Combined temperature-controlled radiofrequency tongue reduction and UPPP in apnea surgery.

Abstract

The objective of this study was to investigate the effects on outcomes and morbidity of combining temperature-controlled radiofrequency (TCRF) tongue reduction with uvulopalatopharyngoplasty (UPPP) as an initial site-directed approach to the surgical treatment of obstructive sleep apnea syndrome (OSAS). This investigation was a prospective, nonrandomized, open-enrollment study of 20 consecutive eligible patients with OSAS. Seven patients had a single-level velopharyngeal obstruction (Fujita type I or IIa), and they were assigned to undergo UPPP only (group 1). Thirteen patients had a multilevel velopharyngeal and retroglossal obstruction (Fujita type Ilb), and they were assigned to undergo TCRF tongue reduction in addition to UPPP (group 2). Patients who had only a retroglossal obstruction (Fujita type III) were not included in this study. Following their initial operation, nine patients in group 2 underwent two subsequent in-office TCRF tongue treatments under local anesthesia. Three patients in group 2 were lost to followup, and one patient underwent only one TCRF procedure at the initial operation; data on the latter patient are included in some of the outcomes measures reported here where indicated, but no data are reported on the three who were lost to followup. The primary post-treatment outcomes measures were the results of comparative polysomnography and clinician and patient evaluations regarding morbidity and symptom improvement. Analysis of these data showed that there was no difference between the two groups in terms of postoperative pain or dysphagia following the initial operative session. The degree of symptom improvement in the two groups was similar. Overall success rates--as measured by the apnea/hypopnea index (group 1: 57.1% success; group 2: 50.0%) and by the apnea index (group 1: 71.4% success; group 2: 70.0%)--were statistically comparable, given the small size of the sample (figures are based on l0 patients in group 2). The author concludes that combining TCRF tongue reduction with UPPP in patients with multilevel obstruction improves response rates to a degree that is comparable to that seen with UPPP alone in patients with single-level velopharyngeal obstruction. Moreover, the combination treatment does not increase the risk of additional pain, morbidity, and complications compared with UPPP alone.

Introduction

Over the past 2 decades, uvulopalatopharyngoplasty (UPPP) has become one of the more frequently used surgical treatments for obstructive sleep apnea syndrome (OSAS). Its relative technical simplicity and the usually uncomplicated postoperative course have made it a popular option for most otolaryngologists. However, despite encouraging reports by patients of subjective quality-of-life improvements, outcomes have been disappointing. Among patients with retropalatal narrowing or collapse (Fujita type I or IIa obstruction), the clinical response to UPPP (as defined by a reduction in the apnea/hypopnea index [AHI] of [greater than or equal to] 50% and a post-treatment AHI of < 20 in patients whose AHI was >20 preoperatively) has been reported to be 52.3%. (1) But for patients with a retrolingual narrowing or collapse, with or without a retropalatal component (Fujita type IIb or III obstruction), the UPPP response rate is only 5.3%. (1) Although the Fujita classification system is an oversimplified representation of airway compromise in OSAS, it does indicate the presence of obstruction at multilevel sites, each of which must be addressed surgically.

Several approaches have been advocated to improve the low UPPP response rate among patients with Fujita type IIb obstruction, who make up the largest group of OSAS patients. They include limited tongue resection, midline glossectomy/epiglottidectomy, lingualplasty, genioglossal advancement, and hyoid myotomy with suspension. (2) Yet most otolaryngologists are reluctant to undertake these additional, highly invasive procedures. Their reasons include a lack of confidence that such a procedure will be successful, a fear of complications and increased morbidity, and the need for a high degree of technical expertise. (3)

One procedure that has been shown to be safe, effective, well tolerated, and technically straightforward in reducing retroglossal obstruction and collapse is temperature-controlled radiofrequency (TCRF) tongue reduction (Somnoplastyr). In a study of 18 patients who underwent TCRF tongue reduction, Powell et al reported a mean decrease of 17% in tongue volume following an average delivery of 8,490 J to the base of the tongue. (4) The procedure lowered the mean respiratory disturbance index from 39.6 to 17.8 and raised the mean Sa[O.sub.2] nadir from 81.9 to 88.3%.

The technical simplicity and relative tolerability of TCRF tongue reduction could make it an appealing option for practitioners and for patients with multilevel obstruction as an initial site-directed operation in combination with UPPP. If this combination ultimately proves to be successful and increases the percentage of successful initial responders, it would decrease the size of the patient population that requires a graduated treatment plan and more-invasive operations.

The present feasibility study was undertaken to investigate this concept. Its goal was answer two questions:

* Does the addition of TCRF tongue reduction to UPPP add to the risk of adverse effects or to the degree of discomfort?

* Will the combination increase the treatment success rate in patients with multilevel retropalatal and retro-glossal obstruction to a level that is comparable to that achieved by UPPP alone in patients with single-level retropalatal obstruction?

Patients and methods

The present investigation was a prospective, nonrandomized, open-enrollment study of 20 consecutive patients who were seen at the author's community-based general otolaryngology practice. Eligible patients had polysomnographically documented OSAS, were seeking a surgical treatment option, and met the inclusion criteria established by an institutional review board-approved protocol. Inclusion was not restricted by the severity of OSAS or by the patient's weight, age, or sex.

Based on physical examination with supine pharyngoscopy and cephalometric x-ray imaging, the study population was divided into two groups. Seven patients (mean age: 35.8 yr; mean body-mass index [BMI]: 32.3; six men) had primarily single-level velopharyngeal obstruction (Fujita type I or IIa), and they were designated as group 1. Thirteen patients (mean age: 51.5 yr; mean BMI: 27.9; 12 men) had multilevel velopharyngeal and retroglossal obstruction (Fujita type IIb), and they were designated as group 2. Patients who had only retroglossal obstruction (Fujita type III) were not included in this study.

The criteria used to define a significant retroglossal obstruction included (1) an inability on the part of the surgeon to visualize no more than the arytenoids because of tongue enlargement during unforced end-expiration on supine pharyngoscopy with the endoscope in the oropharynx (5) and (2) the presence of a posterior airway space of 10 mm or less on end-expiration cephalometric x-ray analysis. (6) Over the course of the study, three patients in group 2 were lost to followup.

Treatment consisted of an initial operative procedure under general anesthesia in which all patients underwent traditional surgical UPPP, along with removal of the tonsils (if present), septoplasty (if indicated), and turbinate reduction (either surgical submucosal resection or TCRF tongue--reduction, depending on the type of turbinate obstruction). In addition, patients in group 2 underwent their first TCRF tongue-reduction procedure during that initial operative session. All patients were hospitalized overnight.

Subsequently, nine of the 10 remaining patients in group 2 underwent two additional in-office TCRF tongue reductions under local anesthesia; these repeat TCRF tongue-reduction procedures were spaced 3 to 6 weeks apart. (The remaining patient underwent only the initial TCRF procedure [4,000 J] and required no further treatment; the data on this patient are included in some outcomes measures where indicated.) During all TCRF procedures, two to six submucosal lesions per session were created by delivering 1,000 J of energy to each site at 85[degrees] C with an S2 control unit (set to the maximum power setting of 10 W) and a commercially available model 1200 tongue handpiece (Somnus Medical Technologies; Sunnyvale, Calif.). Each of the nine patients received a cumulative dose of approximately 12,000 J.

Initial followup examinations and questionnaires were administered 1 to 3 days following each phase of treatment, and again at 1, 4, and 8 weeks. The questionnaires gathered information on treatment morbidity and the status of OSAS symptoms. Responses were quantified by using the Epworth sleepiness scale and visual analog scales for daytime sleepiness, snoring, dysphonia, dysphagia, and throat pain (assessments of pain in general as well as pain that was perceived to be localized to the palate, nose, and tongue). Attended sleep studies were repeated 2 months following the completion of all treatment.

Results

There were no clinically significant complications in either group. Neither group experienced hemorrhage, clinically significant infections, or respiratory problems, and no patient required hospitalization beyond the initial postoperative short stay. The two groups exhibited similar degrees of temporary nasal congestion, which is typical following nasal procedures, as well as initial oropharyngeal pain and dysphagia.

Subjective assessments of outcome. Following the initial operative session, there was generally no significant difference between the two groups in the degree of postoperative discomfort (table). The one exception wasthat group 2 experienced more palatal pain at postoperative week 1, although not before or after.

The subsequent in-office TCRF tongue reductions administered to the nine patients in group 2 under local anesthesia were well tolerated, and most of them were able to return to normal activity within 24 hours (table). All of these patients experienced mild tongue edema and pain 1 to 3 days following their additional treatments, but these disturbances resolved within I week.

All patients in both groups, including the one patient in group 2 who underwent only one TCRF procedure, reported subjective improvement in their OSAS symptoms (figure). The difference in improvement between the two groups was not statistically significant, given the small sample size.

[FIGURE OMITTED]

Apnea/hypopnea index. Objective assessments of response were based on individual AHI and apnea index (AI) values. A successful response in terms of AHI was defined as a reduction of at least 50% in AHI and the attainment of a post-treatment AHI of less than 20 in those patients whose preoperative AHI was greater than 20.

As determined by the AHI, success rates in the two groups were comparable; tour of the seven patients in group 1 (57.1%) and five of the 10 in group 2 (50.0%) were considered to have experienced a successful response to therapy. The mean AHI values in group 1 fell from 47.2 ([+ or -]29.1) before treatment to 20.5 ([+ or -]23.6) afterward; the corresponding decline in group 2 was from 29.5 ([+ or -]14.8) to 18.8 ([+ or -]14.6). The difference between the two groups in the mean post-treatment AH1 was not statistically significant. The difference between the mean pre- and post-treatment AHI values approached statistical significance in both groups (p = 0.07).

Apnea index. Although efforts were made to use the same sleep facility for pre- and postoperative testing for each patient, this was not possible for two of the 17 patients. This is noteworthy because hypopneas can be scored differently by technicians in different facilities--and indeed even by individual technicians within the same facility--which can skew the AHI value. Therefore, each patient was also evaluated according to his or her AI value. A successful response to treatment was defined as a reduction in AI of at least 50% and the attainment of a post-treatment AI of less than 10 in those patients whose preoperative AI was greater than 10.

As determined by the AI, success rates in the two groups were comparable; five of the seven patients in group 1 (71.4%) and seven of the 10 patients in group 2 (70.0%) were considered to have experienced a successful response to therapy. The mean Al values in group I fell from 23.8 ([+ or -]24.2) before treatment to 4.6 ([+ or -]5.3) afterward; the corresponding decline in group 2 was from 8.7 ([+ or -]6.4) to 3.7 ([+ or -]4.9). The difference between the two groups in the mean post-treatment AI was not statistically significant. The difference between the mean pre- and post-treatment AI was significant in group 2 (p = 0.05), but not in group 1, probably because of the small size of the sample.

Discussion

This study was undertaken to test the feasibility of a concept. Although the small number of patients and the demographic variance between the two groups limit both statistical significance and confidence boundaries in validating results, certain trends are evident. Foremost among them is that the addition of TCRF tongue reduction to UPPP does not appear to significantly increase the degree of pain and morbidity or the incidence of complications compared with UPPP alone. In this study population--which is fairly typical of a community-based general otolaryngology practice--combining TCRF tongue reduction with UPPP as an initial operative approach appeared to improve success rates in patients with both velopharyngeal and retroglossal obstruction to levels that are comparable to those achieved with UPPP alone in patients with single-level velopharyngeal obstruction.

TCRF tongue reduction is technically straightforward, and it can be easily mastered by the practicing otolaryngologist with a minimum of additional instruction. Moreover, because patients who undergo UPPP are routinely hospitalized, adding thhe TCRF procedure to UPPP provides maximal use of the hospital stay, and patients can be observed for the occurrence of any adverse effects that might influence the success or failure of any future in-office treatment. Because it is minimally invasive and well tolerated, TCRF tongue reduction might be an acceptable adjunct to already-planned surgery. Its use could encourage patients and practitioners to undertake a more inclusive site-directed initial approach to treatment. Ultimately, the addition of TCRF tongue reduction to UPPP could improve initial response rates in the largest segment of the OSAS population--those with multilevel obstructions and would decrease the size of the patient population that requires more-costly and more-morbid interventions.

Table. Patients' assessments of postoperative pain by perceived
anatomic area on a 10-point visual analog scale (mean)

                                      Tx 1

                           Palate               Nose

1 to 3 days postop
Group 1 (n = 7)       7.0 [+ or -] 0.8    2.4 [+ or -] 1.2
Group 2 (n = 9)       7.3 [+ or -] 2.5    3.5 [+ or -] 3.5

1 wk postop
Group 1 (n = 7)       3.2 [+ or -] 2.0    1.5 [+ or -] 0.8
Group 2 (n = 9)       6.9 [+ or -] 2.2    3.1 [+ or -] 3.1

                            Tx 1                Tx 2

                           Tongue              Tongue

1 to 3 days postop
Group 1 (n = 7)       4.5 [+ or -] 0.5          N/A
Group 2 (n = 9)       4.7 [+ or -] 3.4    4.3 [+ or -] 3.4

1 wk postop
Group 1 (n = 7)       1.9 [+ or -] 1.2          N/A
Group 2 (n = 9)       4.0 [+ or -] 2.9    1.9 [+ or -] 2.6

                            Tx 3

                           Tongue

1 to 3 days postop
Group 1 (n = 7)             N/A
Group 2 (n = 9)       1.9 [+ or -] 1.2

1 wk postop
Group 1 (n = 7)             N/A
Group 2 (n = 9)       0.5 [+ or -] 1.1

Acknowledgments

The author thanks Michelle Boytim, PhD, and Kathryn Grenier, MT (ASCP), CLS, for their assistance with this study.

References

(1.) Sher AE, Schechtman KB, Piccirillo JF. The efficacy of surgical modifications of the upper airway in adults with obstructive sleep apnea syndrome. Sleep 1996; 19:156-77.

(2.) Practice parameters for the treatment of obstructive sleep apnea in adults: The efficacy of surgical modifications of the upper airway. Report of the American Sleep Disorders Association. Sleep 1996:19:152-5.

(3.) Loube DI. Technologic advances in the treatment of obstructive sleep apnea syndrome. Chest 1999:116:1426-33.

(4.) Powell NB, Riley RW, Guilleminault C. Radiofrequency tongue base reduction in sleep-disordered breathing: A pilot study. Otolaryngol Head Neck Surg 1999; 120:656-64.

(5.) Woodson BT, Naganuma H. Comparison of methods of airway evaluation in obstructive sleep apnea syndrome. Otolaryngol Head Neck Surg 1999;120:460-3.

(6.) Riley RW, Powell NB, Guilleminault C. Obstructive sleep apnea syndrome: A review of 306 consecutively treated surgical patients. Otolaryngol Head Neck Surg 1993; 108:117-25.

Dr. Nelson is an otolaryngologist-head and neck surgeon in private practice in San Jose, Calif.

Reprint requests: Lionel M. Nelson, MD, 2505 Samaritan Dr., Suite 510, San Jose, CA 95124. Phone: [phone]; fax: [phone]; e-mail: [e-mail]

Presented at the annual meeting of the American Academy of Otolaryngology-Head and Neck Surgery; Washington, D.C.; Sept. 25, 2000.

This study was supported by Somnus Medical Technologies of Sunnyvale, Calif. Dr. Nelson is a paid consultant of and shareholder in Somnus Medical Technologies.