© 2003 American Thoracic Society
A Randomized Trial of Laser-assisted Uvulopalatoplasty in the Treatment of Mild Obstructive Sleep ApneaDepartments of Medicine and Otolaryngology, University of Western Ontario, London, Ontario, Canada Correspondence and requests for reprints should be addressed to Dr. Kathleen A. Ferguson, London Health Sciences Centre, 375 South Street, London, ON, N6A 4G5, Canada. E-mail: kafergus{at}uwo.ca
Laser-assisted uvulopalatoplasty (LAUP) is an outpatient surgical treatment for snoring and obstructive sleep apnea (OSA), but to date, no controlled trials have been published. Forty-five subjects with mild OSA (apnea/hypopnea index [AHI], 10–27 per hour) were randomized to LAUP or to no treatment (control group). The AHI post-LAUP was reduced by 21% overall and to 10 or less per hour in 5 of 21 subjects (24%). Four of 24 subjects in the control group (16.7%) had an AHI of 10 or less per hour at outcome. The AHI decreased with the LAUP compared with no change with the control group at outcome. Ten subjects (48%) reported significantly improved snoring after the LAUP. There was no improvement in excessive daytime sleepiness, but there was a small improvement in quality of life (unless side effects were included in the quality of life score). Side effects were common, but serious complications did not occur. LAUP surgery is effective in some subjects with mild OSA for the treatment of snoring, but the reduction in AHI and the level of symptomatic improvement were minor overall.
Key Words: obstructive sleep apnea • randomized trial • laser surgery • laser-assisted uvulopalatoplasty
There are a number of options for the treatment of snoring and obstructive sleep apnea (OSA). Patients with moderate to severe OSA are usually treated with nasal continuous positive airway pressure (1). Although continuous positive airway pressure eliminates apnea in virtually all cases, there are problems with patient acceptance of continuous positive airway pressure therapy, particularly in patients with mild OSA. Oral appliances are safe and reasonably effective in patients with mild OSA, but they also represent a control and not a cure (2, 3). Patients who desire a cure for their problem generally look to surgical treatment. The two most common surgical procedures are uvulopalatopharyngoplasty (UPPP) and laser-assisted uvulopalatoplasty (LAUP). Kamami developed the LAUP procedure as an alternative to UPPP (4). He first used it in patients who complained of snoring after UPPP and who still had residual vibrating tissue on their soft palate (5). Subsequently, the procedure was expanded to patients who wanted their snoring treated without undergoing a major procedure under general anesthesia. Despite its widespread use, LAUP has not been established as an effective therapy for OSA in the setting of controlled clinical trials, and in uncontrolled trials, the results are variable (6). Many clinicians and patients are convinced of its effectiveness in relieving snoring based on their personal experience, but rarely is treatment success objectively documented. A published guideline focused on the lack of controlled studies and recommended that patients should be told that the risks, benefits, and complications of the procedure are not known (7). In an update published in 2000, the authors stated that no controlled trials were found in the peer-reviewed literature, and they recommended that LAUP not be done for the treatment of OSA (8). However, it could be used for the treatment of snoring. In general, the studies published to date have been retrospective case series or prospective precomparisons and postcomparisons without control groups and often with incomplete follow-up. Problems with many of these studies include poorly detailed methods and selection criteria, inadequate overnight monitoring (before and after surgery), mainly subjective follow-up, and incompletely defined outcome criteria. Other deficiencies of this literature have been well described previously (9). The objective of this study was to determine with a randomized, controlled trial the efficacy of LAUP in relieving OSA in patients with mild symptomatic sleep apnea. The secondary objectives were to determine the effect of LAUP on symptoms of OSA, including snoring, daytime sleepiness, and quality of life.
Forty-five subjects who had mild OSA (apnea/hypopnea index [AHI], 10.1–25) and complained of loud snoring were recruited. Weight, body mass index, blood pressure, and neck circumference were recorded at baseline. A detailed questionnaire was administered along with the Epworth Sleepiness Scale (10) and the Calgary Sleep Apnea Quality of Life Index (SAQLI) (11). The five domains of the SAQLI include domain A daily functioning, domain B social interactions, domain C emotional functioning, domain D symptoms, and domain E treatment-related symptoms (side effects and complications). Snoring intensity and frequency were rated on a modified four-point Likert Scale (never, softly, moderately, loudly and never, rarely, sometimes, often respectively). The bed partner or household member rated snoring intensity (0 = none to 10 = loud) and frequency (0 = never to 10 = always) with a visual analog scale. Using a table of random numbers, subjects were randomized to either the LAUP group or the no-treatment group (control group). The control subjects were not offered any therapy but were offered LAUP at the end of the study. Questionnaires, scales, and the polysomnogram were repeated 3 months after the last LAUP procedure or 6 months after baseline in the control group. The frequency and severity of treatment side effects and complications in the LAUP group were recorded.
Overnight Polysomnography
LAUP Procedure Subjects received preoperative pain medication. Topical anesthesia was applied liberally, and lidocaine was injected into the uvula and soft palate. A series of full-thickness vertical trenches were created with the CO2 laser on the free edge of the soft palate on either side of the uvula. The uvula was shortened and thinned to a fraction of its former size. The edge of the soft palate was also reduced. The tonsils were not treated. A backstop on the laser attachment was used to prevent injury to the posterior pharyngeal wall. The laser was activated during exhalation, and suction was used to remove the smoke from the surgical field. Subjects received postoperative antibiotics, analgesics, antiinflammatories and dilute hydrogen peroxide gargles for 7 days.
Outcome Measurements
The sample size was calculated with data from published studies of LAUP (12–16). The study had a 90% power to detect a 25% reduction in AHI in the treatment group compared with no reduction in AHI in the control group with an
Statistical Analysis
Forty-six subjects with mild OSA (AHI, 10.6 to 26.9 per hour) (data as mean ± SD [range] unless otherwise specified) were recruited for this study, including 35 men and 11 women. There were 21 subjects randomly assigned to the LAUP group and 25 subjects to the control group. One woman assigned to the control group dropped out because of the development of a serious unrelated illness. Data are reported on the remaining 45 subjects (21 LAUP and 24 control subjects), all of whom completed the study and had follow-up polysomnography. The subjects in general were middle aged (44.6 ± 8.1 years [31 to 65]), overweight (body mass index, 31.6 ± 4.5 kg/m2 [23.5 to 43.5]), and had complaints of excessive snoring. There were no differences between the LAUP and control groups at baseline with respect to demographic variables or symptoms. One patient in the control group and one patient in the LAUP group had previously used nasal continuous positive airway pressure without success. No patient had previously had palatal surgery. The LAUP subjects had an average of 2.4 ± 0.98 procedures (one to four). The outcome sleep study and final evaluation were performed 8.2 ± 2.7 months (3.7 to 14.1) after the baseline study in the control subjects and 15.4 ± 5.5 months (8.2 to 32.2) after the baseline study in the LAUP subjects. This was 7.2 ± 5.9 months (2 to 27.4) after the final laser procedure.
Clinical Features and Symptoms
Quality of Life There was no change in quality of life in the control subjects from baseline to outcome. The SAQLI score for domains A, B, C, and D increased from 4.17 ± 0.81 to 4.88 ± 0.8 (p = 0.003) in the LAUP subjects after treatment. The outcome SAQLI was measured at the time of the follow-up polysomnogram (in the LAUP subjects, 7.2 ± 5.9 months after the final procedure). There were statistically significant increases in the daily functioning domain A and social interaction domain B (mean increases 0.55, p < 0.005, and 0.56, p = 0.03, respectively) but no change in emotional functioning domain C. The absolute increases in domains A and B are small. There was a significant improvement in the symptom domain D with a mean increase of 1.26 (p = 0.001) that would be considered a clinically important improvement. In the control subjects, there was a 0.1 increase in the symptom domain D that was less than the increase in the LAUP group (p = 0.04). When side effects or complications of LAUP (domain E) were taken into consideration, the post-LAUP SAQLI score was increased to 4.56 ± 0.9 from baseline—a nonsignificant increase (p = 0.09) (Table 2). There was no change in the SAQLI in the control group.
Polysomnographic Variables
In the control group (no therapeutic intervention), 4 of 24 (16.7%) were treatment successes (AHI, 10 or less), and 17 of 24 (70.8%) were treatment failures (AHI, more than 10). Although the AHI was lower at outcome in some patients with no intervention, no patient had a resolution of symptoms; hence, there were no complete responses in the control group. The pretreatment AHI was 16.1 ± 4.0 per hour. At outcome in the control group, it increased to 22.7 ± 15.2 per hour, but this increase was not significant (Table 2 and Figure 1). There was no difference in sleep architecture, awakenings, or sleep efficiency with the LAUP or in the control group compared with pretreatment values and no difference in the magnitude of the changes.
Side Effects and Complications
Patient Satisfaction
The results of this randomized controlled trial of LAUP in mild OSA indicate that LAUP is effective in some patients for the treatment of mild OSA. Twenty-four percent of subjects had a reduction in AHI to 10 or less, and 4 of 21 (19%) had a complete response (in terms of AHI and a satisfactory resolution of symptoms. Overall, the reduction in AHI was small. Subjective snoring was significantly better in some subjects, but excessive daytime sleepiness and quality of life were not improved (when treatment related side effects were taken into consideration). Untreated subjects in the control group had no change in the AHI. If the reduction in AHI were related to night-to-night variability or to regression to the mean, then a similar change would have been expected in the control group. This study does not support the routine use of LAUP in patients with mild OSA because of the limited efficacy of this treatment. There are other ways to present surgical success rates. When success is defined as a more than 50% reduction in AHI, then LAUP was effective in 19% of our subjects. Kamami reported one of the earliest studies on LAUP in OSA on a group of 63 patients (14). The number and extent of the LAUP procedures was not given, and the length of follow-up was not stated. An unspecified number of subjects underwent laser turbinectomy in addition to LAUP. The study reported that 55 of 63 subjects (87%) had a more than 50% reduction in AHI with surgery. Other studies have not replicated this high success rate. Most report that 31 to 58% of subjects have a more than 50% reduction in AHI (17–20). Clearly, this is not an optimal definition of success because the post-treatment AHI could still be clinically important even if it is reduced 50% from baseline. In studies that reported the number of subjects with an AHI of less than 10 per hour after LAUP, the range of success was between 24 and 44% (6, 17, 18, 21). The results of this randomized study are similar to these precomparison and postcomparison studies with 24% of the subjects having a post-LAUP AHI of 10 or less per hour. Our success rate with snoring (48% improved) was lower than in many other LAUP studies. Most studies report, at least initially, that approximately 50 to 95% of patients have a major improvement in or a cure of their snoring (12–15, 17, 22, 23). In some of these studies, patients had additional surgery to LAUP, and how snoring improvement was measured was not always defined. One potential explanation for our relatively low snoring success rate was that we followed up with all subjects in the trial, and our follow-up was on average 7.2 months after the last LAUP procedure. In the immediate postsurgical period, many subjects in our study reported major reductions in snoring intensity and frequency, but this was not always sustained. This has been shown in other studies where long-term follow-up took place (17, 20, 23, 24). It is possible that success of the LAUP procedure in this study may have been less beneficial for snoring and the reduction in AHI if we had followed up with patients 1 to 2 years after the procedure. It is a limitation of our study that we did not objectively measure snoring. However, even if there was an objective improvement in snoring intensity, it is more relevant clinically that the bed partner is satisfied with the reduction in snoring, and in this study, snoring was rated as improved in 48% of subjects after LAUP. In one study where snoring was objectively measured, the snoring index was not significantly different in the group after LAUP (6). It is known that there may be a significant discrepancy between the objective and subjective outcome of UPPP for snoring improvement when measured many months after UPPP (25). Miljeteig and colleagues sent questionnaires to 100 patients after UPPP (25). Sixty-nine patients replied to the questionnaire, and 78% reported that their snoring was reduced or cured. When snoring was measured during polysomnography, there was no difference in measured snoring (loudness or snoring index) or AHI an average of 13 ± 15 months after the UPPP. Patients may therefore report better snoring results than have actually occurred. Only one patient (5%) had a significant worsening of OSA after laser surgery, and this was in the setting of a large weight gain (7 kg). Other studies have found that a significant proportion of patients, approximately 20 to 25%, have worsening of the AHI after the procedure (6, 15, 16, 18, 22, 23). This is a concerning result, as many patients do not undergo adequate objective follow-up and there is too heavy of a reliance on patient self-report. Follow-up sleep studies are therefore essential to establish objectively treatment outcome after LAUP. A potential criticism of this study is that it had a small sample size. However, the study was adequately powered to detect a 25% improvement in AHI in the LAUP group and follow-up was complete. Any less of an improvement is unlikely to be clinically relevant. Another potential criticism is that the study evaluated only one surgical technique. We did not find that the transpalatal trench approach with subtotal uvulectomy and a reduction of soft palate was an effective treatment for OSA, and the effect on snoring was modest. The patients did have an average of 2.4 ± 1 procedures each. Six subjects had a poor clinical response, and the surgeon felt that no more tissue could be safely removed. Three patients had a poor result and refused more surgery. Perhaps more surgery might have helped some of the patients with a suboptimal result. It is possible that the LAUP procedure that is more similar to UPPP would have better results. Our subjects complained vigorously about the amount of pain from surgery during the first several days after surgery and complications such as a change in swallowing occurred even with the lesser surgical procedure. The more extensive procedures are likely to have more side effects and complications than the more limited approach. Furthermore, the overall success of UPPP for OSA is only 40.7% when response was defined as a reduction in AHI of 50% with a postoperative AHI of less than 20 or a postoperative AHI of less than 10 (26). It is unlikely that the more aggressive LAUP procedure would equal or exceed the results of UPPP.
Conclusion
Supported by the Ontario Thoracic Society; the Department of Medicine, University of Western Ontario; and the Department of Otolaryngology, University of Western Ontario, London, Canada. This article has an online supplement, which is accessible from this issue's table of contents online at www.atsjournals.org Received in original form August 10, 2001; accepted in final form August 6, 2002
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
of 0.05. The University Of Western Ontario Review Board for Health Sciences Research Involving Human Subjects approved the study. 
