An Investigation of Upper Airway Changes Associated with Mandibular Advancement Device Using Sleep Videofluoroscopy in Patients with Obstructive Sleep Apnea

Home , Sleep induction

ORIGINAL ARTICLE An Investigation of Upper Airway Changes Associated With Mandibular Advancement Device Using Sleep Videofluoroscopy in Patients With Obstructive Sleep Apnea

Chul Hee Lee, MD, PhD; Jeong-Whun Kim, MD, PhD; Hyun Jong Lee, MD, PhD; Pil-Young Yun, DDS; Dong-Young Kim, MD, PhD; Beom Seok Seo, MD; In-Young Yoon, MD, PhD; Ji-Hun Mo, MD, PhD

Objective: To quantitatively evaluate the effects of the Results: Without the MAD, the length of the soft palate mandibular advancement device (MAD) on changes in and the angle of mouth opening increased during sleep the upper respiratory tract during sleep using sleep vid- events, especially in desaturation sleep, compared with eofluoroscopy (SVF) in patients with obstructive sleep the awake state. The retropalatal space and retrolingual apnea (OSA). space became much narrower during sleep compared with the awake state. The MAD had marked effects on Design: Retrospective analysis. the length of the soft palate, retropalatal space, retrolingual space, and angle of mouth opening. The retropala- Setting: Academic tertiary referral center. tal space and retrolingual space were widened, and the length of the soft palate was decreased. The MAD kept Patients: Seventy-six patients (68 men and 8 women) the mouth closed. who were treated with the MAD for OSA were included from September 1, 2005, through August 31, 2008. Conclusions: Sleep videofluoroscopy showed dynamic upper airway changes in patients with OSA, and the Intervention: All patients underwent nocturnal polysomnography and SVF before and at least 3 months af- MAD exerted multiple effects on the size and configura- ter receipt of the custom-made MAD. Sleep videofluo- tion of the airway. Sleep videofluoroscopy demon- roscopy was performed before and after sleep induction strated the mechanism of action of the MAD in patients and was analyzed during 3 states of awakeness, normoxy- with OSA. The MAD increased the retropalatal and ret- genation sleep, and desaturation sleep. rolingual spaces and decreased the length of the soft palate and the angle of mouth opening, resulting in Main Outcome Measures: Changes in the length of improvement of OSA. the soft palate, retropalatal space, retrolingual space, and angle of mouth opening were evaluated during sleep events with or without the MAD. Arch Otolaryngol Head Neck Surg. 2009;135(9):910-914

HE MANDIBULAR ADVANCE- Most studies are limited in that they ment device (MAD) is an es- used static images instead of dynamic tablished treatment option images. Furthermore, images were ob- for mild to moderate obtained during awakeness instead of dur- structive sleep apnea (OSA) ing sleep and with the patient in upright andT is recommended as first-line therapy for postures instead of supine. Few investi- Author Affiliations: mild OSA and as second-line therapy for gations have obtained dynamic images Departments of moderate to severe OSA.1 Its underlying from the supine position during sleep. Otorhinolaryngology–Head and mechanism of action for apnea improve- Sleep videofluoroscopy (SVF) is a good Neck Surgery (Drs C. H. Lee, ment has been extensively studied since its modality to visualize dynamic airway J.-W. Kim, H. J. Lee, D.-Y. Kim, introduction, and the effects of the MAD on changes,2-5 and the efficacy of SVF for air- Seo, and Mo), Oral Surgery the upper respiratory tract have been vari- way evaluation in patients with OSA has (Dr Yun), and Psychiatry 6 (Dr Yoon), Seoul National able. These discrepancies may result from been demonstrated. In this study, we evalu- University Bundang Hospital, differences in the imaging modality used, ated dynamic upper airway changes with or Seoul National University position of the patient’s body (upright or su- without the MAD during sleep and as- College of Medicine, pine), sleep status (awake or asleep), and sessed the mechanism of action of the MAD Seongnam, Korea. degree of protrusion by the MAD. by SVF.

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 135 (NO. 9), SEP 2009 WWW.ARCHOTO.COM 910

Angle of mouth opening

Length Retropalatal of soft palate space

Retrolingual space

Figure 1. Lateral cephalometry showing the variables measured. The length of the soft palate was defined as the distance from the posteronasal spine to the uvula tip (A), the retropalatal space as the narrowest posterior airway space at the level of the soft palate (A), the retrolingual space as the narrowest posterior airway space at the level of the tongue base (A), and the angle of mouth opening as the angle formed by the intersection of lines drawn from the maxillary incisor to the glenoid fossa and from the glenoid fossa to the mandibular incisor (B).

METHODS the sleep event examinations started. Although oxygen saturation does not decrease, a 15-second respiratory state was recorded as a normoxygenation sleep event. When oxygen satu- PATIENTS ration dropped by 4% or more, two 15-second desaturation sleep Seventy-six patients (68 male, with a mean [SD] age of 51.7 events were recorded. [10.3] years [age range, 21-69 years]) who visited the Sleep Cen- ter at Seoul National University Bundang Hospital, Seong- EVALUATION OF SVF VARIABLES nam, Korea, from September 1, 2005, through August 31, 2008, were retrospectively included in this study. All patients under- The length of the soft palate, retropalatal space, retrolingual space, went full-night nocturnal polysomnography and were diag- and angle of mouth opening were measured and evaluated dur- nosed as having OSA. They were referred to a single dentist ing 3 different states (awakeness, normoxygenation sleep, and (P.-Y.Y.), and a custom-made MAD was fabricated for each pa- desaturation sleep). The length of the soft palate was defined as tient. The MAD was designed as a monobloc that holds the man- the distance from the posteronasal spine to the uvula tip, the ret- dible fixed at 60% of maximal protrusion without open bites. ropalatal space as the narrowest posterior airway space at the level Patients underwent a second full-night nocturnal polysomnog- of the soft palate, the retrolingual space as the narrowest posterior raphy at least 3 months after receipt of the custom-made MAD. airway space at the level of the tongue base, and the angle of mouth The mean (SD) apnea-hypopnea indexes of patients were 38.9 opening as the angle formed by the intersection of lines drawn (19.7) without the MAD and 12.3 (11.4) with the MAD. (Pa- from the maxillary incisor to the glenoid fossa and from the gle- tients with an apnea-hypopnea index of Ն5 are considered to noid fossa to the mandibular incisor (Figure 1). have OSA: 5-20 indicates mild OSA; 21-40, moderate OSA; and Ն41, severe OSA.) Their mean (SD) body mass index (calcu- STATISTICAL ANALYSIS lated as weight in kilograms divided by height in meters squared) was 25.6 (2.6), and the lowest mean (SD) oxygen saturation was The t test was used to analyze differences among awakeness 79.2% (7.8%). This study was approved by the Institutional Re- and sleep events. Paired t test was used to analyze differences view Board of Seoul National University Bundang Hospital. between variables with or without the MAD. All results were expressed as the mean (SD). Statistical significance was as- Ͻ SLEEP VIDEOFLUOROSCOPY sumed at P .05 for all variables.

All patients underwent SVF with or without the MAD as pre- RESULTS viously described.6 In brief, patients were in the supine position on a C-arm table with their head on a pillow. They were instructed to breathe in and out naturally. Oxygen saturation EFFECTS ON THE LENGTH OF THE SOFT PALATE was monitored throughout the examination. During normal res- piration before sedation, an awake event was recorded for 15 When the length of the soft palate was compared among seconds. Thereafter, sleep was induced by intravenous admin- events without the MAD, it was longer during sleep events istration with midazolam (2 mg). After the patient fell asleep, than during awakeness. It was longest in desaturation

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 135 (NO. 9), SEP 2009 WWW.ARCHOTO.COM 911

©2009 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/05/2021 Table. Length of the Soft Palate, Retropalatal and Retrolingual Spaces, and the Angle of the Mouth Opening With the MAD During Different Events

Event Without MAD With MAD P Value Length of Soft Palate, mm Awakeness, mean (SD) 40.4 (4.8) 39.2 (4.8) .06 Normoxygenation sleep, mean (SD) 46.5 (6.3) 42.0 (5.7) Ͻ.001 Desaturation sleep, mean (SD) 48.3 (6.3) 43.2 (6.2) Ͻ.001 P value Ͻ.001 Ͼ.05 . . . Retropalatal Space, mm Awakeness, mean (SD) 4.7 (1.4) 5.5 (2.0) .003 Normoxygenation sleep, mean (SD) 2.1 (1.5) 3.2 (2.3) .004 Desaturation sleep, mean (SD) 1.2 (1.4) 2.6 (2.7) .002 P value Ͻ.001 Ͻ.001 . . . Retrolingual Space, mm Awakeness, mean (SD) 10.2 (2.4) 11.8 (2.8) Ͻ.001 Normoxygenation sleep, mean (SD) 7.5 (3.4) 10.1 (4.0) Ͻ.001 Desaturation sleep, mean (SD) 5.1 (3.7) 10.0 (4.1) Ͻ.001 P value Ͻ.001 Ͼ.05 . . . Angle of Mouth Opening, Degrees Awakeness, mean (SD) 0.58 (1.19) 1.22 (0.73) .001 Normoxygenation sleep, mean (SD) 3.77 (1.79) 1.42 (0.96) Ͻ.001 Desaturation sleep, mean (SD) 4.94 (2.00) 1.56 (1.28) Ͻ.001 P value Ͻ.001 Ͼ.05 . . .

Abbreviations: Ellipses, not applicable; MAD, mandibular advancement device.

sleep, although no significant difference was noted be- to 10.0 (4.1) mm in desaturation sleep. However, the ret- tween the 2 sleep events. The MAD decreased the length rolingual space with the MAD did not change signifi- of the soft palate from 46.5 (6.3) mm to 42.0 (5.7) mm cantly during different sleep events (PϾ.05), showing in normoxygenation sleep and from 48.3 (6.3) mm to 43.2 relative constant preservation of the retrolingual space. (6.2) mm in desaturation sleep (P Ͻ .001 for both) (Table). The MAD had no significant effect on the length EFFECTS ON THE ANGLE OF MOUTH OPENING of the soft palate in awake events (P=.06). In awake events, the angle of mouth opening was 0.58° EFFECTS ON THE RETROPALATAL SPACE (1.19°) without the MAD, showing that the mouth was almost closed during awakeness (Table). During sleep, Sleep videofluoroscopy performed without the MAD it increased to 3.77° (1.79°) in normoxygenation sleep showed dramatic changes in the retropalatal space asso- and to 4.94° (2.00°) in desaturation sleep. With the MAD, ciated with different sleep events. The retropalatal space the angle of mouth openings were 1.22° (0.73°) during in awake events was 4.7 (1.4) mm and decreased to 2.1 awakeness, 1.42° (0.96°) in normoxygenation sleep, and (1.5) mm during normoxygenation sleep and to 1.2 (1.4) 1.56° (1.28°) in desaturation sleep. The angle of mouth mm during desaturation sleep, demonstrating narrow- opening did not differ significantly with various awake ing of the upper airway during sleep in patients with OSA or sleep events, indicating that the MAD prevented mouth Ͻ (P .001) (Table). The MAD increased the retropalatal opening during sleep and maintained a constant angle Ͻ space significantly during sleep events (P .001 for both). of opening (Figure 2). It increased from 4.7 (1.4) mm to 5.5 (2.0) mm in awake events, from 2.1 (1.5) mm to 3.2 (2.3) mm in normoxygenation sleep, and from 1.2 (1.4) mm to 2.6 (2.7) mm COMMENT in desaturation sleep. In this study, SVF showed dynamic upper airway changes EFFECTS ON THE RETROLINGUAL SPACE during different sleep events and demonstrated the mechanism of action of the MAD on dynamics of the upper air- The retrolingual space during SVF without the MAD also way in patients with OSA. The upper airway size changed showed dramatic changes during different sleep events. dramatically during sleep, especially in desaturation sleep, It was 10.2 (2.4) mm in awake events and decreased to compared with awakeness. During desaturation sleep, the 5.1 (3.7) mm in desaturation sleep (Table). With the length of the soft palate increased, the retropalatal and MAD, the retrolingual space widened significantly in retrolingual spaces narrowed, and the angle of mouth op- awake and sleep events (PϽ.001). The MAD increased ening increased in patients with OSA. The MAD had an the retrolingual space from 10.2 (2.4) mm to 11.8 (2.8) important role in the dynamics of the upper airway and mm in awake events, from 7.5 (3.4) mm to 10.1 (4.0) counteracted the aforementioned changes during sleep. mm in normoxygenation sleep, and from 5.1 (3.7) mm In other words, it decreased the length of the soft palate,

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 135 (NO. 9), SEP 2009 WWW.ARCHOTO.COM 912

©2009 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/05/2021 Awake event Normoxygenation sleep event Desaturation sleep event With MAD application Without MAD application

Figure 2. The angle of mouth opening with or without the mandibular advancement device (MAD) application during different sleep events as assessed by sleep videofluoroscopy. Without the MAD application, the angle of mouth opening increased significantly in sleep events compared with awake events. With the MAD, the angle of mouth opening stabilized at 2°.

widened the retropalatal and retrolingual spaces, and de- venting mouth opening and widening the retropalatal and creased the angle of mouth opening during sleep. retrolingual spaces, the MAD decreases the length of the The mechanism of action of the MAD on the upper soft palate. In general, the MAD has been thought to be airway size has been studied using various methods, in- mainly effective in widening the retropalatal and retro- cluding cephalometry, computed tomography, and mag- lingual spaces. However, the present study shows that netic resonance imaging, and results vary slightly among the MAD applies tension to the soft palate, preventing studies.7-13 The velopharynx (retropalatal space)10,13 or the collapse of the retropalatal space. hypopharynx8,12 was claimed to be widened in those stud- This study has some limitations, which are the same ies. Variable findings may be attributed to different study as those discussed in a previous study.6 Briefly, SVF is a methods, body position, or sleep status among patients. superimposed 2-dimensional image of 3-dimensional Although several studies7,8,10,11 have described the ef- structures. Therefore, it cannot explain lateral move- fects of the MAD on upper airway changes, most have limi- ment of the upper airway. In addition, full-night sleep tations in that the data were obtained while patients were was not included for SVF, and sleep was induced by drug awake, or the studies were based on static images of pa- administration, although another study14 proved the va- tients in the supine position. To overcome these limita- lidity of drug-induced sleep as representative of normal tions, we used SVF to evaluate the mechanism of action of sleep. Despite those limitations, our study provides novel the MAD. Sleep videofluoroscopy has several advantages. information about the mechanism of action of the MAD First, it provides dynamic images while patients are asleep on the upper airway in patients with OSA. in the supine position for a short period. Therefore, it can In conclusion, the upper airway changes dynami- be used to detect dynamic airway changes during sleep. Sec- cally during awake and sleep events. Sleep videofluoros- ond, it easily shows the mechanism of action of the ana- copy showed dynamic upper airway changes in patients tomical structures outside of the pharyngeal airway. For with OSA, and the MAD exerted multiple effects on the instance, movements of the cervical spine, mouth open- size and configuration of the upper airway. The mecha- ing, tongue, hyoid bone, and jaw can be directly ob- nism of action of the MAD in patients with OSA in- served. Third, SVF detects upper airway changes not only cludes widening the retrolingual space, decreasing the in awake events but also in sleep events (normoxygen- length of the soft palate, and narrowing the angle of mouth ation sleep and desaturation sleep). Because desaturation opening. sleep corresponds to the period of sleep apnea when the upper airway changes most severely and dynamically, analy- Submitted for Publication: February 16, 2009; final sis of desaturation sleep events can elucidate the exact revision received April 10, 2009; accepted April 23, mechanism of action of the MAD during apneic periods, 2009. which cannot be evaluated by other methods. Correspondence: Ji-Hun Mo, MD, PhD, Department of Indeed, our results provide useful information about Otorhinolaryngology–Head and Neck Surgery, Seoul Na- the mechanism of action of the MAD. In addition to pre- tional University Bundang Hospital, Seoul National Uni-

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 135 (NO. 9), SEP 2009 WWW.ARCHOTO.COM 913

©2009 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/05/2021 versity College of Medicine, 300 Gumi-dong Bundang- 3. Walsh JK, Katsantonis GP, Schweitzer PK, Verde JN, Muehlbach M. Somnofluo- gu, Seongnam 464-707, Korea ([email protected]). roscopy: cineradiographic observation of obstructive sleep apnea. Sleep. 1985; 8(3):294-297. Author Contributions: Drs C. H. Lee, J.-W. Kim, H. J. 4. Pepin JL, Ferretti G, Veale D, et al. Somnofluoroscopy, computed tomography, Lee, and Mo had full access to all the data in the study and cephalometry in the assessment of the airway in obstructive sleep apnoea. and take responsibility for the integrity of the data and Thorax. 1992;47(3):150-156. the accuracy of the data analysis. Drs C. H. Lee and J.-W. 5. Hillarp B, Nylander G, Rose´n I, Wickström O. Videoradiography of patients with Kim equally contributed to this work. Study concept and habitual snoring and/or sleep apnea: technical description and presentation of videoradiographic results during sleep concerning occurrence of apnea, type of design: J.-W. Kim, H. J. Lee, Yun, D.-Y. Kim, Yoon, and apnea, and site of obstruction. Acta Radiol. 1996;37(3, pt 1):307-314. Mo. Acquisition of data: H. J. Lee, Yun, D.-Y. Kim, Seo, 6. Lee CH, Mo JH, Kim BJ, et al. Evaluation of the soft palate changes using sleep Yoon, and Mo. Analysis and interpretation of data: J.-W. videofluoroscopy in patients with obstructive sleep apnea [published correction Kim, Seo, and Mo. Drafting of the manuscript: Mo. Criti- appears in Arch Otolaryngol Head Neck Surg. 2009;135(4):354]. Arch Otolaryn- cal revision of the manuscript for important intellectual con- gol Head Neck Surg. 2009;135(2):168-172. tent: C. H. Lee, J.-W. Kim, H. J. Lee, Yun, D.-Y. Kim, Seo, 7. Liu Y, Zeng X, Fu M, Huang X, Lowe AA. Effects of a mandibular repositioner on obstructive sleep apnea. Am J Orthod Dentofacial Orthop. 2000;118(3):248- and Yoon. Statistical analysis: H. J. Lee, Seo, and Mo. Ob- 256. tained funding: J.-W. Kim, Yun, and Mo. Administrative, 8. Mayer G, Meier-Ewert K. Cephalometric predictors for orthopaedic mandibular technical, and material support: C. H. Lee, J.-W. Kim, Yun, advancement in obstructive sleep apnoea. Eur J Orthod. 1995;17(1):35-43. D.-Y. Kim, Yoon, and Mo. Study supervision: C. H. Lee, 9. Bernhold M, Bondemark L. A magnetic appliance for treatment of snoring pa- D.-Y. Kim, and Yoon. tients with and without obstructive sleep apnea. Am J Orthod Dentofacial Orthop. 1998;113(2):144-155. Financial Disclosure: None reported. 10. Tsuiki S, Hiyama S, Ono T, et al. Effects of a titratable oral appliance on supine Funding/Support: This study was supported by grant 11- airway size in awake non-apneic individuals. Sleep. 2001;24(5):554-560. 2008-011 from the Seoul National University Bundang 11. Gale DJ, Sawyer RH, Woodcock A, Stone P, Thompson R, O’Brien K. Do oral Hospital Research Fund. appliances enlarge the airway in patients with obstructive sleep apnoea? a pro- spective computerized tomographic study. Eur J Orthod. 2000;22(2):159-168. 12. Gao XM, Zeng XL, Fu MK, Huang XZ. Magnetic resonance imaging of the upper REFERENCES airway in obstructive sleep apnea before and after oral appliance therapy. Chin J Dent Res. 1999;2(2):27-35. 1. American Sleep Disorders Association. Practice parameters for the treatment of 13. Ishida M, Inoue Y, Suto Y, et al. Mechanism of action and therapeutic indication snoring and obstructive sleep apnea with oral appliances. Sleep. 1995;18(6): of prosthetic mandibular advancement in obstructive sleep apnea syndrome. Psy- 511-513. chiatry Clin Neurosci. 1998;52(2):227-229. 2. Suratt PM, Dee P, Atkinson RL, Armstrong P, Wilhoit SC. Fluoroscopic and com- 14. Sadaoka T, Kakitsuba N, Fujiwara Y, Kanai R, Takahashi H. The value of sleep puted tomographic features of the pharyngeal airway in obstructive sleep apnea. nasendoscopy in the evaluation of patients with suspected sleep-related breath- Am Rev Respir Dis. 1983;127(4):487-492. ing disorders. Clin Otolaryngol Allied Sci. 1996;21(6):485-489.

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 135 (NO. 9), SEP 2009 WWW.ARCHOTO.COM 914