Chest ACCP Education Calendar
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Guest Access | Sign In via User Name/Password
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Millman, R. P.
Right arrow Articles by Kramer, N. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Millman, R. P.
Right arrow Articles by Kramer, N. R.
(Chest. 2000;118:1025-1030.)
© 2000 American College of Chest Physicians

Simple Predictors of Uvulopalatopharyngoplasty Outcome in the Treatment of Obstructive Sleep Apnea*

Richard P. Millman, MD, FCCP; Carol C. Carlisle, RN; Cynthia Rosenberg, DDS; Douglas Kahn, DO; Robert McRae, MD and Naomi R. Kramer, MD

* From the Departments of Medicine (Drs. Millman, Kahn, Kramer, and Ms. Carlisle), Otorhinolaryngology (Dr. McRae), and Plastic Surgery (Dr. Rosenberg), Brown University School of Medicine, Providence, RI.

Correspondence to: Richard P. Millman, MD, FCCP, Division of Pulmonary, Sleep, and Critical Care Medicine, APC 479A, Rhode Island Hospital, 593 Eddy St, Providence, RI 02903; e-mail: rmillman{at}lifespan.org


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Study objective: Our objective was to determine whether baseline polysomnography, cephalometry, and anthropometry data could predict uvulopalatopharyngoplasty (UPPP) success or failure.

Design: We retrospectively reviewed polysomnography, cephalometry, and anthropometry data from patients who underwent UPPP for obstructive sleep apnea (OSA).

Setting: A university medical center.

Patients: OSA was diagnosed by polysomnography in 46 patients who underwent UPPP surgery for their sleep disorder.

Interventions: UPPP surgery with/or without tonsillectomy.

Measurements and results: The mean patient age was 43 years, and the mean body mass index was 32.5 kg/m2. The mean presurgical apnea-hypopnea index (AHI) was 45, and the mean baseline nadir oxygen saturation was 81%. Successful surgery was defined as a reduction in AHI to < 10 or to < 20 with a 50% reduction from the patient’s baseline AHI. Of the 46 patients, 16 were successfully treated and 30 did not respond to surgical treatment. A mandibular-hyoid distance (MP-H) > 20 mm was found to be significantly (p = 0.05) predictive of failure of UPPP. When stepwise regression analysis was performed utilizing postsurgical AHI as the dependent variable and presurgical AHI, age, body mass index, baseline nadir O2 saturation, and five cephalometric measurements as independent variables, MP-H distance significantly (r = 0.524; p = 0.01) correlated positively with postsurgical AHI. The distance between the superior point of a line-constructed plane of the sphenoidale (parallel to Frankfort horizontal) and a point at the intersection of the palatal plane perpendicular to the hyoid correlated negatively with postsurgical AHI (r = 0.586; p = 0.05). By creating a logistic model of this data, an MP-H distance < 21 mm, an angle created by point A to the nasion to point B < 3°, and the presence of a baseline AHI < 38 enhanced the predictability of UPPP success.

Conclusions: The presence of a baseline AHI < 38 and an MP-H <= 20 mm, and the absence of retrognathia are predictors of improvement after UPPP. Based on these findings, we would advocate the continued evaluation of cephalometric measurements and careful consideration of surgical treatment options for OSA.

Key Words: apnea-hypopnea index • obstructive sleep apnea • outcome • uvulopalatopharyngoplasty


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Uvulopalatopharyngoplasty (UPPP) has been used as a surgical procedure to treat obstructive sleep apnea (OSA) since it was first described by Fujita et al.1 Although subjectively it is thought to be a curative procedure by many otorhinolaryngologists and their patients, the actual success rate for the procedure has been found to be only 40.7%.2 Multiple techniques have been employed to try to predict surgical success, but there is no consensus available on whether any of them are significantly predictive.2

We analyzed cephalometric, anthropometric, and polysomnographic variables obtained prior to UPPP to determine whether any variables were helpful in predicting the response of patients in our population to surgery.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Patient Selection
Our group retrospectively reviewed the medical records of consecutive patients from the Rhode Island Hospital Sleep Disorders Center who had undergone polysomnography and UPPP. These records were taken from our clinical database for the years 1991 to 1999. Ninety patients were identified who had undergone some form of surgery for their OSA and had completed baseline cephalometric and polysomnographic studies. Of these patients, 21 were excluded for lack of a follow-up sleep study. Fourteen patients had undergone UPPP at the same time as a genioglossal/mandibular advancement procedure. Nine patients with a baseline apnea-hypopnea index (AHI) < 10 who had been operated on primarily for snoring were not thought to have sleep apnea significant enough to be included in the analysis. Therefore, complete patient information was available on 46 patients who had undergone UPPP. All patients had either refused nasal continuous positive airway pressure or had tried it and not been able to tolerate it.

Polysomnography
Polysomnography included modified EEG monitoring, a bilateral electrooculogram, and a submental, intercostal, and anterior tibialis electromyogram. Airflow was monitored by oral and nasal thermistors. Respiratory effort was monitored with chest wall and abdominal piezoelectric bands as well as with intercostal electromyogram. Cardiac monitoring was conducted via modified V1 telemetry, and continuous oxygen saturation was monitored by pulse oximetry. We defined events as previously described by our center.3 Obstructive apneas were defined as the total cessation of airflow for >= 10 s with continued respiratory effort. An obstructive hypopnea was defined as an event lasting > 10 s that was associated with two of the following conditions: a decrease in airflow of 50%; an EEG arousal; or oxygen desaturation >= 2%. Postsurgical sleep studies were repeated approximately 4 months after surgery.

Cephalometric Measurements
Standard radiographic cephalometry was done with the patient standing in an upright standing position. A lateral projection of the skull was taken (Oralix Ceph; Philips Dental Systems; Shelton, CT). The films were taken at a film focus distance of 5 feet, with the patient’s head secured in a cephalostat. All cephalograms were recorded with the patient in a natural head posture and using a mirror-eye reference position. Exposures were made at 75 kV (peak) and 10 mA at an exposure time of 0.8 to 1.2 s. Patients were instructed not to breathe or swallow during this procedure. All cephalometric radiographs were analyzed by a single orthodontist (CR) who was blinded to the severity of the sleep apnea and surgical outcomes. Selected skeletal and soft-tissue measurements were made. Posterior airway length and other skeletal landmarks were measured by the method of Woodson et al,4 as previously described. Retrognathia was defined as an angle created by point A (see Fig 1 for the cephalometric variables used in this study) to the nasion to point B (ANB) >= 3°.



View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1.. Cephalometric variables. PNS-P = distance from posterior nasal spine (PNS) to point P (soft palate length); PAS = posterior airway space; SNA = angle created from the sella (S) to the nasion (N) to point A; SNB = angle created from the sella to the nasion to point B.

 
Data Analysis
All patient information about the surgical candidates is kept in a dedicated database. All data are presented as mean ± SEM. For analysis, if a patient had multiple sleep studies before undergoing UPPP, the study date closest to the UPPP procedure was used unless it was a continuous positive airway pressure trial. Linear univariate and stepwise regression using postsurgical AHI as the dependent variable was analyzed against baseline polysomnographic, cephalometric, and selected anthropometric variables. Linear regression was performed among all cephalometric and anthropometric variables.

Patients were classified as having a total response to surgery (AHI, < 10) or a partial response (AHI, < 20 [ with a minimum of a 50% reduction in AHI from baseline]). Total and partial responders to surgery were combined into a responder group for all statistical comparisons. Nonresponders were all patients who did not fit into the above categories. Cephalometric and anthropometric measurements were compared between the responder and nonresponder groups using unpaired t tests. p Values determined by Fisher’s Exact Test and {chi}2 test were used to compare the frequency of response vs nonresponse.

Univariate {chi}2 and logistic regression analyses were performed on independent variables against the outcome variable (response vs nonresponse). Variables resulting in univariate p values of <= 0.15 were entered into a multivariate logistic regression model. This resulted in the formation of dichotomous variables for the presence of retrognathia (yes/no), a distance from the hyoid to the mandible perpendicular to the mandibular plane (MP-H) <= 20 mm (or > 20 mm), and a presurgical AHI < 38 (or >= 38). This was based on the natural cutoff for each variable entered.

The Wald test and the likelihood ratio test then were employed to eliminate noncontributory variables from our model. Analyses were performed using statistical software (Stata Statistical Software; Stata Corp; College Station, TX, and Statview, version 4.0 for Macintosh; Abacus Concepts; Berkeley, CA).


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
There were 43 men and 3 women in our study population. Their mean age was 43 ± 1.5 years, and their mean body mass index was 32.5 ± 0.9 kg/m2. The baseline AHI was 44.8 ± 4.3, and the baseline O2 saturation nadir was 80.7 ± 1.4%. These data are similar to those from other studies previously reporting data for our sleep apnea patient population.5 6 Based on the stringent criteria of a postsurgical AHI < 10, our overall positive response to UPPP surgery was 22% (10 of 46 patients). Including those patients included by the definition of partial response, as defined previously, our positive response rate improved to 35% (16 of 46 patients). For the group as a whole, the presurgical to postsurgical AHI statistically improved from 44.8 ± 4.3 to 32.6 ± 4.6, respectively (p < 0.05), and O2 saturation nadir statistically improved from 80.7 ± 1.3 to 85.3 ± 1.1%, respectively (p = 0.0001). Also, for the group as a whole, the presurgical arousal index was 43.3 ± 4.8, and postsurgical arousal index was 31.8 ± 4.9 (difference not significant). For the group with a positive response to surgery, the presurgical arousal index was 37.8 ± 6.9, and postsurgical arousal index was 10.5 ± 1.3 (p = 0.003). However, sleep-stage distribution did not change.

We then attempted to correlate baseline polysomnographic, cephalometric, and anthropometric variables with the response to UPPP. First, using univariate regression, we found that there was a positive correlation between surgical outcome, expressed as post AHI and MP-H, H2-H3, pre-AHI and baseline O2 nadir (Table 1 ). There was no correlation with any other cephalometric variable or with age or body mass index. Table 2 reflects the interrelationships and strong correlations among cephalometric variables MP-H, the distance between the superior point of a line-constructed plane (parallel to the Frankfort horizontal) (H1) and a point at the intersection of the palatal plane perpendicular the hyoid (H2), and the distance from H2 to the hyoidale (H3). Using a stepwise regression model and selected cephalometric variables, 34% of the variability in the postsurgical AHI could be explained by a positive correlation with MP-H and a negative relationship to the H1-H2 distance.


View this table:
[in this window]
[in a new window]

  		Table 1.. Dependent Variable (Postsurgical AHI) Compared to Cephalometric Polysomnographic, and Anthropometric Variables*

 

View this table:
[in this window]
[in a new window]

  		Table 2.. Regression Analysis Between Cephalometric Variables*

 
When categorizing the patients in the group as responders or nonresponders, we found that the MP-H distance was significantly longer in the nonresponder group than in the responder group (26.0 ± 1.0 mm vs 20.8 ± 1.6 mm, respectively; p = 0.006). All other cephalometric and anthropometric variables were not statistically different (Table 3 ).


View this table:
[in this window]
[in a new window]

  		Table 3.. Response vs Nonresponse to UPPP*

 
Seventy-four percent of patients studied (34 of 46 patients) had an abnormal hyoid to MP-H distance > 20 mm. We found the positive response rate to surgery of 20.6% (7 of 34 patients) was significantly smaller in this group when compared to an MP-H <= 20 mm (p = 0.001) where the positive response rate was 75% (9 of 12 patients).

Baseline disease severity had an impact on the surgical response to UPPP as well. In patients with more severe disease who had a presurgical AHI >= 38, the overall response rate was 18% (4 of 22 patients) compared to 50% (12 of 24 patients) in those patients with less severe disease (p < 0.02).

Sixty-five percent (30 of 46 patients) of this patient population had retrognathia as defined by an ANB angle >= 3°. Of these patients, 27% (8 of 30 patients) were treatment responders. Thirty-five percent (16 of 46 patients) did not have retrognathia, and their response rate to surgery was 50% (8 of 16 patients). This difference neared statistical significance (p = 0.11).

For the group as a whole, using multivariate logistic regression, an MP-H <= 20 mm was the only significant predictor of a positive response to surgery (Table 4 ). Although the odds ratios for a presurgical AHI < 38 and the absence of retrognathia as defined by an ANB angle < 3° were not statistically significant, these variables were retained in the model for reasons of clinical judgment and criteria of likelihood testing, and they serve as adjusting variables for MP-H. Using this model, the result of the Hosmer-Lemeshow goodness-of-fit summary {chi}2 test was 0.28 (p = 0.96).


View this table:
[in this window]
[in a new window]

  		Table 4.. Logistic Regression Contributors to Successful Response in UPPP Surgery*

 
Questions were asked of patients regarding weight change, snoring, and daytime sleepiness in follow-up examinations with their sleep physicians. In a majority of patients, snoring was lessened or diminished. One half of the patients reported less or no excessive daytime sleepiness, whether they had received a surgical cure or not. Also, there was no change in weight for the group as a whole (weight prior to surgery, 32.5 ± 0.9 kg/m2; weight after surgery, 32.7 ± 1.0 kg/m2). As was the case with many patients, weight reduction had previously been attempted.


   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
In this article, we reviewed our surgical results in 46 unselected patients who had undergone UPPP (with tonsillectomy if tonsils were present). Thirty-five percent of our patients met our criteria for response based on a postsurgical AHI < 10 or an AHI < 20 with at least a 50% reduction from the baseline index. These results are similar to the 40.7% success rate for the UPPP reported by Sher et al2 in their review of the literature for surgery for OSA. They used a similar definition for success, specifically a postsurgical apnea index < 10 or an AHI < 20 with at least a 50% reduction from baseline.

In our series, the positive response to surgery was directly related to the MP-H distance. In addition, an MP-H <= 20 mm was the single most significant predictor of surgical success. This finding is consistent with the results of some other studies in the literature. Guilleminault et al7 were the first to suggest that a low MP-H distance was consistent with a poor response to the UPPP. That group reported five patients who had undergone UPPP. The MP-H distance in the three patients who did not respond to surgery ranged from 25 to 35 mm. Petri et al8 also showed that an increase in the MP-H distance was associated with a poor surgical outcome.

Woodson and his colleagues4 reported 66 consecutive patients treated by UPPP using the same definition of response as we employed. They were able to separate their group into 32 responders and 34 nonresponders. There was no difference in the MP-H distance between responders and nonresponders. There was a significant difference in posterior airway measurements (H1-H3 distance and H2-H3 distance). Nonresponders were found to have longer posterior airway measurements.

Since H3 is the most anterior superior point on the body of the hyoid, the measurements of H1-H3 and H2-H3 distances are very similar to those of the MP-H distance. In fact, we found a statistical correlation between measurements for MP-H distance and those for H1-H2 and H2-H3 distances (Table 2) . Longer distances of MP-H, H1-H3, and H2-H3 all reflect an inferior displacement of the hyoid bone. Although our results and those of Woodson and his colleagues4 appear to be different, the results of both studies essentially depict the same finding: the lower the hyoid bone, the greater the potential for collapse of the tongue into the hypopharynx during sleep, and the less likely it is that UPPP alone will be a sufficient treatment for the patient’s sleep apnea.

Ryan and his colleagues9 found that alternative cephalometric measurements were helpful in predicting UPPP success. It is difficult to compare that study to the present one since they recorded only apneas and did not report hypopneas.

Other studies have found the cephalometric radiograph to be less useful in predicting surgical success. Gislason and colleagues10 found that no cephalometric variable helped to predict surgical outcome, although the MP-H distance tended to be longer in the nonresponders. Doghramji et al11 did not find any relationship between MP-H distance and surgical outcome, although soft palate length tended to be longer in patients in the responder group. Both studies used much less stringent criteria for surgical response, which could have affected their findings. The UPPP was thought to be successful if the postsurgical AHI was reduced by at least 50%. Thus, patients with persistent moderate disease could have been labeled as responders.

The baseline severity of OSA also had an impact on surgical response. Using an AHI >= 38 as consistent with severe disease, response rates were 18% compared to 50% in patients with less severe disease. Although an early study suggested that patients with an apnea index > 70 were the most appropriate patients for the UPPP procedure,12 subsequent studies have shown that the more severe OSA at baseline, the more negative the impact on surgical outcome, no matter what the definition of cure that was employed.8 10 13

Sixty-five percent of our patient population was retrognathic, which was defined as an ANB >= 3°. The presence of retrognathia also had a negative impact on surgical outcome. The surgical success rate was only 27% in retrognathic patients compared to 50% in the nonretrognathic group. One other study found a 30% success rate in retrognathic patients compared to 60% in nonretrognathic patients, using a 50% reduction in AHI as the definition of success.14

Clearly, UPPP alone does not lead to a good response in many patients with OSA. Most patients, independent of surgical success, reported a decrease in snoring intensity. One half of the total group reported less or no excessive sleepiness. There was improvement in this regard in nonresponders independent of changes in AHI or arousal index, which has been reported previously.15 Persistent sleepiness in responders could perhaps be explained by insufficient sleep or depression since the AHI and arousal index were normalized in these patients.

Although it continues to be very difficult to preoperatively predict UPPP efficacy, this study presents "red flags" that clinicians could look for prior to recommending a surgical option to their patients. Specifically, in our population of patients there was a high chance of failure of UPPP (with or without tonsillectomy) as the sole surgical procedure if the MP-H distance was > 20 mm, if there was cephalometric evidence of retrognathia, or if severe OSA (AHI, >= 38) was present. A cephalometric radiograph is an inexpensive screening tool, and prospective studies at other centers would be helpful to assess its efficacy in other groups of patients.

Patients found to be at high risk for failure might benefit from a genioglossal advancement procedure at the same time that they undergo UPPP.16 Patients who do not respond to UPPP might be candidates for subsequent genioglossal/mandibular advancement surgery or insertion of an adjustable oral appliance.3


   Acknowledgements
 
We recognize Steve Reinert for his assistance with the biostatistics for this study.


   Footnotes
 
Abbreviations: AHI = apnea-hypopnea index; ANB = angle created by point A to the nasion to point B; H1 = superior point of a line constructed plane of the sphenoidale (parallel to the Frankfort horizontal); H2 = point at the intersection of the palatal plane perpendicular to the hyoid; H3 = hyoidale; MP-H = hyoid to mandible perpendicular to the mandibular plane; OSA = obstructive sleep apnea; UPPP = uvulopalatopharyngoplasty

Received for publication October 22, 1999. Accepted for publication May 5, 2000.


   References
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Fujita, S, Conway, W, Zorick, F, et al (1981) Surgical correction of anatomic abnormalities in obstructive sleep apnea syndrome: uvulopalatopharyngoplasty. Otolaryngol Head Neck Surg 89,923-934[ISI][Medline]
  2. Sher, AE, Schechtman, KB, Piccirillo, JF (1995) The efficacy of surgical modifications of the upper-airway in adults with obstructive sleep apnea syndrome. Sleep 18,659-666[Medline]
  3. Millman, RP, Rosenberg, CL, Carlisle, CC, et al (1998) The efficacy of oral appliances in the treatment of persistent sleep apnea after uvulopalatopharyngoplasty. Chest 113,992-996[Abstract/Free Full Text]
  4. Woodson, BT, Conley, SF, Dohse, A, et al (1997) Posterior cephalometric analysis in obstructive sleep apnea. Ann Otol Rhinol Laryngol 106,310-313[Medline]
  5. Millman, RP, Redline, S, Carlisle, CC, et al (1991) Daytime hypertension in obstructive sleep apnea. Chest 99,861-866[Abstract/Free Full Text]
  6. Levinson, PD, McGarvey, ST, Carlisle, CC, et al (1993) Adiposity and cardiovascular risk factors in men with obstructive sleep apnea. Chest 103,1336-1342[Abstract/Free Full Text]
  7. Guilleminault, C, Riley, R, Powell, N (1984) Obstructive sleep apnea and abnormal cephalometric measurements. Chest 86,793-794[Abstract/Free Full Text]
  8. Petri, N, Suadicani, P, Wildschiodtz, G, et al (1994) Predictive value of Muller maneuver, cephalometry and clinical features for the outcome of uvulopalatopharyngoplasty: evaluation of predictive factors using discriminant analysis in 30 sleep apnea patients. Acta Otolaryngol 114,565-571[Medline]
  9. Ryan, CF, Dickson, RI, Lowe, AA, et al (1990) Upper airway measurements predict response to uvulopalatopharyngoplasty in obstructive sleep apnea. Laryngoscope 100,248-253[Medline]
  10. Gislason, T, Lindholm, C, Almqvst, M, et al (1988) Uvulopalatopharyngoplasty in the sleep apnea syndrome. Arch Otolaryngol Head Neck Surg 114,45-51
  11. Doghramji, K, Jabourian, ZH, Pilla, M, et al (1995) Predictors of outcome for uvulopalatopharyngoplasty. Laryngoscope 105,311-314[Medline]
  12. Caldarelli, DD, Cartwright, RD, Lilie, JK (1985) Obstructive sleep apnea: variations in surgical management. Laryngoscope 95,1070-1073[Medline]
  13. Janson, C, Gislason, T, Bengtsson, H, et al (1997) Long-term follow-up of patients with obstructive sleep apnea treated with uvulopalatopharyngoplasty. Arch Otolaryngol Head Neck Surg 123,257-262
  14. Woodson, BT, Conley, SF (1997) Prediction of uvulopalatopharyngoplasty response using cephalometric radiographs. Am J Otolaryngol 18,179-184[Medline]
  15. Miljeteig, H, Mateika, S, Haight, JS, et al (1994) Subjective and objective assessment of uvulopalatopharyngoplasty for treatment of snoring and obstructive sleep apnea. Am J Respir Crit Care Med 150,1286-1290[Abstract]
  16. Riley, RW, Powell, NB, Guilleminault, C (1993) Obstructive sleep apnea syndrome: a review of 306 consecutively treated surgical patients. Otolaryngol Head Neck Surg 108,117-125[ISI][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Millman, R. P.
Right arrow Articles by Kramer, N. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Millman, R. P.
Right arrow Articles by Kramer, N. R.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS