Customize your JAMA Network experience by selecting one or more topics from the list below.
Vilaseca I, Morelló A, Montserrat JM, Santamaría J, Iranzo A. Usefulness of Uvulopalatopharyngoplasty With Genioglossus and Hyoid Advancement in the Treatment of Obstructive Sleep Apnea. Arch Otolaryngol Head Neck Surg. 2002;128(4):435–440. doi:10.1001/archotol.128.4.435
Copyright 2002 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2002
To evaluate the usefulness of uvulopalatopharyngoplasty plus mandibular osteotomy with genioglossus and hyoid advancement in the treatment of obstructive sleep apnea syndrome (OSAS).
Prospective study of 20 consecutive patients with OSAS.
University medical center.
Patients and Interventions
Twenty OSAS patients with multilevel upper airway obstruction who refused continuous positive airway pressure treatment. All patients were evaluated before and 6 months after surgery by clinical history, the Epworth Sleepiness Scale, physical examination, fiberoptic nasopharyngoscopy combined with the Müller maneuver, cephalometric analysis, nocturnal polysomnography, and a second-night polysomnography with upper airway pressure recording during sleep. Surgery procedures were uvulopalatopharyngoplasty plus mandibular osteotomy with genioglossus and hyoid advancement. Surgical successful outcome was defined as an apnea-hypopnea index (AHI) lower than 20 plus subjective resolution of daytime symptoms.
Main Outcome Measure
Surgical success rate.
Mean ± SD AHI decreased from 60.5 ± 16.5 to 44.6 ± 27 (P = .007), and CT90 (percentage of time with oxyhemoglobin saturation below 90%) decreased from 39.5% ± 26% to 25.1% ± 26.4% (P = .002). The overall surgical success rate was 35% but increased to 57% in patients with moderate OSAS (AHI, 41-60) and to 100% in mild OSAS (AHI, 21-40). In the group of severe OSAS, the success rate was 9%. Predictors of surgical outcome success were the AHI, CT90, stages 2 and 3-4 sleep percentages, and the cephalometric ANB angle (angle formed from the deepest point on the maxillary outer contour to the nasion to the deepest point on the outer mandibular contour).
Patients with mild and moderate OSAS and multilevel obstruction in the upper airway may benefit from uvulopalatopharyngoplasty plus genioglossus and hyoid advancement.
OBSTRUCTIVE sleep apnea syndrome (OSAS) is caused by repetitive occlusion of the upper airway during sleep and may be treated with continuous positive airway pressure (CPAP) or with several surgical techniques. Surgery is used to correct any anatomic abnormality that potentially narrows the upper airway, which may be disclosed clinically or by other means (eg, cephalometry and fiberoptic nasopharyngoscopy).1
Uvulopalatopharyngoplasty (UPPP) is used to eliminate the upper airway obstruction selectively at the level of the oropharynx by removing a portion of the soft palate and the uvula; it is effective in 33% to 77% of patients.2 The association of UPPP plus mandibular osteotomy with genioglossus and hyoid advancement (multi-level reconstruction phase 1) was introduced by Riley et al3 as a surgical alternative to isolated UPPP in patients with multilevel obstruction. When this procedure fails, maxillomandibular advancement (multi-level reconstruction phase 2) may follow phase 1. This surgical strategy was adopted by several centers maintaining the gradual 2-phase surgical procedure proposed by Riley et al,3 who showed that the multi-level reconstruction phase 2 success rate was similar to CPAP efficacy. However, the usefulness of multi-level phase 1 reconstruction remains unclear because apart from those by the Stanford group,3 the studies published are few, with small numbers of patients and varying rates of success.4-10 The aim of our study was to evaluate the usefulness of a multilevel phase 1 surgical procedure in a group of OSAS patients with multilevel obstruction.
We prospectively observed 20 consecutive OSAS patients with UPPP failure and multilevel obstruction in the upper airway. Inclusion criteria were (1) persistent symptoms suggestive of OSAS and post-UPPP apnea-hypopnea index (AHI) (total number of apneas and hypopneas per hour of sleep) greater than 20, as determined by polysomnography (PSG); (2) obstruction in both the oropharynx and the hypopharynx, and (3) CPAP refusal or intolerance. Exclusion criteria were (1) age older than 65 years, (2) chronic pulmonary disease and/or (3) poor health condition. Presurgical evaluation included clinical history, Epworth Sleepiness Scale (ESS) evaluation,11 physical examination, fiberoptic pharyngolaryngoscopy, cephalometry, nocturnal PSG, and second-night PSG with upper airway pressure measurement during sleep.
In all patients, the surgical procedure included UPPP and hyoid advancement. Inferior mandibular osteotomy with genioglossus advancement was also performed in 11 subjects according to cephalometric criteria of mandibular deficiency. To evaluate the effectiveness of the surgical procedure, we repeated the presurgical evaluation 6 months after surgery. A procedure was considered successful (and the patient, a responder) if the postoperative PSG demonstrated an AHI lower than 20 and the patient reported significant clinical improvement. This study was approved by the ethics committee at our institution and written informed consent was obtained from each patient.
Patients were evaluated by a comprehensive clinical history that covered their sleep habits and the occurrence of sleep disturbances. Excessive daytime sleepiness was estimated by the ESS: a score greater than 10 was considered indicative of hypersomnia.11 Physical examination included measurement of the body mass index (BMI) and neck circumference and evaluation of the anatomic characteristics and abnormalities of the upper airway. A combination of nasopharyngolaryngoscopy with the Müller maneuver12 was used to evaluate the upper airway compliance.
Lateral cephalometric radiographs were performed according to standard procedure.13 Before the exposure, patients swallowed contrast to outline the pharynx soft structures. We determined and evaluated the SNA angle (maxilla to cranial base, position of the maxilla), SNB angle (mandible to cranial base, postion of the mandible), PAS (posterior airway space, distance between the base of the tongue and the dorsal pharyngeal wall), PNS-P (distance from the posterior nasal spine to the tip of the soft palate, length of soft palate), point A (deepest point on the maxillary outer contour), point B (deepest point on the outer mandibular contour), point G (palate thickness), point P (inferior tip of the palate), ANB angle (angle formed from point A to nasion to point B), and MP-H (distance from inferior mandible plane to hyoid bone; position of the hyoid bone).14,15 Palatal obstruction was defined when cephalometric analysis showed a PNS-P greater than 40 mm.16 Hypopharyngeal obstruction was defined when cephalometric analysis showed mandibular deficiency (SNB≤77°) and narrowing of the airway space at the base of the tongue (PAS≤10 mm).17 After surgery, mandibular advancement was calculated according to the depth of the mandibular osteotomy.
All night PSG studies included 4 electroencephalogram leads (C3-A1, C4-A2, O1-A1, O2-A2), electrooculogram, chin and left and right anterior tibial surface electromyogram, electrocardiogram, nasal and oral airflow measure (thermistors), thoracic and abdominal movement determination, and continuous oxyhemoglobin saturation evaluation by a pulse oximeter. Sleep stages were scored according to criteria set out by Rechtschaffen and Kales.18 Obstructive apneas were defined as the absence of airflow with respiratory effort for at least 10 seconds. Hypopneas were defined by a greater than 50% reduction of airflow accompanied by an oxygen desaturation of more than 4 points or occurrence of an arousal from sleep. The percentage of time with oxyhemoglobin saturation below 90% (CT90) and lowest saturation level were also calculated.
Pressure recordings at the velopharynx, oropharynx, and hypopharynx were evaluated during the PSG studies before and after surgery using the esophagus pressure as reference. We used a 3-lumen catheter adapted to an esophageal balloon that was connected to a pressure transducer. The right upper airway catheter placement was checked under visual inspection using catheter marks. The detection of changes and cessation of the signal during an obstructive apnea allowed us to detect the precise location of the collapse in the pharynx. When no obstruction was demonstrated, all registers were parallel to the esophagus signal.
All 20 patients had experienced UPPP failures. A second surgical phase included hyoid advancement in all 20 patients and mandibular osteotomy with genioglossus advancement in the 11 patients in whom the SNB angle was less than 78°, indicating a retroposition of the mandible.
The UPPP was performed as previously described by Simmons et al.19 Hyoid advancement technique differed from the original description by Riley et al20 in that we advanced the hyoid bone to an anterior and downward direction over to the thyroid laminae where it was fixed with Gore-Tex (WL Gore & Associates, Flagstaff, Ariz) pushing the tongue ahead and consequently increasing the PAS. The mandibular osteotomy with genioglossus advancement procedure was performed as described by Riley et al3 advancing the tongue at the genioid tubercle of the mandible to relieve the obstruction at the hypopharyngeal level. Surgery was performed under general anesthesia, and during the 24 hours following surgery patients were monitored in the intensive care unit.
Data are presented as mean ± SD. The Wilcoxon test was used to compare the preoperative and postoperative results. The Mann-Whitney test was used to analyze the differences between responders and nonresponders. Differences were also tested with analysis of covariance, with disease severity as covariable. Data were analyzed using SPSS, Windows version 6.1.3 (SPSS Inc, Chicago, Ill), establishing P<.05 as statistically significant.
The main clinical, polysomnographic, and cephalometric characteristics before and after surgery are summarized in Table 1. All 20 patients were men with a mean ± SD age of 44.7 ± 5.7 years (range, 34-58 years). The BMI was 27.8 ± 3.3 kg/m2, and neck circumference, 40.9 ± 2.1 cm. After 6 months of follow-up, the BMI and neck circumference showed no significant differences.
All 20 patients were habitual and heavy snorers and reported daytime fatigue and somnolence; 16 (80%) felt not refreshed on awakening. Postoperatively, snoring was eliminated in 15 patients (75%), decreased in 5 (25%), and early morning fatigue persisted in 6 (30%).
Before surgery, the ESS score was greater than 10 in 11 subjects, and following treatment the score improved in all but 1 (Table 2). The ESS score improved from 12 ± 5.8 (range, 0-24) to 7.9 ± 5 (range, 0-18) but this difference was not statistically significant (P = .64).
Mean baseline PSG parameters were characteristic of severe OSAS, showing a high AHI (60.5 ± 16.5), severe oxygen desaturation (mean CT90, 39.6% ± 26.0%; lowest mean saturation, 68.3% ± 10.1%), and sleep fragmentation induced by a high number of arousals and brief awakenings. Total sleep time and sleep efficiency did not differ between preoperative and postoperative studies. There was a statistically significant reduction in the AHI (P<.001), with a trend toward a decrease in the apnea index and an increase in the hypopnea index. Postoperative studies showed a reduction in the AHI in 16 patients (80%), with an AHI lower than 20 in 7, and 50% or more AHI reduction in 8 (Table 2). In 4 patients, the postoperative AHI was slightly higher than before surgery. Oximetric analysis showed a significant decrease in CT90 (P = .002) and increase in lowest oxyhemoglobin saturation (P = .001) between preoperative and postoperative studies. After surgery, sleep architecture improved, showing a significant decrease in stage 2 sleep (P = .01) and a tendency toward an increase in slow-wave sleep (P = .06) and REM (rapid eye movement) sleep (P = .07).
Before surgery, cephalometric analysis showed long and thick soft palates, narrowing in the PAS behind the tongue, mild mandibular and maxillar deficiency, and lowered position of the hyoid bone. Eleven patients (55%) had mandibular deficiency (SNB<78°).
There were significant differences between preoperative and postoperative findings in the palate parameters PNS-P and point G, the hyoid bone position MP-H, and the SNB angle. Seventeen patients (85%) experienced significant increase in MP-H from 35.3 ± 9.7 mm to 42.5 ± 8.7 mm (P = .005). In those with mandibular osteotomy with genioglossus advancement, the mean depth of osteotomy was 11.5 ± 2.4 mm, resulting in a significant SNB angle increase of 9.4° ± 2.1°. Preoperative PAS was less than 11 mm in 7 subjects (35%), and after surgery increased from 10.9 ± 4.1 mm to 12.8 ± 3.8 mm (P = .07) and enlarged in 14 patients (70%). The 6 subjects who had no increase in their PAS had undergone mandibular osteotomy with genioglossus advancement, and only 1 of them had a preoperative PAS smaller than 11 mm (Table 3).
Preoperative pressures of the upper airway during sleep showed multilevel obstruction in all patients, although a reduced sleep efficiency related to discomfort was disclosed in 2. Postoperative pressure could not be assessed in 8 patients because they refused the procedure. Complete preoperative and postoperative studies to measure pressures were done and well tolerated in 10 nonresponders. In these 10 patients, multilevel obstruction was still demonstrated after surgery.
There was no correlation between postoperative AHI changes and postoperative variations in BMI, neck circumference, ESS score, sleep stage percentages, and the cephalographic parameters PNS-P, point G, PAS, and MP-H. Changes in AHI were correlated with postoperative changes in ANB (P = .01) and SNB (P = .01).
Seven patients (35%) met our polysomnographic and clinical criteria for surgical success. Statistical analysis identified AHI, CT90, stages 2 and 3-4 percentages (sleep quality), and the ANB angle as predictors of surgical outcome. Before surgery, the successful group had significantly fewer apneas and oxyhemoglobin desaturations, lower stage 2 sleep percentage, higher slow-wave sleep percentages, and lower maxillary/mandible discrepancy. Patients with severe OSAS (AHI, >60) had a low incidence of success. Surgical outcome rate was higher in patients with moderate (AHI, 41-60) and mild (AHI, 21-40) OSAS:
Patients who were not successfully treated tended to have bigger neck circumferences (P = .09) and higher ESS scores (P = .08) than responders. There were no statistical differences between the 2 groups in BMI, age, and all the cephalometric parameters except ANB angle (Table 4).
None of the patients in this series had postoperative bleeding, infection, difficulty swallowing, persistent rhinolalia and nasal regurgitation, or significant pain. Two days after surgery, 1 patient with a preoperative history of coronary heart disease had an angina pectoris, which resolved with adequate medical treatment.
This study showed that UPPP plus mandibular osteotomy with genioglossus and hyoid advancement was effective in 35% of OSAS patients with multilevel obstruction. Patients in the successful group reported elimination of snoring and were free of daytime symptoms. Treatment success predictors were AHI, oxyhemoglobin saturation, sleep architecture, and the ANB angle. The highest treatment effect was obtained in patients with mild to moderate AHI, no significant desaturations, no sleep disruption, and slight mandibular deficiency.
Our success rate is similar to that in 2 previous publications, which showed a success rate between 27% and 42%,8,9 but is lower than that of the Stanford group3 and other researchers who reported a 50% to 78% rate of success.4-7,10 There are several possible explanations for these differences in results. First, our patients might have changed weight during the 6 months after surgery and subsequently increased their AHIs and CT90s. However, this speculation was not confirmed; statistical analysis showed that there were no differences between BMI and neck circumference before and 6 months after surgery.
Second, our patients had moderately severe OSAS with a mean AHI of 60, while other studies selected patients with lower AHIs (range, 45-58).4,5,7,8 Because success rate is predicted by the AHI, it can be speculated that our surgical success would have been higher if we had selected patients with lower severity.
Third, our definition of surgical success included not only an AHI lower than 20, but also daytime clinical resolution. Because some studies considered an AHI lower than 20 as the sole indicative parameter of surgical success, some of our patients with AHIs lower than 20 but without complete disappearance of daytime symptoms would have been classified as responders. However, all patients with remaining postsurgery symptoms in our study also showed AHIs higher than 20 and hypersomnia, and consequently were classified in the failure group. Furthermore, if our definition of success had included a 50% reduction in the AHI,4 then the success rate of our study would have increased to 40%.
Fourth, we found a postoperative reduction in the number of apneas at the expense of an increase in the number of hypopneas. This finding was also observed by Bettega et al.8 Because the definition criteria for a hypopnea and the methods used to register the oral and nasal airflow may differ depending on the sleep laboratory,21 a meticulous measurement of respiratory events is necessary to avoid underestimation of residual hypopneas after surgery.
Finally, our surgical protocol differed slightly from that of previous studies,3 and the procedures were not the same for all the patients in our series. For example, we performed hyothyropexia instead of hyoid suspension, and genioglossus advancement was only performed in 11 patients (55%). However, our cephalometric analysis showed significant differences (as have previous publications4,9,10) in most of the parameters between preoperative and postoperative studies, suggesting a significant effect of our surgical protocol in the anatomic structures responsible for the obstruction during sleep. Six of the 7 patients in the successful group showed an increase in PAS. Furthermore, postoperative changes in AHI were significantly correlated with postoperative changes in ANB and SNB angles. However, while all patients with isolated hyoid advancement experienced increased PAS, 6 (55%) of 11 patients with genioglossus plus hyoid advancement did not, suggesting that the combination of genioglossus and hyoid advancement may not always reduce the hypopharyngeal obstruction (Table 3).
In all of our patients, clinical examination, fiberoptic nasopharyngoscopy, cephalometry, and upper airway manometry during sleep disclosed preoperative obstruction in both the soft palate and the base of the tongue. Since all patients were symptomatic and rejected CPAP, the multilevel surgery approach was proposed and accepted. The procedure was designed to remove the obstruction in the oropharynx by UPPP and increase the size of the airway at the hypopharnyx by mandibular osteotomy with genioglossus advancement and hyoid advancement. After 6 months of follow-up, 66% of patients with preoperative mild to moderate sleep apnea (mean AHI, 46) had successful outcome, while most subjects with severe sleep apnea (mean AHI, 68) were still symptomatic and had an AHI higher than 20, continued desaturations, altered sleep architecture, and remnant oropharyngeal and hypopharyngeal obstruction demonstrated by endoscopy and upper airway manometry recording during sleep.
Our study shows the importance of routine postoperative PSG studies and clinical follow-up. The finding that only OSAS severity predicted the surgical outcome in our study suggests that genioglossal and hyoid advancement cannot be recomended on the basis of cephalometrics alone. Patients for whom phase 1 failed were still symptomatic, had a significant number of apneas, retained multilevel obstruction, and consequently required additional treatment. Therapeutic alternatives in these patients may be bimaxillary advancement22,23 and another nasal CPAP attempt. In conclusion, although limited by the small number of patients, our study suggests that after UPPP, only patients with mild to moderate OSAS with oropharyngeal and hypopharyngeal obstruction may benefit from a multi-level reconstruction phase 1 surgical approach.
Accepted for publication September 14, 2001.
This work was supported by grant FISS 94/0434 from the Fondo de Investigaciones Sanitarias, Madrid, Spain (Dr Morelló).
Corresponding author and reprints: Isabel Vilaseca, MD, Otorhinolaryngology Service, Hospital Clínic de Barcelona, C/ Villarroel 170, Barcelona 08036, Spain (e-mail: email@example.com).