Introduction

Obstructive sleep apnea syndrome (OSAS) is highly prevalent and associated with cardiovascular morbidity and deaths [1–3]. Positive airway pressure (PAP) devices are the gold standard treatment. The American Academy of Sleep Medicine (AASM) considers abnormal an apnea/hypopnea index (AHI) above five events per hour, and to bring an OSAS patient to an AHI below five events per hour is one of the major goals when using PAPs [4]. Besides, highly efficacious PAPs are not fully accepted by the patients, and those who start treatment have a low compliance, meaning that they do not use the device all days in the week, and also around 4 h per night [5, 6]. Such poor compliance with PAP treatment may be the reason why a recent study suggests that PAPs for OSAS patients were not fully effective in preventing stroke and deaths of any causes [7].

Among the treatment options for OSAS, the mandibular repositioning appliances (MRAs) with a titration mechanism (tMRAs) are very promising, but important data demonstrating their role have appeared only in specialized journals. Actually, many studies regarding MRAs are published in journals where the main issue is dentistry, preventing the majority of the physicians from accessing essential information already available about MRAs in the treatment of OSAS [8–10].

tMRAs are designed to gradually advance (titration) the mandible forward. And the sleep specialist dentist works weekly with the patient to achieve the mandibular position that is comfortable for the patient to sleep with, and at the same time resolutive regarding the AHI that has to be less than five events per hour, considering AASM criterion [4, 9, 11, 12].

There are few studies about OSAS treatment with MRAs showing outcomes with AHI below five events per hour. Therefore, the main goal of this study is to verify the efficacy of two tMRAs (ARMIO™ and Klearway™) in reducing AHI to a level below five events per hour [13–15].

Materials and methods

Patients

A total of 300 consecutive patients with a polysomnographic diagnosis of OSAS were referred for treatment with tMRA (260 were treated with ARMIO™ and 40 with Klearway™) between 2000 and 2003. Thirty patients were excluded from the treatment due to clinical–dental problems such as an insufficient number of teeth for stabilization of the appliance in the mouth, periodontal disease, and problems in the temporomandibular joints, and eight patients discontinued the use of the appliance already at the beginning of the protocol. Thus, 262 patients received a tMRA using a titration protocol that lasted 4 months. At the end of the titration period, the patient was referred for clinical reassessment and control polysomnography with the tMRA in place. After medical reassessment, 83 patients (62 treated with ARMIO™ and 21 with Klearway™) participated in a new polysomnographic exam (Fig. 1).

Fig. 1
figure 1

Fluxogram

Full size image

The mean age of the 83 patients (13 women and 70 men) was 50.6 ± 11.2 years. OSAS was mild (AHI = 5–<15 episodes per hour) in 20 patients, moderate (AHI = 15–30 episodes per hour) in 40, and severe (AHI > 30 episodes per hour) in 23 (Table 1).

Table 1 Demographic data of the patients and pre- and post-treatment AHI
Full size table

Polysomnography

Standard polysomnograms were obtained with a Neurotec apparatus (model EQSA-400), and the following parameters were recorded: electroencephalogram (C3, C4, A1, A2); right and left electrooculograms; submental and anterior tibial electromyogram; oral and nasal flow (thermistor); respiratory effort (piezoelectric belt sensors on the abdomen and chest); and electrocardiogram (modified D1). Sleep was staged according to the international criteria of Rechtschaffen and Kales [16]. Respiratory sleep disorders were diagnosed based on the criteria of the AASM [4].

Treatment

During the 4 months of the titration protocol, the patients were seen weekly at the office for titration of the tMRA according to their report of weekly progress. The tMRAs were titrated on an increasing scale until maximal mandibular advancement was achieved with comfort for the patient. Termination of the protocol was defined as the end of the fourth month of titration even when the patient still presented daytime symptoms, or before if the patient reported complete and subjective disappearance of the classical symptoms of OSAS. When symptoms of excessive daytime sleepiness or variable subjective complaints persisted even after the end of the titration period, the patients were referred to the responsible physician for reassessment and exclusion of other causes associated with the symptoms such as narcolepsy, restless leg syndrome, periodic limb movements during sleep, insomnia, and other clinical diseases [17].

After titration of the tMRAs, all patients were referred to their physicians for control polysomnographic reassessment with the appliance in place to determine the efficacy of treatment. During this phase, 83 patients returned for control polysomnography. After the end of the protocol, these patients returned to the dental office every 6 months on average for follow-up, revision of the appliance, and/or small repairs.

Patients presenting an AHI <5 events per hour upon control polysomnography were considered to have been treated successfully [4]. Although conceptually incorrect and not the therapeutic approach for our patients, treatment was also analyzed using a criterion called liberal, which was an AHI <10/h as widely used in other published studies.

Titratable mandibular repositioner appliances

The Brazilian titratable mandibular repositioning appliance ARMIO™ and the Klearway™ are custom oral appliances fabricated using the same criteria as described elsewhere [12, 17–19]. In short, these appliances were designed to maintain the mandible in a more anteriorized position than usual during sleep, increasing airway dimensions. Although the two appliances have different designs, both of them consist of two acrylic plates (ARMIO™ is made of hard acrylic and Klearway™ is made of thermoactive acrylic) completely covering the upper and lower arch, and a specific stainless-steel threaded screw mechanism of each appliance connected the two plates allowing only lateral movements for ARMIO™ and lateral and limited vertical movements for Klearway™. The possibility of weekly titration in tenths of millimeters during the treatment protocol is the major characteristic of these two appliances.

Variables studied

The explanatory variables used were gender (male and female), body mass index (BMI), pre- and post-treatment AHI, and age above and below 60 years.

Statistical analysis

The mean polysomnographic results were compared before and after treatment with the tMRAs by the t test. For descriptive analysis of the data, the absolute (AD = AHIpre − AHIpost) and relative differences [RD = (AHIpre − AHIpost) ÷ AHIpre] before and after treatment were considered, where AD is the absolute difference between polysomnographic recordings indicating how much the AHI decreased after treatment, and RD is the relative difference between polysomnographic recordings indicating the percent decrease in the AHI.

In addition to these variables, the pre-treatment AHI was classified as follows: 5 to <15/h = mild apnea; 15 to 30/h = moderate apnea; and >30/h = severe apnea [4].

A multiple linear regression model was used to determine the correlation between the post-treatment AHI (AHIpost) and the pre-treatment AHI (AHIpre), gender, age (years), and BMI (kg/m2) [20]. To adequately fit the model, logarithmic transformation of the pre- and post-treatment AHI was necessary, resulting in the following model: ln(AHIpost + 1) = ln(AHIpre + 1) + b 1 + b 2 + b 3 + K + Ý, where b 1 = age, b 2 = BMI, b 3 = gender, ln = natural logarithm, K = constant, and Ý = random error.

The final model shows the following relationship: AHIpost = (AHIpre)0,445. This formula is not a predictive equation but is used only to understand the results of the descriptive analysis, i.e., the poorer the initial state of the patient, the more effective the appliances.

To measure the efficacy of the appliances, treatment outcome was classified as follows: treated, patients presenting an AHI <5/h, and untreated, patients presenting an AHI ≥5/h.

Results

The mean pre- and post-treatment AHI were 26 ± 17.7 and 4.8 ± 5.3 events per hour (p < 0.00005). Fifty-three (63.8%) of the 83 patients were treated successfully (AHI < 5/h).

AASM criterion (AHI <5 events per hour)

Severe apnea

Twenty-three patients had severe OSAS, and 12 (52.1%) of them were successfully treated (Table 2). The mean number of obstructive events decreased from 39.3 ± 8.5/h to 1.5 ± 1.2/h in this group of treated patients, whereas in patients considered to be untreated, the mean number of events fell from 55.0 ± 26.4/h to 13.2 ± 6.9/h (Table 3).

Table 2 Outcome of treatment with the tMRAs evaluated based on different treatment criteria according to the severity of sleep apnea
Full size table
Table 3 AHI (events/h) obtained for treated and untreated patients according to the severity of sleep apnea (AASM criterion)
Full size table

Moderate apnea

Forty patients had moderate OSAS, and 26 (65%) were successfully treated (Table 2). The mean AHI fell from 22.2 ± 4.6/h to 1.8 ± 1.6/h in the treated group and from 21.6 ± 4.1/h to 9.1 ± 3.7/h in the untreated group (Table 3).

Mild apnea

Twenty patients had mild OSAS, and 15 of them (75%) were successfully treated (Table 2). The mean AHI fell from 10.4 ± 2.8/h to 1.9 ± 1.3/h in the treated group and from 12.5 ± 2/h to 7 ± 1.3/h in the untreated group (Table 3).

Liberal criterion (AHI <10 events per hour)

Severe apnea

Seventeen (65%) of the 23 patients with severe OSAS were treated successfully.

Moderate apnea

Thirty-five (86%) of the 40 patients with moderate OSAS were treated successfully.

Mild apnea

All (100%) of the 20 patients with mild OSAS were treated successfully with the tMRAs (Table 2).

The mean decrease in the AHI was 21 units in absolute terms, and the mean relative decrease was 80%.

The multiple linear regression model showed that only the pre-treatment AHI was significantly correlated with post-treatment AHI (p = 0.01). No correlation was observed between the post-treatment AHI and BMI, age, or gender.

Considering the categorized pre-treatment AHI, the box plots of the absolute and relative differences clearly show that the more severe the pre-treatment apnea, the greater the absolute and percent decrease in the AHI after treatment (Fig. 2). Taking into account the degree of apnea severity at the beginning of treatment, a decline in the percentage of treated subjects was observed with advancing disease.

Fig. 2
figure 2

Box plots of the AD and RD according to the severity of pre-treatment apnea

Full size image

Considering the whole sample, the proportion of treated patients was estimated to be 64%, with a 95% confidence interval of 0.53 to 0.74.

Discussion

We have shown in this study that intra-oral appliances are efficacious in treating OSAS. These appliances bring 75% of patients with mild OSAS, 65% with moderate OSAS, and 52% with severe OSAS to an AHI that is considered normal. Information on the role of intra-oral appliances in treating OSAS is not yet readily available in the medical literature, but many patients all around the world would benefit greatly from this intervention.

One important point that may define the degree of success of this type of treatment is the final titration of the mandibular protrusion. This must always be done individually, exploring the maximum mandibular advancement possible, while at the same time respecting each patient’s limits [21]. In contrast with this principle of individuality, most randomized controlled studies that utilized MRAs for OSAS have reported a fixed and predetermined advancement for the mandible [22–26]. This may have been the reason why some studies demonstrated unsatisfactory results. In the present study, we chose the maximum mandibular advancement tolerated individually by each patient, which ranged from 8 to 11 mm in most patients [15].

Intra-oral appliances have good acceptance among patients, which is not the case with devices that apply PAP [18, 27–30]. The compliance to treatment using positive pressure varies greatly, and despite the documented efficacy of continuous positive airway pressure (CPAP), it has been estimated that more than 50% of the patients who start treatment with CPAP will not be using it 1 year later [18].

Studies have shown that the mean compliance to CPAP may vary from 3.4 to 4.7 h per night [6, 31, 5], whereas the compliance to tMRAs is around 6.8 h per night [18]. Thus, most patients with OSAS who are treated with CPAP spend 2 to 3 h per night without any treatment: a period of time that might have received treatment if another method had been used [3].

Although the real compliance to treatment with CPAP is not as good as expected, even this suboptimal treatment improves sleep structure and reduces O2 desaturation. On the other hand, it does not improve blood pressure, and this emphasizes the importance of ideal treatment that reduces cardiovascular sequelae [3, 7].

Why then is positive pressure therapy so popular and intra-oral appliances are not? The main reason is the high therapeutic efficacy associated with positive pressure. The AHI can be normalized in almost all patients after one night of titration using CPAP. This efficacy is undoubtedly necessary, but unfortunately, it cannot be confounded with efficiency. Nonetheless, the high efficacy of positive pressure is a message that is so strong that physicians almost forget about the other side of the treatment process: the patient. The patient behind the CPAP mask reacts to the side effects of this therapy and faces problems within his social environment and relationships through using CPAP, yet often hardly presents any perceptible signs and symptoms that could justify its daily use. Despite being made aware of the morbidity, mortality, and risk factors associated with OSAS, in most cases, patients cannot be convinced to adhere completely to treatment with CPAP.

Another strong reason for the popularity of CPAP is the economic question surrounding this technology. Companies strive to promote the best solution for each problem or complication, with the objective of bringing greater comfort in using a device that is hard to tolerate on a day-to-day basis. This continual effort towards improving the equipment promotes an image of care and concern for each patient’s health. Although this is true, such efforts cannot fail to constitute a strong advertising message that dazzles patients and suggests that they have no option other than positive pressure therapy. Moreover, the message suggests that certain difficulties can be attenuated through desensitization, more adequate masks, accessories for the positive pressure device, and cognitive approaches, among other measures. Economic interests and the low involvement of many physicians who feel that their work ends with the prescription account for this incredible diffusion of CPAP treatment. Nonetheless, this treatment only really works for patients with significant diurnal symptoms such as somnolence, and is ineffective for asymptomatic patients (the majority), who are recognized to be at greater cardiovascular risk because of the respiratory disturbance to sleep that they present [1–3, 7].

Efficiency (E) is not a concept that is widely understood. It is often confounded with and taken to be the same as efficacy (I). The efficiency of a type of treatment also takes into account the precision of the diagnosis (D) for the disease undergoing treatment, the compliance (C), and the access to treatment (A). Efficiency can therefore be calculated according to the following equation: E = I × D × C × A.

The efficiency associated with treatment using PAP, and particularly CPAP, would be (without considering access): E = 100% × 100% × 25% = 25%, where I = 100%, D = 100%, and C = 25%.

The efficiency for tMRAs, taking the mean adherence to be 80% [32, 33] and the efficacy of 64% from the present study, would be (again, without considering access): E = 64% × 100% × 80% = 51%, where I = 64%, D = 100%, and C = 80%.

The access factor was not taken into consideration in these calculations because it is very variable and depends on the socioeconomic development of different countries.

Surprisingly, the efficiency of intra-oral appliances is twice the efficiency of CPAP treatment. However, intra-oral appliances are not publicized among physicians. The principal consequence of this is that many patients are not receiving adequate treatment and continue to be exposed to the risks of OSAS and disease progression. Patients with low AHI, few diurnal symptoms, and healthy teeth and temporomandibular joints would fare better with intra-oral appliances. On the other hand, patients with severe apnea, excessive diurnal somnolence, and no claustrophobia or other significant symptoms relating to pressure, flows, and masks would fare better with positive pressure therapy.

In conclusion, titratable intra-oral appliances are efficacious and potentially effective for treating OSAS. They need to be more widely publicized among physicians involved in diagnosing and treating OSAS, and we suggest that well-designed studies for analyzing the efficiency of tMRAs should be conducted within future research.