstructive if thoraco-abdominal movement was present during the apnea; (ii) central if thoraco-abdominal movement was absent; and (iii) mixed if thoraco-abdominal movements were both present and absent during the period of airflow cessation.

Coned beam CT. To determine the specific craniofacial anatomy that predicts OSA in Asians with HFrEF, the patients (regardless of the polysomnography results) will undergo a coned beam CT for craniofacial profiling at the Faculty of Dentistry. Image acquisition will be performed during quiet tidal breathing with subjects supine and with a neutral head position. Landmarks will be identified and measurements performed as previously reported. Tongue area will be measured by tracing the contours on the sagittal plane, whereas lateral wall thickness and parapharyngeal fat will be quantified on the axial plane. To determine tongue volume, the tongue limits will be identified and traced manually on each axial image obtained. Volumetric reconstructions will be performed to determine mandible, tongue, and airway volume.

Randomization. After the initial evaluation, the patients found to have OSA (estimated prevalence 50%, n=100) with an AHI of ≥ 15 events/hr will be randomized to (i) MAD followed by sham MAD (non-advanced device) or (ii) sham MAD followed by MAD, with a 2-week washout period in between. A dentist, who is otherwise not involved in HF care, will open a sealed envelope at the time of randomization and program the MAD device accordingly. Each mode will be maintained for 3 months, with comprehensive testing (polysomnography, echo, and biomarker assays) at the end of each 3-month period. The patients would also keep a daily record of time overnight wearing the device (adherence).

Mandibular Advancement Devices. MADs are designed to improve upper airway configuration and prevent collapse through alteration of jaw and tongue position. MADs have been shown in randomized trials to improve OSA and its symptoms. A significant reduction in blood pressure, improvement in endothelial reactivity and a fall in plasma NT-pro BNP have been reported with MAD treatment. Side effects from the treatment, such as pain from the teeth and jaws, are generally mild and transient. In a randomized trial, MAD and continuous positive airway pressure therapy were found to be equally effective in reducing blood pressure in patients with moderate-to-severe OSA. Given that MADs target the mandible by mechanically protruding the lower jaw and enlarging the upper airway, we hypothesize that they could be particularly effective in Asian patients.

Biomarkers. Serum levels of NT-proBNP, TnT, hs-CRP, and ST2 will be evaluated serially. NT-proBNP is released from cardiac myocytes in response to pressure or volume overload, tachyarrhythmia and hypoxaemia, with the predominant stimulus end-diastolic wall stress. NT-proBNP provides independent prognostic information on disease progression, hospital re-admission and mortality. TnT is typically used to diagnose myocardial infarction, but elevated troponin is commonly present in patients with HF. Troponin above the upper reference limit is associated with more severe HF. Inflammation is an important mechanism in the progression of HF. In a recent study of patients with HF, the prognostic value of hs-CRP was investigated and it was found that mortality rate increased from the lowest to highest hs-CRP quartiles. After adjustment for confounders, hs-CRP was found to be an independent predictor of 12-month mortality. ST2 is an interleukin (IL)-1 receptor family member that is secreted into the circulation and functions as a "decoy" receptor for IL-33, inhibiting IL-33/ST2 signaling. In patients with HF, plasma ST2 is strongly associated with measures of HF severity and poor outcome.

Inclusion Criteria:

• Asians (Chinese, Malay or Indians) living in Singapore
• Symptomatic HF (NYHA classes II-III) due to ischemic cardiomyopathy
• LVEF <45% (as determined by echocardiography)
• Clinical stability for ≥1 month
• Optimal medical therapy including a diuretic, an angiotensin-converting enzyme inhibitor or an angiotensin receptor blocker, and a beta-blocker. The doses of these background medications should be stable for ≥1 month.

Exclusion Criteria:

• Known OSA on treatment
• Acute coronary syndrome within the preceding 3 months
• Severe valvular heart disease
• Sustained ventricular arrhythmia
• Stroke with residual neurological deficits
• Contraindication to CMRsuch as a pacemaker or cardioverter-defibrillator implant.