Mandibular advancement device treatment and reverse left ventricular hypertrophic remodeling in patients with obstructive sleep apnea

Abstract

Study Objectives:

Obstructive sleep apnea (OSA) is associated with cardiovascular comorbidities such as left ventricular (LV) hypertrophy. Whether OSA is an independent etiological factor for this hypertrophic remodeling is yet unknown. Continuous positive airway pressure partially reverses this hypertrophy, but data regarding the effect of mandibular advancement devices on LV remodeling are scarce. The aim of this prospective trial is to evaluate the effect of mandibular advancement device therapy on LV geometry and function in patients with OSA.

Methods:

At baseline and 6-month follow-up, participants underwent a home sleep apnea test, 24-hour ambulatory blood pressure monitoring and a 2-dimensional Doppler and tissue Doppler echocardiography.

Results:

Sixty-three patients (age: 49 ± 11 years; body mass index: 27.0 ± 3.4 kg/m²; baseline apnea-hypopnea index home sleep apnea test: 11.7 [8.2; 24.9] events/h) completed the 6-month follow-up visit. Overall, blood pressure values and parameters of LV function were within normal ranges at baseline and did not change under mandibular advancement device therapy. In contrast, the interventricular septum thickness was at the upper limits of normal at baseline and showed a significant decrease at 6-month follow-up (11.1 ± 2.1 mm vs 10.6 ± 2.0 mm, P = .03). This significant improvement is only found in responders but not in nonresponders. There was no correlation between the decrease of interventricular septum thickness and the change in blood pressure.

Conclusions:

In mildly obese, normotensive patients with OSA we observed significant reverse hypertrophic remodeling after 6 months of successful mandibular advancement device therapy, with maintained normotensive systemic blood pressure. This suggests that OSA is an independent factor in the pathophysiology of LV hypertrophy in these patients.

Clinical Trial Registration:

Registry: ClinicalTrials.gov; Name: Evaluation of the Cardiovascular Effects of the MAS in the Treatment of Obstructive Sleep Apnea; URL: https://clinicaltrials.gov/ct2/show/NCT02320877; Identifier: NCT02320877.

Citation:

Dieltjens M, Vanderveken OM, Shivalkar B, et al. Mandibular advancement device treatment and reverse left ventricular hypertrophic remodeling in patients with obstructive sleep apnea. J Clin Sleep Med. 2022;18(3):903–909.

BRIEF SUMMARY

Current Knowledge/Study Rationale: Obstructive sleep apnea is associated with cardiovascular comorbidities such as left ventricular hypertrophy. Data regarding the effect of mandibular advancement devices on this left ventricular remodeling are scarce. The aim of this prospective trial was to evaluate the effect of mandibular advancement device therapy on left ventricular geometry and function in patients with obstructive sleep apnea.

Study Impact: Significant reverse hypertrophic remodeling was observed after 6 months of successful mandibular advancement device therapy, with maintained normotensive systemic blood pressure. This suggests that obstructive sleep apnea is an independent factor in the pathophysiology of left ventricular hypertrophy in these patients.

INTRODUCTION

Obstructive sleep apnea (OSA) is a prevalent public health issue, affecting up to 17% of adult women and 34% of adult men.¹ It is characterized by repetitive episodes of partial (hypopneas) or complete (apneas) upper airway obstruction during sleep, leading to nocturnal hypoxemia and sleep fragmentation.² The consequences associated with untreated OSA include excessive daytime sleepiness, impaired cognitive performance, and reduced quality of life.³ Additionally, epidemiological studies provide objective evidence that OSA is an independent risk factor for hypertension, endothelial dysfunction, and mortality.^4–7 OSA is found to be associated with an increased incidence of left ventricular hypertrophy (LVH) with systolic and diastolic dysfunction.^8–11 Whether LVH results from OSA, or from associated comorbidities such as hypertension and/or obesity, remains controversial.¹²

Due to the high prevalence, as well as the related morbidity and mortality, adequate treatment of OSA is of utmost importance. The standard treatment for patients with moderate-to-severe OSA is continuous positive airway pressure (CPAP), applying pressurized air throughout the respiratory cycle to keep the upper airway patent.¹³ CPAP is highly efficacious in reducing OSA severity, but overall clinical effectiveness of CPAP is often limited by low patient acceptance and suboptimal compliance.¹⁴ It has been shown, however, that CPAP is able to remediate hypertension and reverse the structural and functional cardiac consequences of OSA.^7,10

Another noninvasive device therapy for sleep-disordered breathing is the use of a mandibular advancement device (MAD) during sleep, in order to reduce upper airway collapse by protruding the mandible. Many designs exist, but a custom-made MAD that includes a titratable mechanism allowing for gradual protrusion is currently the recommended MAD to be prescribed.^15,16 Overall, the therapeutic effectiveness of MAD therapy is characterized by a suboptimal efficacy in terms of reduction of the apnea-hypopnea index (AHI) but favored by an optimal compliance.^17,18 In a systematic review and recent meta-analysis it was shown that MAD therapy had a minor positive effect on both systolic and diastolic blood pressure (BP).^19,20

To the best of our knowledge, only one study has investigated the effect of MAD therapy on echocardiographic parameters of left ventricular structure and function in patients with OSA.²¹ This study failed to show any improvements in cardiac parameters; this particular study was limited by lower power with a limited number of patients (n = 12) with a short follow-up period (2 to 3 months).

Therefore, the objective of this study was to evaluate the effect of MAD therapy on left ventricular geometry and function in patients with OSA after 6 months of follow-up.

METHODS

Study population

Patients were prospectively recruited by the OSA multidisciplinary team including a dental sleep professional; an ear, nose, and throat, head and neck surgeon; and a board-certified sleep specialist. All patients were diagnosed, on a recent type I full-night attended polysomnography, showing an AHI ≥ 15 and ≤ 50 events/h of sleep. In addition, patients had to be excluded if they were not judged suitable for MAD therapy (preexisting active temporomandibular joint dysfunction, fully edentulous, or a dental status or periodontal health that precluded them from wearing an oral appliance). Furthermore, patients with uncontrolled hypertension, known cardiomyopathy, significant coronary artery disease, valvular heart disease, or congenital heart disease were not included. Based on the paper of Shivalkar et al,¹⁰ we conservatively anticipated an effect size of 0.4. With this effect size, 55 patients are needed to achieve 80% power at a significance level of .05. Assuming a dropout of 15%, at least 63 patients had to be included.

The study was approved by the institutional review board of the Antwerp University Hospital and University of Antwerp and registered at clinicaltrials.gov (Identifier: NCT02320877). Informed consent was obtained from all participants.

Home sleep apnea testing

A 2-night home sleep apnea testing (HSAT) was performed at baseline as well as 6 months after MAD fitting. The validated HSAT device (MediByte, Braebon Medical Corporation, Kanata, Ontario, Canada) used in this study analyzed both respiration and body position.²² Two belts measuring respiratory effort were attached around the chest and the abdomen, with the device itself positioned at the sternum. Nasal airflow was measured using an external thermistor and nasal pressure by means of a nasal pressure cannula. Oxygen saturation and heart rate were monitored using a finger pulse oximeter. A microphone was taped to the patient’s forehead to register snoring and body position was observed with a piezoelectric sensor. Data were read out using software allowing data analysis to be conducted. Respiratory events were scored following the American Academy of Sleep Medicine 2014 criteria with hypopnea defined as a reduction in airflow ≥ 30% from baseline associated with a drop in oxygen saturation level (SpO₂) ≥ 3% during at least 10 seconds.²² Treatment response was defined as a ≥ 50% reduction in AHI or a treatment AHI < 5 events/h.

MAD therapy

All patients were treated with a custom-made, titratable MAD (SomnoDent Flex, SomnoMed Ltd, Sydney, New South Wales, Australia) that has been validated extensively in randomized controlled trials.^23,24

The MAD was fitted in the maximum comfortable protrusion and titration was guided by self-reported relief of cardinal symptoms like snoring and daytime sleepiness.

Adherence

Adherence to MAD therapy was assessed subjectively by self-reported use as well as objectively by monitoring through an embedded microsensor thermometer (TheraMon, IFT Handels- und Entwicklungsgesellschaft GmbH, Handelsagentur Gschladt, Hargelsberg, Austria).^18,25 Temperature was recorded by the microsensor at a sampling rate of 1 measurement per 15 minutes, allowing data acquisition for a consecutive 100-day period. Objective adherence was based on the assumption that the MAD was worn when a temperature > 35 °C was recorded. Patients were considered adherent when the mean MAD use was ≥ 4 h/night, and “regular users” were defined as patients who used the MAD ≥ 4 h/night on ≥ 70% of all nights.²⁶

Cardiovascular parameters

All patients underwent an ambulatory BP monitoring over a 24-hour time period and comprehensive 2-dimensional echocardiography at baseline and at 6-month follow-up with a Vivid E95 ultrasound machine (GE Healthcare, Chicago, Illinois). Measurements of interventricular septum (IVS) thickness, left ventricular dimensions, and posterior wall thickness were made by 2-dimensional guided M-mode in the parasternal long axis view (Figure 1). Left ventricular ejection fraction was determined by biplane Simpson’s method. Mitral valve inflow velocities were measured by pulsed wave Doppler at the mitral valve leaflet tips (early rapid filling wave, E; late filling wave, A; deceleration time, DT). Pulsed wave tissue Doppler systolic (s′) and diastolic (early, e′) velocities (unit of centimeters per second) were obtained at the medial and lateral mitral valve annulus. Gain settings were adjusted carefully, and the direction of motion was aligned with the scan line direction. Signals were obtained from three end-expiratory cycles, and averages were made for the systolic and diastolic velocities. The left atrial volume was obtained by the area-length method and indexed for body surface area. All measurements were made according to current international guidelines.^27,28

Figure 1. Echocardiographic parasternal long-axis view with measurement of the IVS thickness in a 49-year-old male patient with severe obstructive sleep apnea (AHI = 45.3 events/h).

The patient had a normal body weight (body mass index 25 kg/m²) and was normotensive on 24-hour blood pressure measurement. At baseline (A), IVS thickness was at the upper limits of normal (11.5 mm). Six months after regular mandibular advancement device treatment (B), there was a clear reduction of obstructive sleep apnea severity (AHI = 3.7 events/h) and a 0.7-mm decrease of IVS thickness. AHI = apnea-hypopnea index, IVS = interventricular septum.

Statistical analysis

Data were statistically analyzed using SPSS (SPSS V.26, SPSS Inc., Chicago, Illinois). Measurements of continuous variables at baseline were compared with measurements after 6 months of MAD therapy using a paired t test if the data were normally distributed and the nonparametric Wilcoxon signed rank test if not normally distributed.

RESULTS

Study population

Seventy-seven patients with moderate to severe OSA (AHI ≥ 15 events/h) on the baseline polysomnography were included in this study protocol and underwent a baseline HSAT, a 24-hour BP monitoring, and echocardiography. At 6-month follow-up, 13 patients were lost to follow-up mainly due to the time-consuming protocol and 1 patient was excluded from analysis due to aortic valve stenosis, resulting in a final dataset of 63 patients. Table 1 presents the baseline characteristics of these patients, showing a study population that is predominantly male, slightly middle-aged and slightly overweight, with a moderate-to-severe OSA as diagnosed based on a type I polysomnography. Seventeen patients were using antihypertensive drugs at baseline: 6 patients were using angiotensin-converting enzyme inhibitors, while 3 patients were on betablockers, 1 patient on calcium antagonists, and 1 patient on combination of betablockers and diuretics. There was no change in these cardiovascular medications during the follow-up period of 6 months.

Table 1.

Baseline characteristics (n = 63).

Parameter	Values
Age, y	49 ± 11
Sex (% male)	71%
Body mass index, kg/m2	27.0 ± 3.4
Smoking (% smoker)	20%
Diabetes mellitus (% of patients)	8%
Hypertension (% of patients)	17%
Hypercholesterolemia (% of patients)	14%
Systolic blood pressure, average 24 hours, mmHg	125.8 ± 14.1
Diastolic blood pressure, average 24 hours, mmHg	77.3 ± 7.2
Systolic blood pressure, average day, mmHg	132.9 ± 14.0
Diastolic blood pressure, average day, mmHg	83.0 ± 7.4
Systolic blood pressure, average night, mmHg	113.0 ± 12.9
Diastolic blood pressure, average night, mmHg	67.4 ± 6.6
Use of antihypertensive medications (number of patients)	11 (17%)
Beta-blockers (number of patients)	3
Angiotensin-converting enzyme inhibitors (number of patients)	6
Calcium antagonists (number of patients)	1
Beta-blockers + diuretics (number of patients)	1
Apnea-hypopnea index on PSG, AHI-PSG (events/h)	21.3 [17.3, 28.3]
Apnea-hypopnea index on HSAT, AHI-HSAT (events/h)	11.7 [8.2, 24.9]
Oxygen desaturation index (events/h)	6.5 [2.7, 14.6]
Mean saturation, %	95.4 [93.9, 96.0]
Minimal saturation, %	87.5 [84.0, 90.0]

Values presented as median [Quartile 1; Quartile 3] or mean ± standard deviation unless otherwise specified. AHI = apnea-hypopnea index, HSAT = home sleep apnea testing, PSG = polysomnography.

Effectiveness of oral appliance therapy

Sixty-one patients completed the baseline HSAT as well as the HSAT with MAD at 6-month follow-up successfully. The HSAT showed an overall baseline AHI of 11.7 [8.2; 24.9] events/h (median [quartile 1; quartile 3]) that was significantly decreased with MAD to 5.6 [2.9; 13.1] events/h (P < .05). Overall, at 6-month follow-up, the median MAD efficacy was 56 [7, 72]% with a treatment response rate (Δ AHI ≥ 50%) of 69%.

The objective MAD adherence was available in 54 patients at 1-month and in 50 patients at 3-month follow-up. At 1-month follow-up, the objective MAD use was 6.7 [5.2; 7.9] hours per night with the MAD being used on average 91.0 [75.3; 97.2)]% of the nights. Forty-seven patients (72%) could be considered compliant with mean MAD use of ≥ 4 h/night, and 41 patients (63%) met the “regular user” criteria with MAD use of ≥ 4 h/night on ≥ 70% of all nights. This relatively high adherence was sustained at 3-month follow-up with an objective MAD use of 7.1 [4.9; 8.1] hours per night, a compliant user’s rate of 66%, and 57% of patients fulfilling the regular user criteria.

Cardiovascular outcomes of oral appliance therapy

Overall, systolic and diastolic BP values and parameters of left ventricular systolic and diastolic function were within normal ranges at baseline and did not change under MAD therapy (Table 2 and Table S1 in the supplemental material).

Table 2.

Echocardiographic parameters at baseline and 6-month follow-up (n = 61).

	Baseline	6 Months	P
Systolic blood pressure, mmHg	125 ± 13	125 ± 14	.66
Diastolic blood pressure, mmHg	77 ± 7	78 ± 7	.32
Mean arterial pressure, mmHg	90 ± 8	92 ± 7	.13
Pulse pressure, mmHg	48 ± 10	47 ± 10	.28
Heart rate (bpm)	72 ± 6	72 ± 7	.71
LV geometry and LV mass
IVS thickness, mm	11.1 ± 2.1	10.6 ± 2.0	.03
LVIDd, mm	47.9 ± 6.0	47.8 ± 5.6	.80
LVIDs, mm	31.3 ± 5.2	30.8 ± 4.0	.31
PW thickness, mm	8.8 ± 1.7	8.6 ± 2.7	.09
LV mass, g	173.9 ± 51.6	160.5 ± 43.3	.01
LV systolic function
LVEF (Simpson’s method), %	60.3 ± 6.1	60.8 ± 6.4	.51
TDI med s′, cm/s	7.4 ± 1.5	7.6 ± 1.7	.37
TDI lat s′, cm/s	9.1 ± 2.1	9.0 ± 2.3	.52
LV diastolic function
E/A	1.0 ± 0.3	1.0 ± 0.3	.52
E/e′	9.0 ± 2.2	9.3 ± 2.3	.12
LA vol index, mL/m2	29.4 ± 7.6	29.9 ± 8.3	.66

Values are presented as mean ± standard deviation. A = velocity A wave, bpm = beats per minute, E = velocity E wave, IVS = interventricular septum thickness, LA vol index = left atrium indexed volume, LV = left ventricle, LVEF = left ventricular ejection fraction, LVIDd = left ventricular internal diameter end diastole, LVIDs = left ventricular internal diameter end systole, PW = posterior wall thickness, TDI lat = tissue doppler imaging lateral wall, TDI med = tissue doppler imaging medial wall. e′ = tissue Doppler derived early diastolic velocity, s′ = tissue Doppler derived systolic velocity.

In contrast, the LV mass and IVS thickness were at the upper limits of normal at baseline and showed a significant decrease at 6-month follow-up (173.9 ± 51.6 g vs 160.5 ± 43.3 g, P = .01 for LV mass and 11.1 ± 2.1 mm vs 10.6 ± 2.0 mm, P = .03 for IVS) (n = 61). This significant improvement in IVS thickness is only found in the responders (n = 40) (11.1 ± 2.0 mm vs 10.5 ± 1.8 mm, P = .022 but not in the nonresponders (n = 19) (10.8 ± 2.2 mm vs 10.8 ± 2.3 mm, P = .872) (Figure 2).

Figure 2. Evolution of the interventricular septum thickness after 6 months of MAD therapy in responders vs nonresponders.

The IVS thickness improved significantly after 6 months of MAD therapy in the responders (n = 40) (11.1 ± 2.0 mm vs 10.5 ± 1.8 mm, P = .022) but not in the nonresponders (n = 19) (10.8 ± 2.2 mm vs 10.8 ± 2.3 mm, P = .872). IVS = interventricular septum, MAD = mandibular advancement device.

There was no correlation between the decrease of IVS thickness and the change in BP during 6-month follow-up. All other parameters of LV geometry and systolic and diastolic function were within normal limits at baseline and did not change significantly after 6 months of treatment.

DISCUSSION

The present study is the largest to date to comprehensively assess the impact of MAD therapy on structural and functional parameters of the left ventricle. In our mildly obese, normotensive study cohort with OSA we observed hypertrophic remodeling of the IVS at baseline, which improved significantly 6 months after successful MAD therapy.

LVH and cardiovascular outcome

It is well known that OSA is associated with LVH, which is an established risk factor for cardiovascular mortality.²⁹ In conditions which impose a chronically increased workload on the heart—such as hypertension, obesity, and probably OSA syndrome—the heart will show adaptive remodeling in the form of LVH.³⁰ In hypertension, IVS thickness and LV mass are often used as surrogate risk markers, as they are strongly related with the risk of adverse cardiac outcome³¹ and generally precede systolic and diastolic dysfunction. LVH is not only an established marker of organ damage, it is also a reversible risk factor, and cardiovascular outcomes improve if antihypertensive treatment results in regression of LVH. A meta-analysis by Verdecchia et al showed that reverse remodeling resulted in a 59% risk reduction regarding cardiovascular events, as compared with no improvement or worsening of LVH.³² In OSA syndrome, it is suggested that hypertension and obesity play a central role, but the precise underlying pathophysiologic mechanisms are not fully understood.¹²

Impact of MAD therapy on LVH

Previous studies in patients with severe OSA and significant LVH found a significant correlation between IVS thickness and the AHI, which was partially reversible after successful CPAP treatment.^10,33 In a recent animal study, Liu et al found that MAD therapy prevented LVH in an OSA rabbit model.³⁴ Only one small study evaluated the effect of MAD therapy on left ventricular structure and function in patients with OSA.^21,35 This study could not demonstrate a significant change in IVS thickness after MAD, nor after CPAP, but only included a limited number of patients (data on IVS thickness in 8 patients in the MAD group). In contrast, in our larger cohort we noted reverse remodeling of the IVS after 6 months of successful MAD treatment, regardless of BP. This latter finding suggests that OSA is an independent factor in the pathophysiology of LVH in these patients, even in patients with moderate OSA. Interestingly, the finding that IVS thickness improved under MAD therapy was only true in the responders but not in the nonresponders, emphasizing the need to adequately treat patients who are diagnosed with OSA. Furthermore, in the responder group a high percentage of patients were compliant users (88%) or regular users (79%). This finding may corroborate the idea that the favorable health outcomes of MAD therapy can be explained by a high adherence rate that counterbalances the suboptimal efficacy in terms of AHI reduction of the therapy.¹⁷ In previous literature, it was indeed suggested that MAD therapy and CPAP are equally effective in reducing cardiovascular death, likely explained by the greater efficacy of CPAP’s being offset by inferior compliance relative to MAD.^17,35–38 Therefore, even in patients with severe OSA who are intolerant to CPAP, MAD can be considered as an alternative treatment option.

Impact of MAD therapy on left ventricular function

Besides altered left ventricular geometry, OSA may compromise left ventricular systolic and diastolic function as well, mainly due to a state of intermittent hypoxia, increased sympathetic drive, increased oxygen demand, and mechanical and hemodynamic perturbations during apneic events.¹⁰ Furthermore, a clear relationship was found between the severity of the apneic episodes and functional changes of the left ventricle.¹⁰ In patients with severe OSA, consistent improvement in contractile cardiac parameters was found 6 months after successful CPAP treatment.¹⁰

In our study, most patients had moderate OSA and had normal left ventricular systolic and diastolic function at baseline, limiting potential improvement. Therefore, further research is required to establish that MAD therapy might have a beneficial impact on LVH in patients with more advanced OSA, with repercussions on systolic or diastolic function.

Limitations

Baseline 24-hour BP values and systolic and diastolic function were within normal limits at baseline, so we could not expect a great improvement in these parameters under MAD therapy. However, our results are in line with current literature, suggesting no improvement in BP values after short-term use of MAD therapy.³⁹ The patients referred for MAD therapy in the current trial are mainly male, less obese, and have mild to moderate OSA as compared to patients treated with CPAP in previous trials, possibly resulting in a lower effect of MAD therapy on cardiovascular outcome parameters.¹⁰ Furthermore, all patients included in this study were Caucasian of origin, affecting the generalizability of the results. Additionally, we acknowledge the potentially increased type I error due to multiple hypothesis testing. However, given the exploratory nature of the study, no correction for multiple testing was applied, but results should be interpreted with care. Future prospective studies are needed to further explore our findings in a hypertensive patient population with more severe OSA.

Due to the limited memory capacity (100 days) of the microsensor thermometer, the objective adherence data are limited to 1-month and 3-month follow-up, compared to the overall 6-month follow-up for MAD efficacy and the cardiovascular parameters. However, relatively high adherence rates were observed at 1-month and 3-month follow-up. Based on a recent cluster analysis, showing that long-term usage patterns can be identified in the first 20 days of MAD treatment, we assume similar high adherence rates at 6-month follow-up.⁴⁰

CONCLUSIONS

Overall, the results of this clinical trial showed that successful MAD therapy is able to reverse hypertrophic remodeling of the IVS at 6-month follow-up, regardless of systemic BP, suggesting that OSA is an independent factor in the pathophysiology of LVH in these patients.

DISCLOSURE STATEMENT

All authors have seen and approved this manuscript. This research project was partly supported by a grant at Antwerp University Hospital from SomnoMed, Sydney, New South Wales, Australia. M.D. holds a Postdoctoral Fellowship at the Research Foundation Flanders (FWO: 12H4520N). O.M.V. and M.C.B. report grants from SomnoMed at the Antwerp University Hospital. OMV and MJB report grants from SomnoMed at the Antwerp University Hospital and both authors are member of the advisory board of ResMed and SomnoMed. O.M.V. reports research support outside the submitted work from Philips and Inspire Medical Systems at the Antwerp University Hospital and consultancy for Zephyr Sleep Technologies and GSK. O.M.V. holds a Senior Clinical Investigator Fellowship from the Research Foundation Flanders (FWO: 1833517N). The other authors report no conflicts of interest.

ABBREVIATIONS

AHI

apnea-hypopnea index

BP

blood pressure

CPAP

continuous positive airway pressure

HSAT

home sleep apnea testing

IVS

interventricular septum

LV

left ventricular

LVH

left ventricular hypertrophy

MAD

mandibular advancement device

OSA

obstructive sleep apnea