To Stent or Not to Stent? A Meta-analysis of Endonasal Congenital Bilateral Choanal Atresia Repair

Transcription

1 The Laryngoscope VC 2015 The American Laryngological, Rhinological and Otological Society, Inc. Systematic Review To Stent or Not to Stent? A Meta-analysis of Endonasal Congenital Bilateral Choanal Atresia Repair Julie E. Strychowsky, MD, FRCSC; Kosuke Kawai, ScD; Ethan Moritz, BS; Reza Rahbar, MD; Eelam A. Adil, MD Objectives: The use of nasal stents as a postoperative adjunct following repair of choanal atresia remains controversial. The study objective was to systematically review the literature regarding the efficacy and safety of stenting following transnasal endoscopic repair of bilateral choanal atresia. Study Design: Systematic review with meta-analysis. Methods: A comprehensive search in PubMed, EMBASE, CINAHL, and the Cochrane Library was conducted. Inclusion criteria included articles written in the English language with five or more subjects and clear intervention data and outcomes. Two independent reviewers screened studies for eligibility, appraised the level of evidence, extracted data, and resolved discrepancies by consensus. Successful surgery was defined as the absence of restenosis. Results: Of 154 identified studies, 15 met inclusion criteria. Levels of evidence varied from level 3 to 4. Mean age at surgery ranged from 5 days to 25 months. Thirteen studies (n patients) included patients who were stented; the weighted pooled proportion of successful surgery was 65% (95% confidence interval [CI], 49 76%). Mean duration of stenting ranged from 48 hours to 16 weeks. Six studies (n 5 42) evaluated patients who were not stented; the weighted pooled proportion of successful surgery was 64% (95% CI, 42 84%). Complications associated with stenting included alar injury, vestibular stenosis, columellar tear, and stent dislodgement or blockage. Conclusions: Success rates for bilateral choanal atresia repair were similar with and without the use of nasal stents. The use of nasal stents may be associated with more complications. There is insufficient data to determine if mitomycin C is a useful therapeutic adjunct. Key Words: Choanal atresia, stent, mitomycin, CHARGE, restenosis. Level of Evidence: NA Laryngoscope, 126: , 2016 INTRODUCTION Choanal atresia is defined as congenital obstruction of the posterior nasal choanae. 1,2 There are several theories regarding its embryologic origin, but abnormal neural crest cell migration is the most widely accepted. 2 The reported incidence varies between 1 in 5,000 to 1 in 9,000 live births. 1 The majority of cases are unilateral, and there is a higher prevalence among females. 1,2 From the Department of Otolaryngology and Communication Enhancement (J.E.S., K.K., E.M., R.R., E.A.A), Boston Children s Hospital, Boston, Massachusetts; and the Department of Otology and Laryngology (R.R., E.A.A.), Harvard Medical School, Boston, Massachusetts, U.S.A. Editor s Note: This Manuscript was accepted for publication April 27, Presented at the American Society of Pediatric Otolaryngology section of the Combined Otolaryngology Sections Meeting, Boston, Massachusetts, U.S.A., April 26, The authors have no funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Eelam A. Adil, MD, Department of Otolaryngology and Communication Enhancement, Boston Children s Hospital, 300 Longwood Avenue, LO-367, Boston, MA eelam.adil@ childrens.harvard.edu. DOI: /lary Mixed bony-membranous obstruction is found in 70% of patients. Pure bony obstruction is seen in 30% of patients and occurs secondary to an enlarged vomer and/or medialized pterygoid plates. 2 Congenital malformations such as CHARGE (coloboma, heart defects, atresia of the choanae, retardation of growth and development, genital and urinary abnormalities, ear abnormalities and/or hearing loss) and craniofacial syndromes such as Treacher Collins are commonly associated with bilateral choanal atresia. 2 Bilateral choanal atresia classically presents with respiratory distress at birth. Because newborns are obligate nasal breathers, bilateral nasal obstruction results in immediate neonatal respiratory difficulties that are alleviated once the newborn begins to cry. Patients often require emergent intubation, and repair is scheduled in the early neonatal period. In contrast, patients with unilateral choanal atresia may not present until later in life with chronic, foul-smelling, unilateral rhinorrhea. Various surgical techniques for choanal atresia repair have been described. Historically, blind puncture of the posterior choanae involved placing a trocar through the nostril and perforating the posterior 218

2 Fig. 1. Literature search results. [Color figure can be viewed in the online issue, which is available at choanae. The risk of skull base injury and advent of newer instrumentation led to development of transpalatal, and more recently, transnasal endoscopic techniques. 1 3 Restenosis is a potential complication following surgical repair via any method. 1,2 Intraoperative and postoperative adjuncts such as nasal stenting, mitomycin C, 4 balloon dilation, 5 among others have been suggested to improve outcomes. 1 3 The use of these adjuncts, nasal stents in particular, remains controversial. Given the rarity of this pathology, the literature regarding nasal stenting following surgical repair versus nonstenting is limited to small case series. A randomized controlled study comparing postsurgical adjuncts would be nearly impossible. Therefore, the primary objective of this systematic review with meta-analysis was to summarize the current literature to assess the efficacy and safety of nasal stenting for the repair of bilateral choanal atresia. METHODS This review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta- Analyses guideline. 6 Literature Search Strategy A comprehensive literature search including PubMed, EMBASE, CINAHL, and the Cochrane Library was conducted on September 3, The electronic database search combined disease-specific terms (choanal atresia) with interventionspecific terms (repair, endoscopic surgery, surgery) with adjunct-specific terms (stent, mitomycin C, dilation, balloon, laser). To ensure that all relevant published articles were captured, the search was not limited for publication date or study design. Relevant articles and abstracts were selected and reviewed. Reference lists from these sources and recent review articles were searched for additional publications. All studies written in the English language were considered for inclusion. Study Selection Criteria Articles were assessed for eligibility independently by two review authors (J.E.S., E.A.) and included if they were prospective or retrospective studies that reported surgical outcomes of pediatric patients (age <19 years) with bilateral congenital choanal atresia who underwent transnasal endoscopic surgical repair. The primary outcome measure was successful surgery both with and without nasal stenting. Success was defined as postoperative bilateral nasal patency without evidence of restenosis or need for revision surgery. Secondary outcome measures included restenosis, need for revision surgery, length of hospital 219

3 TABLE I. Study Characteristics and Patient Demographics. Study, Year and Country Study Design, Study Years Group No. of Patients Age Average Age Range Weight at Surgery Male Gender Other Anomalies CHARGE Bony Mixed Comparative studies (stent vs. nonstent) Saafan, 2013 Egypt 10 Prospective RCT, Eladl, 2010 Egypt 11 Retrospective review, Al-Ammar, 2008 Saudi Arabia 12 Retrospective review, Zuckerman et al., 2008 USA 13 Retrospective review, Stent studies Kinis et al., 2014 Turkey 14 Retrospective review, De Freitas et al., 2012 Australia 15 Retrospective review, Uzomefuna et al., 2012 Ireland 16 Retrospective review, Bozkurt et al., 2010 Turkey 17 Retrospective review, Romeh et al., 2010 Egypt 18 Retrospective review, Nour et al., 2008 Egypt 19 Prospective case series, Teissier et al., 2008 France 20 Retrospective review, Schraff et al., 2006 USA 21 Retrospective review, Khafagy, 2002 Egypt 22 Prospective case series, Nonstent studies El-Ahl et al., 2012 Egypt 23 Retrospective review, Ibrahim et al., 2010 Egypt 24 Retrospective review, Stent 10 4 days* NR* 3.1 kg (range, kg)* Nonstent 10 Stent 7 13 days 3 27 days 3.0 kg (range, kg) Nonstent 3 7 days 5 9 days 3.0 kg (range, kg) 7* 0* 0* 8* 12* 3 NR NR NR NR 2 1 Stent 5 NR NR NR NR NR NR NR NR Nonstent 6 NR NR NR NR NR NR NR NR Stent 4 8 days* 5 15 days* NR NR NR Nonstent 5 NR NR NR Stent days NR Stent days 0 32 days 2.7 kg (range, kg) Stent days 1 day 12 weeks NR Stent days days 3 7 days NR 4 NR NR NR NR Stent days 7 days 12 years NR 7 NR Stent days 2 25 days NR 4 NR Stent months SD 51.4 months NR NR Yes, Yes, 3 2 NR NR Stent 18 NR NR NR NR 3 9* NR NR Stent days NR Range, kg Nonstent 7 8 days 4 15 days NR Nonstent 11 NR 2 27 days NR *Not reported separately for stented versus nonstented patients. Not reported separately for patients with bilateral choanal atresia alone (patients with unilateral included). This group compared CHARGE versus non-charge patients. Results only for CHARGE versus non-charge and no differentiated by type of repair; study also reported outcomes for transpalatal repair. CHARGE 5 coloboma, heart defects, atresia of the choanae, retardation of growth and development, genital and urinary abnormalities, ear abnormalities and/or hearing loss; NR 5 not reported; RCT 5randomized control trial; SD 5 standard deviation. 220

4 TABLE II. Quality Assessment. Detsky Quality Scale for Randomized Trials Study, Year Randomization Outcome Exclusion Therapy Statistics Total Saafan, / 4 3 / 4 2 / 4 4 / 4 2 / 5 14 / 21 Newcastle-Ottawa Scale for Cohort Studies Author, Year Selection (Maximum 4 Asterisks) Comparability (Maximum 2 Asterisks) Exposure (Maximum 3 Asterisks) Total (Maximum 9 Asterisks) Comparative studies (stent vs. nonstent) Eladl, 2010 **** ** *** ********* Al-Ammar, 2008 **** ** *** ********* Zuckerman et al., 2008 **** ** *** ********* Stent studies Kinis et al., 2014 *** * *** ******* De Freitas et al., 2012 *** * *** ******* Uzomefuna et al., 2012 **** ** *** ********* Bozkurt et al., 2010 *** * *** ******* Romeh et al., 2010 *** * *** ******* Nour et al., 2008 *** * *** ******* Teissier et al., 2008 *** * *** ******* Schraff et al., 2006 **** ** *** ********* Khafagy, 2002 *** * *** ******* Nonstent studies El-Ahl et al., 2012 *** * *** ******* Ibrahim et al., 2010 *** * *** ******* stay, and complications. Other outcomes included the use of intraoperative adjuncts such as mitomycin C, balloon dilation, and laser. Studies including patients with unilateral choanal atresia repair were excluded unless bilateral choanal atresia patients were included in the same series and their results reported separately. Technical reports, case reports, and small case series (five patients or less) were excluded. CHARGE syndrome; however, we were able to examine only duration of stenting. We examined I 2 heterogeneity, which determines the percentage of the total variation among studies due to heterogeneity rather than chance. Publication bias was assessed by visual inspection of the funnel plot and Begg s rank correlation test. Stata version 13 (StataCorp, College Station, TX) was used for all analyses. Data Analysis Methodological quality of identified studies was appraised using the Oxford Center for Evidence-Based Medicine (OCEBM) 2011 Levels of Evidence, 7 the Detsky quality scale for randomized trials, 8 and the Newcastle-Ottawa scale for cohort studies. 9 Relevant data were extracted from fully published reports by two independent review authors (J.E.S., K.K.) following prescribed tables developed a priori. Disagreement was resolved by consensus. Descriptive statistics were extracted or calculated for success of transnasal endoscopic repair. Other data extracted included study characteristics, patient demographics (age and weight at time of surgery, gender, associated comorbidities, associated CHARGE syndrome, type of atresia), and outcome measures as previously described. A qualitative synthesis of results was performed when applicable. Meta-analysis was performed by calculating the pooled proportion of successful surgery both with and without nasal stents using DerSimonian- Laird weights for the random-effects model. We used Freeman- Tukey (double arcsine) transformation for proportions. Subgroup analyses were developed a priori and planned to include duration of stenting, use of mitomycin C, and association with RESULTS Two hundred ten studies were identified during the initial literature search (Fig. 1). One hundred fifty-four studies remained for consideration after duplicate publications were removed. Titles and abstracts of these studies were screened for the inclusion and exclusion criteria. The search strategy did not include a limit for study design; therefore, some studies excluded were technique articles, small case reports, or review articles. The remaining 84 full-text articles were reviewed in their entirety. Fifteen studies (n patients) satisfied the inclusion criteria and were included in the analysis Four studies compared outcomes for patients who were stented postoperatively versus those who were not stented Thirteen studies reported outcomes on patients who were stented ; six studies reported outcomes on patients who were not stented ,23,24 OCEBM levels of evidence varied from level 3 (one study) 10 to 4 (14 studies) Study characteristics and 221

5 analysis measured an I 2 of 65.6% (P 5.001). We found some evidence of publication bias in the funnel plot asymmetry plot and Begg s test (P 5.01); however, this is most likely due to similarly small sample sizes. Stent Versus Nonstent (Comparative Studies Only) Four studies (n 5 50 patients) compared success of surgery for patients with and without the use of nasal stents A pooled relative risk from the randomeffects model was 1.02 (0.41, 2.50), suggesting that there is no difference in the outcomes between patients who were stented and nonstented (Fig. 3). Heterogeneity analysis measured an I 2 of 32.5% (P 5.218). There was no evidence of publication bias. Fig. 2. Pooled analysis of successful surgery of comparative studies (stent versus nonstent). CI 5 confidence interval; ES 5 effect size. [Color figure can be viewed in the online issue, which is available at patient demographics are reported in Table I. Quality assessment is reported in Table II. Stent Versus Nonstent (All Studies) Thirteen studies (n patients; Fig. 2) included patients who were stented ; the weighted pooled proportion of successful surgery was 65% (95% confidence interval [CI], 49 79%) Six studies (n 5 42) evaluated patients who were not stented 10 13,23,24 ; the weighted pooled proportion of successful surgery was 64% (95% CI, 42 84%). There was no statistically significant difference between patients who were stented and those who were not (P 5.943). Overall heterogeneity Duration of Stenting The mean duration of stenting ranged from 2 days to 16 weeks We considered a stent duration of 7 days or less as short-term stenting versus stenting for longer than 7 days as long-term stenting. Most studies reported the use of modified endotracheal tubes for stenting. One study used nasopharyngeal tubes or calcium alginate packing; this study stented for 2 postoperative days. 20 A second study stented for a short time period of 5 to 7 days. 18 The remaining studies reported stenting patients for longer than 7 days (Table III) 10 17,19,21,22 We performed a subgroup analysis to compare short versus long duration of stenting (Fig. 4). Two studies that used short duration (7 days) of nasal stenting tended to show a higher success rate of 82% (69%-92%) compared to other studies that used longer duration (>7 days) of nasal stenting (60%; 95% CI, 41% 77%; P 5.03) Mitomycin C Two studies reported the use of mitomycin C among a portion of their patients stented after bilateral choanal Fig. 3. Pooled analysis of successful surgery of all studies (stent vs. nonstent). CI 5 confidence interval; ID 5 identification; RR 5 relative risk. [Color figure can be viewed in the online issue, which is available at www. laryngoscope.com.] 222

6 TABLE III. Outcomes Study, Year Group No. of Patients Success Restenosis Revision Stent Duration, Type of Stent MMC, Dose Duration of Hospital Stay Follow-Up Comparative studies (stent vs. nonstent) Saafan, Stent NR 4 weeks, modified ETT None NR Median 3.2 years (range, 1 5 years) Nonstent NR N/A None NR Median 3.2 years (range, 1 5 years) Eladl, Stent * 2.7 weeks (range, 1 4 weeks), polyethylene ETT None 6.9 days (range, 3 8 days) Nonstent * N/A None 5.3 days (range, 4 6 days) 19.7 months (range, 8 35 months) 15.7 months (range, 5 24 months) Al-Ammar, Stent NR 4 6 weeks, polyvinylchloride ETT None NR 2.8 years (range, 1 6 years) Nonstent NR N/A None NR 2.8 years (range, 1 6 years) Zuckerman et al., Stent weeks, ETT None NR 7.5 months Nonstent N/A None NR 2.5 months Stent studies Kinis et al., Stent NR days (range, 3 21 days), ETT stents De Freitas et al., Stent NR 6 75 days (range, days), modified ETT Uzomefuna et al., Stent weeks, polyvinylchloride stents None 14.3 days months (range, 6 61 months) None NR 22.2 months (range, months) 3 patients, 0.4 mg/ ml 3 5 minutes then irrigated with NS NR 4 weeks 3 years Bozkurt et al., Stent weeks (mean 31 days days; range, days), NR 4 patients,0.4 mg/ NR months ml 3 3 minutes Romeh et al., Stent days, Portex ETT None NR months (range, months) Nour et al., Stent weeks, Portex ETT None NR months (range, months) Teissier et al., None NR NR Stent hours, nasopharyngeal tubes or calcium alginate packing Schraff et al Stent weeks (range, 4 16 weeks) CHARGE and 6 weeks (range, 2 16 weeks) non-charge, but both include all types of surgeries, NR NR 24.6 months CHARGE all surgical techniques and 24.9 months non-charge all surgical techniques Khafagy, Stent weeks, Portex ETT None NR months 223

7 TABLE III. (Continued) Duration of Hospital Stay Follow-Up No. of Patients Success Restenosis Revision Stent Duration, Type of Stent MMC, Dose Study, Year Group Nonstent studies El-Ahl et al., Nonstent N/A None months Ibrahim et al., Nonstent N/A None NR 6 18 months *Minor debridement. There were 1/3 patients with MMC who remained open and 2/9 patients without MMC who remained open. There were 4/4 patients with MMC who remained open and 5/8 patients without MMC who remained open. This group compared CHARGE versus non-charge patients. CHARGE 5 coloboma, heart defects, atresia of the choanae, retardation of growth and development, genital and urinary abnormalities, ear abnormalities and/or hearing loss; ETT 5endotracheal tube; MMC 5 mitomycin C; N/A 5 not applicable; NR 5 not reported; NS 5 normal saline. atresia repair. 16,17 Uzomefuna and colleagues treated three out of 12 patients with 0.4 mg/ml on a cottonoid pledget for 5 minutes; the authors reported that the choanae of one out of three patients remained opened versus two out of nine managed without the medication. 16 Bozkurt et al. demonstrated the favorable outcome of successful surgery among patients managed with mitomycin C as an adjunct to stenting (four out of four patent versus five out of eight patent without). 17 The same dose was used but over a duration of 3 minutes. CHARGE Syndrome Association Patients with CHARGE syndrome were included in six studies ,19,21 The majority of studies did not report outcomes separately for those with CHARGE syndrome ,19 One study reported surgical success for patients with and without the syndrome. 21 Schraff and colleagues reported a prevalence of bilateral choanal atresia in 64% of patients with CHARGE versus 19% of those without CHARGE (P 5.006). 21 They also reported a nonsignificant but slightly longer median duration of stent placement among patients with CHARGE; however, this was reported for patients with bilateral and unilateral choanal atresia combined. Among patients with CHARGE who underwent transnasal repair, three out of six (50%) required revision surgery versus two out of seven (28%) for patients without CHARGE. 21 When all patients were included in the analysis (unilateral and bilateral atresia, and transnasal and transpalatal repairs), CHARGE children compared with non- CHARGE children were more likely to be significantly younger at the time of repair and have bilateral atresia. Duration of stenting and need for revision surgery was not statistically significantly different between groups. Fig. 4. Pooled analysis of successful for surgery by stent duration (short versus long). CI 5 confidence interval; ES 5 effect size. [Color figure can be viewed in the online issue, which is available at 224

8 TABLE IV. Complications. Study, Year Group No. of Patients Complications (No.) Comparative studies (stent vs. nonstent) Saafan, Stent 10 Stent dislodgement (n 5 1), premature extrusion (n 5 1), erosion of nares (n 5 1) Nonstent 10 Hemorrhage (n 5 1) Eladl, Stent 7 Granulation (n 5 2), left vestibular stenosis (n 5 1) Nonstent 3 Bleeding (n 5 1), granulation (n 5 1) Al-Ammar, Stent 5 NR Nonstent 6 NR Zuckerman et al., Stent 4 NR Nonstent 5 NR Stent studies Kinis et al., Stent 13 Soft palate perforation (n 5 1) De Freitas et al., Stent 20 Operative complications: palatal laceration (n 5 2), alar laceration (n 5 1);Postoperative complications: stent blockage (n 5 4), alar pressure ulcer (n 5 1), stent dislodgement (n 5 1) Uzomefuna et al., Stent 12 NR Bozkurt et al., Stent 12 NR Romeh et al., Stent 22 Columellar tear from stitch that healed (n 5 1) Nour et al., Stent 14 Bleeding requiring endoscopic control (n 5 1); soft palate perforation intraoperatively unrelated to stent (n 5 1) Teissier et al., Stent 27 NR Schraff et al., Stent 18* NR Khafagy, Stent 9 Death from profuse intraoperative bleeding, presumed hemorrhagic disease of the newborn (n 5 1) Nonstent studies El-Ahl et al., Nonstent 7 Small hard palate perforation that healed conservatively (n 5 1) Ibrahim et al., Nonstent 11 Midnasal synechiae not impacting breathing (n 5 3) *This group compared CHARGE versus non-charge patients. CHARGE 5 coloboma, heart defects, atresia of the choanae, retardation of growth and development, genital and urinary abnormalities, ear abnormalities and/or hearing loss; NR 5 not reported. Length of Stay Duration of hospital stay was reported in two studies. 11,14 Eladl compared patients with and without stents and demonstrated a shorter length of stay in hospital for patients who were nonstented (6.9 days [3 8 days] vs. 5.3 days [4 6 days]). The second study reported a length of stay of 14.3 days among 13 patients, all of whom were stented. 14 It is important to note that there are multiple confounding comorbidities that could extend hospital length of stay irrespective of the bilateral choanal atresia. Complications Complications were reported in nine studies and are outlined in Table IV. 10,11,14,15,18,19,22 24 Most complications were associated with the use of nasal stents in the postoperative period. DISCUSSION To our knowledge, this is the first systematic review and meta-analysis to report the pooled success of endoscopic repair of bilateral congenital choanal atresia in pediatric patients with and without the use of nasal stents. Analysis of the 13 studies that included 173 patients who were stented demonstrated that the weighted pooled proportion of successful surgery was similar and not significantly different when compared with the six studies that evaluated 42 patients who were not stented (65% [49% 79%] vs. 64% [42% 84%]). From an analysis of four comparative studies including 50 patients, the pooled relative risk of 1.02 (0.41, 2.50) suggests that there is no difference in surgical success between groups. One of the major complications of choanal atresia repair is restenosis of the posterior choanae. However, the etiopathology of restenosis is incompletely understood. The use of the endoscope has driven surgical approaches away from a blind puncture technique. In addition to enhanced visualization, this may allow for the creation of a wider opening in the atretic plates. Many surgeons advocate removing the posterior vomer to create a single neochoanae. This posterior septectomy may help to reduce the likelihood of restenosis given the larger size of the opening. There is controversy surrounding the use of nasal stents, which is the thrust of this article and discussed herein. The use of mitomycin C as an inhibitor of fibroblast proliferation is also discussed. Patient comorbidities, which may inhibit 225

9 wound healing, may also contribute to the likelihood of restenosis. There was variability in duration of stenting among studies, ranging from 48 hours to 16 weeks The two studies 18,20 that used short duration (7 days) of nasal stenting showed a higher success rate of 82% (69% 92%) compared to the other studies that used longer duration (60%; 95% CI, 41% 77%; P 5.03). This may suggest that shorter-duration stenting is associated with better surgical outcomes when compared with longerduration stenting. However, this conclusion is based on the pooled analysis of only two studies and should be interpreted with caution. There was heterogeneity when examining the mean age at time of surgery (12.4 days vs months) as well as duration of stenting (5 7 days vs. 48 hours) among these studies. The use of mitomycin C as an adjunct warrants discussion. Mitomycin is an antibiotic produced by Streptomyces caespitosus and has both antineoplastic and antiproliferative properties. 4 The latter allows for the inhibition of fibroblast proliferation and decrease in scar formation. Reduced scar formation following topical application of mitomycin C in the pediatric aerodigestive tract has been described in the literature. 25 In this current analysis of 15 studies involving 215 patients, two studies reported outcomes associated with the use of mitomycin C. Within these two studies combined, only seven out of 24 patients total were treated with this topical adjunct. It is difficult to offer conclusions as to the role of mitomycin C in the setting of bilateral choanal atresia repair with such limited data. There is a paucity of published literature on the management of choanal atresia in patients with CHARGE syndrome. It has been demonstrated that the prevalence of bilateral choanal atresia is higher among patients with CHARGE syndrome when compared with isolated cases. 21 Surgical management may be more challenging given the associated medical comorbidities, specifically cardiac abnormalities such as tetralogy of Fallot, patent ductus arteriosus, and ventricular and atrial septal defects. Although six studies included patients with CHARGE syndrome, 13 16,19,21 only one study directly compared outcomes for CHARGE versus non-charge patients. 21 Schraff and colleagues demonstrated a higher need for revision surgery among CHARGE patients who underwent transnasal repair, although the case series was small. 21 Complications were associated with and without nasal stenting; however, there were fewer and less serious complications reported when nasal stents were not used. Those associated with stenting included alar laceration and erosion, vestibular stenosis, columellar tear, and stent dislodgement or blockage. Alar erosion and vestibular stenosis are worrisome due to the potential long-term functional and aesthetic sequelae. These complications were only reported by studies that used longer duration stenting. A definite correlation is difficult to determine given the limited data. As is the case with all observational studies, the evaluations included in this meta-analysis may have been biased due to lack of comparability between stent 226 and nonstent groups with respect to patient characteristics (e.g., age, type of atresia, and other coexisting anomalies), surgical technique, or the use of intraoperative and postoperative adjuncts. For example, stenting may be preferentially used in neonates with greater severity of disease and at high risk of restenosis. To reduce potential source of heterogeneity, we employed a strict inclusion criteria, which only included patients with bilateral choanal atresia and those managed via a transnasal endoscopic surgical approach. Given the rarity of this congenital anomaly, most studies had a small sample size, which may have partially contributed to the heterogeneity in the estimate. We conducted an analysis of the four studies that had comparison groups for stenting versus nonstenting However, these four studies did not adjust for potential confounding factors in the statistical analysis due to limited sample size. For surgical technique, we limited our inclusion criteria to studies reporting a purely transnasal endoscopic approach and thus excluded studies that used a combined transoral approach or microscopic approach. Surgeon preferences beyond this limitation persist. We excluded studies that did not report outcomes separately for patients with bilateral and unilateral choanal atresia. This was to limit the introduction of bias from artificially improved outcomes that may be associated with the inclusion of unilateral choanal atresia repair patients who are often older at the time of surgical repair. It is possible that confounding from stent type and duration cannot be excluded. Given the small number of patients who were treated with adjunctive topical mitomycin C, we were unable to perform a subgroup analysis. Patients who received mitomycin C were included in their respective stent and nonstent group, which has the potential to introduce confounding. Case series were the most common type of study; this study type satisfied the criteria for level 4 evidence according to the OCEBM. Future studies could be aimed toward higher levels of evidence; however, this may not be possible given the rarity of bilateral choanal atresia. Larger cohort studies would also add to the current literature; a quality assessment tool such as the Newcastle- Ottawa scale for cohort studies could be referenced to improve study quality. In preparing future publications, clinician researchers could consider reporting outcomes such as length of time from initial repair to restenosis when applicable, number of revision surgeries, bony versus mixed bony-membranous atresia, use of mitomycin C, length/type of stenting, and hospital length of stay separately for patients with unilateral versus bilateral choanal atresia as well as for those with CHARGE syndrome. Postoperative management is often variable and at the discretion of the surgeon. These may include the use of nasal saline irrigations, antibiotic and/or corticosteroid drops, and planned return to the operating room for an examination under anesthesia with possible debridement or balloon dilation as warranted. Confounding from variability in this postoperative management cannot be excluded. Further studies are needed to explore possible benefits of these interventions.

10 CONCLUSION Similar success rates for bilateral choanal atresia repair were reported with and without the use of nasal stents. The use of nasal stents may be associated with more frequent and more severe complications. Perioperative adjuncts such as nasal stenting should be considered in the context of the individual patient and surgeon preference. If the surgeon decides to stent, a duration of 1 week or shorter may be associated with better surgical outcomes; however, the evidence is limited. It appears that there is insufficient data to determine if mitomycin C is a useful therapeutic adjunct. BIBLIOGRAPHY 1. Ramsden JD, Campisi P, Forte V. Choanal atresia and choanal stenosis. Otolaryngol Clin North Am 2009;42: Corrales CE, Koltai PJ. Choanal atresia: current concepts and controversies. Curr Opin Otolaryngol Head Neck Surg 2009;17: Bedwell JR, Choi SS. Are stents necessary after choanal atresia repair? Laryngoscope 2012;122: Rahbar R, Jones DT, Nuss RC, et al. The role of mitomycin in the prevention and treatment of scar formation in the pediatric aerodigestive tract: friend or foe? Arch Otolaryngol Head Neck Surg 2002;128: Reipl R, Scheithauer M, Hoffmann TK, Rotter N. Transnasal endoscopic treatment of bilateral choanal atresia in newborns using balloon dilatation: own results and review of the literature. Int J Pediatr Otorhinolaryngol 2014;78: Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol 2009;62:e1 e OCEBM Levels of Evidence Working Group. The Oxford 2011 levels of evidence. Oxford Centre for Evidence-Based Medicine. Available at: Accessed March 24, Detsky AS, Naylor CD, O Rourke K, McGeer AJ, L Abbe K. Incorporating variations in the quality of individual randomized trials into meta-analysis. J Clin Epidemiol 1992;45: Wells G, O Connel D, Peterson J, Welch V, Losos M, Tugwell P. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomized studies in meta-analysis. Available at: clinical_epidemiology/oxford.asp. Accessed August 19, Saafan ME. Endoscopic management of congenital bilateral posterior choanal atresia: value of using stents. Eur Arch Otorhinolaryngol 2013;270: Eladl HM. Transnasal endoscopic repair of bilateral congenital choanal atresia: Controversies. J Laryngol Otol 2010;124: Al-Ammar AY. The use of nasal stent for choanal atresia. Saudi Med J 2008;29: Zuckerman JD, Zapata S, Sobol SE. Single-stage choanal atresia repair in the neonate. Arch Otolaryngol Head Neck Surg 2008;134: Kinis V, Ozbay M, Akdag M, et al. Patients with congenital choanal atresia treated by transnasal endoscopic surgery. J Craniofac Surg 2014;25: De Freitas RP, Berkowitz RG. Bilateral choanal atresia repair in neonates a single surgeon experience. Int J Pediatr Otorhinolaryngol 2012;76: Uzomefuna V, Glynn F, Al-Omari B, Hone S, Russell J. Transnasal endoscopic repair of choanal atresia in a tertiary care centre: a review of outcomes. Int J Pediatr Otorhinolaryngol 2012;76: Bozkurt MK, Keles B, Azimov A, Ozturk K, Arbag H. The use of adjunctive topical mitomycin in endoscopic congenital choanal atresia repair. Int J Pediatr Otorhinolaryngol 2010;74: Romeh HE, Albirmawy OA. A 13-year experience and predictors for success in transnasal endoscopic repair of congenital choanal obliteration. Int J Pediatr Otorhinolaryngol 2010;74: Nour YA, Foad H. Swinging door flap technique for endoscopic transeptal repair of bilateral choanal atresia. Eur Arch Otorhinolaryngol 2008;265: Teissier N, Kaguelidou F, Couloigner V, François M, Van Den Abbeele T. Predictive factors for success after transnasal endoscopic treatment of choanal atresia. Arch Otolaryngol Head Neck Surg 2008;134: Schraff SA, Vijayasekaran S, Meinzen-Derr J, Myer CM. Management of choanal atresia in CHARGE association patients: a retrospective review. Int J Pediatr Otorhinolaryngol 2006;70: Khafagy YW. Endoscopic repair of bilateral congenital choanal atresia. Laryngoscope 2002;112: El-Ahl MA, El-Anwar MW. Stentless endoscopic transnasal repair of bilateral choanal atresia starting with resection of vomer. Int J Pediatr Otorhinolaryngol 2012;76: Ibrahim AA, Magdy EA, Hassab MH. Endoscopic choanoplasty without stenting for congenital choanal atresia repair. Int J Pediatr Otorhinolaryngol 2010;74: