- Email: [email protected]
Multicenter study of a novel adjustable tongue-advancement device for obstructive sleep apnea
Otolaryngology–Head and Neck Surgery (2010) 143, 585-590
ORIGINAL RESEARCH–SLEEP MEDICINE
Multicenter study of a novel adjustable tongue-advancement device for obstructive sleep apnea B. Tucker Woodson, MD, David L. Steward, MD, Samuel Mickelson, MD, Tod Huntley, MD, and Andrew Goldberg, MD, Milwaukee, WI; Cincinnati, OH; Atlanta, GA; Carmel, IN; and San Francisco, CA Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article. ABSTRACT OBJECTIVE: The safety and clinical effect of a new surgical device for tongue suspension for obstructive sleep apnea (OSA) was assessed. STUDY DESIGN: Multicenter phase 2 prospective case series. SETTING: Multicenter academic and private. SUBJECTS AND METHODS: Surgically naïve patients with moderate-to-severe OSA and tongue base obstruction (body mass index ⬍ 32, apnea/hypopnea index [AHI] 15-60) underwent surgical insertion of a midline tissue anchor into the posterior tongue and connected to an adjustable mandibular bone anchor with a flexible tether. Outcomes included changes in AHI, sleepiness (Epworth Sleepiness Scale), sleep-related quality-of-life (Functional Outcomes of Sleep Questionnaire), snoring, swallowing, speech, and pain (0-10 visual analog scale). RESULTS: After the implant, 42 patients (mean age 50 years, body mass index 28) noted improvement at six months for AHI (mean [SD]: 35.5 [20.4] to 27.3 [18.8]), Epworth Sleepiness Scale (11.5 [3.9] to 7.8 [4.7]), and Functional Outcomes of Sleep Questionnaire (15.5 [2.6] to 17.5 [2.6], all P ⬍ 0.01). Snoring VAS scores improved (7.3 [2.1] to 4.7 [2.9], P ⬍ 0.01). Postimplant pain scores were mild to moderate (4.4) at day one and resolved by day five. Post-titration pain scores were mild (⬍ 2). Devicerelated adverse events included wound infection (7%) and edema or seroma (5%), which resolved. However, in 31 percent of patients, asymptomatic tissue anchor barb fractures were observed radiographically. CONCLUSION: The tissue anchor failure rate of the tested device precludes its clinical use; however, the study results support that a titratable, tongue-suspension device with low direct surgical morbidity in patients with moderate-to-severe sleep apnea significantly improves multiple measures of sleep apnea. Further investigation is warranted. © 2010 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved.
O
ne treatment option for patients with obstructive sleep apnea (OSA) is upper airway surgery. Reconstructive surgery has the potential of reducing airway obstruction by
altering, supporting, or opening the upper airway. Traditional surgery for OSA may include uvulopalatopharyngoplasty for the retropalatal component of the pharyngeal collapse combined with anterior mandibular osteotomy and genioglossus advancement for the hypopharyngeal component.1 However, the perceived invasiveness of an osteotomy, postoperative pain, and dental side effects likely limit both patient and surgeon enthusiasm for this approach despite its potential benefits.2 Tongue base suspension is conceptually appealing because it may support and stabilize the hypopharyngeal airway without need for mandibular osteotomy as required for traditional genioglossus advancement.3,4 A commercially available tongue suture suspension device (Repose, Medtronic ENT, Jacksonville, FL) has shown promise in some studies5-9 but does not allow for adjustment after the initial surgery to account for individual anatomy, optimization of tongue position, or potential disease relapse. It is also associated with significant acute postoperative pain and dysphagia. The Advance System (Aspire Medical, Sunnyvale, CA) incorporates a mandibular bone anchor with an adjustable spool tethered to a tongue base tissue anchor implanted into the midline posterior tongue muscle just deep to the mucosal surface (Fig 1). The bone and tissue anchors are placed via a 2-cm submental incision during a first-stage implantation procedure with only the slack removed from the tether line. In the second stage, a titration procedure is performed two to four weeks after implantation. Tension on the tissue anchor is adjusted via a small titration needle, which engages the spool on the bone anchor, advancing the tongue anteriorly or relaxing it to fall more posteriorly. This is done percutaneously or through the same submental incision under local anesthesia. It is widely accepted that multiple structures of the upper airway contribute to sleep apnea and that successful treatment often requires addressing multiple structures to obtain an acceptable reduction in disease severity. Although a small number of patients may benefit from isolated treat-
Received August 13, 2009; revised April 16, 2010; accepted June 10, 2010.
0194-5998/$36.00 © 2010 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved. doi:10.1016/j.otohns.2010.06.902
586
Otolaryngology–Head and Neck Surgery, Vol 143, No 4, October 2010
Figure 1 Advance System for tongue suspension and advancement is depicted. (A) System components, including sheath, trochar, insertion device, and adjustment needle shown. (B) Tissue and bone anchors are shown, enlarged.
ment of individual structures, there is no demonstrated or even widely accepted method of selecting these patients or procedures. To achieve a definitive clinical outcome, therefore, multiple surgical procedures are likely required. Multiple procedures, however, confound attempts to assess the contribution of individual procedures. An alternative approach to assess surgical procedures is to measure whether procedures are effective rather than curative. Procedures, especially when they are of low invasiveness and morbidity, may be combined in a stepwise algorithm until an adequate clinical effect has been achieved. With this concept in mind, we designed a multicenter prospective phase 2 study to assess the safety, feasibility, and therapeutic effectiveness of the Advance System for patients with moderate-to-severe OSA.
Methods Study Design This is a prospective, multicenter phase 2 study.
Human Subjects This study was approved by the local institutional review boards at the four study sites and performed in compliance with the institutional review board regulations, good clinical practice guidelines, and Health Insurance Portability and Accountability Act. Subjects were recruited from private and academic otolaryngology office settings and from newspaper advertisements. The study was monitored by a data safety monitoring board, and the study was conducted and monitored in compliance with 21 Code of Federal Regulations parts 50, 54, 56, and 812 and the 18th World Medical Assembly, Helsinki, Finland, 1964, and later revisions.
Inclusion/Exclusion Criteria Adult patients (18-65 years) with moderate-to-severe OSA (apnea hypopnea index [AHI] ⫽ 15-60) were included 1) on the basis of diagnostic polysomnography within six months of enrollment or on repeat baseline sleep studies with the use of WatchPat (Itimar Medical, Haifa, Israel); 2) clinical findings consistent with tongue-related hypopharyngeal up-
per airway obstruction (investigator clinical impression on the basis of physical examination and upper airway endoscopy); or 3) failure or refusal of medical treatment with nasal continuous positive airway pressure (CPAP). Exclusion criteria included 1) major obesity (body mass index [BMI] ⬍ 32 kg/m2), 2) previous upper airway surgery exclusive of tonsillectomy, adenoidectomy, or nasal surgery, 3) enlarged tonsils (3⫹ and 4⫹); 4) anatomically unable to accommodate implant; 5) severe mandibular deficiency; 6) severe nasal obstruction; 7) systemic infection; 8) history of any of the following: head and neck or respiratory tract cancer, radiation therapy to head and neck, dysphagia, major cardiovascular disease, major pulmonary disease including chronic obstructive pulmonary disease, inadequately treated depression, or falling asleep while driving or motor vehicle accidents secondary to sleepiness; or 9) anesthesia class (American Society of Anesthesiologists) IV or V. Baseline and six-month full night diagnostic polysomnography were performed separately from studies used for inclusion/exclusion and included electroencephalography, electrooculography, chin and leg muscle electromyography, electrocardiography, measures of oronasal airflow, thoracic and abdominal efforts, body position, and pulse oximetry. Each site used the same sleep laboratory facility for all studies. Apnea was defined as cessation of inspiratory airflow of at least 10 seconds. Hypopnea was defined as a 30 percent reduction of inspiratory airflow of at least 10 seconds, with an associated four percent decrease in oxyhemoglobin saturation. Home sleep studies (WatchPat; Itimar Medical, Haifa, Israel) were performed at multiple time points, including screening, after implantation, after titration, and at three months’ follow-up.
Implant Procedure/Advance System The Advance System includes a trochar through which a tissue anchor composed of a nickel titanium memory metal and an attached flexible tether is inserted and a bone anchor with an adjustable spool that allows changes in tether length transcutaneously by the use of a small needle-like adjustment tool. For this study, the device was placed under general anesthesia by the use of nasal intubation. Patients were prepped for a chin procedure, and a 2-cm transverse incision was used to access the posterior mentum. The trochar was inserted into the genioglossus muscle near the mandible and advanced to the posterior margin of the tongue. The tissue anchor was deposited within the tongue. Position was confirmed by palpation or, in some patients, by the use of fluoroscopy. The bone anchor was fixed to the inferoposterior aspect of the mandible at the midline by the use of three self-tapping bone screws. The tissue anchor was attached to the bone anchor via the tether line, and the bone anchor spool was adjusted by the physician to remove slack from the tether line and suspend the tongue, often under fluoroscopy. Patients were observed overnight in the hospital. Preoperative and seven days of postoperative antibiotics
Woodson et al
Multicenter study of a novel adjustable . . .
587
(amoxicillin or clindamycin) and analgesics were prescribed. Two to four weeks after the procedure, titration was performed in the operating room under local anesthesia with or without sedation according to the surgeon and patient preference. A small incision was made on the inferior surface of the chin proximate to the bone anchor. The adjustment tool was then used to access the spool adjustment mechanism on the bone anchor. The tongue base tissue anchor was advanced anteriorly via the tether line by use of the spool and titrated to speech and swallowing tolerance and/or fluoroscopically visualized tongue base location relative to the posterior pharyngeal wall. In two patients, a second titration procedure was performed. The reported measured advancement was arbitrarily defined as the distance of tether shortening.
Results Forty-two subjects (4 female, 38 male) were enrolled and implanted. No patient was lost to follow-up. Patient demographics are shown in Table 1. On average, subjects were predominately middle-aged, overweight men with moderate-to-severe sleep apnea and excessive daytime sleepiness. The vast majority (⬎ 90%) had modified Mallampati class II or III oropharyngeal airways (free edge of palate not visible without phonation or tongue depressor use) and absent or 1⫹ tonsil size. Mean surgical implant time was 43 minutes (SD, 20 minutes), and mean surgical titration time was 14 minutes (SD, 8 minutes). The mean amount of tongue advancement was 1.3 cm (SD, 0.3 cm). After implantation and subsequent titration, statistically significant improvements in AHI (Fig 2), as well as Epworth Sleepiness Scale and Functional Outcomes of Sleep Questionnaire were noted at three and six months after the procedure (Table 2). Snoring VAS scores were also significantly improved at three and six months after titration. Implant procedure side effects and morbidity were studied. Pain on postimplant day one was moderate (4.4 on a 0-10 visual analog scale) and then decreased steadily through day five. Post-titration procedure pain was mild
Table 1 Baseline demographic data Advance System (n ⫽ 42) Age (yrs) % Male BMI, kg/m2 AHI, events/h ESS FOSQ
50.1 (9.1) 90 27.8 (2.1) 35.5 (20.4) 11.5 (3.9) 15.5 (2.6)
Values are mean (SD) or proportion of subjects. BMI, body mass index; AHI, apnea/hypopnea index; ESS, Epworth Sleepiness Scale (⬍ 10 is normal); FOSQ, Functional Outcomes of Sleep Questionnaire (⬎ 17 is normal).
Figure 2 Individual AHI results, including independent postinclusion baseline AHI, are shown. Screening to baseline AHI did not demonstrate a change and was not significantly different (P ⫽ NS). Baseline to six months was significantly reduced (P ⬍ 0.01).
(⬍ 2.0). Pain scores at three and six months were negligible (Fig 3). Swallowing, speech, taste, and throat irritation scores demonstrated mild and transient change and are shown in Figure 4. Complications were significant. In one patient, failure of the bone anchor locking mechanism prevented titration, and one bone anchor was repositioned under local anesthesia because of surgical error and placement off midline. Tongue base tissue anchor barb fracture occurred in 13 patients (31%). Fractures were not observed perioperatively or immediately after implantation or titration but were identified on radiographs taken at three, six, or 12 months (Fig 5). Other device-related complications were transient and included wound infection or cellulitis in seven percent, who required a longer than protocol course of antibiotics, and edema/seroma in five percent. No device was removed, but after device fractures were identified, implantation of additional implants were halted.
588
Otolaryngology–Head and Neck Surgery, Vol 143, No 4, October 2010
Table 2 Baseline and postoperative outcome measures Screening AHI ESS FOSQ Snore VAS
36.7 (14.1)
Baseline 35.5 11.5 15.5 7.3
(20.4) (3.9) (2.6) (2.1)
Three-month
7.5 (4.9) 17.9 (2.1) 3.6 (2.8)
Six-month
P value
27.3 7.8 17.5 4.7
⬍0.002 ⬍0.001 ⬍0.001 ⬍0.001
(18.8) (4.7) (2.6) (2.9)
Values are mean (SD). AHI, apnea/hypopnea index; ESS, Epworth Sleepiness Scale; FOSQ, Functional Outcomes of Sleep Questionnaire; VAS, visual analog scale (0-10).
Discussion The current study was designed to demonstrate the safety and therapeutic effect of the Advance System for patients with evidence of hypopharyngeal tongue base obstruction and moderate-to-severe OSA syndrome. However, the procedure was complicated by unexpected device tissue anchor failures in 13 of 42 patients. Failures were asymptomatic and identified during radiographic protocol monitoring. As a result, all subjects were offered the option of device removal at no cost. No patient with or without device fracture or malfunction elected for device removal during the course of the study. Failure involved fracture of one of eight nitinol tissue barbs on the anchor (one patient with two barb fractures) and occurred between the anchor hub and the tissue barb consistent with excess loading forces at this point (Fig 5). The exact timing of the fractures are not known but were not observed at the first postimplant and pretitration radiograph. They were identified between one and three months. At six-month follow-up, no additional fractures were observed. Future tongue suspension or advancement device development will need to take into account the significant forces generated on a fixed point in the posterior superior tongue during swallowing estimated to exceed five pounds. Surgical treatment of the tongue base and hypopharynx remains a significant surgical challenge in the treatment of OSA.2 Although multiple procedures exist to treat this segment, few are widely used. Tongue suspension has been described as an alternative to maxillofacial surgical techniques, with the goal of supporting the tongue and preventing retroglossal airway collapse with techniques of lower morbidity and complexity.5-9 Conceptually, tongue suspension potentially is less morbid and may be effective by
Figure 3
Post-implant and titration pain: visual analog scores.
dilating the upper airway by applying a vector of force on the soft tissue surrounding the airway lumen. This idea is in contrast to nasal CPAP, which dilates the airway by applying a vector of force to the luminal airway surface. The force to treat the airway in many patients does not exceed common pressures of 5 to 15 cm/H2O used with CPAP. However, the forces exerted during other movements of the tongue, such as swallowing, may be vastly different. It is likely these forces that cause failure not those to treat disease. Previous methods of tongue suspension have used many different materials to support the soft tissue, including suture and fascia. With these, tongue tissue is triangulated, and the suspension material is anchored to the mandible with screw or wire. Data suggest that suspension improves clinical symptoms and may reduce clinical severity of disease and that morbidity may be less than with traditional limited mandibular osteotomy.5-10 However, current tongue suspension methods cannot be easily adjusted and still have significant perioperative morbidity with pain and swallowing dysfunction. The Advance System was designed to incorporate the following goals: minimally invasive, adjustable, and revers-
Figure 4 Post-implant and titration swallowing, speech, taste, and throat irritation visual analog scores.
Woodson et al
Multicenter study of a novel adjustable . . .
589
Figure 5 Post-implant (A) and post-titration (B) lateral cephalometric x-rays are shown. Device fracture is marked by arrow in post-titration x-ray.
ible. The design tested in this study met these objectives. The study used general anesthesia per protocol, but implantation of the device is a rapid minor procedure that potentially may be performed under local anesthesia. The device was readily titratable under local anesthesia with little or no sedation required and was well tolerated by patients. By allowing several weeks for tissues to adjust to the placement of the tissue and bone anchor, subsequent adjustments were performed with little acute changes in speech, swallowing, taste, or throat irritation with no significant changes in these measures from baseline or pretitration levels. Transient dysarthria was noted by some patients manifesting with pronunciation of “cks” or “x” sounds, which require tongue base retraction for correct pronunciation. Tongue advancement under straight local anesthesia permitted titration based up until or just before detection of subtle changes in these sounds. The amount of titration on average was 1.3 cm, measured on the basis of the amount of tether line shorted during titration. The ideal amount of titration to maximize effect and minimize side effects or risk of device failure was not determined by this study; however, a significant clinical effect was observed with this change. In addition to being of low surgical morbidity and being titratable, the device can be removed by the use of a specially designed removal system. Removal is important in the potential of infection, patient discomfort or concerns, and because the long-term effects of a tongue implant are unknown. Of interest, the device not only reduced AHI, a global metric of sleep apnea, but also snoring. A common perception is that snoring is a palatal phenomenon and how tongue advancement may affect this may provide insight into its clinical effect. Physiologically, snoring represents the pharynx as a Starling resister. A Starling resister exists when there is a luminal pressure difference across a collapsible tube. When pharyngeal transmural pressure (which determines upper airway stability of the airway) is below upstream (i.e., nasal) and above downstream (i.e., tracheal) pressure, a condition of airway instability and flutter creates snoring. Changes in snoring reflect changes in airway stability, which can occur either via a stiffening in airway wall
characteristics (increases in transmural pressure) or by alteration and downstream (tracheal) negative pressures by reduction in airway resistance and the work of breathing. This lowers downstream (tracheal) pressures.11 Data from this study could not determine the cause of the change in snoring, but because the mechanical efforts of Advance are likely isolated to the tongue, it is unlikely the device changed pharyngeal transmural pressure. It is speculated that a change in snoring reflects a change in airway resistance. This study supports that tongue stabilization reduces airway obstruction and OSA severity. Although not a comparative trial, the methodology of the study reduced the potential influences of both regression to the mean and placebo effect. Patients were required to have a preexisting diagnosis of sleep apnea, to have failed medical treatment, and to have had a recent inclusion criteria polysomnogram. A separate level 1 polysomnogram was then performed as a baseline measure. In contrast to all reported previous studies, multiple follow-up sleep studies were performed at longitudinal follow-up times. The likelihood that observed changes were by chance is small. In addition, the consistent improvement of all metrics related to sleep apnea and their significant correlation to AHI supports that the findings were clinically relevant and valid. Some may argue that the observed objective AHI changes were small in magnitude. We agree that the changes in surrogate outcome measures were small. However, the changes in clinical measures were both statistically and clinically significant. Epworth Sleepiness Scale and Functional Outcomes of Sleep Questionnaire scores were normalized in the group. It may be speculated that reducing elevated upper airway resistance alone, independent of AHI, may affect sleep quality and snoring. These should not be casually dismissed, especially because the a priori goal of the study was not definitive treatment with this device. Patient selection criteria were intentionally broad and did not exclude palatal obstruction, which is likely the most common contributor to OSA. Rather, the hypothesis was that lower pharyngeal airway obstruction is also common and contributes to apnea severity. Reduction in this obstruc-
590
Otolaryngology–Head and Neck Surgery, Vol 143, No 4, October 2010
tion, therefore, should reduce this severity, which is what we observed. Because there is no validated method of identifying tongue base obstruction, we did not use any specific criteria to define lower airway obstruction. Lower airway and tongue base obstruction may likely be common, and the use of arbitrary selection criteria could skew results. Patients were surgically naïve to sleep apnea surgery, excluding tonsillectomy and nasal surgery. This study noted a remarkable reduction in both clinically relevant symptoms and objectively measured disease by only treating an isolated tongue airway segment. Those unfamiliar with upper airway structure and function might dismiss the results of the study. When evaluated not as a single procedure but as part of multiple interventions to stabilize the airway, the current findings are significant. Not only does residual obstruction affect surgical interventions, but residual obstruction occurs with the use of oral appliances and nasal CPAP. Data suggest that this residual obstruction may occur in the lower pharynx. In summary, the current study identified that the Advance System has measurable effects on AHI, sleepiness, functional outcomes of sleep, and snoring at six months. The device, implantation, and titration were well tolerated by patients; however, a high and unacceptable device tissue anchor failure rate occurred. This problem makes the current system unacceptable for current clinical use. Therefore, although demonstrating success of the concept, the Advance System as currently designed failed as a clinical device. Results, however, are important in providing data that support the concept of tongue advancement and stabilization to treat both obstructive sleep apnea and potentially even snoring. However, future modification of this or other systems must balance the anterior forces necessary to prevent pharyngeal collapse with the posterior forces necessary for deglutition.
Author Information From the Department of Otolaryngology, Medical College of Wisconsin (Dr. Woodson), Milwaukee, WI; Department of Otolaryngology, University of Cincinnati Academic Health Center (Dr. Steward), Cincinnati, OH; Advanced Ear, Nose & Throat Associates, PC (Dr. Mickelson), Atlanta, GA; Head & Neck Surgery Associates Inc. (Dr. Huntley), Carmel, IN; and University of California, San Francisco (Dr. Goldberg), San Francisco, CA. Corresponding author: B. Tucker Woodson, MD, Department of Otolaryngology, Medical College of Wisconsin, 9200 W. Wisconsin Avenue, Milwaukee, WI 53226. E-mail address: [email protected].
Author Contributions B. Tucker Woodson, contributed to design of study, collection, interpretation, and analysis of data, and writing and final approval of manuscript; David L. Steward, contributed to collection, interpretation, and analysis of data, and revision and final approval of manuscript; Samuel Mickelson, contributed to collection, interpretation, and analysis of data, and revision and final approval of manuscript; Tod Huntley, contributed to collection, interpretation, and analysis of data, and revision and final approval of manuscript; Andrew Goldberg, contributed to collection, interpretation, and analysis of data, and revision and final approval of manuscript.
Disclosures Competing interests: B. Tucker Woodson, involved with patent assignment without royalty or compensation, served on the scientific advisory board: Aspire Medical; consultant: Medtronics, Inspire Medical, ResMed, Johnson & Johnson; David L. Steward, study funding: Aspire Medical; Andrew Goldberg, stockholder: Apnicure, Inc.; consultant: Carbylan Biosurgery. Sponsorships: Aspire Medical (funding source) was involved in study design, as well as study conduct and collection and analysis of data.
References 1. Riley RW, Powell NB, Guilleminault C. Obstructive sleep apnea syndrome: a review of 306 consecutively treated surgical patients. Otolaryngol Head Neck Surg 1993;108:117–25. 2. Kezirian EJ, Goldberg AN. Hypopharygeal surgery in obstructive sleep apnea: an evidence-based medicine review. Arch Otolaryngol Head Neck Surg 2006;132:206 –13. 3. Li KK, Riley RW, Powell NB, et al. Obstructive sleep apnea surgery: genioglossus advancement revisited. J Oral Maxillofac Surg 2001;59: 1181– 4. 4. Lewis MR, Duclic Y. Genioglossus muscle advancement with the genioglossus bone advancement technique for base of tongue obstruction. J Otolaryngol 2003;32:168 –73. 5. Woodson BT, Derowe A, Hawke M, et al. Pharyngeal suspension suture with repose bone screw for obstructive sleep apnea. Otolaryngol Head Neck Surg 2000;122:395– 401. 6. Woodson BT. A tongue suspension suture for obstructive sleep apnea and snorers. Otolaryngol Head Neck Surg 2001;124:297–303. 7. DeRowe A, Gunther E, Fibbi A, et al. Tongue base suspension with a soft tissue-to-bone anchor for obstructive sleep apnea: preliminary clinical results of a new minimally invasive technique. Otolaryngol Head Neck Surg 2000;122:100 –3. 8. Sorrenti G, Piccin O, Latini G, et al. Tongue base suspension technique in obstructive sleep apnea: personal experience. Acta Otorhinolaryngol Ital 2003;23:274 – 80. 9. Vicente E, Marin JM, Carrizo S, et al. Tongue-base suspension in conjunction with uvulopalatopharyngoplasty for treatment of severe obstructive sleep apnea: long-term follow-up results. Laryngoscope 2006;116:1223–7. 10. Thomas AJ, Chavoya M, Terris DJ. Preliminary findings from a prospective, randomized trial of two tongue-base surgeries for sleepdisordered breathing. Otolaryngol Head Neck Surg 2003;129:539 – 46. 11. Oliven A, Tov N, Geitini L, et al. Effect of genioglossus contraction on pharyngeal lumen and airflow in sleep apnea patients. Eur Respir J 2007;30:748 –58.
ORIGINAL RESEARCH–SLEEP MEDICINE
Multicenter study of a novel adjustable tongue-advancement device for obstructive sleep apnea B. Tucker Woodson, MD, David L. Steward, MD, Samuel Mickelson, MD, Tod Huntley, MD, and Andrew Goldberg, MD, Milwaukee, WI; Cincinnati, OH; Atlanta, GA; Carmel, IN; and San Francisco, CA Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article. ABSTRACT OBJECTIVE: The safety and clinical effect of a new surgical device for tongue suspension for obstructive sleep apnea (OSA) was assessed. STUDY DESIGN: Multicenter phase 2 prospective case series. SETTING: Multicenter academic and private. SUBJECTS AND METHODS: Surgically naïve patients with moderate-to-severe OSA and tongue base obstruction (body mass index ⬍ 32, apnea/hypopnea index [AHI] 15-60) underwent surgical insertion of a midline tissue anchor into the posterior tongue and connected to an adjustable mandibular bone anchor with a flexible tether. Outcomes included changes in AHI, sleepiness (Epworth Sleepiness Scale), sleep-related quality-of-life (Functional Outcomes of Sleep Questionnaire), snoring, swallowing, speech, and pain (0-10 visual analog scale). RESULTS: After the implant, 42 patients (mean age 50 years, body mass index 28) noted improvement at six months for AHI (mean [SD]: 35.5 [20.4] to 27.3 [18.8]), Epworth Sleepiness Scale (11.5 [3.9] to 7.8 [4.7]), and Functional Outcomes of Sleep Questionnaire (15.5 [2.6] to 17.5 [2.6], all P ⬍ 0.01). Snoring VAS scores improved (7.3 [2.1] to 4.7 [2.9], P ⬍ 0.01). Postimplant pain scores were mild to moderate (4.4) at day one and resolved by day five. Post-titration pain scores were mild (⬍ 2). Devicerelated adverse events included wound infection (7%) and edema or seroma (5%), which resolved. However, in 31 percent of patients, asymptomatic tissue anchor barb fractures were observed radiographically. CONCLUSION: The tissue anchor failure rate of the tested device precludes its clinical use; however, the study results support that a titratable, tongue-suspension device with low direct surgical morbidity in patients with moderate-to-severe sleep apnea significantly improves multiple measures of sleep apnea. Further investigation is warranted. © 2010 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved.
O
ne treatment option for patients with obstructive sleep apnea (OSA) is upper airway surgery. Reconstructive surgery has the potential of reducing airway obstruction by
altering, supporting, or opening the upper airway. Traditional surgery for OSA may include uvulopalatopharyngoplasty for the retropalatal component of the pharyngeal collapse combined with anterior mandibular osteotomy and genioglossus advancement for the hypopharyngeal component.1 However, the perceived invasiveness of an osteotomy, postoperative pain, and dental side effects likely limit both patient and surgeon enthusiasm for this approach despite its potential benefits.2 Tongue base suspension is conceptually appealing because it may support and stabilize the hypopharyngeal airway without need for mandibular osteotomy as required for traditional genioglossus advancement.3,4 A commercially available tongue suture suspension device (Repose, Medtronic ENT, Jacksonville, FL) has shown promise in some studies5-9 but does not allow for adjustment after the initial surgery to account for individual anatomy, optimization of tongue position, or potential disease relapse. It is also associated with significant acute postoperative pain and dysphagia. The Advance System (Aspire Medical, Sunnyvale, CA) incorporates a mandibular bone anchor with an adjustable spool tethered to a tongue base tissue anchor implanted into the midline posterior tongue muscle just deep to the mucosal surface (Fig 1). The bone and tissue anchors are placed via a 2-cm submental incision during a first-stage implantation procedure with only the slack removed from the tether line. In the second stage, a titration procedure is performed two to four weeks after implantation. Tension on the tissue anchor is adjusted via a small titration needle, which engages the spool on the bone anchor, advancing the tongue anteriorly or relaxing it to fall more posteriorly. This is done percutaneously or through the same submental incision under local anesthesia. It is widely accepted that multiple structures of the upper airway contribute to sleep apnea and that successful treatment often requires addressing multiple structures to obtain an acceptable reduction in disease severity. Although a small number of patients may benefit from isolated treat-
Received August 13, 2009; revised April 16, 2010; accepted June 10, 2010.
0194-5998/$36.00 © 2010 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved. doi:10.1016/j.otohns.2010.06.902
586
Otolaryngology–Head and Neck Surgery, Vol 143, No 4, October 2010
Figure 1 Advance System for tongue suspension and advancement is depicted. (A) System components, including sheath, trochar, insertion device, and adjustment needle shown. (B) Tissue and bone anchors are shown, enlarged.
ment of individual structures, there is no demonstrated or even widely accepted method of selecting these patients or procedures. To achieve a definitive clinical outcome, therefore, multiple surgical procedures are likely required. Multiple procedures, however, confound attempts to assess the contribution of individual procedures. An alternative approach to assess surgical procedures is to measure whether procedures are effective rather than curative. Procedures, especially when they are of low invasiveness and morbidity, may be combined in a stepwise algorithm until an adequate clinical effect has been achieved. With this concept in mind, we designed a multicenter prospective phase 2 study to assess the safety, feasibility, and therapeutic effectiveness of the Advance System for patients with moderate-to-severe OSA.
Methods Study Design This is a prospective, multicenter phase 2 study.
Human Subjects This study was approved by the local institutional review boards at the four study sites and performed in compliance with the institutional review board regulations, good clinical practice guidelines, and Health Insurance Portability and Accountability Act. Subjects were recruited from private and academic otolaryngology office settings and from newspaper advertisements. The study was monitored by a data safety monitoring board, and the study was conducted and monitored in compliance with 21 Code of Federal Regulations parts 50, 54, 56, and 812 and the 18th World Medical Assembly, Helsinki, Finland, 1964, and later revisions.
Inclusion/Exclusion Criteria Adult patients (18-65 years) with moderate-to-severe OSA (apnea hypopnea index [AHI] ⫽ 15-60) were included 1) on the basis of diagnostic polysomnography within six months of enrollment or on repeat baseline sleep studies with the use of WatchPat (Itimar Medical, Haifa, Israel); 2) clinical findings consistent with tongue-related hypopharyngeal up-
per airway obstruction (investigator clinical impression on the basis of physical examination and upper airway endoscopy); or 3) failure or refusal of medical treatment with nasal continuous positive airway pressure (CPAP). Exclusion criteria included 1) major obesity (body mass index [BMI] ⬍ 32 kg/m2), 2) previous upper airway surgery exclusive of tonsillectomy, adenoidectomy, or nasal surgery, 3) enlarged tonsils (3⫹ and 4⫹); 4) anatomically unable to accommodate implant; 5) severe mandibular deficiency; 6) severe nasal obstruction; 7) systemic infection; 8) history of any of the following: head and neck or respiratory tract cancer, radiation therapy to head and neck, dysphagia, major cardiovascular disease, major pulmonary disease including chronic obstructive pulmonary disease, inadequately treated depression, or falling asleep while driving or motor vehicle accidents secondary to sleepiness; or 9) anesthesia class (American Society of Anesthesiologists) IV or V. Baseline and six-month full night diagnostic polysomnography were performed separately from studies used for inclusion/exclusion and included electroencephalography, electrooculography, chin and leg muscle electromyography, electrocardiography, measures of oronasal airflow, thoracic and abdominal efforts, body position, and pulse oximetry. Each site used the same sleep laboratory facility for all studies. Apnea was defined as cessation of inspiratory airflow of at least 10 seconds. Hypopnea was defined as a 30 percent reduction of inspiratory airflow of at least 10 seconds, with an associated four percent decrease in oxyhemoglobin saturation. Home sleep studies (WatchPat; Itimar Medical, Haifa, Israel) were performed at multiple time points, including screening, after implantation, after titration, and at three months’ follow-up.
Implant Procedure/Advance System The Advance System includes a trochar through which a tissue anchor composed of a nickel titanium memory metal and an attached flexible tether is inserted and a bone anchor with an adjustable spool that allows changes in tether length transcutaneously by the use of a small needle-like adjustment tool. For this study, the device was placed under general anesthesia by the use of nasal intubation. Patients were prepped for a chin procedure, and a 2-cm transverse incision was used to access the posterior mentum. The trochar was inserted into the genioglossus muscle near the mandible and advanced to the posterior margin of the tongue. The tissue anchor was deposited within the tongue. Position was confirmed by palpation or, in some patients, by the use of fluoroscopy. The bone anchor was fixed to the inferoposterior aspect of the mandible at the midline by the use of three self-tapping bone screws. The tissue anchor was attached to the bone anchor via the tether line, and the bone anchor spool was adjusted by the physician to remove slack from the tether line and suspend the tongue, often under fluoroscopy. Patients were observed overnight in the hospital. Preoperative and seven days of postoperative antibiotics
Woodson et al
Multicenter study of a novel adjustable . . .
587
(amoxicillin or clindamycin) and analgesics were prescribed. Two to four weeks after the procedure, titration was performed in the operating room under local anesthesia with or without sedation according to the surgeon and patient preference. A small incision was made on the inferior surface of the chin proximate to the bone anchor. The adjustment tool was then used to access the spool adjustment mechanism on the bone anchor. The tongue base tissue anchor was advanced anteriorly via the tether line by use of the spool and titrated to speech and swallowing tolerance and/or fluoroscopically visualized tongue base location relative to the posterior pharyngeal wall. In two patients, a second titration procedure was performed. The reported measured advancement was arbitrarily defined as the distance of tether shortening.
Results Forty-two subjects (4 female, 38 male) were enrolled and implanted. No patient was lost to follow-up. Patient demographics are shown in Table 1. On average, subjects were predominately middle-aged, overweight men with moderate-to-severe sleep apnea and excessive daytime sleepiness. The vast majority (⬎ 90%) had modified Mallampati class II or III oropharyngeal airways (free edge of palate not visible without phonation or tongue depressor use) and absent or 1⫹ tonsil size. Mean surgical implant time was 43 minutes (SD, 20 minutes), and mean surgical titration time was 14 minutes (SD, 8 minutes). The mean amount of tongue advancement was 1.3 cm (SD, 0.3 cm). After implantation and subsequent titration, statistically significant improvements in AHI (Fig 2), as well as Epworth Sleepiness Scale and Functional Outcomes of Sleep Questionnaire were noted at three and six months after the procedure (Table 2). Snoring VAS scores were also significantly improved at three and six months after titration. Implant procedure side effects and morbidity were studied. Pain on postimplant day one was moderate (4.4 on a 0-10 visual analog scale) and then decreased steadily through day five. Post-titration procedure pain was mild
Table 1 Baseline demographic data Advance System (n ⫽ 42) Age (yrs) % Male BMI, kg/m2 AHI, events/h ESS FOSQ
50.1 (9.1) 90 27.8 (2.1) 35.5 (20.4) 11.5 (3.9) 15.5 (2.6)
Values are mean (SD) or proportion of subjects. BMI, body mass index; AHI, apnea/hypopnea index; ESS, Epworth Sleepiness Scale (⬍ 10 is normal); FOSQ, Functional Outcomes of Sleep Questionnaire (⬎ 17 is normal).
Figure 2 Individual AHI results, including independent postinclusion baseline AHI, are shown. Screening to baseline AHI did not demonstrate a change and was not significantly different (P ⫽ NS). Baseline to six months was significantly reduced (P ⬍ 0.01).
(⬍ 2.0). Pain scores at three and six months were negligible (Fig 3). Swallowing, speech, taste, and throat irritation scores demonstrated mild and transient change and are shown in Figure 4. Complications were significant. In one patient, failure of the bone anchor locking mechanism prevented titration, and one bone anchor was repositioned under local anesthesia because of surgical error and placement off midline. Tongue base tissue anchor barb fracture occurred in 13 patients (31%). Fractures were not observed perioperatively or immediately after implantation or titration but were identified on radiographs taken at three, six, or 12 months (Fig 5). Other device-related complications were transient and included wound infection or cellulitis in seven percent, who required a longer than protocol course of antibiotics, and edema/seroma in five percent. No device was removed, but after device fractures were identified, implantation of additional implants were halted.
588
Otolaryngology–Head and Neck Surgery, Vol 143, No 4, October 2010
Table 2 Baseline and postoperative outcome measures Screening AHI ESS FOSQ Snore VAS
36.7 (14.1)
Baseline 35.5 11.5 15.5 7.3
(20.4) (3.9) (2.6) (2.1)
Three-month
7.5 (4.9) 17.9 (2.1) 3.6 (2.8)
Six-month
P value
27.3 7.8 17.5 4.7
⬍0.002 ⬍0.001 ⬍0.001 ⬍0.001
(18.8) (4.7) (2.6) (2.9)
Values are mean (SD). AHI, apnea/hypopnea index; ESS, Epworth Sleepiness Scale; FOSQ, Functional Outcomes of Sleep Questionnaire; VAS, visual analog scale (0-10).
Discussion The current study was designed to demonstrate the safety and therapeutic effect of the Advance System for patients with evidence of hypopharyngeal tongue base obstruction and moderate-to-severe OSA syndrome. However, the procedure was complicated by unexpected device tissue anchor failures in 13 of 42 patients. Failures were asymptomatic and identified during radiographic protocol monitoring. As a result, all subjects were offered the option of device removal at no cost. No patient with or without device fracture or malfunction elected for device removal during the course of the study. Failure involved fracture of one of eight nitinol tissue barbs on the anchor (one patient with two barb fractures) and occurred between the anchor hub and the tissue barb consistent with excess loading forces at this point (Fig 5). The exact timing of the fractures are not known but were not observed at the first postimplant and pretitration radiograph. They were identified between one and three months. At six-month follow-up, no additional fractures were observed. Future tongue suspension or advancement device development will need to take into account the significant forces generated on a fixed point in the posterior superior tongue during swallowing estimated to exceed five pounds. Surgical treatment of the tongue base and hypopharynx remains a significant surgical challenge in the treatment of OSA.2 Although multiple procedures exist to treat this segment, few are widely used. Tongue suspension has been described as an alternative to maxillofacial surgical techniques, with the goal of supporting the tongue and preventing retroglossal airway collapse with techniques of lower morbidity and complexity.5-9 Conceptually, tongue suspension potentially is less morbid and may be effective by
Figure 3
Post-implant and titration pain: visual analog scores.
dilating the upper airway by applying a vector of force on the soft tissue surrounding the airway lumen. This idea is in contrast to nasal CPAP, which dilates the airway by applying a vector of force to the luminal airway surface. The force to treat the airway in many patients does not exceed common pressures of 5 to 15 cm/H2O used with CPAP. However, the forces exerted during other movements of the tongue, such as swallowing, may be vastly different. It is likely these forces that cause failure not those to treat disease. Previous methods of tongue suspension have used many different materials to support the soft tissue, including suture and fascia. With these, tongue tissue is triangulated, and the suspension material is anchored to the mandible with screw or wire. Data suggest that suspension improves clinical symptoms and may reduce clinical severity of disease and that morbidity may be less than with traditional limited mandibular osteotomy.5-10 However, current tongue suspension methods cannot be easily adjusted and still have significant perioperative morbidity with pain and swallowing dysfunction. The Advance System was designed to incorporate the following goals: minimally invasive, adjustable, and revers-
Figure 4 Post-implant and titration swallowing, speech, taste, and throat irritation visual analog scores.
Woodson et al
Multicenter study of a novel adjustable . . .
589
Figure 5 Post-implant (A) and post-titration (B) lateral cephalometric x-rays are shown. Device fracture is marked by arrow in post-titration x-ray.
ible. The design tested in this study met these objectives. The study used general anesthesia per protocol, but implantation of the device is a rapid minor procedure that potentially may be performed under local anesthesia. The device was readily titratable under local anesthesia with little or no sedation required and was well tolerated by patients. By allowing several weeks for tissues to adjust to the placement of the tissue and bone anchor, subsequent adjustments were performed with little acute changes in speech, swallowing, taste, or throat irritation with no significant changes in these measures from baseline or pretitration levels. Transient dysarthria was noted by some patients manifesting with pronunciation of “cks” or “x” sounds, which require tongue base retraction for correct pronunciation. Tongue advancement under straight local anesthesia permitted titration based up until or just before detection of subtle changes in these sounds. The amount of titration on average was 1.3 cm, measured on the basis of the amount of tether line shorted during titration. The ideal amount of titration to maximize effect and minimize side effects or risk of device failure was not determined by this study; however, a significant clinical effect was observed with this change. In addition to being of low surgical morbidity and being titratable, the device can be removed by the use of a specially designed removal system. Removal is important in the potential of infection, patient discomfort or concerns, and because the long-term effects of a tongue implant are unknown. Of interest, the device not only reduced AHI, a global metric of sleep apnea, but also snoring. A common perception is that snoring is a palatal phenomenon and how tongue advancement may affect this may provide insight into its clinical effect. Physiologically, snoring represents the pharynx as a Starling resister. A Starling resister exists when there is a luminal pressure difference across a collapsible tube. When pharyngeal transmural pressure (which determines upper airway stability of the airway) is below upstream (i.e., nasal) and above downstream (i.e., tracheal) pressure, a condition of airway instability and flutter creates snoring. Changes in snoring reflect changes in airway stability, which can occur either via a stiffening in airway wall
characteristics (increases in transmural pressure) or by alteration and downstream (tracheal) negative pressures by reduction in airway resistance and the work of breathing. This lowers downstream (tracheal) pressures.11 Data from this study could not determine the cause of the change in snoring, but because the mechanical efforts of Advance are likely isolated to the tongue, it is unlikely the device changed pharyngeal transmural pressure. It is speculated that a change in snoring reflects a change in airway resistance. This study supports that tongue stabilization reduces airway obstruction and OSA severity. Although not a comparative trial, the methodology of the study reduced the potential influences of both regression to the mean and placebo effect. Patients were required to have a preexisting diagnosis of sleep apnea, to have failed medical treatment, and to have had a recent inclusion criteria polysomnogram. A separate level 1 polysomnogram was then performed as a baseline measure. In contrast to all reported previous studies, multiple follow-up sleep studies were performed at longitudinal follow-up times. The likelihood that observed changes were by chance is small. In addition, the consistent improvement of all metrics related to sleep apnea and their significant correlation to AHI supports that the findings were clinically relevant and valid. Some may argue that the observed objective AHI changes were small in magnitude. We agree that the changes in surrogate outcome measures were small. However, the changes in clinical measures were both statistically and clinically significant. Epworth Sleepiness Scale and Functional Outcomes of Sleep Questionnaire scores were normalized in the group. It may be speculated that reducing elevated upper airway resistance alone, independent of AHI, may affect sleep quality and snoring. These should not be casually dismissed, especially because the a priori goal of the study was not definitive treatment with this device. Patient selection criteria were intentionally broad and did not exclude palatal obstruction, which is likely the most common contributor to OSA. Rather, the hypothesis was that lower pharyngeal airway obstruction is also common and contributes to apnea severity. Reduction in this obstruc-
590
Otolaryngology–Head and Neck Surgery, Vol 143, No 4, October 2010
tion, therefore, should reduce this severity, which is what we observed. Because there is no validated method of identifying tongue base obstruction, we did not use any specific criteria to define lower airway obstruction. Lower airway and tongue base obstruction may likely be common, and the use of arbitrary selection criteria could skew results. Patients were surgically naïve to sleep apnea surgery, excluding tonsillectomy and nasal surgery. This study noted a remarkable reduction in both clinically relevant symptoms and objectively measured disease by only treating an isolated tongue airway segment. Those unfamiliar with upper airway structure and function might dismiss the results of the study. When evaluated not as a single procedure but as part of multiple interventions to stabilize the airway, the current findings are significant. Not only does residual obstruction affect surgical interventions, but residual obstruction occurs with the use of oral appliances and nasal CPAP. Data suggest that this residual obstruction may occur in the lower pharynx. In summary, the current study identified that the Advance System has measurable effects on AHI, sleepiness, functional outcomes of sleep, and snoring at six months. The device, implantation, and titration were well tolerated by patients; however, a high and unacceptable device tissue anchor failure rate occurred. This problem makes the current system unacceptable for current clinical use. Therefore, although demonstrating success of the concept, the Advance System as currently designed failed as a clinical device. Results, however, are important in providing data that support the concept of tongue advancement and stabilization to treat both obstructive sleep apnea and potentially even snoring. However, future modification of this or other systems must balance the anterior forces necessary to prevent pharyngeal collapse with the posterior forces necessary for deglutition.
Author Information From the Department of Otolaryngology, Medical College of Wisconsin (Dr. Woodson), Milwaukee, WI; Department of Otolaryngology, University of Cincinnati Academic Health Center (Dr. Steward), Cincinnati, OH; Advanced Ear, Nose & Throat Associates, PC (Dr. Mickelson), Atlanta, GA; Head & Neck Surgery Associates Inc. (Dr. Huntley), Carmel, IN; and University of California, San Francisco (Dr. Goldberg), San Francisco, CA. Corresponding author: B. Tucker Woodson, MD, Department of Otolaryngology, Medical College of Wisconsin, 9200 W. Wisconsin Avenue, Milwaukee, WI 53226. E-mail address: [email protected].
Author Contributions B. Tucker Woodson, contributed to design of study, collection, interpretation, and analysis of data, and writing and final approval of manuscript; David L. Steward, contributed to collection, interpretation, and analysis of data, and revision and final approval of manuscript; Samuel Mickelson, contributed to collection, interpretation, and analysis of data, and revision and final approval of manuscript; Tod Huntley, contributed to collection, interpretation, and analysis of data, and revision and final approval of manuscript; Andrew Goldberg, contributed to collection, interpretation, and analysis of data, and revision and final approval of manuscript.
Disclosures Competing interests: B. Tucker Woodson, involved with patent assignment without royalty or compensation, served on the scientific advisory board: Aspire Medical; consultant: Medtronics, Inspire Medical, ResMed, Johnson & Johnson; David L. Steward, study funding: Aspire Medical; Andrew Goldberg, stockholder: Apnicure, Inc.; consultant: Carbylan Biosurgery. Sponsorships: Aspire Medical (funding source) was involved in study design, as well as study conduct and collection and analysis of data.
References 1. Riley RW, Powell NB, Guilleminault C. Obstructive sleep apnea syndrome: a review of 306 consecutively treated surgical patients. Otolaryngol Head Neck Surg 1993;108:117–25. 2. Kezirian EJ, Goldberg AN. Hypopharygeal surgery in obstructive sleep apnea: an evidence-based medicine review. Arch Otolaryngol Head Neck Surg 2006;132:206 –13. 3. Li KK, Riley RW, Powell NB, et al. Obstructive sleep apnea surgery: genioglossus advancement revisited. J Oral Maxillofac Surg 2001;59: 1181– 4. 4. Lewis MR, Duclic Y. Genioglossus muscle advancement with the genioglossus bone advancement technique for base of tongue obstruction. J Otolaryngol 2003;32:168 –73. 5. Woodson BT, Derowe A, Hawke M, et al. Pharyngeal suspension suture with repose bone screw for obstructive sleep apnea. Otolaryngol Head Neck Surg 2000;122:395– 401. 6. Woodson BT. A tongue suspension suture for obstructive sleep apnea and snorers. Otolaryngol Head Neck Surg 2001;124:297–303. 7. DeRowe A, Gunther E, Fibbi A, et al. Tongue base suspension with a soft tissue-to-bone anchor for obstructive sleep apnea: preliminary clinical results of a new minimally invasive technique. Otolaryngol Head Neck Surg 2000;122:100 –3. 8. Sorrenti G, Piccin O, Latini G, et al. Tongue base suspension technique in obstructive sleep apnea: personal experience. Acta Otorhinolaryngol Ital 2003;23:274 – 80. 9. Vicente E, Marin JM, Carrizo S, et al. Tongue-base suspension in conjunction with uvulopalatopharyngoplasty for treatment of severe obstructive sleep apnea: long-term follow-up results. Laryngoscope 2006;116:1223–7. 10. Thomas AJ, Chavoya M, Terris DJ. Preliminary findings from a prospective, randomized trial of two tongue-base surgeries for sleepdisordered breathing. Otolaryngol Head Neck Surg 2003;129:539 – 46. 11. Oliven A, Tov N, Geitini L, et al. Effect of genioglossus contraction on pharyngeal lumen and airflow in sleep apnea patients. Eur Respir J 2007;30:748 –58.