Uvulopalatopharyngoplasty in Obstructive Apnea Value of Preoperative Localization of Site of Upper Airway Narrowing during Sleep1-3
DAVID W. HUDGEL, THERESA HARASICK, ROBERT L. KATZ, WILLIAM J. WITT, and TOM I. ABELSON
Introduction
Previously we demonstrated that inspiratory obstruction during sleep occurred within one of two regions of the upper airway in patients with obstructive sleep apnea (OSA). Obstruction occurred within either the transpalatal or the hypopharyngeal airway (1). Since one-half of the subjects studied by us and by Chaban and colleagues (2) reproducibly obstructed at or above the level of the palate, and sinceuvulopalatopharyngoplasty is successfulapproximately 50010 of the time (3-11), we hypothesized that the patients with transpalatal airway obstruction would be the individuals most likely to experience a favorable response to uvulopalatopharyngoplasty. This hypothesis was tested in those obstructive apnea patients seen in our Sleep Disorders Center who elected uvulopalatopharyngoplasty. Methods Subjects From March 1, 1987 to February 1, 1989, 14 patients with symptoms of heavy snoring and daytime hypersomnolence who had obstructive sleep apnea on polysomnogram elected uvulopalatopharyngoplasty and agreed to participate in this protocol. Surgery was not specifically recommended for any patient; there was no attempt made to encourage those with tonsillar hypertrophy or a narrow upper airway to undergo surgery. All patients who volunteered for the study returned for postoperative evaluation. Pressure and Flow Determination The bias-flow, multiple-catheter technique used in our original study of GSA patients (1) was modified for this study: (1) the esophageal balloon was not used and (2) a tight-fitting face mask was used to record inspiratory flow quantitatively. A plastic catheter, 1.4 mm ID, 30 em long with a plugged end and eight 17 F holes in the distal 1 em, was used for pharyngeal pressure measurements. After light local anesthesia spray to the nose and pharynx, three catheters were inserted transnasallyto 16to 18, 10to 12,and 8 em, respectively.The tip and pressure ports 942
SUMMARY We hypothesized that those obstructive sleep apnea (OSA) patients with upper airway collapse during sleep within the transpalatal airway would have a more favorable response to uvulopalatopharyngoplasty (UPP) than those patients with obstruction within the hypopharyngeal airway. Wetested this hypothesis In seven OSApatients with transpalatal and sevenwith hypopharyngeal obstruction undergoing UPP. Preoperatively the apnea/hypopnea Index (AHI) was different between palatal and hypopharyngeal obstructors, 37.8 ± 6.0 (± SEM) and 63.9 ± 6.3, respectively (p < 0.05), but the apnea-associated arterial oxygen desaturatlon and the lowest sleep saturation level were not different between the two groups. Postoperatively the AHI was 17.6 ± 7.2In the palatal obstructors and 40.3 ± 15.6 In the hypopharyngeal obstructors (both p < 0.05 from preoperative AHI). The palatal obstructors had a significant decrease In the percentage of sleep time spent apneic and the hypopharyngeal obstructors had a significant decrease In the hypopnea, but not apnea, time following surgery. The palatal obstructors had a significantly higher postoperative arterial oxygen saturation than the hypopharyngeal obstructors. 1\vo hypopharyngeal obstructors worsened postoperatively. In addition we found that regardless of the site of the obstruction preoperatively, all obstructions occurred at the level of the palate postoperatively. We conclude that patients with preoperative transpalatal obstruction had diminution In obstructive apneas and those with hypopharyngeal obstruction had diminution In hypopneas but not apneas. Oxygenation was better postoperatively In the palatal obstructors, and none worsened postoperatively. These results suggest that Identification of the site of upper airway obstruction In OSA may be beneficial. We anticipate that these findings will lead to further study of the benefit of obstruction site Identification and the role of this principle In designing specific operative approaches to OSA at different pharynAM REV RESPIR DIS 1991; 143:942-946 geal sites.
of the most distal catheter, the hypopharyngeal catheter, were located below the tip of the epiglottis, as verified previously (12).The pressure ports of the middle catheter, the oropharyngeal catheter, could be visualized just below the soft palate in the oropharynx; and the upper catheter tip, the nasopharyngeal catheter, was located in the nasopharynx posterior to the nasal turbinates. The frequency-response characteristics of our catheters and recording system have been described elsewhere (13). A bias flow of compressed air of 0.1 L/min was directed in a retrograde manner through each catheter. Catheters were introduced the same distance postoperatively as preoperatively in each individual.
Protocol Subjects reported to the sleep laboratory approximately 1.5 h before their usual bedtime. Electrodes were applied for sleep staging by standard techniques (14). A Nellcor finger oximeter (Nellcor, Inc, Hayward, CA) measured oxygensaturation. Inspiratory effort wasmeasured by chest wall electromyogram (EMG) with electrodes located at the right costal margin in the anterior axillary line (15). Patients slept supine for site of obstruction determi-
nation. If the patient awakened after at least 1 h of sleep during which the site of obstruction had been determined, the catheters and mask were removed and nasal and oral thermistors werepositioned to record inspiratory flow. The subject was allowed to sleep the remainder of the night or through one REM period in a body position of his or her choice. This same protocol was repeated from 1 to 4 months following surgery in all patients except one who was studied 9 months following surgery. In all studies the body position was the same as in the initial study. Patients were positioned, and they slept in the same body position as during the preoperative study. (Received in original form March 7, 1990 and in revised form November 21, 1990). 1 From the Departments of Medicine and Otolaryngology, Case Western Reserve University and Cleveland Metropolitan General Hospital, Cleveland, Ohio. . 2 Supported by Grants AG 04391 and HL 42215 from the National Institutes of Health. 3 Correspondence and requests for reprints should be addressed to D. W. Hudgel, M.D., Pulmonary Division, MetroHealth Medical Center, 3395 Scranton Road, Cleveland, OH 44109.
PREOPERATIVE LOCALIZATION OF OBSTRUCTION SITE IN SLEEP APNEA
Uvulopalatopharyngoplasty The surgical technique of Fujita and coworkers was followed (16, 17).This procedure was a modification of the original Ikematsu technique for the treatment of snoring (18). Six different otolaryngologists who were blind to the preoperative findings performed the uvulopalatopharyngoplasty procedures on these patients. Corticosteroids were routinely administered during the immediate postoperative period to decrease local edema and inflammation. A tracheostomy was performed in conjunction with the uvulopalatopharyngoplasty in one patient. The postoperative sleep study was performed following removal of the tracheostomy in this individual. No operative complications occurred in any patient.
11
943
rlJl
rI\ n
--.JUIU~\ r\
L/SEC
FLOW
NASOPHARYNGEAL PRESSURE
120 CM H20 -
OROPHARYNGEAL PRESSURE
120 C\.1 H20
HYPOPHARYNGEAL PRESSURE 120 C\.1 H20
Sleep RecordAnalysis Sleep staging was performed in the standard fashion (14). An obstructive apnea was recorded if inspiratory flow was absent for at least 10s and inspiratory effort, as detected by chest wall EMG, persisted during this interval. No chest wall inspiratory EMG activity existed if an apnea was central. A mixed apnea was taken to be a combination of a central and an obstructive apnea. An obstructive hypopnea was recorded if inspiratory flow decreased by at least 50070 but did not cease for at least 10s while chest wall inspiratory EMG activity persisted. In a desaturation episode the arterial oxygen saturation (Sa02) decreased by at least 4% from the preeventlevel.The lowest Sa02value for each desaturation episode was termed the apneaor hypopneasaturation nadir. An apneic or hypopneic arousal was recorded if an arousal accompanied termination of a hypopneic or apneic event. Sleep stage, ventilation, and oxygenation data were entered into a computer summation program that provided total sleep time (TST), number of apneas and hypopneas, apnea and hypopnea time, number of desaturation episodes, apnea or hypopnea saturation nadir, and number of arousals. Apnea and hypopnea indices were calculated as the number of events per total sleep time. The site of narrowing (hypopnea) or obstruction was determined to be in the transpalatal airway if inspiratory pressure swings decreased or ceased in the nasopharyngeal catheter recording but persisted in the oroand hypopharyngeal recordings during the diminution or absence of inspiratory flow. Hypopharyngeal narrowing or obstruction was said to be present if inspiratory pressure swings were decreased or eliminated from the naso- and oropharyngeal catheter tracings but persisted in the hypopharyngeal catheter tracing during the diminution or absence of inspiratory flow.
PRE-UPP
POST-UPP
Fig. 1. Pre- and postoperative recordings in a palatal obstructor. During obstructive apneas there is persistence of inspiratory-expiratory intraluminal pressure excursions in the oropharyngeal and hypopharyngeal catheters but not in the nasopharyngeal catheter. Thus the obstruction occurs between the upper two catheters in the transpalatal airway. The same pattern is present postoperatively.
different pre- and postoperatively between the transpalatal and hypo pharyngeal obstructor groups. Within each group the change in each variable from the pre- to the postoperative state was determined by the univariate t test. The general linear models procedure allowed normalization of preoperative differences in variables and comparison of the normalized changes that occurred with surgery between groups. The statistical significance level was set at 5%.
Results
Preoperative Findings Of the 14patients studied, 7 obstructed
within the transpalatal airway and 7 obstructed within the hypopharynx. The left panels of figures 1 and 2 illustrate preoperative transpalatal and hypopharyngeal obstruction, respectively. Morphometric and sleep characteristics for each group are provided in table 1. Thirteen subjects weremales. The one female was a palatal obstructor. Palatal and hypopharyngeal obstructor groups were of equivalent age, weight, and height. Two palatal obstructors had obviously enlarged tonsils (Subjects 1 and 5), and one had an elongated soft palate and
FLOW
I
I LISEe
NASOPHARYNGEAL PRESSURE
OROPHARYNGEAL PRESSURE
HYPOPHARYNGEAL PRESSURE
Data Analysis The univariate procedure was used to test normality of data distribution. In general data were not normally distributed. A nonparametric analysis of variance was therefore used to determine if the variables measured were
PRE-UPP
POST-UPP
Fig. 2. Pre- and postoperative recordings in a hypopharyngeal obstructor. During obstructive apneas there is persistence of inspiratory-expiratory pressure excursions in the hypopharyngeal catheter only. Thus the obstruction occurs between the lower two catheters in the hypopharyngeal airway. The site of obstruction shifts to the transpalatal airway postoperatively.
944
HUDGEL, HARASICK, KATZ, WITT, AND ABELSON
TABLE 1
had an apnea/hypopnea index of 63.9 ± 6.3 events/h and were apneic or hypopneic 37.8 ± 4.3070 of their sleep time preoperatively (both different from palatal obstructors, p < 0.03, respectively). The number of desaturation episodes, the apnea and hypopnea arterial oxygen saturation nadir, and the number of apneic or hypopneic arousals were not different between the palatal and hypopharyngeal obstructors preoperatively (table 3). These patients had a total sleep time of 176 ± 35 min preoperatively.
PREOPERATIVE SUBJECT CHARACTERISTICS* Hypopharyngeal Obstructors (7M)
Palatal Obstructors (6 M, 1 F) Age Weight, Height, % TST % TST
kg cm non-REM REM
42.4 109 178 87 6.5
37.3 114 178 83 6.6
2.9 4 4 4 ± 2.8
± ± ± ±
± ± ± ± ±
1.2 6 2 4 3.2
• Mean ± SEM.
uvula (Subject 6); no other subjects had pharyngeal anatomic abnormalities evident on physical examination. Those with palatal obstruction had an apnea/hypopnea index of 37.8 ± 6.0 (mean ± SEM) events/h and spent 20.5 ± 4.3 % of their sleep time apneic or hypopneic preoperatively (tables 2 and 3). Total sleep time for these patients was 159 ± 18 min. Those with hypopharyngeal obstruction
Postoperative Findings Subjects in each group lost a similar amount of weight from the pre- to the postoperative sleep study, 3 and 4 kg, respectively. The palatal and hypopharyngeal obstructors slept 141 ± 24 and 142 ± 26 min, respectively. Individual changes in the pattern of apneas and
TABLE 2 INDIVIDUAL CHANGES IN APNEAlHYPOPNEA PATIERN WITH UVULOPALATOPHARYNGOPLASTY Obstructive Apneas (%Sleep Time)
AlH Index* Subject Number Palatal obstructors 1t 2 3 4 5t 6t 7 Hypopharyngeal obstructors 1 2 3 4 5 6 7
Obstructive Hypopneas (%Sleep Time)
Sum (%Sleep Time)
Preop
Postop
Preop
Postop
Preop
Postop
Preop
Postop
26.1 31.5 35.0 12.8 55.1 52.8 50.6
6.7 23.2 7.2 0 47.1 39.1 0
8.4 12.4 21.6 0.5 13.7 19.0 4.1
1.6 1.1 1.2 0 6.8 1.5 0
3.9 6.0 20.8 5.5 6.4 2.1 19.6
3.5 17.3 5.8 0 18.2 15.9 0
12.3 18.2 42.4 5.8 20.1 21.1 23.7
5.1 18.4 7.0 0 25.0 17.4 0
58.8 33.4 84.7 77.4 60.7 60.0 72.6
0 5.7 22.2 42.0 78.3 21.1 112.6
29.3 5.4 3.5 24.1 4.0 18.3 47.6
0 0 0.7 0.9 28.7 5.1 61.3
11.4 17.8 51.4 14.7 25.1 11.0 0.8
0 2.7 16.4 16.5 6.7 4.9 7.9
40.7 23.2 54.9 38.8 29.1 29.3 48.4
0 2.7 17.1 17.4 35.6 10.3 69.1
• Apnealhypopnea index. Anatomically narrow upper airway.
t
TABLE 3 CHANGES IN RESPIRATORY AND OXYGENATION VARIABLES BEFORE AND AFTER UVULOPALATOPHARYNGOPLASTY Palatal Obstructors Preop Apnealhypopnea index % Sleep time apneic or hypopneic Apneic Hypopneic No. desaturation episodes/h sleep Low Sa02 with apneas or hypopneas No. arousals/h sleep
37.8 20.5 11.4 9.2 31.5 86 19
± 6.0 ± 4.3
± 2.9 ± 2.9 ± 8.0 ± 2 ± 5
Postop
± 7.2* ± 3.7 ± 0.9* ± 3.1 ± 6.5* ± 1 9 ± 3*
17.6 10.4 1.7 8.7 14.2 89
• Postoperative versus preoperative, p < 0.05. Preoperative hypopharyngeal versus preoperative palatal, p < 0.05. ;j: Postoperative hypopharyngeal versus postoperative palatal, p < 0.05.
t
Hypopharyngeal Obstructors Preop 63.9 37.8 18.9 18.9 48.7 81 25
± ± ± ± ± ± ±
6.3t 4.3t 6.2 6.1 7.9
2 5
Postop 40.3 21.7 13.8 7.9 30.6 78 11
± ± ± ± ± ± ±
15.6 9.1 8.8 2.4* 11.8*
5* 3*
hypopneas from the pre- to the postoperative state are shown in table 2.
Pre- to Postoperative Changes Palatalobstructors. 1\\'0 palatal obstructors had no apneas nor hypopneas postoperatively. For the group, variables that improved significantly following uvulopalatopharyngoplasty were apnea/ hypopnea index (p < 0.01), percentage of total sleep time apneic and hypopneic (p = 0.05), apneic arousals (p < 0.04), and number of desaturation episodes (p < 0.04). Palatal obstructors experienced improvement in the proportion of sleep time spent apneic, from 11.4 ± 2.9 to 1.7 ± 0.9% (p < 0.05; table 3). However, the hypopnea time did not decrease in this group. There was no association between the presence of tonsillar enlargement or palatal abnormalities observed during the preoperative physical examination and the response to surgery. Hypopharyngeal obstructors. Two hypopharyngeal obstructors had complete resolution of apneas with surgery. One of these subjects had a fewobstructive hypopneas postoperatively (table 2). For the group, variables that significantly improved following surgery were hypopnea time (p < 0.05), number of desaturation episodes (p < 0.05), and the number of apneic arousals (p < 0.01; tables 2 and 3). In general the hypopharyngeal obstructors had improvement in hypopneas but not apneas. Two patients worsened postoperatively (Subjects 5 and 7).
Relative Change between Palatal and Hypopharyngeal Obstructors following Surgery Since the apnea/hypopnea index was significantly different between the two experimental groups preoperatively, the general linear models procedure was used to correct for the preoperative difference. With this analysis there was no difference in the improvement seen in this variable between the palatal and hypopharyngeal obstructors with surgery. Although the Sa02 was not different preoperatively, the postoperative apnea and hypopnea Sao, nadir was significantly higher in the palatal obstructors, 89 ± 1 versus 78 ± 5% in the hypopharyngeal group (p < 0.04) (table 3).
Change in Site of Obstruction' In all 11 patients who had apneas or hypopneas postoperatively the narrowing or obstruction occurred within the transpalatal airway even if the obstruction was in the hypopharynx preopera-
PREOPERATIVE LOCALIZATION OF OBSTRUCTION SITE IN SLEEP APNEA
tively. This phenomenon is illustrated in figures 1 and 2, which show that the palatal obstructor obstructs at the same level postoperatively but the hypo pharyngeal obstructor obstructs within the transpalatal airway. Discussion
We hypothesized that those obstructive sleep apnea patients with upper airway collapse during sleep within the transpalatal airway would have a more favorable response to uvulopalatopharyngoplasty than those patients with obstruction within the hypopharyngeal airway. The findings of this study partially support this hypothesis. (1) Palatal 0 bstructors had significantly fewer apneas and less apnea time postoperatively without a change in the number of hypopneas and hypopnea time. (2) In contrast, hypopharyngeal obstructors had improvement in the hypopneic time but no improvement in the apnea time. (3) This difference in apnea/hypopnea pattern in response to surgery between the two groups resulted in a higher apnea- or hypopnea-related Sa02 nadir value in the palatal obstructors. (4) 1\\'0 patients, both hypo pharyngeal obstructors, worsened postoperatively in that their apnea/hypopnea indicies and apneic times were higher following surgery than before surgery. We therefore conclude that although most patients benefited from uvulopalatopharyngoplasty, those OSA patients with obstruction within the transpalatal airway had improvement in the obstructive apnea component of their illness and postoperative arterial oxygen saturation. Those with initial hypopharyngeal obstruction had improvement in their hypopnea time but no change in their apneas and oxygenation postoperatively.
Technical Considerations Crucial to classifying a patient as a palatal or hypopharyngeal obstructor with our technique is the position of the oropharyngeal catheter, which divides the two upper airway segments we assessed. The orifices for pressure measurement in the oropharyngeal catheter's distal 1 em were positioned just below the level of the soft palate. With this placement the pressure ports were located lateral to or below the uvula, depending on an individual's anatomy. Placement ofthe catheter tip was done with the soft palate relaxed and therefore in its most caudal position, thereby assuring that the pressure ports were always caudal to the level of the palate. To prevent catheter movement, some investigators have at-
tached the pharyngeal catheter to an esophageal extension to help maintain catheter position (19). Attaching all three catheters to one esophageal extension would not have allowed for adjustment of any individual catheter, and use of three esophageal extensions surely would have produced too much discomfort. Therefore, we did not use this technique. Regardless, we have found that our catheters do not move once positioned. This statement is supported by two observations. First, visual inspection at the conclusion of the study shows the tip of the oropharyngeal catheter has not been displaced. Second, the pressure differentials recorded from these catheters indicate that the site of obstruction remains identical throughout a whole study on a given individual. Tonsil position may have affected our results. With placement of the oropharyngeal catheter described above, the tonsillar region may have been in either the hypopharyngeal or transpalatal airway segments depending on the patient's pharyngeal anatomy. If a patient with benefit from surgery had significant preoperative narrowing or obstruction at the tonsillar level and this level were in the hypopharyngeal segment, then we would classify this patient as a hypopharyngeal obstructor who improved with uvulopalatopharyngoplasty, This possibility may explain why some hypo pharyngeal obstruction improved with uvulopalatopharyngoplasty. Future studies may need to isolate the tonsillar region more specifically. There is concern whether the whole or a portion of the upper airway collapses in OSA. Differential pressures recorded with our catheters indicate that upper airway narrowing or obstruction can occur exclusively at one specific site or region within the upper airway (1) (figures 1and 2). In our patients the obstruction at either site likely did not extend upward or downward because both the oropharyngeal and hypopharyngeal catheters continued to show atmospheric pressure or inspiratory pressure excursions when they were away from the site of obstruction. If obstruction extended and encompassed either of the catheter tips, the pressure signal from that catheter would have deflected to a high positive pressure level because of accumulating pressure within the catheter due to the bias flow of air into the catheter. Our results wereconfirmed by Chaban and colleagues (2). These investigators slowly withdrew a hypopharyngeal pressure catheter from the nose during sleep in OSA patients. They found that the area of obstruction
945
extended over a region of approximately 1 em in 8 of 10 patients studied. In addition, they found that an equal number of patients obstructed above and below the level ofthe soft palate, similar to our results (1).Using rapid computed tomography during sleep, Stein and coworkers showed that the site of obstruction was usually localized within the transpalatal or oropharyngeal airway over a range of 4 em or less (20). In a minority of the patients the obstruction extended caudally to the hypopharynx. This extension occurred later in the apnea when vigorous inspiratory efforts produced large intraluminal subatmospheric inspiratory pressures. From these studies weconclude that relatively isolated palatal or hypopharyngeal narrowing or obstruction is a common picture in OSA. Because the uvulopalatopharyngoplasty was performed by different surgeons in our study, it is possible the surgical technique varied. However, all surgeons who operated on these patients used the technique described by Fujita and colleagues (16, 17). In addition, no surgeons operated exclusively on palatal or hypopharyngeal obstructors. To have studied patients operated on by only one surgeon would have been ideal, but not practical. With the availability of CPAP therapy, only a small proportion of OSA patients now elect surgery. Therefore, in order to accumulate enough patients to perform this analysis, we had to study patients operated on by more than one surgeon. During the study nights patients slept only 2 to 3 h. With three catheters and a face mask in place, it is difficult for patients, even hypersomnolent patients, to tolerate this degree of instrumentation and sleep for long periods. Even after removal of the catheters and mask it was difficult for patients to sleep, partially because of the pharyngeal irritation produced by the catheters. Fortunately we obtained Stage 2 and REM sleep in all the patients.
Physiologic Changes within the Upper Airway Following Uvulopalatopharyngoplasty Our results indicate that the characteristics of airflow limitation within the upper airway changed following uvulopalatopharyngoplasty. In those with preoperative narrowing or obstruction within the transpalatal airway, primarily the obstructive component of the process, but not the hypopneic portion, improved following surgery. Thus following uvular and palatal resection, the periodic pattern of breathing continued with
946
episodic trans palatal narrowing rather than obstruction. Postoperatively in the hypopharyngeal obstructors the site of narrowing or obstruction shifted to the transpalatal airway. The reason for this transition is unclear. Using computed tomography in awake OSA patients, Shepard and Thawley found the crosssectional area of the most rostral portion of the transpalatal airway to be more narrowed following surgery than before uvulopalatopharyngoplasty (21). Inflammation produced by the operative procedure in this region may result in scarring. Scar retraction would tend to narrow the pharynx at this site. The typical pattern of breathing during sleep was different in the hypopharyngeal obstructors, with a persistence of apneas and improvement in hypopneas. Weassume that those hypopharyngeal obstructors who improved with surgery had preoperative narrowing or obstruction within the oropharynx at the tonsillar level, even though this narrowing was not evident on oral examination. It is possible that the tonsillar area in these patients was located below the oropharyngeal catheter pressure ports; these patients were therefore classified as hypopharyngeal obstructors. If resection of tonsillar tissue was instrumental in their surgical response, they would be hypopharyngeal responders. It is likely that those hypopharyngeal obstructors who did not improve or worsened after surgery had narrowing or obstruction more caudally in the hypopharynx preoperatively. The hypopharyngeal caliber would not be expected to increase with surgery at the palatal level; in fact, Shepard and Thawley showed that the hypopharynx narrowed following uvulopalatopharyngoplasty (21).
Importance of Determining the Site of Obstruction in OSA Management This study raises the possibility that the proportion of patients improving with uvulopalatopharyngoplasty would increase if this procedure were performed primarily on those patients with transpalatal obstruction. It would be easier if this determination could be made during wakefulness. However, studies concluded to date do not support that this determination can be made during wakefulness. Sher and coworkers found that patients who had collapse of their palatal airway with a Muller maneuver had a good response to uvulopalatopharyngoplasty (22). However, no patients with hypopharyngeal obstruction were studied, so that no comparison was made be-
HUDGEL, HARASICK, KATZ, WITT, AND ABELSON
tween the surgical results in the palatal or the hypopharyngeal obstructors. Wittig and colleagues resolved this problem and compared the response to uvulopalatopharyngoplasty in patients who had awake upper airway collapse isolated to the palatal airway with the surgical response of patients with both palatal and hypopharyngeal collapse (23). These investigators found no difference in the surgical results betweenthe groups. Only apnea and apnea/hypopnea indices were evaluated in this study, however. Therefore, to date these fewstudies do not easily differentiate surgical responders from nonresponders during wakefulness. During sleep using somnofluoroscopy, Walshand coworkersfound that those subjects with isolated obstruction at the palatal level had a better response than those with hypopharyngeal obstruction (24). Our study expands on this by incorporating direct measurements of airflow obstruction. Those with specific hypopharyngeal obstruction may respond more favorably to surgical techniques that increase the cross-sectional area of the hypopharynx, such as mandibular or hyoid bone advancement or a combination of these two procedures (25-27); those with transpalatal obstruction should continue to respond to UPP. It is hoped that with improved definition of the site of obstruction during sleep and the use of more specific operative techniques the results of surgical treatment of OSA will improve substantially. References 1. Hudgel DW. Variable site of airway narrowing among obstructive sleep apnea patients. J Appl Physiol 1986; 61:1403-9. 2. Chaban R, Cole P, Hoffstein V. Site of upper airway obstruction in patients with idiopathic obstructive sleep apnea. Laryngoscope 1988;98:641-7. 3. Guilleminault C, Hayes B, Smith L, Simmons FB. Palatopharyngoplasty and obstructive sleep apnea syndrome. Bull Eur Physiopathol Respir 1983; 19:595-9. 4. Silvestri R, Guilleminault C, Simmons FB. Palatopharyngoplasty in the treatment of obstructive sleep apnea. In: Guilleminault C, Lugaresi E, eds. Sleep/wake disorders: natural history, epidemiology and long-term evolution. New York: Raven Press. 1983: 163-70. 5. Simmons FB, Guilleminault C, Silvestri R. Snoring, and some obstructive sleep apnea can be cured by oropharyngeal surgery. Arch Otolaryngol (Stockh) 1983:503-7. 6. Zorick F, Roehrs T, Conway W, Fujita S. Wittig R, Roth T. Effects of uvulopalatopharyngoplasty on the daytime sleepiness associated with sleep apnea syndrome. Bull Eur Physiopathol Respir 1983; 19:600-3. 7. Rivilin J, Hoffstein V,Kalbfleisch J. Upper airway morphology in patients with idiopathic obstructive sleep apnea. Am Rev Respir Dis 1984; 129:355-60. 8. Conway W, Fujita S, Zorick F, et at.
Uvulopalatopharyngoplasty. One-year followup. Chest 1985; 88:385-7. 9. deBerry-Borowiecki B, Kukwa AA, Blanks RHI. Indications for palatopharyngoplasty, Arch Otolaryngol 1985; 111:659-63. 10. Fujita S, Conway WA, Sickleested JM, et al. Evaluation of the effectiveness of uvulopalatopharyngoplasty, Laryngoscope 1985; 95:70-4. 11. Gislason T. Lindholm C-E, Almquist M, et a/. Uvulopalatopharyngoplasty in the sleep apnea syndrome. Arch Otolaryngol Head Neck Surg 1988; 114:45-51. 12. Hudgel DW, Martin RJ, Johnson B, Hill P. Mechanics of the respiratory system and breathing pattern during sleep in normal humans. J Appl Physiol 1984; 56:133-7. 13. Hudgel DW, Hendricks C. Palate and hypopharynx - sites ofinspiratory narrowing of the upper airway during sleep. Am Rev Respir Dis 1988; 138:1542-7. 14. Rechtschaffen A, Kales A. A manual of standardized terminology, technique and scoring system for sleep stages of human subjects. Washington DC: National Institutes of Health, 1986(Publ. 204). 15. Muller N. Volgyesi G. Becker L, Bryan MH, Bryan AC. Diaphragmatic muscle tone. J Appl Physiol 1979; 47:279-84. 16. Fujita S, Conway W, Zorick F, Roth T. Surgical correction of anatomic abnormalities in obstructive sleep apnea syndrome: uvulopalatopharyngoplasty, Otolaryngol Head Neck Surg 1981; 89:923-34. 17. Ikematsu T, Fujita S, Simmons FB. Uvulopalatopharyngoplasty: Variations. In: Fairbanks DNF, Fujita S, Ikematsu T, Simmons FB. eds. Snoring and obstructive sleep apnea. New York: Raven Press, 1987; 129-70. 18. Ikematsu T. Study of snoring, 4th report. Therapy. Jpn Oto-Rhino-Laryngol 1964; 64:434-5. 19. Wiegand L, Zwillich CW, White DP. Collapsibility of the human upper airway during normal sleep. J Appl Physiol 1989; 66:1800-8. 20. Stein MG, Gamsu G. DeGeer G, Golden JA, Crumley RL, Webb WR. Fast-CT in obstructive sleep apnea. AJR 1987; 148:1069-74. 21. Shepard JW Jr, Thawley SE. Evaluation of the upper airway by computerized tomography in patients undergoing uvulopalatopharyngoplasty for obstructive sleep apnea. Am Rev Respir Dis 1989; 140:711-6. 22. Sher AE, Thorpy MJ. Spielman AJ, Shprintzen RJ, Burack B, McGregor PA. Predictive value of Muller maneuver in selection of patients for uvulopalatopharyngoplasty. Laryngoscope 1985; 95:1483-7. 23. Wittig R, Fujita S, Fortier J, Zorick F, Potts G. Roth T. Results of uvulopalatopharyngoplasty (UPPP) in patients with both oropharyngeal and hypopharyngeal collapse on Muller maneuver. Sleep Res 1988; 17:269. 24. Walsh JK, Katsautonis GP. Somnofluoroscopy as a predictor of UPPP efficacy. Sleep Res 1984; 13:21. 25. Bear SE, Priest JH. Sleep apnea syndrome: Correction with surgical advancement of the mandible. J Oral Surg 1980; 38:543-9. 26. Riley R, Powell NB, Guilleminault C. Inferior sagittal osteotomy of the mandible with hyoid myotomy-suspension: a new procedure for obstructive sleep apnea. Otolaryngol Head Neck Surg 1986; 94:589-93. 27. Riley R, Powell NB, Guilleminault C. et al. Maxillary, mandibulory and hyoid advancement: an alternative to tracheostomy in obstructive sleep apnea syndrome. Otolaryngol Head Neck Surg 1986; 94:587-8.
DAVID W. HUDGEL, THERESA HARASICK, ROBERT L. KATZ, WILLIAM J. WITT, and TOM I. ABELSON
Introduction
Previously we demonstrated that inspiratory obstruction during sleep occurred within one of two regions of the upper airway in patients with obstructive sleep apnea (OSA). Obstruction occurred within either the transpalatal or the hypopharyngeal airway (1). Since one-half of the subjects studied by us and by Chaban and colleagues (2) reproducibly obstructed at or above the level of the palate, and sinceuvulopalatopharyngoplasty is successfulapproximately 50010 of the time (3-11), we hypothesized that the patients with transpalatal airway obstruction would be the individuals most likely to experience a favorable response to uvulopalatopharyngoplasty. This hypothesis was tested in those obstructive apnea patients seen in our Sleep Disorders Center who elected uvulopalatopharyngoplasty. Methods Subjects From March 1, 1987 to February 1, 1989, 14 patients with symptoms of heavy snoring and daytime hypersomnolence who had obstructive sleep apnea on polysomnogram elected uvulopalatopharyngoplasty and agreed to participate in this protocol. Surgery was not specifically recommended for any patient; there was no attempt made to encourage those with tonsillar hypertrophy or a narrow upper airway to undergo surgery. All patients who volunteered for the study returned for postoperative evaluation. Pressure and Flow Determination The bias-flow, multiple-catheter technique used in our original study of GSA patients (1) was modified for this study: (1) the esophageal balloon was not used and (2) a tight-fitting face mask was used to record inspiratory flow quantitatively. A plastic catheter, 1.4 mm ID, 30 em long with a plugged end and eight 17 F holes in the distal 1 em, was used for pharyngeal pressure measurements. After light local anesthesia spray to the nose and pharynx, three catheters were inserted transnasallyto 16to 18, 10to 12,and 8 em, respectively.The tip and pressure ports 942
SUMMARY We hypothesized that those obstructive sleep apnea (OSA) patients with upper airway collapse during sleep within the transpalatal airway would have a more favorable response to uvulopalatopharyngoplasty (UPP) than those patients with obstruction within the hypopharyngeal airway. Wetested this hypothesis In seven OSApatients with transpalatal and sevenwith hypopharyngeal obstruction undergoing UPP. Preoperatively the apnea/hypopnea Index (AHI) was different between palatal and hypopharyngeal obstructors, 37.8 ± 6.0 (± SEM) and 63.9 ± 6.3, respectively (p < 0.05), but the apnea-associated arterial oxygen desaturatlon and the lowest sleep saturation level were not different between the two groups. Postoperatively the AHI was 17.6 ± 7.2In the palatal obstructors and 40.3 ± 15.6 In the hypopharyngeal obstructors (both p < 0.05 from preoperative AHI). The palatal obstructors had a significant decrease In the percentage of sleep time spent apneic and the hypopharyngeal obstructors had a significant decrease In the hypopnea, but not apnea, time following surgery. The palatal obstructors had a significantly higher postoperative arterial oxygen saturation than the hypopharyngeal obstructors. 1\vo hypopharyngeal obstructors worsened postoperatively. In addition we found that regardless of the site of the obstruction preoperatively, all obstructions occurred at the level of the palate postoperatively. We conclude that patients with preoperative transpalatal obstruction had diminution In obstructive apneas and those with hypopharyngeal obstruction had diminution In hypopneas but not apneas. Oxygenation was better postoperatively In the palatal obstructors, and none worsened postoperatively. These results suggest that Identification of the site of upper airway obstruction In OSA may be beneficial. We anticipate that these findings will lead to further study of the benefit of obstruction site Identification and the role of this principle In designing specific operative approaches to OSA at different pharynAM REV RESPIR DIS 1991; 143:942-946 geal sites.
of the most distal catheter, the hypopharyngeal catheter, were located below the tip of the epiglottis, as verified previously (12).The pressure ports of the middle catheter, the oropharyngeal catheter, could be visualized just below the soft palate in the oropharynx; and the upper catheter tip, the nasopharyngeal catheter, was located in the nasopharynx posterior to the nasal turbinates. The frequency-response characteristics of our catheters and recording system have been described elsewhere (13). A bias flow of compressed air of 0.1 L/min was directed in a retrograde manner through each catheter. Catheters were introduced the same distance postoperatively as preoperatively in each individual.
Protocol Subjects reported to the sleep laboratory approximately 1.5 h before their usual bedtime. Electrodes were applied for sleep staging by standard techniques (14). A Nellcor finger oximeter (Nellcor, Inc, Hayward, CA) measured oxygensaturation. Inspiratory effort wasmeasured by chest wall electromyogram (EMG) with electrodes located at the right costal margin in the anterior axillary line (15). Patients slept supine for site of obstruction determi-
nation. If the patient awakened after at least 1 h of sleep during which the site of obstruction had been determined, the catheters and mask were removed and nasal and oral thermistors werepositioned to record inspiratory flow. The subject was allowed to sleep the remainder of the night or through one REM period in a body position of his or her choice. This same protocol was repeated from 1 to 4 months following surgery in all patients except one who was studied 9 months following surgery. In all studies the body position was the same as in the initial study. Patients were positioned, and they slept in the same body position as during the preoperative study. (Received in original form March 7, 1990 and in revised form November 21, 1990). 1 From the Departments of Medicine and Otolaryngology, Case Western Reserve University and Cleveland Metropolitan General Hospital, Cleveland, Ohio. . 2 Supported by Grants AG 04391 and HL 42215 from the National Institutes of Health. 3 Correspondence and requests for reprints should be addressed to D. W. Hudgel, M.D., Pulmonary Division, MetroHealth Medical Center, 3395 Scranton Road, Cleveland, OH 44109.
PREOPERATIVE LOCALIZATION OF OBSTRUCTION SITE IN SLEEP APNEA
Uvulopalatopharyngoplasty The surgical technique of Fujita and coworkers was followed (16, 17).This procedure was a modification of the original Ikematsu technique for the treatment of snoring (18). Six different otolaryngologists who were blind to the preoperative findings performed the uvulopalatopharyngoplasty procedures on these patients. Corticosteroids were routinely administered during the immediate postoperative period to decrease local edema and inflammation. A tracheostomy was performed in conjunction with the uvulopalatopharyngoplasty in one patient. The postoperative sleep study was performed following removal of the tracheostomy in this individual. No operative complications occurred in any patient.
11
943
rlJl
rI\ n
--.JUIU~\ r\
L/SEC
FLOW
NASOPHARYNGEAL PRESSURE
120 CM H20 -
OROPHARYNGEAL PRESSURE
120 C\.1 H20
HYPOPHARYNGEAL PRESSURE 120 C\.1 H20
Sleep RecordAnalysis Sleep staging was performed in the standard fashion (14). An obstructive apnea was recorded if inspiratory flow was absent for at least 10s and inspiratory effort, as detected by chest wall EMG, persisted during this interval. No chest wall inspiratory EMG activity existed if an apnea was central. A mixed apnea was taken to be a combination of a central and an obstructive apnea. An obstructive hypopnea was recorded if inspiratory flow decreased by at least 50070 but did not cease for at least 10s while chest wall inspiratory EMG activity persisted. In a desaturation episode the arterial oxygen saturation (Sa02) decreased by at least 4% from the preeventlevel.The lowest Sa02value for each desaturation episode was termed the apneaor hypopneasaturation nadir. An apneic or hypopneic arousal was recorded if an arousal accompanied termination of a hypopneic or apneic event. Sleep stage, ventilation, and oxygenation data were entered into a computer summation program that provided total sleep time (TST), number of apneas and hypopneas, apnea and hypopnea time, number of desaturation episodes, apnea or hypopnea saturation nadir, and number of arousals. Apnea and hypopnea indices were calculated as the number of events per total sleep time. The site of narrowing (hypopnea) or obstruction was determined to be in the transpalatal airway if inspiratory pressure swings decreased or ceased in the nasopharyngeal catheter recording but persisted in the oroand hypopharyngeal recordings during the diminution or absence of inspiratory flow. Hypopharyngeal narrowing or obstruction was said to be present if inspiratory pressure swings were decreased or eliminated from the naso- and oropharyngeal catheter tracings but persisted in the hypopharyngeal catheter tracing during the diminution or absence of inspiratory flow.
PRE-UPP
POST-UPP
Fig. 1. Pre- and postoperative recordings in a palatal obstructor. During obstructive apneas there is persistence of inspiratory-expiratory intraluminal pressure excursions in the oropharyngeal and hypopharyngeal catheters but not in the nasopharyngeal catheter. Thus the obstruction occurs between the upper two catheters in the transpalatal airway. The same pattern is present postoperatively.
different pre- and postoperatively between the transpalatal and hypo pharyngeal obstructor groups. Within each group the change in each variable from the pre- to the postoperative state was determined by the univariate t test. The general linear models procedure allowed normalization of preoperative differences in variables and comparison of the normalized changes that occurred with surgery between groups. The statistical significance level was set at 5%.
Results
Preoperative Findings Of the 14patients studied, 7 obstructed
within the transpalatal airway and 7 obstructed within the hypopharynx. The left panels of figures 1 and 2 illustrate preoperative transpalatal and hypopharyngeal obstruction, respectively. Morphometric and sleep characteristics for each group are provided in table 1. Thirteen subjects weremales. The one female was a palatal obstructor. Palatal and hypopharyngeal obstructor groups were of equivalent age, weight, and height. Two palatal obstructors had obviously enlarged tonsils (Subjects 1 and 5), and one had an elongated soft palate and
FLOW
I
I LISEe
NASOPHARYNGEAL PRESSURE
OROPHARYNGEAL PRESSURE
HYPOPHARYNGEAL PRESSURE
Data Analysis The univariate procedure was used to test normality of data distribution. In general data were not normally distributed. A nonparametric analysis of variance was therefore used to determine if the variables measured were
PRE-UPP
POST-UPP
Fig. 2. Pre- and postoperative recordings in a hypopharyngeal obstructor. During obstructive apneas there is persistence of inspiratory-expiratory pressure excursions in the hypopharyngeal catheter only. Thus the obstruction occurs between the lower two catheters in the hypopharyngeal airway. The site of obstruction shifts to the transpalatal airway postoperatively.
944
HUDGEL, HARASICK, KATZ, WITT, AND ABELSON
TABLE 1
had an apnea/hypopnea index of 63.9 ± 6.3 events/h and were apneic or hypopneic 37.8 ± 4.3070 of their sleep time preoperatively (both different from palatal obstructors, p < 0.03, respectively). The number of desaturation episodes, the apnea and hypopnea arterial oxygen saturation nadir, and the number of apneic or hypopneic arousals were not different between the palatal and hypopharyngeal obstructors preoperatively (table 3). These patients had a total sleep time of 176 ± 35 min preoperatively.
PREOPERATIVE SUBJECT CHARACTERISTICS* Hypopharyngeal Obstructors (7M)
Palatal Obstructors (6 M, 1 F) Age Weight, Height, % TST % TST
kg cm non-REM REM
42.4 109 178 87 6.5
37.3 114 178 83 6.6
2.9 4 4 4 ± 2.8
± ± ± ±
± ± ± ± ±
1.2 6 2 4 3.2
• Mean ± SEM.
uvula (Subject 6); no other subjects had pharyngeal anatomic abnormalities evident on physical examination. Those with palatal obstruction had an apnea/hypopnea index of 37.8 ± 6.0 (mean ± SEM) events/h and spent 20.5 ± 4.3 % of their sleep time apneic or hypopneic preoperatively (tables 2 and 3). Total sleep time for these patients was 159 ± 18 min. Those with hypopharyngeal obstruction
Postoperative Findings Subjects in each group lost a similar amount of weight from the pre- to the postoperative sleep study, 3 and 4 kg, respectively. The palatal and hypopharyngeal obstructors slept 141 ± 24 and 142 ± 26 min, respectively. Individual changes in the pattern of apneas and
TABLE 2 INDIVIDUAL CHANGES IN APNEAlHYPOPNEA PATIERN WITH UVULOPALATOPHARYNGOPLASTY Obstructive Apneas (%Sleep Time)
AlH Index* Subject Number Palatal obstructors 1t 2 3 4 5t 6t 7 Hypopharyngeal obstructors 1 2 3 4 5 6 7
Obstructive Hypopneas (%Sleep Time)
Sum (%Sleep Time)
Preop
Postop
Preop
Postop
Preop
Postop
Preop
Postop
26.1 31.5 35.0 12.8 55.1 52.8 50.6
6.7 23.2 7.2 0 47.1 39.1 0
8.4 12.4 21.6 0.5 13.7 19.0 4.1
1.6 1.1 1.2 0 6.8 1.5 0
3.9 6.0 20.8 5.5 6.4 2.1 19.6
3.5 17.3 5.8 0 18.2 15.9 0
12.3 18.2 42.4 5.8 20.1 21.1 23.7
5.1 18.4 7.0 0 25.0 17.4 0
58.8 33.4 84.7 77.4 60.7 60.0 72.6
0 5.7 22.2 42.0 78.3 21.1 112.6
29.3 5.4 3.5 24.1 4.0 18.3 47.6
0 0 0.7 0.9 28.7 5.1 61.3
11.4 17.8 51.4 14.7 25.1 11.0 0.8
0 2.7 16.4 16.5 6.7 4.9 7.9
40.7 23.2 54.9 38.8 29.1 29.3 48.4
0 2.7 17.1 17.4 35.6 10.3 69.1
• Apnealhypopnea index. Anatomically narrow upper airway.
t
TABLE 3 CHANGES IN RESPIRATORY AND OXYGENATION VARIABLES BEFORE AND AFTER UVULOPALATOPHARYNGOPLASTY Palatal Obstructors Preop Apnealhypopnea index % Sleep time apneic or hypopneic Apneic Hypopneic No. desaturation episodes/h sleep Low Sa02 with apneas or hypopneas No. arousals/h sleep
37.8 20.5 11.4 9.2 31.5 86 19
± 6.0 ± 4.3
± 2.9 ± 2.9 ± 8.0 ± 2 ± 5
Postop
± 7.2* ± 3.7 ± 0.9* ± 3.1 ± 6.5* ± 1 9 ± 3*
17.6 10.4 1.7 8.7 14.2 89
• Postoperative versus preoperative, p < 0.05. Preoperative hypopharyngeal versus preoperative palatal, p < 0.05. ;j: Postoperative hypopharyngeal versus postoperative palatal, p < 0.05.
t
Hypopharyngeal Obstructors Preop 63.9 37.8 18.9 18.9 48.7 81 25
± ± ± ± ± ± ±
6.3t 4.3t 6.2 6.1 7.9
2 5
Postop 40.3 21.7 13.8 7.9 30.6 78 11
± ± ± ± ± ± ±
15.6 9.1 8.8 2.4* 11.8*
5* 3*
hypopneas from the pre- to the postoperative state are shown in table 2.
Pre- to Postoperative Changes Palatalobstructors. 1\\'0 palatal obstructors had no apneas nor hypopneas postoperatively. For the group, variables that improved significantly following uvulopalatopharyngoplasty were apnea/ hypopnea index (p < 0.01), percentage of total sleep time apneic and hypopneic (p = 0.05), apneic arousals (p < 0.04), and number of desaturation episodes (p < 0.04). Palatal obstructors experienced improvement in the proportion of sleep time spent apneic, from 11.4 ± 2.9 to 1.7 ± 0.9% (p < 0.05; table 3). However, the hypopnea time did not decrease in this group. There was no association between the presence of tonsillar enlargement or palatal abnormalities observed during the preoperative physical examination and the response to surgery. Hypopharyngeal obstructors. Two hypopharyngeal obstructors had complete resolution of apneas with surgery. One of these subjects had a fewobstructive hypopneas postoperatively (table 2). For the group, variables that significantly improved following surgery were hypopnea time (p < 0.05), number of desaturation episodes (p < 0.05), and the number of apneic arousals (p < 0.01; tables 2 and 3). In general the hypopharyngeal obstructors had improvement in hypopneas but not apneas. Two patients worsened postoperatively (Subjects 5 and 7).
Relative Change between Palatal and Hypopharyngeal Obstructors following Surgery Since the apnea/hypopnea index was significantly different between the two experimental groups preoperatively, the general linear models procedure was used to correct for the preoperative difference. With this analysis there was no difference in the improvement seen in this variable between the palatal and hypopharyngeal obstructors with surgery. Although the Sa02 was not different preoperatively, the postoperative apnea and hypopnea Sao, nadir was significantly higher in the palatal obstructors, 89 ± 1 versus 78 ± 5% in the hypopharyngeal group (p < 0.04) (table 3).
Change in Site of Obstruction' In all 11 patients who had apneas or hypopneas postoperatively the narrowing or obstruction occurred within the transpalatal airway even if the obstruction was in the hypopharynx preopera-
PREOPERATIVE LOCALIZATION OF OBSTRUCTION SITE IN SLEEP APNEA
tively. This phenomenon is illustrated in figures 1 and 2, which show that the palatal obstructor obstructs at the same level postoperatively but the hypo pharyngeal obstructor obstructs within the transpalatal airway. Discussion
We hypothesized that those obstructive sleep apnea patients with upper airway collapse during sleep within the transpalatal airway would have a more favorable response to uvulopalatopharyngoplasty than those patients with obstruction within the hypopharyngeal airway. The findings of this study partially support this hypothesis. (1) Palatal 0 bstructors had significantly fewer apneas and less apnea time postoperatively without a change in the number of hypopneas and hypopnea time. (2) In contrast, hypopharyngeal obstructors had improvement in the hypopneic time but no improvement in the apnea time. (3) This difference in apnea/hypopnea pattern in response to surgery between the two groups resulted in a higher apnea- or hypopnea-related Sa02 nadir value in the palatal obstructors. (4) 1\\'0 patients, both hypo pharyngeal obstructors, worsened postoperatively in that their apnea/hypopnea indicies and apneic times were higher following surgery than before surgery. We therefore conclude that although most patients benefited from uvulopalatopharyngoplasty, those OSA patients with obstruction within the transpalatal airway had improvement in the obstructive apnea component of their illness and postoperative arterial oxygen saturation. Those with initial hypopharyngeal obstruction had improvement in their hypopnea time but no change in their apneas and oxygenation postoperatively.
Technical Considerations Crucial to classifying a patient as a palatal or hypopharyngeal obstructor with our technique is the position of the oropharyngeal catheter, which divides the two upper airway segments we assessed. The orifices for pressure measurement in the oropharyngeal catheter's distal 1 em were positioned just below the level of the soft palate. With this placement the pressure ports were located lateral to or below the uvula, depending on an individual's anatomy. Placement ofthe catheter tip was done with the soft palate relaxed and therefore in its most caudal position, thereby assuring that the pressure ports were always caudal to the level of the palate. To prevent catheter movement, some investigators have at-
tached the pharyngeal catheter to an esophageal extension to help maintain catheter position (19). Attaching all three catheters to one esophageal extension would not have allowed for adjustment of any individual catheter, and use of three esophageal extensions surely would have produced too much discomfort. Therefore, we did not use this technique. Regardless, we have found that our catheters do not move once positioned. This statement is supported by two observations. First, visual inspection at the conclusion of the study shows the tip of the oropharyngeal catheter has not been displaced. Second, the pressure differentials recorded from these catheters indicate that the site of obstruction remains identical throughout a whole study on a given individual. Tonsil position may have affected our results. With placement of the oropharyngeal catheter described above, the tonsillar region may have been in either the hypopharyngeal or transpalatal airway segments depending on the patient's pharyngeal anatomy. If a patient with benefit from surgery had significant preoperative narrowing or obstruction at the tonsillar level and this level were in the hypopharyngeal segment, then we would classify this patient as a hypopharyngeal obstructor who improved with uvulopalatopharyngoplasty, This possibility may explain why some hypo pharyngeal obstruction improved with uvulopalatopharyngoplasty. Future studies may need to isolate the tonsillar region more specifically. There is concern whether the whole or a portion of the upper airway collapses in OSA. Differential pressures recorded with our catheters indicate that upper airway narrowing or obstruction can occur exclusively at one specific site or region within the upper airway (1) (figures 1and 2). In our patients the obstruction at either site likely did not extend upward or downward because both the oropharyngeal and hypopharyngeal catheters continued to show atmospheric pressure or inspiratory pressure excursions when they were away from the site of obstruction. If obstruction extended and encompassed either of the catheter tips, the pressure signal from that catheter would have deflected to a high positive pressure level because of accumulating pressure within the catheter due to the bias flow of air into the catheter. Our results wereconfirmed by Chaban and colleagues (2). These investigators slowly withdrew a hypopharyngeal pressure catheter from the nose during sleep in OSA patients. They found that the area of obstruction
945
extended over a region of approximately 1 em in 8 of 10 patients studied. In addition, they found that an equal number of patients obstructed above and below the level ofthe soft palate, similar to our results (1).Using rapid computed tomography during sleep, Stein and coworkers showed that the site of obstruction was usually localized within the transpalatal or oropharyngeal airway over a range of 4 em or less (20). In a minority of the patients the obstruction extended caudally to the hypopharynx. This extension occurred later in the apnea when vigorous inspiratory efforts produced large intraluminal subatmospheric inspiratory pressures. From these studies weconclude that relatively isolated palatal or hypopharyngeal narrowing or obstruction is a common picture in OSA. Because the uvulopalatopharyngoplasty was performed by different surgeons in our study, it is possible the surgical technique varied. However, all surgeons who operated on these patients used the technique described by Fujita and colleagues (16, 17). In addition, no surgeons operated exclusively on palatal or hypopharyngeal obstructors. To have studied patients operated on by only one surgeon would have been ideal, but not practical. With the availability of CPAP therapy, only a small proportion of OSA patients now elect surgery. Therefore, in order to accumulate enough patients to perform this analysis, we had to study patients operated on by more than one surgeon. During the study nights patients slept only 2 to 3 h. With three catheters and a face mask in place, it is difficult for patients, even hypersomnolent patients, to tolerate this degree of instrumentation and sleep for long periods. Even after removal of the catheters and mask it was difficult for patients to sleep, partially because of the pharyngeal irritation produced by the catheters. Fortunately we obtained Stage 2 and REM sleep in all the patients.
Physiologic Changes within the Upper Airway Following Uvulopalatopharyngoplasty Our results indicate that the characteristics of airflow limitation within the upper airway changed following uvulopalatopharyngoplasty. In those with preoperative narrowing or obstruction within the transpalatal airway, primarily the obstructive component of the process, but not the hypopneic portion, improved following surgery. Thus following uvular and palatal resection, the periodic pattern of breathing continued with
946
episodic trans palatal narrowing rather than obstruction. Postoperatively in the hypopharyngeal obstructors the site of narrowing or obstruction shifted to the transpalatal airway. The reason for this transition is unclear. Using computed tomography in awake OSA patients, Shepard and Thawley found the crosssectional area of the most rostral portion of the transpalatal airway to be more narrowed following surgery than before uvulopalatopharyngoplasty (21). Inflammation produced by the operative procedure in this region may result in scarring. Scar retraction would tend to narrow the pharynx at this site. The typical pattern of breathing during sleep was different in the hypopharyngeal obstructors, with a persistence of apneas and improvement in hypopneas. Weassume that those hypopharyngeal obstructors who improved with surgery had preoperative narrowing or obstruction within the oropharynx at the tonsillar level, even though this narrowing was not evident on oral examination. It is possible that the tonsillar area in these patients was located below the oropharyngeal catheter pressure ports; these patients were therefore classified as hypopharyngeal obstructors. If resection of tonsillar tissue was instrumental in their surgical response, they would be hypopharyngeal responders. It is likely that those hypopharyngeal obstructors who did not improve or worsened after surgery had narrowing or obstruction more caudally in the hypopharynx preoperatively. The hypopharyngeal caliber would not be expected to increase with surgery at the palatal level; in fact, Shepard and Thawley showed that the hypopharynx narrowed following uvulopalatopharyngoplasty (21).
Importance of Determining the Site of Obstruction in OSA Management This study raises the possibility that the proportion of patients improving with uvulopalatopharyngoplasty would increase if this procedure were performed primarily on those patients with transpalatal obstruction. It would be easier if this determination could be made during wakefulness. However, studies concluded to date do not support that this determination can be made during wakefulness. Sher and coworkers found that patients who had collapse of their palatal airway with a Muller maneuver had a good response to uvulopalatopharyngoplasty (22). However, no patients with hypopharyngeal obstruction were studied, so that no comparison was made be-
HUDGEL, HARASICK, KATZ, WITT, AND ABELSON
tween the surgical results in the palatal or the hypopharyngeal obstructors. Wittig and colleagues resolved this problem and compared the response to uvulopalatopharyngoplasty in patients who had awake upper airway collapse isolated to the palatal airway with the surgical response of patients with both palatal and hypopharyngeal collapse (23). These investigators found no difference in the surgical results betweenthe groups. Only apnea and apnea/hypopnea indices were evaluated in this study, however. Therefore, to date these fewstudies do not easily differentiate surgical responders from nonresponders during wakefulness. During sleep using somnofluoroscopy, Walshand coworkersfound that those subjects with isolated obstruction at the palatal level had a better response than those with hypopharyngeal obstruction (24). Our study expands on this by incorporating direct measurements of airflow obstruction. Those with specific hypopharyngeal obstruction may respond more favorably to surgical techniques that increase the cross-sectional area of the hypopharynx, such as mandibular or hyoid bone advancement or a combination of these two procedures (25-27); those with transpalatal obstruction should continue to respond to UPP. It is hoped that with improved definition of the site of obstruction during sleep and the use of more specific operative techniques the results of surgical treatment of OSA will improve substantially. References 1. Hudgel DW. Variable site of airway narrowing among obstructive sleep apnea patients. J Appl Physiol 1986; 61:1403-9. 2. Chaban R, Cole P, Hoffstein V. Site of upper airway obstruction in patients with idiopathic obstructive sleep apnea. Laryngoscope 1988;98:641-7. 3. Guilleminault C, Hayes B, Smith L, Simmons FB. Palatopharyngoplasty and obstructive sleep apnea syndrome. Bull Eur Physiopathol Respir 1983; 19:595-9. 4. Silvestri R, Guilleminault C, Simmons FB. Palatopharyngoplasty in the treatment of obstructive sleep apnea. In: Guilleminault C, Lugaresi E, eds. Sleep/wake disorders: natural history, epidemiology and long-term evolution. New York: Raven Press. 1983: 163-70. 5. Simmons FB, Guilleminault C, Silvestri R. Snoring, and some obstructive sleep apnea can be cured by oropharyngeal surgery. Arch Otolaryngol (Stockh) 1983:503-7. 6. Zorick F, Roehrs T, Conway W, Fujita S. Wittig R, Roth T. Effects of uvulopalatopharyngoplasty on the daytime sleepiness associated with sleep apnea syndrome. Bull Eur Physiopathol Respir 1983; 19:600-3. 7. Rivilin J, Hoffstein V,Kalbfleisch J. Upper airway morphology in patients with idiopathic obstructive sleep apnea. Am Rev Respir Dis 1984; 129:355-60. 8. Conway W, Fujita S, Zorick F, et at.
Uvulopalatopharyngoplasty. One-year followup. Chest 1985; 88:385-7. 9. deBerry-Borowiecki B, Kukwa AA, Blanks RHI. Indications for palatopharyngoplasty, Arch Otolaryngol 1985; 111:659-63. 10. Fujita S, Conway WA, Sickleested JM, et al. Evaluation of the effectiveness of uvulopalatopharyngoplasty, Laryngoscope 1985; 95:70-4. 11. Gislason T. Lindholm C-E, Almquist M, et a/. Uvulopalatopharyngoplasty in the sleep apnea syndrome. Arch Otolaryngol Head Neck Surg 1988; 114:45-51. 12. Hudgel DW, Martin RJ, Johnson B, Hill P. Mechanics of the respiratory system and breathing pattern during sleep in normal humans. J Appl Physiol 1984; 56:133-7. 13. Hudgel DW, Hendricks C. Palate and hypopharynx - sites ofinspiratory narrowing of the upper airway during sleep. Am Rev Respir Dis 1988; 138:1542-7. 14. Rechtschaffen A, Kales A. A manual of standardized terminology, technique and scoring system for sleep stages of human subjects. Washington DC: National Institutes of Health, 1986(Publ. 204). 15. Muller N. Volgyesi G. Becker L, Bryan MH, Bryan AC. Diaphragmatic muscle tone. J Appl Physiol 1979; 47:279-84. 16. Fujita S, Conway W, Zorick F, Roth T. Surgical correction of anatomic abnormalities in obstructive sleep apnea syndrome: uvulopalatopharyngoplasty, Otolaryngol Head Neck Surg 1981; 89:923-34. 17. Ikematsu T, Fujita S, Simmons FB. Uvulopalatopharyngoplasty: Variations. In: Fairbanks DNF, Fujita S, Ikematsu T, Simmons FB. eds. Snoring and obstructive sleep apnea. New York: Raven Press, 1987; 129-70. 18. Ikematsu T. Study of snoring, 4th report. Therapy. Jpn Oto-Rhino-Laryngol 1964; 64:434-5. 19. Wiegand L, Zwillich CW, White DP. Collapsibility of the human upper airway during normal sleep. J Appl Physiol 1989; 66:1800-8. 20. Stein MG, Gamsu G. DeGeer G, Golden JA, Crumley RL, Webb WR. Fast-CT in obstructive sleep apnea. AJR 1987; 148:1069-74. 21. Shepard JW Jr, Thawley SE. Evaluation of the upper airway by computerized tomography in patients undergoing uvulopalatopharyngoplasty for obstructive sleep apnea. Am Rev Respir Dis 1989; 140:711-6. 22. Sher AE, Thorpy MJ. Spielman AJ, Shprintzen RJ, Burack B, McGregor PA. Predictive value of Muller maneuver in selection of patients for uvulopalatopharyngoplasty. Laryngoscope 1985; 95:1483-7. 23. Wittig R, Fujita S, Fortier J, Zorick F, Potts G. Roth T. Results of uvulopalatopharyngoplasty (UPPP) in patients with both oropharyngeal and hypopharyngeal collapse on Muller maneuver. Sleep Res 1988; 17:269. 24. Walsh JK, Katsautonis GP. Somnofluoroscopy as a predictor of UPPP efficacy. Sleep Res 1984; 13:21. 25. Bear SE, Priest JH. Sleep apnea syndrome: Correction with surgical advancement of the mandible. J Oral Surg 1980; 38:543-9. 26. Riley R, Powell NB, Guilleminault C. Inferior sagittal osteotomy of the mandible with hyoid myotomy-suspension: a new procedure for obstructive sleep apnea. Otolaryngol Head Neck Surg 1986; 94:589-93. 27. Riley R, Powell NB, Guilleminault C. et al. Maxillary, mandibulory and hyoid advancement: an alternative to tracheostomy in obstructive sleep apnea syndrome. Otolaryngol Head Neck Surg 1986; 94:587-8.
