Systematic Review
The Effect of Glossectomy for Obstructive Sleep Apnea: A Systematic Review and Meta-analysis
Otolaryngology– Head and Neck Surgery 1–9 Ó American Academy of Otolaryngology—Head and Neck Surgery Foundation 2015 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0194599815594347 http://otojournal.org
Alexander W. Murphey, MD1, Jessica A. Kandl1, Shaun A. Nguyen, MD1, Aimee C. Weber, MA1, and M. Boyd Gillespie, MD1
No sponsorships or competing interests have been disclosed for this article.
Abstract Objective. Determine the effect of glossectomy as part of multilevel sleep surgery on sleep-related outcomes in patients with obstructive sleep apnea. Data Sources. PubMED, Scopus. Review Methods. Two independent researchers conducted the review using PubMed-NCBI and Scopus literature databases. Studies on glossectomy for obstructive sleep apnea that reported pre- and postoperative apnea-hypopnea index (AHI) score with 10 or more patients were included. Results. A total of 18 articles with 522 patients treated with 3 glossectomy techniques (midline glossectomy, lingualplasty, and submucosal minimally invasive lingual excision) met inclusion criteria. Pooled analyses (baseline vs post surgery) showed a significant improvement in AHI (48.1 6 22.01 to 19.05 6 15.46, P \ .0001), Epworth Sleepiness Scale (ESS; 11.41 6 4.38 to 5.66 6 3.29, P \ .0001), snoring visual analog scale (VAS; 9.08 6 1.21 to 3.14 6 2.41, P \.0001), and Lowest O2 saturation (76.67 6 10.58 to 84.09 6 7.90, P \ .0001). Surgical success rate was 59.6% (95% CI, 53.0%-65.9%) and surgical cure was achieved in 22.5% (95% CI, 11.26%-36.26%) of cases. Acute complications occurred in 16.4% (79/481) of reported patients. Glossectomy was used as a standalone therapy in 24 patients. In this limited cohort, significant reductions in AHI (41.84 6 32.05 to 25.02 6 20.43, P = .0354) and ESS (12.35 6 5.05 to 6.99 6 3.84, P \ .0001) were likewise observed. Conclusion. Glossectomy significantly improves sleep outcomes as part of multilevel surgery in adult patients with OSA. Currently, there is insufficient evidence to analyze the role of glossectomy as a standalone procedure for the treatment of sleep apnea, although the evidence suggests positive outcomes in select patients. Keywords glossectomy, midline glossectomy, lingualplasty, submucosal minimally invasive lingual excision, SMILE, obstructive sleep apnea, OSA, sleep surgery
Received February 24, 2015; revised June 2, 2015; accepted June 12, 2015.
Introduction Obstructive sleep apnea (OSA) is a common and morbid disease affecting approximately 9% of women and 24% of men aged 30 to 60 years.1,2 The disorder is characterized by narrowing and collapse of the pharyngeal airway during sleep that leads to reductions (hypopnea) and cessations (apneas) of airflow. The gold standard for the treatment of moderate to severe disease continues to be continuous positive airway pressure (CPAP); however, long-term compliance rates are poor despite the health benefits of CPAP therapy. Adherence is defined as .4 hours use/night. Of patients, 46% to 83% are labeled as nonadherent, and compliance rates significantly decrease after 1 year of therapy.3,4 While generally not as effective as CPAP, upper airway surgery is an option as salvage therapy in CPAP failures, reducing the physiological burden of the disorder while improving symptoms. Uvulopalatopharyngoplasty (UPPP) was widely popularized by Fujita as treatment for OSA in 1980 and continues to be the mainstay of surgical therapy.5,6 However, studies have shown modest results with an overall rate of success of 40% in unselected patients.7 The failures of UPPP lie in the pathophysiology of OSA. Obstructions can occur at many areas along the upper airway, and 58% to 87% of patients with OSA have multilevel collapse.8-10 Residual obstructions along the tongue base account for 17% to 33% of upper airway collapse and is especially prominent in the obese and severe OSA (apnea-hypopnea index [AHI] .30) 1
Department of Otolaryngology-Head and Neck Surgery, University of South Carolina, Charleston, South Carolina, USA
Medical
Presented at Triological Society, Combined Sections Meeting; January 22-24, 2015; Coronado Island, California, USA. Corresponding Author: Alexander W. Murphey, MD, Clinical Research Fellow, Department of Otolaryngology-Head and Neck Surgery, Medical University of South Carolina, 135 Rutledge Avenue, MSC 550, Charleston, SC 29425, USA. Email: [email protected]
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
2
Otolaryngology–Head and Neck Surgery
populations.9-11 Therefore, adjunct therapies were developed with a focus on the tongue base and hypopharynx. Specific treatments, such as partial glossectomy, hyoid suspension, genioglossus advancement, and radiofrequency, have been utilized alone or as part of multilevel surgery with palate procedures in order to improve the rate of surgical success. Although various techniques of partial glossectomy for OSA have been described, for the purposes of this review, partial glossectomy includes any procedure that removes tongue tissue in an effort to treat OSA. Sleep surgeries that fit this criterion include midline glossectomy (MLG), submucosal minimally invasive lingual excision (SMILE), and lingualplasty (LP). Fujita was the first to describe removal of tongue tissue for OSA with the MLG.12 MLG utilized as a salvage procedure in patients who failed to respond to UPPP was very successful and resulted in a mean reduction of respiratory disturbance index (RDI) of 19. Woodson and others have developed further modifications of MLG, such as LP, in an effort to improve outcomes results. LP extends tissue removal from MLG posteriorly and laterally. The resulting defect is sutured posterior to anterior, which advances and modifies the dorsal profile of the tongue base.13 Due to morbidity of the initial techniques and concerns for the safety and preservation of the lingual artery and hypoglossal nerve, glossectomy has changed extensively over the years. Recent focus has been on minimally invasive techniques, such as SMILE, in which a coblator is inserted into the tongue through a midline incision and tissue is removed submucosally.14 However, MLG is gaining popularity again using transoral robotic surgery (TORS), which improves visualization of the surgical field, thereby maximizing tissue removal while minimizing risk to vital structures. The goal of this review is to determine the sleeprelated outcomes of partial glossectomy as part of multilevel surgery and when possible its role as a single, standalone surgery for OSA.
Materials and Methods Literature Search Two independent comprehensive literature searches using PubMed-NCBI and SCOPUS databases were performed using the following previously defined terms: (1) glossectomy AND ‘‘sleep apnea,’’ (2) ‘‘midline glossectomy’’ AND ‘‘sleep apnea,’’ (3) lingualplasty AND ‘‘sleep apnea, (4) ‘‘lingual excision’’ AND ‘‘sleep apnea,’’ and (5) ‘‘tongue base’’ AND ‘‘sleep apnea.’’ Results were limited to the English language. Articles were screened according to their titles, and abstracts were chosen for review to determine eligibility for inclusion based on predetermined criteria. The reference lists of all identified articles were examined for additional studies. The literature search strategy and results can be found in the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) compliant literature review diagram (Figure 1).15 Study design was a retrospective literature review; therefore, Institutional Review Board approval was not necessary for this study.
Inclusion Criteria 1. Treated patients carried a previous diagnosis of OSA or diagnosis was confirmed with a preoperative polysomnogram (PSG) demonstrating an AHI (RDI) .5; 2. Pre- and posttreatment data for at least 1 of the established endpoint measures, including AHI, Epworth Sleepiness Scale (ESS), Lowest O2 saturation (LSAT), and/or snoring visual analog scale (VAS); 3. Treatment clearly describes the removal of tongue tissue alone or as part of a multilevel surgery; 4. Study included 10 or more total patients; 5. Adult patients .18 years of age.
Exclusion Criteria 1. Manuscripts not available in English; 2. Any studies that examined tongue reduction without evidence of actual tongue tissue removal, namely, radiofrequency procedures; 3. Any procedures that involved open, cervical approaches to the tongue base or hypopharynx.
Data Extraction Data from studies meeting inclusion criteria were extracted using a standardized form and verified by a second author (J.K). The form was generated prior to article review. The primary outcomes of interest included postoperative changes in AHI, ESS, and LSAT; surgical success; and surgical cure rate. Surgical success was defined as a reduction in AHI by .50% and a final AHI \20, and surgical cure was defined as postoperative AHI \5.7,16 Other study data collected included sample size, patient demographics, length to follow-up, and associated procedures. Prior to 2005, RDI was commonly used interchangeably with AHI. For those articles, methods sections were carefully read to confirm RDI was the same as traditional AHI measurements. Once confirmed, RDI was counted as AHI. The articles after 2005 specifically and consistently use AHI. When both AHI and RDI were listed as outcome measures, only the AHI was used. In some cases, the means and standard deviations (SD) were computed using either postoperative percentage change or per patient statistics. No computational discrepancies between the 2 independent authors occurred. If articles did not contain the correct endpoints or were missing data, 2 attempts were made to contact the corresponding author. If no response was received, the article was excluded. In 1 article, the SD was unavailable for the postoperative changes in AHI. After attempts for contact were unsuccessful, a weighted average was performed using the data from the other manuscripts, and this value was used for analysis.
Stat Section The meta-analysis utilized a pretreatment (baseline) to postglossectomy comparison, with all subjects serving as their
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
Murphey et al
3
Figure 1. Preferred Reporting Items for Systematic Reviews and Meta-analyses compliant literature review diagram.
own controls. Each treatment outcome was assessed and recorded as an absolute value and/or change from baseline using the last time-point data available. Data reported in graphical plots were not extracted for pooled meta-analysis unless numerical points were discernible or the authors of the relevant trial supplied additional data. Meta-analysis of selected studies with a continuous measure (comparison of means and standard deviations between baseline and post-glossectomy groups) was performed with Cochrane Review Manager (RevMan) version 5.3 (The Nordic Cochrane Centre, The Cochrane Collaboration, 2011, Copenhagen, Denmark). Both the fixed effects model and the random effects model were used in this study. Under the fixed effects model, it is assumed that the studies share a common true effect, and the summary effect is an estimate of the common effect size. Under the random effects model, the true effects in the studies are assumed to vary between studies, and the summary effect is the weighted average of the effects reported in the different studies.17 The random effects model provides a more
conservative estimate (ie, with a wider confidence interval [CI]), but the results from the 2 models usually agree when there is no heterogeneity. When heterogeneity was present, the random effects model was the preferred model. This assumption was tested by the heterogeneity test or I2 statistic. If the heterogeneity test yielded a low P value (P \ .05), the fixed effects model was invalid. In this case, the random effects model was more appropriate, in which both the random variation within the studies and the variation between the different studies was incorporated. For this study, the null hypothesis was that there was no difference between baseline and post-glossectomy with respect to AHI, ESS, and LSAT. Effect sizes were calculated using mean differences (MDs) with corresponding 95% CIs. If the value of 0 is not within the 95% CI range, then the MD was statistically significant at the 5% level (P \ .05). In addition, a meta-analysis of proportion was done for surgical success rate. The program MedCalc 14.12.0 (MedCalc Software bvba, Belgium) lists the proportions
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
4
Otolaryngology–Head and Neck Surgery
Table 1. Articles Satisfying Inclusion Criteria.a Author (Year)
No. of Subjects
Fujita Mickelson H.Y. Li (2004) Elasfour Woodson (2012) Suh H.C. Lin (2014) S. Li (2013) Vicinib H.S. Linb Toh Friedman (2012)
12 12 6 18 39 50 35 67 20 12 20 49
Robinsonb Friedman (2008) Babademez
15 48 45
Gunawardena Woodson (1992) H.Y. Li (2015)
27 22 25
Procedure MLG MLG MLG MLG MLG MLG MLG MLG MLG MLG MLG MLG SMILE SMILE SMILE MLG SMILE LP LP LP
Instrument
Associated Procedure
CO2 laser CO2 laser CO2 laser Transoral Transoral Transoral Transoral Transoral TORS TORS TORS TORS Coblator Coblator Coblator Coblator Harmonic scalpel Coblator, scalpel CO2 laser, scalpel Coblator
LT, AE fold reduction, epiglottidectomy LT, epiglottidectomy EUPF UPPP Palatoplasty UPPP, LT Z-palatoplasty UPPP Various nose and palate None Uvulopalatal flap, partial epiglottidectomy, LT Z-palatoplasty UPPP, palate RF, hyoid suspension UPPP, various nasal procedures UPPP UPPP, PA UPPP, LT, epiglottidectomy Relocation pharyngoplasty, LT
Abbreviations: AE, aryepiglottic; EUPF, extended uvulopalatal complex flap; LP, lingualplasty; LT, lingual tonsillectomy; MLG, midline glossectomy; PA, palatal advancement; RF, radiofrequency; SMILE, submucosal minimally invasive lingual excision; TORS, transoral robotic surgery; UPPP, uvulopharyngopalatoplasty. a Technique, instrumentation, and associated procedures for each included study. b Gathered data for subjects undergoing glossectomy as sole therapy.
(expressed as a percentage), with their 95% CI, found in the individual studies included in the meta-analysis. The pooled proportion with 95% CI is given both for the fixed effects model and the random effects model. Each technique was weighted according to the number of patients treated. Analysis of pooled proportions was performed where appropriate. MedCalc used a Freeman-Tukey transformation to calculate the weighted summary proportion under the fixed and random effects model.18,19 Data were presented as weighted proportions with corresponding 95% CIs. Both the fixed effects model and the random effects model were used in this study.
Results The literature search resulted in a total of 1240 manuscripts, of which only 18 met the inclusion and exclusion criteria (Table 1). Robinson et al21 discusses a technique where a coblator was inserted percutaneously through a suprahyoid incision to resect tongue tissue. This was not considered a true cervical approach and was therefore included in analysis. It was later modified to the current SMILE technique and thus considered as such. Two studies investigated more than 1 glossectomy technique. Friedman et al25 analyzes MLG using a SMILE control group, and Babademez et al23 compares results of 3 minimally invasive tongue base procedures (MLG, SMILE with coblator, and SMILE with harmonic scalpel). For both manuscripts, each technique was analyzed as a separate study and labeled in figures (Friedman 2012A, B and Babademez 2011A, B, C, respectively).
The majority of manuscripts include glossectomy as part of a multilevel surgical approach; however, 3 manuscripts, with a total of 24 patients, analyze glossectomy alone without concurrent surgical procedures. H.S. Lin et al20 presents 12 patients undergoing glossectomy via TORS as an isolated surgery. The other 12 patients were obtained from 2 other manuscripts that study both multilevel surgery and isolated glossectomy.21,22 All 3 studies include pre- and postoperative values for AHI, ESS, and LSAT. Separate analysis was conducted using these 24 patients only. Therefore, their data were not included in the meta-analysis for the multilevel surgery group.
Demographics A total of 522 patients underwent glossectomy procedures, 498 multilevel surgery and 24 isolated tongue base surgery. Patients were predominantly male (84.4%) with a mean age of 45.75 years. Mean pre-procedure BMI was 29.4 (mean range, 26.5-36). The median time to follow-up with PSG was 6 months (range, 6 weeks-21 months). Inclusion and exclusion criteria varied among studies, but a common theme was applicable across cohorts. All included subjects had evidence of moderate to severe OSA (AHI .15), CPAP intolerance or failure, and evidence of hypopharyngeal collapse. Hypopharyngeal obstruction was predicted by physical exam, nasal endoscopy, clinical staging, Friedman type 3/4 or Fujita type 2b/3, or with imaging (cephalometry or computed tomography [CT]).
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
Murphey et al
5
Figure 2. Forest plot depicting mean apnea hypopnea index change after glossectomy.
Figure 3. Forest plot depicting mean Epworth Sleepiness Scale change after glossectomy.
Apnea-Hypopnea Index Seventeen studies, with 498 subjects, record pre- and postoperative PSG confirmed AHI for multilevel surgery.12-14,21-34 Comparison of means reveals a statistically significant mean reduction in AHI of 48.1 6 22.01 to 19.05 6 15.46 (P \ .001) from baseline to follow-up. Pooled analysis also shows a significant improvement in AHI of 227.81 (95% CI, –33.00 to 222.62) (Figure 2).
Epworth Sleepiness Scale Ten studies,14,21-23,25-27,29,30,33 totaling 331 patients, record pre- and postoperative ESS scores. Comparison of means
reveals a statistically significant improvement in ESS, 11.41 6 4.38 to 5.66 6 3.29 (P \ .001) from baseline to follow-up. Pooled analysis also shows a significant improvement in ESS, –5.49 (95% CI, –7.17 to 23.81) (Figure 3).
LSAT Fifteen studies,12-14,21,22,24-33 totaling 414 patients, recorded pre- and postoperative LSAT. Comparison of means reveals statistically significant improvement in LSAT, 76.67% 6 10.58% to 84.09% 6 7.90% (P \ .001), from baseline to follow-up. Pooled analysis also shows a statistically significant improvement in LSAT, 7.68 (95% CI, 5.34-10.02) (Figure 4).
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
6
Otolaryngology–Head and Neck Surgery
Figure 4. Forest plot depicting mean lowest oxygen saturation change after glossectomy.
Figure 5. Forest plot depicting mean snoring visual analog scale change after glossectomy.
Snoring VAS 14,23,25,30,33
Five studies, totaling 197 patients, recorded preand postoperative snoring VAS scales. Comparison of means reveals a statistically significant decrease in VAS, 9.08 6 1.21 to 3.14 6 2.41 (P \ .001), from baseline to follow-up. Pooled analysis also shows significant improvement in snoring VAS 25.60 (95% CI, –6.57 to 24.63) (Figure 5).
Surgical Success Fifteen studies,12-14,21,22,24-33 totaling 414 patients, were analyzed for surgical success. Surgical success was considered to be a postoperative AHI \20 and reduction from preoperative AHI .50%. Pooled analysis reveals success was achieved in 59.56% (95% CI, 52.99%-65.96%) of cases (Figure 6).
Surgical Cure Seven studies,21,22,24,29-31,33 totaling 175 patients, were analyzed for surgical cure. Surgical cure was defined as
postoperative AHI \5. Pooled analysis reveals cure was achieved in 22.5% (95% CI, 11.26%-36.26%) of cases.
Glossectomy Only Three studies, totaling 24 patients, analyzed glossectomy alone without concurrent surgical procedures. Comparison of means reveals a statistically significant reduction in both AHI (41.84 6 32.05 to 25.02 6 20.43, P = .04) and ESS (12.35 6 5.05 to 6.99 6 3.84, P = .001) from baseline to follow-up. Pooled analysis also reveals significant reduction in ESS, –5.41 (95% CI, –7.76 to 23.05), from baseline to follow-up. Pooled analysis is not significant for AHI or LSAT. Data were not available for snoring VAS or surgical success.
Complications Complications were published in 16 of 18 analyzed manuscripts. They were broken down into acute (\30 days post procedure) and chronic (.30 days). Acute complications
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
Murphey et al
7
Figure 6. Forest plot depicting proportion of patients achieving surgical success.
included bleeding, wound dehiscence, infection, airway complications, and subcutaneous emphysema. Chronic complications included prolonged dysphagia, loss or change in taste, delayed surgical intervention, globus, and hypoglossal paralysis. Multiple complications in single patients were counted twice if they were deemed separate events. Overall, there were a total of 81 complications in 79 patients, for a complication rate of 16.4% (79/481) in reported patients. Rates of acute and chronic complications were 6.2% (30/ 481) and 10.6% (51/481), respectively. The most common complication was loss or change in taste and was reported in 5.82% (28/481) of patients. Change in taste was rarely permanent and usually cleared between 2 weeks and 2 months. Postoperative bleeding was noted in 4.2% (20/481) of patients. Hypoglossal paresis was noted in 5 patients in 2 studies, with all but 1 achieving full recovery. There were 2 cases of oropharyngeal stenosis, which needed surgical correction.
Discussion Glossectomy has almost exclusively been studied as part of multilevel surgery, and unlike other tongue therapy such as radiofrequency, few direct comparisons to other procedures have been made. This study is one of the first to gather data on the role of glossectomy alone and as part of multilevel surgery on such a large scale. Our analysis showed that glossectomy, as part of multilevel surgery, significantly improves sleep outcomes in adult patients with obstructive sleep apnea. When combined with palate surgery as multilevel surgical treatment, glossectomy appears to be effective. Data indicate that including glossectomy as part of a multilevel surgical plan significantly improved outcomes across 4 independent sleep metrics, including AHI, ESS, LSAT, and VAS. The best results across all outcomes were seen with Li and Shi,29 who utilized Computed Tomography Angiography (CTA) to map out the lingual artery before open midline glossectomy. Results were impressive, showing reduction in AHI (36.4)
and ESS (9.3) and an increase in LSAT of 17%.29 Such impressive results were most likely due to strict inclusion criteria, including preoperative CT-confirmed hypopharyngeal airway space \180 mm. This further demonstrates that in such a site-specific sleep surgery, such as UPPP and glossectomy, determination of site obstruction is imperative. Combined surgical success rate for all studies was 59.6% while surgical cure was achieved in 22.5% of cases. Despite its popularity, the definition used for success, .50% improvement and AHI \20, is far from ideal. It excludes drastic improvements in AHI or improvements in other sleep outcomes. For example, Mickelson and Rosenthal31 only reported a success rate of 25% (3/12). However, mean AHI was reduced by 26.7, which is similar to the mean reduction across all 18 of our reported studies. The LSAT also improved by 19% in the 9 nonresponders. Although other reviews have analyzed glossectomy in the past, many have not fully analyzed outcome parameters that have clinical significance. One review analyzed glossectomy as part of multilevel surgery and compared it to other hypopharyngeal surgeries based on success rate.35 Four of the 5 included studies in the review referenced previously were also included in the present analysis.12,13,28,31 Using the same definition of surgical success, they found that 50% (37/74) of treatments were considered successful, which is slightly less than the success rate found in our study (59.6%). Our success rate is also similar to other multilevel surgeries including genioglossus advancement and hyoid suspension, which at one time was considered firstline therapy for hypopharyngeal obstruction. The present review was only able to identify 24 subjects treated with glossectomy as sole therapy for OSA.20-22 Despite this small population, initial results are promising. Reduction in ESS score, in particular, compared favorably to multilevel surgery, –5.41 to 25.49, respectively. Unfortunately, pooled analysis for LSAT and AHI was not significant. These findings appear to support the general trend for advocating multilevel surgical therapy in patients with moderate to severe OSA. However, this negative finding may be partly explained by the small patient population and large deviation in results between patients in these studies. This is particularly evident with Vicini et al,22 who presented only 2 patients who received TORS glossectomy as a single therapy. One subject’s AHI decreased by 79 (91 to 12), while the other actually increased by 5 (32 to 37). Robinson et al21 did not achieve results until adding other levels of treatment. Their first 10 patients received plasma wand treatment, and only 2 of 10 were considered successful. The final 5 were given multilevel surgery (UPPP, hyoid suspension, and/or geniotubercle advancement), which increased success rate to 80% (4/5). Failures in this case were attributed to poor volume reduction with the plasma wand and residual obstruction mainly in the lateral pharyngeal wall. A recent review on the effectiveness of alternative treatments for OSA stated that they cannot recommend glossectomy as a single treatment option for obese patients with moderate to severe OSA.36 This review did not take
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
8
Otolaryngology–Head and Neck Surgery
the 3 studies we used to assess a single glossectomy treatment into account. Although evidence presented here cannot disagree with that statement, it is certainly possible that under the right circumstances and with identification of the correct subjects, glossectomy is an effective single treatment for OSA. Our study has some important limitations. There is a lack of quality research involving glossectomy with the majority of available published data drawn from small case series without control arms. This has made it difficult to determine the true effectiveness of glossectomy. Similar procedures vary in surgical approach, the inclusion criteria for patient selection, and types of additional procedures, and definitions of and the amount of attention paid to complications in each series, while similar, are not standardized across studies. Additionally, many of the isolated glossectomy patients analyzed received previous palate surgery. This makes it extremely difficult to truly compare treatments and complications. It should also be noted that AHI, while the most popular and universally reported outcome metric used for sleep medicine, has its limitations.37 Future directions for research include further evaluation of isolated glossectomy and direct comparisons of glossectomy to other tongue base surgeries in multilevel surgery.
Conclusion Our study suggests that glossectomy significantly improves sleep outcomes as part of multilevel surgery in select adult patients with OSA. Currently, there is insufficient evidence to analyze the role of glossectomy as a standalone procedure for the treatment of sleep apnea. Author Contributions Alexander W. Murphey, collected data, analyzed data, drafting article, revised article; Jessica A. Kandl, collected data, analyzed data, drafting article; Shaun A. Nguyen, designed study, analyzed and interpreted data, revised article; Aimee C. Weber, collected data, revised article; M. Boyd Gillespie, designed study, interpreted data, drafting article, revised article.
Disclosures Competing interests: None. Sponsorships: None. Funding source: None.
References 1. Epstein LJ, Kristo D, Strollo PJ, Jr, et al. Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults. J Clin Sleep Med. 2009;5:263-276. 2. Young T, Peppard PE, Gottlieb DJ. Epidemiology of obstructive sleep apnea: a population health perspective. Am J Respir Crit Care Med. 2002;165:1217-1239. 3. Chai-Coetzer CL, Luo YM, Antic NA, et al. Predictors of long-term adherence to continuous positive airway pressure therapy in patients with obstructive sleep apnea and cardiovascular disease in the SAVE study. Sleep. 2013;36:1929-1937.
4. Weaver TE, Grunstein RR. Adherence to continuous positive airway pressure therapy: the challenge to effective treatment. Proc Am Thorac Soc. 2008;5:173-178. 5. Fujita S, Conway W, Zorick F, et al. Surgical correction of anatomic azbnormalities in obstructive sleep apnea syndrome: uvulopalatopharyngoplasty. Otolaryngol Head Neck Surg. 1981;89:923-934. 6. Kezirian EJ, Maselli J, Vittinghoff E, et al. Obstructive sleep apnea surgery practice patterns in the United States: 2000 to 2006. Otolaryngol Head Neck Surg. 2010;143:441-447. 7. Sher AE, Schechtman KB, Piccirillo JF. The efficacy of surgical modifications of the upper airway in adults with obstructive sleep apnea syndrome. Sleep. 1996;19:156-177. 8. Abdullah VJ, Wing YK, van Hasselt CA. Video sleep nasendoscopy: the Hong Kong experience. Otolaryngol Clin North Am. 2003;36:461-471. 9. Bachar G, Feinmesser R, Shpitzer T, et al. Laryngeal and hypopharyngeal obstruction in sleep disordered breathing patients, evaluated by sleep endoscopy. Eur Arch Otorhinolaryngol. 2008;265:1397-1402. 10. Kotecha BT, Hannan SA, Khalil HM, et al. Sleep nasendoscopy: a 10-year retrospective audit study. Eur Arch Otorhinolaryngol. 2007;264:1361-1367. 11. Do KL, Ferreyra H, Healy JF, et al. Does tongue size differ between patients with and without sleep-disordered breathing? Laryngoscope. 2000;110:1552-1555. 12. Fujita S, Woodson BT, Clark JL, et al. Laser midline glossectomy as a treatment for obstructive sleep apnea. Laryngoscope. 1991;101:805-809. 13. Woodson BT, Fujita S. Clinical experience with lingualplasty as part of the treatment of severe obstructive sleep apnea. Otolaryngol Head Neck Surg. 1992;107:40-48. 14. Friedman M, Soans R, Gurpinar B, et al. Evaluation of submucosal minimally invasive lingual excision technique for treatment of obstructive sleep apnea/hypopnea syndrome. Otolaryngol Head Neck Surg. 2008;139:378-384. 15. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg. 2010;8:336-341. 16. Elshaug AG, Moss JR, Southcott AM, et al. Redefining success in airway surgery for obstructive sleep apnea: a meta analysis and synthesis of the evidence. Sleep. 2007;30:461467. 17. Borenstein M, Hedges LV, Higgins J, Rothstein HR. Introduction to Meta-analysis. New York: Wiley; 2009. 18. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177-188. 19. Freeman M, Tukey JW. Transformations related to the angular and the square root. Ann Math Stat. 1950;21:607-611. 20. Lin HS, Rowley JA, Badr MS, et al. Transoral robotic surgery for treatment of obstructive sleep apnea-hypopnea syndrome. Laryngoscope. 2013;123:1811-1816. 21. Robinson S, Lewis R, Norton A, et al. Ultrasound-guided radiofrequency submucosal tongue-base excision for sleep apnoea: a preliminary report. Clin Otolaryngol Allied Sci. 2003;28:341-345.
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
Murphey et al
9
22. Vicini C, Dallan I, Canzi P, et al. Transoral robotic surgery of the tongue base in obstructive sleep Apnea-Hypopnea syndrome: anatomic considerations and clinical experience. Head Neck. 2012;34:15-22. 23. Babademez MA, Yorubulut M, Yurekli MF, et al. Comparison of minimally invasive techniques in tongue base surgery in patients with obstructive sleep apnea. Otolaryngol Head Neck Surg. 2011;145:858-864. 24. Elasfour A, Miyazaki S, Itasaka Y, et al. Evaluation of uvulopalatopharyngoplasty in treatment of obstructive sleep apnea syndrome. Acta Otolaryngol Suppl. 1998;537:52-56. 25. Friedman M, Hamilton C, Samuelson CG, et al. Transoral robotic glossectomy for the treatment of obstructive sleep apnea-hypopnea syndrome. Otolaryngol Head Neck Surg. 2012;146:854-862. 26. Gunawardena I, Robinson S, MacKay S, et al. Submucosal lingualplasty for adult obstructive sleep apnea. Otolaryngol Head Neck Surg. 2013;148:157-165. 27. Li HY, Lee LA, Kezirian EJ. Coblation endoscopic lingual lightening (CELL) for obstructive sleep apnea [published online January 13, 2015]. Eur Arch Otorhinolaryngol. PMID: 25577994 28. Li HY, Wang PC, Hsu CY, et al. Same-stage palatopharyngeal and hypopharyngeal surgery for severe obstructive sleep apnea. Acta Otolaryngol. 2004;124:820-826. 29. Li S, Shi H. Lingual artery CTA-guided midline partial glossectomy for treatment of obstructive sleep apnea hypopnea syndrome. Acta Otolaryngol. 2013;133:749-754.
30. Lin HC, Friedman M, Chang HW, et al. Z-palatopharyngoplasty combined with endoscopic coblator open tongue base resection for severe obstructive sleep apnea/hypopnea syndrome. Otolaryngol Head Neck Surg. 2014;150:1078-1085. 31. Mickelson SA, Rosenthal L. Midline glossectomy and epiglottidectomy for obstructive sleep apnea syndrome. Laryngoscope. 1997;107:614-619. 32. Suh GD. Evaluation of open midline glossectomy in the multilevel surgical management of obstructive sleep apnea syndrome. Otolaryngol Head Neck Surg. 2013;148:166-171. 33. Toh ST, Han HJ, Tay HN, et al. Transoral robotic surgery for obstructive sleep apnea in Asian patients: a Singapore sleep centre experience. JAMA Otolaryngol Head Neck Surg. 2014; 140:624-629. 34. Woodson BT, Laohasiriwong S. Lingual tonsillectomy and midline posterior glossectomy for obstructive sleep apnea. Operative Techniques in Otolaryngology. 2012;23:155-161. 35. Kezirian EJ, Goldberg AN. Hypopharyngeal surgery in obstructive sleep apnea: an evidence-based medicine review. Arch Otolaryngol Head Neck Surg. 2006;132:206-213. 36. Keymel S, Kelm M, Randerath WJ. [Non-CPAP therapies in obstructive sleep apnoea: an overview]. Pneumologie. 2013; 67:50-57. 37. Tam S, Woodson BT, Rotenberg B. Outcome measurements in obstructive sleep apnea: beyond the apnea-hypopnea index. Laryngoscope. 2014;124: 337-343.
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
The Effect of Glossectomy for Obstructive Sleep Apnea: A Systematic Review and Meta-analysis
Otolaryngology– Head and Neck Surgery 1–9 Ó American Academy of Otolaryngology—Head and Neck Surgery Foundation 2015 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0194599815594347 http://otojournal.org
Alexander W. Murphey, MD1, Jessica A. Kandl1, Shaun A. Nguyen, MD1, Aimee C. Weber, MA1, and M. Boyd Gillespie, MD1
No sponsorships or competing interests have been disclosed for this article.
Abstract Objective. Determine the effect of glossectomy as part of multilevel sleep surgery on sleep-related outcomes in patients with obstructive sleep apnea. Data Sources. PubMED, Scopus. Review Methods. Two independent researchers conducted the review using PubMed-NCBI and Scopus literature databases. Studies on glossectomy for obstructive sleep apnea that reported pre- and postoperative apnea-hypopnea index (AHI) score with 10 or more patients were included. Results. A total of 18 articles with 522 patients treated with 3 glossectomy techniques (midline glossectomy, lingualplasty, and submucosal minimally invasive lingual excision) met inclusion criteria. Pooled analyses (baseline vs post surgery) showed a significant improvement in AHI (48.1 6 22.01 to 19.05 6 15.46, P \ .0001), Epworth Sleepiness Scale (ESS; 11.41 6 4.38 to 5.66 6 3.29, P \ .0001), snoring visual analog scale (VAS; 9.08 6 1.21 to 3.14 6 2.41, P \.0001), and Lowest O2 saturation (76.67 6 10.58 to 84.09 6 7.90, P \ .0001). Surgical success rate was 59.6% (95% CI, 53.0%-65.9%) and surgical cure was achieved in 22.5% (95% CI, 11.26%-36.26%) of cases. Acute complications occurred in 16.4% (79/481) of reported patients. Glossectomy was used as a standalone therapy in 24 patients. In this limited cohort, significant reductions in AHI (41.84 6 32.05 to 25.02 6 20.43, P = .0354) and ESS (12.35 6 5.05 to 6.99 6 3.84, P \ .0001) were likewise observed. Conclusion. Glossectomy significantly improves sleep outcomes as part of multilevel surgery in adult patients with OSA. Currently, there is insufficient evidence to analyze the role of glossectomy as a standalone procedure for the treatment of sleep apnea, although the evidence suggests positive outcomes in select patients. Keywords glossectomy, midline glossectomy, lingualplasty, submucosal minimally invasive lingual excision, SMILE, obstructive sleep apnea, OSA, sleep surgery
Received February 24, 2015; revised June 2, 2015; accepted June 12, 2015.
Introduction Obstructive sleep apnea (OSA) is a common and morbid disease affecting approximately 9% of women and 24% of men aged 30 to 60 years.1,2 The disorder is characterized by narrowing and collapse of the pharyngeal airway during sleep that leads to reductions (hypopnea) and cessations (apneas) of airflow. The gold standard for the treatment of moderate to severe disease continues to be continuous positive airway pressure (CPAP); however, long-term compliance rates are poor despite the health benefits of CPAP therapy. Adherence is defined as .4 hours use/night. Of patients, 46% to 83% are labeled as nonadherent, and compliance rates significantly decrease after 1 year of therapy.3,4 While generally not as effective as CPAP, upper airway surgery is an option as salvage therapy in CPAP failures, reducing the physiological burden of the disorder while improving symptoms. Uvulopalatopharyngoplasty (UPPP) was widely popularized by Fujita as treatment for OSA in 1980 and continues to be the mainstay of surgical therapy.5,6 However, studies have shown modest results with an overall rate of success of 40% in unselected patients.7 The failures of UPPP lie in the pathophysiology of OSA. Obstructions can occur at many areas along the upper airway, and 58% to 87% of patients with OSA have multilevel collapse.8-10 Residual obstructions along the tongue base account for 17% to 33% of upper airway collapse and is especially prominent in the obese and severe OSA (apnea-hypopnea index [AHI] .30) 1
Department of Otolaryngology-Head and Neck Surgery, University of South Carolina, Charleston, South Carolina, USA
Medical
Presented at Triological Society, Combined Sections Meeting; January 22-24, 2015; Coronado Island, California, USA. Corresponding Author: Alexander W. Murphey, MD, Clinical Research Fellow, Department of Otolaryngology-Head and Neck Surgery, Medical University of South Carolina, 135 Rutledge Avenue, MSC 550, Charleston, SC 29425, USA. Email: [email protected]
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
2
Otolaryngology–Head and Neck Surgery
populations.9-11 Therefore, adjunct therapies were developed with a focus on the tongue base and hypopharynx. Specific treatments, such as partial glossectomy, hyoid suspension, genioglossus advancement, and radiofrequency, have been utilized alone or as part of multilevel surgery with palate procedures in order to improve the rate of surgical success. Although various techniques of partial glossectomy for OSA have been described, for the purposes of this review, partial glossectomy includes any procedure that removes tongue tissue in an effort to treat OSA. Sleep surgeries that fit this criterion include midline glossectomy (MLG), submucosal minimally invasive lingual excision (SMILE), and lingualplasty (LP). Fujita was the first to describe removal of tongue tissue for OSA with the MLG.12 MLG utilized as a salvage procedure in patients who failed to respond to UPPP was very successful and resulted in a mean reduction of respiratory disturbance index (RDI) of 19. Woodson and others have developed further modifications of MLG, such as LP, in an effort to improve outcomes results. LP extends tissue removal from MLG posteriorly and laterally. The resulting defect is sutured posterior to anterior, which advances and modifies the dorsal profile of the tongue base.13 Due to morbidity of the initial techniques and concerns for the safety and preservation of the lingual artery and hypoglossal nerve, glossectomy has changed extensively over the years. Recent focus has been on minimally invasive techniques, such as SMILE, in which a coblator is inserted into the tongue through a midline incision and tissue is removed submucosally.14 However, MLG is gaining popularity again using transoral robotic surgery (TORS), which improves visualization of the surgical field, thereby maximizing tissue removal while minimizing risk to vital structures. The goal of this review is to determine the sleeprelated outcomes of partial glossectomy as part of multilevel surgery and when possible its role as a single, standalone surgery for OSA.
Materials and Methods Literature Search Two independent comprehensive literature searches using PubMed-NCBI and SCOPUS databases were performed using the following previously defined terms: (1) glossectomy AND ‘‘sleep apnea,’’ (2) ‘‘midline glossectomy’’ AND ‘‘sleep apnea,’’ (3) lingualplasty AND ‘‘sleep apnea, (4) ‘‘lingual excision’’ AND ‘‘sleep apnea,’’ and (5) ‘‘tongue base’’ AND ‘‘sleep apnea.’’ Results were limited to the English language. Articles were screened according to their titles, and abstracts were chosen for review to determine eligibility for inclusion based on predetermined criteria. The reference lists of all identified articles were examined for additional studies. The literature search strategy and results can be found in the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) compliant literature review diagram (Figure 1).15 Study design was a retrospective literature review; therefore, Institutional Review Board approval was not necessary for this study.
Inclusion Criteria 1. Treated patients carried a previous diagnosis of OSA or diagnosis was confirmed with a preoperative polysomnogram (PSG) demonstrating an AHI (RDI) .5; 2. Pre- and posttreatment data for at least 1 of the established endpoint measures, including AHI, Epworth Sleepiness Scale (ESS), Lowest O2 saturation (LSAT), and/or snoring visual analog scale (VAS); 3. Treatment clearly describes the removal of tongue tissue alone or as part of a multilevel surgery; 4. Study included 10 or more total patients; 5. Adult patients .18 years of age.
Exclusion Criteria 1. Manuscripts not available in English; 2. Any studies that examined tongue reduction without evidence of actual tongue tissue removal, namely, radiofrequency procedures; 3. Any procedures that involved open, cervical approaches to the tongue base or hypopharynx.
Data Extraction Data from studies meeting inclusion criteria were extracted using a standardized form and verified by a second author (J.K). The form was generated prior to article review. The primary outcomes of interest included postoperative changes in AHI, ESS, and LSAT; surgical success; and surgical cure rate. Surgical success was defined as a reduction in AHI by .50% and a final AHI \20, and surgical cure was defined as postoperative AHI \5.7,16 Other study data collected included sample size, patient demographics, length to follow-up, and associated procedures. Prior to 2005, RDI was commonly used interchangeably with AHI. For those articles, methods sections were carefully read to confirm RDI was the same as traditional AHI measurements. Once confirmed, RDI was counted as AHI. The articles after 2005 specifically and consistently use AHI. When both AHI and RDI were listed as outcome measures, only the AHI was used. In some cases, the means and standard deviations (SD) were computed using either postoperative percentage change or per patient statistics. No computational discrepancies between the 2 independent authors occurred. If articles did not contain the correct endpoints or were missing data, 2 attempts were made to contact the corresponding author. If no response was received, the article was excluded. In 1 article, the SD was unavailable for the postoperative changes in AHI. After attempts for contact were unsuccessful, a weighted average was performed using the data from the other manuscripts, and this value was used for analysis.
Stat Section The meta-analysis utilized a pretreatment (baseline) to postglossectomy comparison, with all subjects serving as their
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
Murphey et al
3
Figure 1. Preferred Reporting Items for Systematic Reviews and Meta-analyses compliant literature review diagram.
own controls. Each treatment outcome was assessed and recorded as an absolute value and/or change from baseline using the last time-point data available. Data reported in graphical plots were not extracted for pooled meta-analysis unless numerical points were discernible or the authors of the relevant trial supplied additional data. Meta-analysis of selected studies with a continuous measure (comparison of means and standard deviations between baseline and post-glossectomy groups) was performed with Cochrane Review Manager (RevMan) version 5.3 (The Nordic Cochrane Centre, The Cochrane Collaboration, 2011, Copenhagen, Denmark). Both the fixed effects model and the random effects model were used in this study. Under the fixed effects model, it is assumed that the studies share a common true effect, and the summary effect is an estimate of the common effect size. Under the random effects model, the true effects in the studies are assumed to vary between studies, and the summary effect is the weighted average of the effects reported in the different studies.17 The random effects model provides a more
conservative estimate (ie, with a wider confidence interval [CI]), but the results from the 2 models usually agree when there is no heterogeneity. When heterogeneity was present, the random effects model was the preferred model. This assumption was tested by the heterogeneity test or I2 statistic. If the heterogeneity test yielded a low P value (P \ .05), the fixed effects model was invalid. In this case, the random effects model was more appropriate, in which both the random variation within the studies and the variation between the different studies was incorporated. For this study, the null hypothesis was that there was no difference between baseline and post-glossectomy with respect to AHI, ESS, and LSAT. Effect sizes were calculated using mean differences (MDs) with corresponding 95% CIs. If the value of 0 is not within the 95% CI range, then the MD was statistically significant at the 5% level (P \ .05). In addition, a meta-analysis of proportion was done for surgical success rate. The program MedCalc 14.12.0 (MedCalc Software bvba, Belgium) lists the proportions
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
4
Otolaryngology–Head and Neck Surgery
Table 1. Articles Satisfying Inclusion Criteria.a Author (Year)
No. of Subjects
Fujita Mickelson H.Y. Li (2004) Elasfour Woodson (2012) Suh H.C. Lin (2014) S. Li (2013) Vicinib H.S. Linb Toh Friedman (2012)
12 12 6 18 39 50 35 67 20 12 20 49
Robinsonb Friedman (2008) Babademez
15 48 45
Gunawardena Woodson (1992) H.Y. Li (2015)
27 22 25
Procedure MLG MLG MLG MLG MLG MLG MLG MLG MLG MLG MLG MLG SMILE SMILE SMILE MLG SMILE LP LP LP
Instrument
Associated Procedure
CO2 laser CO2 laser CO2 laser Transoral Transoral Transoral Transoral Transoral TORS TORS TORS TORS Coblator Coblator Coblator Coblator Harmonic scalpel Coblator, scalpel CO2 laser, scalpel Coblator
LT, AE fold reduction, epiglottidectomy LT, epiglottidectomy EUPF UPPP Palatoplasty UPPP, LT Z-palatoplasty UPPP Various nose and palate None Uvulopalatal flap, partial epiglottidectomy, LT Z-palatoplasty UPPP, palate RF, hyoid suspension UPPP, various nasal procedures UPPP UPPP, PA UPPP, LT, epiglottidectomy Relocation pharyngoplasty, LT
Abbreviations: AE, aryepiglottic; EUPF, extended uvulopalatal complex flap; LP, lingualplasty; LT, lingual tonsillectomy; MLG, midline glossectomy; PA, palatal advancement; RF, radiofrequency; SMILE, submucosal minimally invasive lingual excision; TORS, transoral robotic surgery; UPPP, uvulopharyngopalatoplasty. a Technique, instrumentation, and associated procedures for each included study. b Gathered data for subjects undergoing glossectomy as sole therapy.
(expressed as a percentage), with their 95% CI, found in the individual studies included in the meta-analysis. The pooled proportion with 95% CI is given both for the fixed effects model and the random effects model. Each technique was weighted according to the number of patients treated. Analysis of pooled proportions was performed where appropriate. MedCalc used a Freeman-Tukey transformation to calculate the weighted summary proportion under the fixed and random effects model.18,19 Data were presented as weighted proportions with corresponding 95% CIs. Both the fixed effects model and the random effects model were used in this study.
Results The literature search resulted in a total of 1240 manuscripts, of which only 18 met the inclusion and exclusion criteria (Table 1). Robinson et al21 discusses a technique where a coblator was inserted percutaneously through a suprahyoid incision to resect tongue tissue. This was not considered a true cervical approach and was therefore included in analysis. It was later modified to the current SMILE technique and thus considered as such. Two studies investigated more than 1 glossectomy technique. Friedman et al25 analyzes MLG using a SMILE control group, and Babademez et al23 compares results of 3 minimally invasive tongue base procedures (MLG, SMILE with coblator, and SMILE with harmonic scalpel). For both manuscripts, each technique was analyzed as a separate study and labeled in figures (Friedman 2012A, B and Babademez 2011A, B, C, respectively).
The majority of manuscripts include glossectomy as part of a multilevel surgical approach; however, 3 manuscripts, with a total of 24 patients, analyze glossectomy alone without concurrent surgical procedures. H.S. Lin et al20 presents 12 patients undergoing glossectomy via TORS as an isolated surgery. The other 12 patients were obtained from 2 other manuscripts that study both multilevel surgery and isolated glossectomy.21,22 All 3 studies include pre- and postoperative values for AHI, ESS, and LSAT. Separate analysis was conducted using these 24 patients only. Therefore, their data were not included in the meta-analysis for the multilevel surgery group.
Demographics A total of 522 patients underwent glossectomy procedures, 498 multilevel surgery and 24 isolated tongue base surgery. Patients were predominantly male (84.4%) with a mean age of 45.75 years. Mean pre-procedure BMI was 29.4 (mean range, 26.5-36). The median time to follow-up with PSG was 6 months (range, 6 weeks-21 months). Inclusion and exclusion criteria varied among studies, but a common theme was applicable across cohorts. All included subjects had evidence of moderate to severe OSA (AHI .15), CPAP intolerance or failure, and evidence of hypopharyngeal collapse. Hypopharyngeal obstruction was predicted by physical exam, nasal endoscopy, clinical staging, Friedman type 3/4 or Fujita type 2b/3, or with imaging (cephalometry or computed tomography [CT]).
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
Murphey et al
5
Figure 2. Forest plot depicting mean apnea hypopnea index change after glossectomy.
Figure 3. Forest plot depicting mean Epworth Sleepiness Scale change after glossectomy.
Apnea-Hypopnea Index Seventeen studies, with 498 subjects, record pre- and postoperative PSG confirmed AHI for multilevel surgery.12-14,21-34 Comparison of means reveals a statistically significant mean reduction in AHI of 48.1 6 22.01 to 19.05 6 15.46 (P \ .001) from baseline to follow-up. Pooled analysis also shows a significant improvement in AHI of 227.81 (95% CI, –33.00 to 222.62) (Figure 2).
Epworth Sleepiness Scale Ten studies,14,21-23,25-27,29,30,33 totaling 331 patients, record pre- and postoperative ESS scores. Comparison of means
reveals a statistically significant improvement in ESS, 11.41 6 4.38 to 5.66 6 3.29 (P \ .001) from baseline to follow-up. Pooled analysis also shows a significant improvement in ESS, –5.49 (95% CI, –7.17 to 23.81) (Figure 3).
LSAT Fifteen studies,12-14,21,22,24-33 totaling 414 patients, recorded pre- and postoperative LSAT. Comparison of means reveals statistically significant improvement in LSAT, 76.67% 6 10.58% to 84.09% 6 7.90% (P \ .001), from baseline to follow-up. Pooled analysis also shows a statistically significant improvement in LSAT, 7.68 (95% CI, 5.34-10.02) (Figure 4).
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
6
Otolaryngology–Head and Neck Surgery
Figure 4. Forest plot depicting mean lowest oxygen saturation change after glossectomy.
Figure 5. Forest plot depicting mean snoring visual analog scale change after glossectomy.
Snoring VAS 14,23,25,30,33
Five studies, totaling 197 patients, recorded preand postoperative snoring VAS scales. Comparison of means reveals a statistically significant decrease in VAS, 9.08 6 1.21 to 3.14 6 2.41 (P \ .001), from baseline to follow-up. Pooled analysis also shows significant improvement in snoring VAS 25.60 (95% CI, –6.57 to 24.63) (Figure 5).
Surgical Success Fifteen studies,12-14,21,22,24-33 totaling 414 patients, were analyzed for surgical success. Surgical success was considered to be a postoperative AHI \20 and reduction from preoperative AHI .50%. Pooled analysis reveals success was achieved in 59.56% (95% CI, 52.99%-65.96%) of cases (Figure 6).
Surgical Cure Seven studies,21,22,24,29-31,33 totaling 175 patients, were analyzed for surgical cure. Surgical cure was defined as
postoperative AHI \5. Pooled analysis reveals cure was achieved in 22.5% (95% CI, 11.26%-36.26%) of cases.
Glossectomy Only Three studies, totaling 24 patients, analyzed glossectomy alone without concurrent surgical procedures. Comparison of means reveals a statistically significant reduction in both AHI (41.84 6 32.05 to 25.02 6 20.43, P = .04) and ESS (12.35 6 5.05 to 6.99 6 3.84, P = .001) from baseline to follow-up. Pooled analysis also reveals significant reduction in ESS, –5.41 (95% CI, –7.76 to 23.05), from baseline to follow-up. Pooled analysis is not significant for AHI or LSAT. Data were not available for snoring VAS or surgical success.
Complications Complications were published in 16 of 18 analyzed manuscripts. They were broken down into acute (\30 days post procedure) and chronic (.30 days). Acute complications
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
Murphey et al
7
Figure 6. Forest plot depicting proportion of patients achieving surgical success.
included bleeding, wound dehiscence, infection, airway complications, and subcutaneous emphysema. Chronic complications included prolonged dysphagia, loss or change in taste, delayed surgical intervention, globus, and hypoglossal paralysis. Multiple complications in single patients were counted twice if they were deemed separate events. Overall, there were a total of 81 complications in 79 patients, for a complication rate of 16.4% (79/481) in reported patients. Rates of acute and chronic complications were 6.2% (30/ 481) and 10.6% (51/481), respectively. The most common complication was loss or change in taste and was reported in 5.82% (28/481) of patients. Change in taste was rarely permanent and usually cleared between 2 weeks and 2 months. Postoperative bleeding was noted in 4.2% (20/481) of patients. Hypoglossal paresis was noted in 5 patients in 2 studies, with all but 1 achieving full recovery. There were 2 cases of oropharyngeal stenosis, which needed surgical correction.
Discussion Glossectomy has almost exclusively been studied as part of multilevel surgery, and unlike other tongue therapy such as radiofrequency, few direct comparisons to other procedures have been made. This study is one of the first to gather data on the role of glossectomy alone and as part of multilevel surgery on such a large scale. Our analysis showed that glossectomy, as part of multilevel surgery, significantly improves sleep outcomes in adult patients with obstructive sleep apnea. When combined with palate surgery as multilevel surgical treatment, glossectomy appears to be effective. Data indicate that including glossectomy as part of a multilevel surgical plan significantly improved outcomes across 4 independent sleep metrics, including AHI, ESS, LSAT, and VAS. The best results across all outcomes were seen with Li and Shi,29 who utilized Computed Tomography Angiography (CTA) to map out the lingual artery before open midline glossectomy. Results were impressive, showing reduction in AHI (36.4)
and ESS (9.3) and an increase in LSAT of 17%.29 Such impressive results were most likely due to strict inclusion criteria, including preoperative CT-confirmed hypopharyngeal airway space \180 mm. This further demonstrates that in such a site-specific sleep surgery, such as UPPP and glossectomy, determination of site obstruction is imperative. Combined surgical success rate for all studies was 59.6% while surgical cure was achieved in 22.5% of cases. Despite its popularity, the definition used for success, .50% improvement and AHI \20, is far from ideal. It excludes drastic improvements in AHI or improvements in other sleep outcomes. For example, Mickelson and Rosenthal31 only reported a success rate of 25% (3/12). However, mean AHI was reduced by 26.7, which is similar to the mean reduction across all 18 of our reported studies. The LSAT also improved by 19% in the 9 nonresponders. Although other reviews have analyzed glossectomy in the past, many have not fully analyzed outcome parameters that have clinical significance. One review analyzed glossectomy as part of multilevel surgery and compared it to other hypopharyngeal surgeries based on success rate.35 Four of the 5 included studies in the review referenced previously were also included in the present analysis.12,13,28,31 Using the same definition of surgical success, they found that 50% (37/74) of treatments were considered successful, which is slightly less than the success rate found in our study (59.6%). Our success rate is also similar to other multilevel surgeries including genioglossus advancement and hyoid suspension, which at one time was considered firstline therapy for hypopharyngeal obstruction. The present review was only able to identify 24 subjects treated with glossectomy as sole therapy for OSA.20-22 Despite this small population, initial results are promising. Reduction in ESS score, in particular, compared favorably to multilevel surgery, –5.41 to 25.49, respectively. Unfortunately, pooled analysis for LSAT and AHI was not significant. These findings appear to support the general trend for advocating multilevel surgical therapy in patients with moderate to severe OSA. However, this negative finding may be partly explained by the small patient population and large deviation in results between patients in these studies. This is particularly evident with Vicini et al,22 who presented only 2 patients who received TORS glossectomy as a single therapy. One subject’s AHI decreased by 79 (91 to 12), while the other actually increased by 5 (32 to 37). Robinson et al21 did not achieve results until adding other levels of treatment. Their first 10 patients received plasma wand treatment, and only 2 of 10 were considered successful. The final 5 were given multilevel surgery (UPPP, hyoid suspension, and/or geniotubercle advancement), which increased success rate to 80% (4/5). Failures in this case were attributed to poor volume reduction with the plasma wand and residual obstruction mainly in the lateral pharyngeal wall. A recent review on the effectiveness of alternative treatments for OSA stated that they cannot recommend glossectomy as a single treatment option for obese patients with moderate to severe OSA.36 This review did not take
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
8
Otolaryngology–Head and Neck Surgery
the 3 studies we used to assess a single glossectomy treatment into account. Although evidence presented here cannot disagree with that statement, it is certainly possible that under the right circumstances and with identification of the correct subjects, glossectomy is an effective single treatment for OSA. Our study has some important limitations. There is a lack of quality research involving glossectomy with the majority of available published data drawn from small case series without control arms. This has made it difficult to determine the true effectiveness of glossectomy. Similar procedures vary in surgical approach, the inclusion criteria for patient selection, and types of additional procedures, and definitions of and the amount of attention paid to complications in each series, while similar, are not standardized across studies. Additionally, many of the isolated glossectomy patients analyzed received previous palate surgery. This makes it extremely difficult to truly compare treatments and complications. It should also be noted that AHI, while the most popular and universally reported outcome metric used for sleep medicine, has its limitations.37 Future directions for research include further evaluation of isolated glossectomy and direct comparisons of glossectomy to other tongue base surgeries in multilevel surgery.
Conclusion Our study suggests that glossectomy significantly improves sleep outcomes as part of multilevel surgery in select adult patients with OSA. Currently, there is insufficient evidence to analyze the role of glossectomy as a standalone procedure for the treatment of sleep apnea. Author Contributions Alexander W. Murphey, collected data, analyzed data, drafting article, revised article; Jessica A. Kandl, collected data, analyzed data, drafting article; Shaun A. Nguyen, designed study, analyzed and interpreted data, revised article; Aimee C. Weber, collected data, revised article; M. Boyd Gillespie, designed study, interpreted data, drafting article, revised article.
Disclosures Competing interests: None. Sponsorships: None. Funding source: None.
References 1. Epstein LJ, Kristo D, Strollo PJ, Jr, et al. Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults. J Clin Sleep Med. 2009;5:263-276. 2. Young T, Peppard PE, Gottlieb DJ. Epidemiology of obstructive sleep apnea: a population health perspective. Am J Respir Crit Care Med. 2002;165:1217-1239. 3. Chai-Coetzer CL, Luo YM, Antic NA, et al. Predictors of long-term adherence to continuous positive airway pressure therapy in patients with obstructive sleep apnea and cardiovascular disease in the SAVE study. Sleep. 2013;36:1929-1937.
4. Weaver TE, Grunstein RR. Adherence to continuous positive airway pressure therapy: the challenge to effective treatment. Proc Am Thorac Soc. 2008;5:173-178. 5. Fujita S, Conway W, Zorick F, et al. Surgical correction of anatomic azbnormalities in obstructive sleep apnea syndrome: uvulopalatopharyngoplasty. Otolaryngol Head Neck Surg. 1981;89:923-934. 6. Kezirian EJ, Maselli J, Vittinghoff E, et al. Obstructive sleep apnea surgery practice patterns in the United States: 2000 to 2006. Otolaryngol Head Neck Surg. 2010;143:441-447. 7. Sher AE, Schechtman KB, Piccirillo JF. The efficacy of surgical modifications of the upper airway in adults with obstructive sleep apnea syndrome. Sleep. 1996;19:156-177. 8. Abdullah VJ, Wing YK, van Hasselt CA. Video sleep nasendoscopy: the Hong Kong experience. Otolaryngol Clin North Am. 2003;36:461-471. 9. Bachar G, Feinmesser R, Shpitzer T, et al. Laryngeal and hypopharyngeal obstruction in sleep disordered breathing patients, evaluated by sleep endoscopy. Eur Arch Otorhinolaryngol. 2008;265:1397-1402. 10. Kotecha BT, Hannan SA, Khalil HM, et al. Sleep nasendoscopy: a 10-year retrospective audit study. Eur Arch Otorhinolaryngol. 2007;264:1361-1367. 11. Do KL, Ferreyra H, Healy JF, et al. Does tongue size differ between patients with and without sleep-disordered breathing? Laryngoscope. 2000;110:1552-1555. 12. Fujita S, Woodson BT, Clark JL, et al. Laser midline glossectomy as a treatment for obstructive sleep apnea. Laryngoscope. 1991;101:805-809. 13. Woodson BT, Fujita S. Clinical experience with lingualplasty as part of the treatment of severe obstructive sleep apnea. Otolaryngol Head Neck Surg. 1992;107:40-48. 14. Friedman M, Soans R, Gurpinar B, et al. Evaluation of submucosal minimally invasive lingual excision technique for treatment of obstructive sleep apnea/hypopnea syndrome. Otolaryngol Head Neck Surg. 2008;139:378-384. 15. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg. 2010;8:336-341. 16. Elshaug AG, Moss JR, Southcott AM, et al. Redefining success in airway surgery for obstructive sleep apnea: a meta analysis and synthesis of the evidence. Sleep. 2007;30:461467. 17. Borenstein M, Hedges LV, Higgins J, Rothstein HR. Introduction to Meta-analysis. New York: Wiley; 2009. 18. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177-188. 19. Freeman M, Tukey JW. Transformations related to the angular and the square root. Ann Math Stat. 1950;21:607-611. 20. Lin HS, Rowley JA, Badr MS, et al. Transoral robotic surgery for treatment of obstructive sleep apnea-hypopnea syndrome. Laryngoscope. 2013;123:1811-1816. 21. Robinson S, Lewis R, Norton A, et al. Ultrasound-guided radiofrequency submucosal tongue-base excision for sleep apnoea: a preliminary report. Clin Otolaryngol Allied Sci. 2003;28:341-345.
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
Murphey et al
9
22. Vicini C, Dallan I, Canzi P, et al. Transoral robotic surgery of the tongue base in obstructive sleep Apnea-Hypopnea syndrome: anatomic considerations and clinical experience. Head Neck. 2012;34:15-22. 23. Babademez MA, Yorubulut M, Yurekli MF, et al. Comparison of minimally invasive techniques in tongue base surgery in patients with obstructive sleep apnea. Otolaryngol Head Neck Surg. 2011;145:858-864. 24. Elasfour A, Miyazaki S, Itasaka Y, et al. Evaluation of uvulopalatopharyngoplasty in treatment of obstructive sleep apnea syndrome. Acta Otolaryngol Suppl. 1998;537:52-56. 25. Friedman M, Hamilton C, Samuelson CG, et al. Transoral robotic glossectomy for the treatment of obstructive sleep apnea-hypopnea syndrome. Otolaryngol Head Neck Surg. 2012;146:854-862. 26. Gunawardena I, Robinson S, MacKay S, et al. Submucosal lingualplasty for adult obstructive sleep apnea. Otolaryngol Head Neck Surg. 2013;148:157-165. 27. Li HY, Lee LA, Kezirian EJ. Coblation endoscopic lingual lightening (CELL) for obstructive sleep apnea [published online January 13, 2015]. Eur Arch Otorhinolaryngol. PMID: 25577994 28. Li HY, Wang PC, Hsu CY, et al. Same-stage palatopharyngeal and hypopharyngeal surgery for severe obstructive sleep apnea. Acta Otolaryngol. 2004;124:820-826. 29. Li S, Shi H. Lingual artery CTA-guided midline partial glossectomy for treatment of obstructive sleep apnea hypopnea syndrome. Acta Otolaryngol. 2013;133:749-754.
30. Lin HC, Friedman M, Chang HW, et al. Z-palatopharyngoplasty combined with endoscopic coblator open tongue base resection for severe obstructive sleep apnea/hypopnea syndrome. Otolaryngol Head Neck Surg. 2014;150:1078-1085. 31. Mickelson SA, Rosenthal L. Midline glossectomy and epiglottidectomy for obstructive sleep apnea syndrome. Laryngoscope. 1997;107:614-619. 32. Suh GD. Evaluation of open midline glossectomy in the multilevel surgical management of obstructive sleep apnea syndrome. Otolaryngol Head Neck Surg. 2013;148:166-171. 33. Toh ST, Han HJ, Tay HN, et al. Transoral robotic surgery for obstructive sleep apnea in Asian patients: a Singapore sleep centre experience. JAMA Otolaryngol Head Neck Surg. 2014; 140:624-629. 34. Woodson BT, Laohasiriwong S. Lingual tonsillectomy and midline posterior glossectomy for obstructive sleep apnea. Operative Techniques in Otolaryngology. 2012;23:155-161. 35. Kezirian EJ, Goldberg AN. Hypopharyngeal surgery in obstructive sleep apnea: an evidence-based medicine review. Arch Otolaryngol Head Neck Surg. 2006;132:206-213. 36. Keymel S, Kelm M, Randerath WJ. [Non-CPAP therapies in obstructive sleep apnoea: an overview]. Pneumologie. 2013; 67:50-57. 37. Tam S, Woodson BT, Rotenberg B. Outcome measurements in obstructive sleep apnea: beyond the apnea-hypopnea index. Laryngoscope. 2014;124: 337-343.
Downloaded from oto.sagepub.com at UNIV OF OTTAWA LIBRARY on August 6, 2015
