Drug-induced sedation endoscopy in surgically naive children with Down syndrome and obstructive sleep apnea

Sleep Medicine 24 (2016) 63–70

Contents lists available at ScienceDirect

Sleep Medicine j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / s l e e p

Original Article

Drug-induced sedation endoscopy in surgically naive children with Down syndrome and obstructive sleep apnea Mieke Maris a, Stijn Verhulst b, Vera Saldien c, Paul Van de Heyning a, Marek Wojciechowski b, An Boudewyns a,* a

Department of Otorhinolaryngology, Head and Neck Surgery, University Hospital Antwerp, Belgium-Antwerp University, Belgium Department of Pediatrics, University Hospital Antwerp, Belgium-Antwerp University, Belgium c Department of Anesthesiology, University Hospital Antwerp, Belgium-Antwerp University, Belgium b

A R T I C L E

I N F O

Article history: Received 6 December 2015 Received in revised form 20 May 2016 Accepted 4 June 2016 Available online 22 August 2016 Keywords: Children Down syndrome Obstructive sleep apnea Upper airway Endoscopy

A B S T R A C T

Objective: To describe the pattern of upper airway (UA) obstruction in surgically naive children with Down syndrome and obstructive sleep apnea (OSA), and to evaluate the outcome of drug-induced sedation endoscopy (DISE)-directed treatment. Methods: A prospective study of DISE in surgically naive children with Down syndrome and OSA was performed. Treatment was individually tailored based on the DISE findings and was evaluated by control polysomnography (PGS). Results are presented as median (lower–upper quartile) unless otherwise stated. Results: In 41 children, aged 4.2 years (range, 2.8–6.0) with a body mass z score of 1.04 (−0.55 to 1.82) and obstructive apnea–hypopnea index (oAHI) of 10.1/h (range, 6.3–23.0), DISE was performed. Adeno-/ tonsillar obstruction was found in 75.6% of the patients, and these patients subsequently underwent UA surgery. Seven patients were non-surgically treated, and three received a combined treatment. A multilevel collapse was present in 85.4%. Tongue base obstruction was present in ten patients (24.4%) and epiglottic collapse in 48.8%. Pre- and postoperative PSG data were available for 25 children (adenotonsillectomy, n = 16; tonsillectomy, n = 7; adenoidectomy, n = 2). A significant improvement in oAHI from 11.4/h (range, 7.7– 27.0) to 5.5/h (range, 2.1–7.6) was found. Persistent OSA was present in 52% of the children. No significant association between different DISE findings and persistent OSA could be found. Conclusion: Most patients with Down syndrome and OSA present with multilevel collapse on DISE. Adenotonsillectomy results in a significant improvement of the oAHI; however more than half of the patients had persistent OSA, probably due to multilevel collapse. Upper airway evaluation may provide more insights into the pattern of UA obstruction in patients with persistent OSA. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Children with Down syndrome (DS) are predisposed to a number of health problems affecting their development and quality of life, as described in several earlier reports [1–3]. The prevalence of obstructive sleep apnea (OSA) in children with DS is 30–60%, which is substantially higher compared to 1–4% in a general pediatric population [3]. Multiple factors predispose children with DS to upper airway (UA) obstruction, such as midfacial hypoplasia, relative macroglossia (due to smaller bony framework of mandible and maxilla), hypotonia, obesity, hypothyroidism, an immature immune system with more respiratory infections, and a higher risk of gastroesophageal reflux. Furthermore, laryngomalacia, subglottic stenosis, and

* Corresponding author. Department of Otorhinolaryngology, Head and Neck Surgery, University Hospital Antwerp, Belgium-Antwerp University, Wilrijkstraat 10, 2650 Edegem, Belgium. Fax: +003238214425. E-mail address: [email protected] (A. Boudewyns). http://dx.doi.org/10.1016/j.sleep.2016.06.018 1389-9457/© 2016 Elsevier B.V. All rights reserved.

tracheomalacia are more frequently reported in children with DS [4,5]. All of these factors contribute to the high risk of developing complex UA obstruction in children with DS. Adenotonsillectomy (AT) is the first-line treatment for pediatric OSA, with a success rate of 71–87% in a normally developing pediatric population [6,7]. The success rate of AT in children with DS and OSA is, however, poor and less favorable compared to that in normally developing children. This has been attributed to multilevel UA obstruction in children with DS [8,9]. Persistent OSA has been observed in 30– 50% of children with DS following adenotonsillectomy [8,10]. A thorough and detailed evaluation of the UA should be performed to identify the level and degree of UA obstruction in children known to have multilevel UA collapse, such as those with DS. Depending on these results, an individually tailored treatment may be proposed to optimize treatment outcome. Different techniques are reported in the literature to evaluate the UA in children with OSA. A detailed clinical examination should always be performed, and awake flexible laryngoscopy with or without imaging can be

64

M. Maris et al. / Sleep Medicine 24 (2016) 63–70

done, although neither allows a dynamic evaluation of the airway during sleep. Cine-magnetic resonance imaging (MRI) may provide a high-resolution examination of the dynamic airway, particularly in children with multilevel obstruction [11]. However, experience with this technique is limited mainly to a single center [12]. Druginduced sedation endoscopy (DISE) in children was introduced as a method to evaluate the UA of children with complex UA obstruction or persistent OSA after AT [8,13,14]. It is a safe and costeffective tool to evaluate the level and degree of UA obstruction during drug-induced sleep [15,16]. In contrast to cine-MRI, DISE allows direct visualization of the airway at multiple levels during spontaneous breathing, demonstrating the type and degree of airway obstruction in a clinical setting, and the findings may direct further surgery without the need of another sedation or admission to the hospital [8]. Previous reports in children with complex UA obstruction describe a heterogeneous population including children with DS, but also children with other medical conditions predisposing to UA collapse and children with or without a history of previous UA surgery [9,15,17]. This study aims, first, to describe the pattern of UA obstruction during DISE in a group of children with DS and OSA, without previous UA surgery; and, second, to evaluate the outcome of DISEdirected treatment and to investigate which DISE findings are related to failure of UA surgery. 2. Methods A prospective study was performed including all children with DS who are enrolled in our multidisciplinary Down team (Antwerp University Hospital, Belgium), seen at the Ear Nose Throat (ENT) department and diagnosed with OSA. This study was approved by the local Ethics Committee (B30020108389). All parents gave written informed consent. Children with previous UA surgery [(adeno)tonsillectomy] were excluded from the study. In each child, a clinical examination was performed, and tonsil size was scored according to the Brodsky score [18]. Body mass index (BMI) z scores were calculated according to Flanders growth curves for boys and girls. Diagnosis of OSA was based upon full-overnight polysomnography (PSG), performed at the Pediatric Sleep Disorders Center of the Antwerp University Hospital, Belgium. The following variables were continuously measured and recorded by a computerized polysomnography device (Brain RT, OSG, Rumst, Belgium): electroencephalography (C4/Al, C3/A2, F3/A2, F4/A1), electro-oculography, electromyography of anterior tibialis and chin muscles, and electrocardiography. Respiratory effort was measured by respiratory inductance plethysmography and oxygen saturation by a finger probe connected to a pulse oximeter. Airflow was measured by means of nasal pressure cannula and thermistor, and snoring was detected by means of a microphone at the suprasternal notch. Children were also monitored on audio/videotape using an infrared camera. If snoring was present during the night, this was reported (as present or absent). Polysomnograms were manually scored by certified technicians according to international guidelines [19]. A diagnosis of OSA is established with an obstructive apnea– hypopnea index (oAHI) ≥ 2/h [20]. Obstructive sleep apnea severity was defined as mild with oAHI ≥2/h and <5/h, moderate with oAHI ≥5/h and <10/h, and severe OSA with an oAHI ≥10/h. Drug-induced sedation endoscopy was performed in the operating theater by a single pediatric ear–nose–throat surgeon, as previously described [21]. The children were sedated on the operating table with full cardiovascular monitoring by a pediatric anesthesiologist, using facial mask with a mixture of sevoflurane and oxygen. When intravenous access was obtained, sevoflurane was stopped and intravenous propofol was administered with a bolus injection of 1–2 mg, followed by continuous infusion according to

body weight (6–10 mg/kg/h) to obtain the desired level of sedation and to maintain spontaneous breathing. Once a stable breathing pattern was obtained, a flexible fiberoptic laryngoscope was passed through a swivel adaptor on the mask and introduced into one nostril up to the level of the nasopharynx. No local anesthesia was used, and care was taken to avoid any pressure from the mask on the patient’s face. The examination was performed with the child lying supine and the head in a neutral position. From the nasopharynx, the scope was gently passed toward the oral cavity, hypopharynx, and larynx. At each level, the pattern of UA collapse was scored according to a standard protocol [21]. The presence and degree of fixed and dynamic airway obstruction were noted. At each UA level, the degree of obstruction was described. At the level of the nasopharynx, adenoid hypertrophy was graded as follows: 0, no adenoids; 1, adenoids occupying <50% of the lumen; 2, adenoids occupying 50–75% of the lumen; and 3, adenoids >75% obstruction of the nasopharynx by adenoid tissue. A score ≥2 was considered significant adenoid hypertrophy. Tonsillar obstruction was graded as 0 when there were no tonsils present; 1, with <50% collapse of the tonsils; 2, with 50–90% collapse of the tonsils; and 3, with tonsils touching at the midline. A score ≥2 was considered as indicating significant tonsillar hypertrophy. Tongue base obstruction was always in anteroposterior direction and scored as absent (0), partial collapse (1), or complete collapse (2). Dynamic obstructions include hypotonia, defined as follows: a circumferential collapse at the entire oropharyngeal and/or hypopharyngeal level; anteroposterior palatal collapse or flutter; an anteroposterior collapse of the epiglottis, which is sucked against the posterior pharyngeal wall; and findings of late-onset laryngomalacia. The latter is characterized by redundant mucosa of the aryepiglottic folds being pulled into the airway and causing UA obstruction during forceful inspiration (type 1 laryngomalacia). Dynamic collapse was either absent (0) or present (1) for the following parameters: anteroposterior palatal collapse, collapse of the epiglottis, laryngomalacia, and hypotonia at oropharyngeal of hypopharyngeal level [21]. Multi-level obstruction was defined as the presence of one or more UA abnormalities outside the adenotonsillar region. When DISE showed clinically relevant obstruction at the level of the adenoids or tonsils, the child was intubated and ventilated at the end of the endoscopic evaluation, and UA airway surgery, ie, (adeno)/tonsillectomy, was performed during the same anesthesia. Surgery was performed with cold instruments; in case of tonsillectomy, the anterior and posterior tonsillar pillars were sutured with resorbable sutures. If no surgical intervention was required, the child was awakened following the DISE examination and transferred to the recovery room. All children undergoing surgery were scheduled for a control PSG at least three months postoperatively. Surgical intervention was considered successful with a postoperative oAHI <5/h, and a complete cure was defined with a postoperative oAHI <2/h. An overview of the decision-making process is depicted in Fig. 1. In the case of persistent disease (oAHI ≥5/h), additional treatment was based on OSA severity and preoperative DISE findings. A third PSG was scheduled to evaluate additional treatment outcome. Statistical analysis was performed using IBM SPSS statistical version 20. Data are reported as median value with lower and upper quartiles unless otherwise stated. Correlations between variables were calculated using the Spearman correlation coefficient, and preand postoperative data were compared by the Wilcoxon signedrank test. Statistical significance was obtained at a p value of <0.05. 3. Results A total of 41 children with DS who met the inclusion criteria underwent DISE between September 2010 and June 2015. The study group consisted of 18 boys and 23 girls, aged 4.2 years (range,

M. Maris et al. / Sleep Medicine 24 (2016) 63–70

65

Fig. 1. Overview of the treatment decision-making process.

2.8–6.0), BMI z score 1.04 (range, −0.55 to 1.82). Patients mostly had moderate to severe OSA with a median oAHI of 10.1/h (range, 6.3– 23.0), mean oxygen saturation (meanSat) 96.2% (range, 94.4– 97.0%), and minimum oxygen saturation (minSat) 86.0% (81.5– 90.3%). Only 22% of the patients had mild OSA. 3.1. DISE findings All DISE examinations were performed without adverse effects. In 35 children (85.4%), a multilevel obstruction was found. More than half of the children presented with significant adenoid hypertrophy (53.7%). Significant tonsillar obstruction is found in 65.9% of the children, and in 17 children a significant combined collapse at the level of the adenoids and tonsils was reported (41.4%). Only two children presented without any tonsillar collapse (4.9%), and in 12 children there was less than 50% obstruction (29.3%). Tongue base obstruction was present in a minority of the children (24.4%; three with complete collapse and seven with partial anteroposterior collapse). Dynamic obstruction of the UA was less frequently reported, and in almost half of the children epiglottis collapse was present (48.8%). Late-onset laryngomalacia was found in five children with DS during DISE (12.2%). The median age of the children who presented with late-onset laryngomalacia was 3.1 years (range, 1.9–5.2 years), which was not significantly younger than children without laryngomalacia (median age, 4.3 years; range, 2.8–7.1 years), (p = 0.4). Hypotonia was described in 22% of children. An overview of the DISE findings at each UA level is presented in Figs. 2 and 3. There was no significant difference in OSA severity (oAHI) between the children with a multilevel UA obstruction (9.3/h [range, 5.8–22.5]) and those with UA obstruction only at the level of the tonsils and/or adenoids (14.2/h [range, 6.1–33.3]) (p = 0.5). There was no correlation between the Brodsky score or tonsillar obstruction

during DISE and the oAHI. However, a significant correlation was found between the degree of tonsillar obstruction during DISE and the Brodsky score (p < 0.001, r = 0.74). 3.2. DISE-directed treatment An overview of the different DISE-directed treatment modalities is presented in Fig. 4. Most of the children underwent surgical treatment. Of the children, 31 underwent an (adeno)tonsillectomy following the DISE examination (75.6%). A total of 26 children undergoing UA surgery had multilevel obstruction (83.9%). A combined therapy of UA surgery with medication/continuous positive airway pressure (CPAP)/ orthodontics was proposed in three children. One child had a tonsillectomy, combined with orthodontics and montelukast, and another child underwent adenoidectomy followed by CPAP therapy. In one child, adenoidectomy combined with partial resection of the inferior turbinates was proposed; however, this child was lost to follow-up. For the other seven children, treatment was started with montelukast, whether or not it was combined with a nasal corticosteroid spray, CPAP, or orthodontics. 3.3. Outcome of DISE-directed treatment 3.3.1. Upper airway surgery Pre- and postoperative PSG data were available in 25 of the 31 children after UA airway surgery. Of these children, 16 underwent an adenotonsillectomy, seven a tonsillectomy, and two an adenoidectomy. The median postoperative time for the postoperative PSG was 3.8 (range, 3.1–5.3) months. There was a significant decrease in the oAHI from 11.4/h (range, 7.7–27.0 months) preoperative to 5.6/h (range, 2.1–7.6) postoperative (p = 0.001), as presented in Fig. 5, and a significant increase in meanSat from 96.1% (range,

66

M. Maris et al. / Sleep Medicine 24 (2016) 63–70

Fig. 2. Drug-induced sedation endoscopy findings for upper airway (UA) obstruction at the level of the adenoids, tonsils, and tongue base (fixed UA obstruction).

94.6–96.8) to 96.8% (range 95.1–98.1) (p = 0.03). The minSat increased from 84.5% (range, 79.2–89.0) to 88.5% (range, 83.9–90.1), but this change was not significant (p = 0.06). There was no significant change in BMI z score postoperatively. After treatment, snoring was significantly less registered during postoperative PSG (p = 0.01). Adverse effects of UA surgery were reported in four children. One child had respiratory problems during the first night postoperatively, and short-term use of oxygen support was required. Three other patients were readmitted to the hospital because of vomiting during the first week postoperatively, and intravenous rehydration therapy was started.

Fig. 3. Drug-induced sedation endoscopy findings for upper airway (UA) obstruction at the level of the palate and epiglottis, and the presence of hypotonia and laryngomalacia (dynamic UA obstruction).

When the success rate of adenotonsillectomy was evaluated, 12 of the 25 surgically treated patients obtained an oAHI <5/h postoperatively (oAHI 2.1/h [range, 0.7–3.0]). Thirteen children (52%) had persistent OSA after (adeno)tonsillectomy oAHI of 7.2/h (range, 6.6– 14.8). A multilevel collapse during preoperative DISE was found in 84.6% of the children with persistent OSA after UA surgery and in 75.0% with successful treatment. No significant association between multilevel collapse and persistent OSA was found in our study population (p = 0.6); however, only five of the 25 children presented without multilevel collapse during DISE (20%). No significant association between different DISE findings and persistent OSA could be found. An overview of the management of the children with persistent OSA is summarized in Table 1. Treatment of persistent OSA was guided by preoperative DISE findings and OSA severity. A control PSG was scheduled to evaluate supplementary treatment outcome. Patient 1 had persistent OSA after adenoidectomy at the age of one year. Because of his young age, no tonsillectomy was performed. A significant improvement in OSA severity was found postoperatively (oAHI 7.2/h) and he was treated with montelukast for residual disease. A third PSG (29 months after surgery) revealed no persistent OSA (oAHI 1/h). Patient 2 showed positional OSA after tonsillectomy and was successfully treated with a sleep positioner trainer (SPT). In patient 3, the use of an SPT (because of positional OSA on PSG) was not successful, and montelukast was prescribed. Persistent OSA was found on the third PSG 12 months after surgery (oAHI 13.5/h). This child was scheduled for reevaluation of the UA by DISE. Patient 4 had a cold during the postoperative PSG, and a third PSG 11 months after surgery showed successful treatment, with an oAHI of 2.6/h. Patient 5 was treated with montelukast, as preoperative DISE findings did not show collapse outside the level of the tonsils, adenoids, and epiglottis. Patient 6 presented with a collapse at the level of the adenoids and tonsils, with laryngomalacia. He was treated with montelukast and nasal corticosteroid spray. Supraglottoplasty was considered; however, a third PSG showed complete cure. Patient 7 was initially treated by

M. Maris et al. / Sleep Medicine 24 (2016) 63–70

67

Fig. 4. Drug-induced sedation endoscopy suggested treatment in children with Down syndrome and obstructive sleep apnea (OSA) (n = 41).

montelukast because of persistent OSA after AT. However, at the third PSG, a remarkable increase in OSA severity was observed and bilevel non-invasive ventilation was started. In patient 8, an obstruction at the level of the palate, adenoids, and tonsils was found on preoperative DISE. She was treated with montelukast and nasal corticosteroid spray but stopped the medication a couple of weeks before the third PSG, which showed moderately severe OSA. The parents of patient 9 refused long-term use of medication, so further clinical follow-up was advised, as no other levels of UA obstruction outside the adenotonsillar level and epiglottis were found on preoperative DISE. Patient 10 had initially very severe OSA (preoperative oAHI 70/h), which was lowered to oAHI of 6.3/h postoperatively with mostly hypopneas, and no obstruction outside the adenotonsillar level on preoperative DISE. Clinical follow-up under nasal corticosteroid spray was proposed. Continuous positive airway pressure (CPAP) was started in patient 11

because of severe OSA on postoperative PSG (oAHI 21.5/h) and hypotonia on preoperative DISE. Because of poor tolerance, this treatment was stopped. A third PSG (without CPAP) showed mild OSA that was treated with nasal corticosteroid spray. For subject 12, an orthodontic treatment was proposed based on tongue base obstruction on preoperative DISE, but this patient was lost to follow-up. Patient 13 presented with laryngomalacia and tongue base collapse during DISE and was treated with omeprazole 10 mg twice daily, combined with montelukast. A third PSG showed persistent OSA (oAHI 8.9/h), and on subsequent DISE obvious lingual tonsillar hypertrophy was present. A lingual tonsillectomy by means of transoral robotic surgery was proposed. For nine patients, a third PSG was available, and in five patients (55.6%), an oAHI <5/h was obtained, with a decrease in median oAHI from 7.2/h (range, 6.6–11.2) to 3.0/h (range, 1.9–13.9). This difference was not significant (p = 0.7). 3.3.2. Medical treatment Seven children received nonsurgical treatment. Four children underwent a second PSG after nonsurgical treatment. Polysomnographic and DISE findings are summarized in Table 2. Although this study group was small, we could not find an improvement in oAHI after nonsurgical treatment (oAHI pretreatment 7.9/h [range, 5.1–23.9] vs oAHI posttreatment 6.9/h [range, 1.6–9.1]). Pre- and posttreatment oAHI for all of the children after DISEdirected treatment (surgical, nonsurgical, and combined therapy) showed a significant improvement in oAHI from oAHI 11.4/h (range, 7.5–25.2) before treatment vs oAHI 5.8/h (2.1–7.7) posttreatment (p = 0.001). Overall, persistent OSA was found in 53.3% of the children after DISE-directed treatment. 4. Discussion

Fig. 5. Pre- and postoperative obstructive apnea–hypopnea index (oAHI) in 25 patients with Down syndrome. (Left) Median values with upper and lower quartiles. (Right) Values for individual patients.

To the best of our knowledge, this is the first report on DISE finding in a homogeneous group of children with DS and OSA

68

M. Maris et al. / Sleep Medicine 24 (2016) 63–70

Table 1 Overview of the patients with persistent obstructive sleep apnea (OSA) after upper airway (UA) surgery. DISE findings Subject oAHI first PSG preoperatively (n/h)

Treatment oAHI second PSG postoperatively (n/h)

Surgery

1 2

51.6 61.3

A3 P0 T1 TB0 E0 H1 LM0 Adenoidectomy A1 P 0 T3 TB0 E0 H0 LM0 Tonsillectomy

7.2 6.8

3

8.7

A1 P0 T2 TB2 E0 H0 LM0

Tonsillectomy

4 5 6

5.8 3.5 9.3

A0 P0 T1 TB0 E1 H0 LM1 A2 P0 T3 TB0 E1 H0 LM0 A2 P0 T0 TB0 E0 H0 LM1

Tonsillectomy 8.1 Adenotonsillectomy 6.8 Adenoidectomy 14.2

7 8

23.8 11.8

A2 P1 T3 TB0 E0 H0 LM0 A2 P1 T3 TB0 E0 H0 LM0

Adenotonsillectomy Adenotonsillectomy

9 10 11 12 13

35.8 70 30 11.1 11.1

A3 P0 T3 TB0 E1 H0 LM0 A2 P1 T3 TB0 E0 H0 LM0 A1 P0 T1 TB0 E0 H1 LM0 A0 P0 T3 TB1 E0 H0 LM0 A3 P0 T3 TB0 E0 H0 LM0

Adenotonsillectomy 5.9 Adenotonsillectomy 6.3 Tonsillectomy 21.5 Tonsillectomy 21.6 Adenotonsillectomy 7.2

15.4

8.2 5.6

oAHI third Final treatment PSG (n/h)

Montelukast Sleep position trainer NightBalance Sleep position trainer NightBalance Watchful waiting Montelukast Montelukast + nasal corticosteroid spray Montelukast Montelukast + nasal corticosteroid spray Montelukast Nasal corticosteroid spray CPAP Orthodontics + montelukast Montelukast

1/h 2 – 2.6 1.2 1.8

Adenotonsillectomy – Montelukast – – –

35.3 6.3

BiPAP –

– 6.3 3 – 8.9

– – Nasal corticosteroid spray – TORS, proton pump inhibitors

Abbreviations: CPAP, continuous positive airway pressure; DISE, drug-induced sedation endoscopy; DISE findings: A, adenoids; E, epiglottis; H, hypotonia; LM, laryngomalacia; P, palate; T, tonsils; TB, tongue base. A/T: 1 = <50% collapse, 2 = 50–75%; 3 = >75% obstruction/tonsils touch at midline; P/TB/E/H/LM: 0 = no collapse, 1 = collapse present. oAHI, obstructive apnea–hypopnea index; oAHI at first PSG (preoperative); oAHI at second PSG (postoperative), supplementary treatment; oAHI at third PSG, final treatment recommendation; PSG, polysomnography; TORS, trans-oral robotic surgery.

without previous UA surgery. The outcomes of DISE-directed treatment were documented by postoperative PSG. In this study, all DISE examinations were performed without adverse effects. Overall, in 85.4% of the children, a multilevel obstruction was found. More than half of the children presented with adenoid hypertrophy, and in 65.9% a significant obstruction at the level of the tonsils was found. In our study, 53.7% of the children had significant adenoid hypertrophy, which is remarkably lower compared to 75% of normally developing children in the study reported by Boudewyns et al. [21]. Fung et al. reported similar findings in their case–control study [8]. Only 24% of the patients had tongue base collapse (7% complete and 17% partial) in our study. Other authors reported that children with DS have relative rather than true macroglossia, due to a relatively large tongue compared to the small size of their bony framework formed by a small maxilla and mandible [4,22]. A complete tongue base collapse was seen on DISE in only three children. Two were referred for orthodontics, and it is planned that the third child will undergo transoral robotic resection of lingual tonsils. Besides fixed UA obstruction, most children with DS presented with dynamic UA obstruction during DISE examination. We found a high prevalence of epiglottic collapse in children with DS (49%) as compared to normally developing children (<25%) [21]. Hypotonia was reported in only 22% of the DS children, comparable to the data reported by Boudewyns et al. in normally developing children [21]. Laryngomalacia was found in 12% of the patients with DS in this study. In a retrospective analysis, Mitchell

et al. found laryngomalacia in 43% of children with DS [23]. However, in that study, laryngomalacia was seen especially in young children (<1 month of age) and frequently associated with the presence of gastroesophageal reflux disease. This refers to “congenital laryngomalacia,” which should be differentiated from late-onset laryngomalacia. Revell et al. described late-onset laryngomalacia as a cause for pediatric OSA, in which children do not present clinically with symptoms of stridor but, rather, with typical symptoms of OSA [24]. The predominant feature in these patients is redundant mucosa of the aryepiglottic folds, which is pulled into the airway, causing obstruction during forceful inspiration. Laryngomalacia in our population may be classified as late-onset laryngomalacia. The contribution of late-onset laryngomalacia in UA obstruction in children with DS and OSA may be underestimated. Future studies are required to elucidate the role of lateonset laryngomalacia in children with DS and OSA later in childhood. In this study, no correlation was found between Brodsky score or tonsillar obstruction during DISE and OSA severity (oAHI). This might be explained by the relative low number of subjects and a nonnormal distribution. However, Mitchell et al. evaluated the predictive value of clinical factors for OSA severity in a large study population of children with OSA (N = 450) [25]. In that study, no correlation between tonsil score and OSA severity was found, which is in line with our findings. Although tonsil score did not correlate with OSA severity, the Brodsky score was significantly correlated to the degree of tonsillar obstruction during DISE in this study.

Table 2 Overview of results for patients undergoing drug-induced sedation endoscopy (DISE)-directed nonsurgical treatment. Subject

oAHI first PSG (n/h)

DISE findings

Treatment

oAHI second PSG (n/h)

Treatment

1 2 3 4

7.7 4.2 29.1 8.1

A1 P0 T1 TB0 E1 H1 LM0 A3 P0 T3 TB0 E0 H0 LM0 A1 P0 T1 TB0 E1 H1 LM1 A1 P0 T1 TB2 E0 H0 LM0

Nasal corticosteroid spray + montelukast Montelukast CPAP + nasal corticosteroid spray SPT + nasal corticosteroid spray + montelukast

6.2 9.6 0 (with CPAP) 7.7

Lost to follow-up Adenotonsillectomy Lost to follow-up Nasal corticosteroid spray + montelukast

Abbreviations: CPAP, continuous positive airway pressure; DISE findings: A, adenoids; E, epiglottis; H, hypotonia; LM, laryngomalacia. P, palate; T, tonsils; TB, tongue base. oAHI, obstructive apnea–hypopnea index; oAHI pre, oAHI before treatment; oAHI post, oAHI posttreatment. P/TB/E/H/LM: 0 = no collapse, 1 = collapse present; SPT, sleep position trainer. A/T: 1 = <50% collapse, 2 = 50–75%, 3 = >75% obstruction/tonsils touch at midline.

M. Maris et al. / Sleep Medicine 24 (2016) 63–70

Based on the DISE findings, the majority of the surgically naive children with DS (75.6%) underwent UA surgery during the same anesthesia. Postoperatively, a significant decrease in oAHI was obtained, although 52% of the children had persistent OSA (oAHI >5/h). This finding indicates that (adeno)tonsillectomy may be a first treatment in what becomes a process of surveillance in DS children. Our success rate with AT is comparable to the results of Shete et al., in which 55% of the children needed CPAP or surgery after AT [6]. In that study, however, only 11 patients with DS, with a higher mean age (8.5 years), were included, and clinical data (eg, Brodsky score) were not reported. Multilevel collapse was found in 85.4% of the children, which is high compared to 56.7% in normally developing children as reported by Boudewyns et al. [21]. Also, in 84.6% of the children with persistent OSA multilevel collapse on DISE was found. This might explain the rather low percentage of success of AT compared to normally developing children. We did not find a significant correlation between multilevel collapse on DISE and persistent OSA. However the number of children without multilevel collapse was low in our study population (five of 25 surgically treated patients; 20%). Future studies should include more children in both groups (ie, with and without multilevel collapse) to provide a more reliable evaluation of the role of multilevel collapse in DISE in persistent OSA. A sample size of 37 children in each group would be necessary to achieve a power of 80%. In the case of persistent OSA, further treatment was based on the DISE findings and the severity of persistent OSA. A nonsurgical treatment was proposed based on the DISE findings in seven children. A combined treatment seemed appropriate in three children. In more than half of the patients with persistent OSA, an oAHI <5/h was obtained after additional treatment (55.6%) based on the preoperative DISE findings and OSA severity. Only limited data were available to evaluate the outcome of treatment in the non-surgically treated group with no significant effects. Based upon these data, we cannot draw conclusions regarding the role of nonsurgical treatment in DS children with OSA. In this study, we did not find an association between preoperative DISE findings in surgically naive children with DS and OSA, and persistent OSA after (adeno)tonsillectomy. These findings suggest that we could not predict the risk of persistent disease based on preoperative findings, although the number of children presenting with a specific site of UA obstruction may be rather low for statistical analysis. In previous studies, persistent OSA after (adeno)tonsillectomy was associated with tongue base obstruction [26,27]. A possible explanation might be that previous studies on UA findings in children with persistent OSA were performed several months after the initial surgery, and that changes in UA anatomy took place in the meantime, such as lingual tonsillar hypertrophy (after removal of adenotonsillar tissue). One could speculate that lingual tonsillar hypertrophy (causing tongue base obstruction) was less common in our study population because these children did not have previous UA surgery (and were therefore less likely to develop compensatory hypertrophy of lymphoid tissues at other airway levels such as the tongue base). One child (patient 13) in this study population had a postoperative DISE that revealed lingual tonsillar hypertrophy. Although we did not repeat DISE in all of the children with persistent OSA, we believe that this may have additional value and should be included in future management programs for persistent OSA. Comparing pre- and postoperative DISE findings may provide more insights into changes in UA morphology after surgery for OSA. Besides compensatory lingual tonsillar hypertrophy after (adeno)tonsillectomy, children with DS have been found to present more commonly with lingual tonsillar hypertrophy compared to normally developing children, increasing with age [28]. Our study population was rather young, which may also explain the low number of tongue base collapses on DISE. In the literature, DISE has been reported to be a valuable tool in the evaluation of UA obstruction. A recently published study

69

proposes to perform DISE prior to surgery, irrespective of any previous UA surgery, to select those patients who would benefit the most from surgery and to optimize treatment outcome [21]. In addition, Goldberg et al. suggested DISE prior to surgery to increase the success rate of UA surgery [29]. These authors concluded that DISE is a useful tool in the evaluation of OSA in children, especially in those with persistent OSA after AT, those without (adeno)tonsillar hypertrophy, and those with significant comorbidities. In this study, DISE allowed us to describe the pattern of UA obstruction in a cohort of children with DS without previous UA surgery. The majority of these children present with multilevel UA obstruction, which was not always combined with obstruction at the level of the adenoids or the tonsils. This resulted in a nonsurgical treatment in 17.1% of the children, which may suggest that also in children with DS, an a priori selection of the patients who would benefit most from, surgery can be made based on DISE findings. As discussed before, limited data were available for analysis of the non-surgically treated patients. When DISE is performed at the time of the first surgery (usually adeno/tonsillectomy), it provides information regarding the different sites of UA collapse. This information may then be used to guide treatment in patients with persistent disease following adenotonsillectomy. If one limited DISE to patients with persistent OSA, this would require an additional anesthesia. In children with persistent OSA following adenotonsillectomy, UA findings are usually more complex, and then a third anesthesia might be required to perform complementary surgery such as tongue base surgery, or supraglottoplasty, as these treatment modalities require extensive counseling of the parents. This scenario is also likely to cause a substantial delay before achieving a complete treatment response. There are a few studies objectively evaluating outcome measurement of complex surgery for pediatric OSA. CPAP remains the gold standard in treating OSA, although this treatment is not always tolerated by children with DS, requiring alternative treatment. Our study has some limitations. First, discussion remains regarding the drug of choice during DISE to induce sedation. In this study, sevoflurane inhalation was used, followed by intravenous propofol administration as described before [21]. Dexmedetomidine is proposed as an alternative drug to induce sedation during DISE, resulting in non–rapid eye movement (NREM) sleep without significant respiratory depression [30]. Mahmoud et al. examined the dose–response effects of dexmedetomidine and propofol on airway morphology in children and adolescents with a history of OSA. Although propofol may reduce the cross-sectional area of the entire airway in a dose-dependent way, these authors found that both agents may provide acceptable levels of anesthesia in sleep studies for patients with OSA (during MRI) with nonsignificant changes in airway dimension [31]. The use of dexmedetomidine in children is not licensed for this purpose in Europe, and thus propofol may be considered as an alternative anesthetic choice. In this study, care was taken to avoid high-dose–dependent effects of propofol on UA dynamics by using standard drug doses according to body weight in a continuous infusion. Furthermore, not all children had postoperative PSG data available, and there was only a small group of children who were nonsurgically treated. A larger study group is needed to evaluate the DISE-directed treatment outcome in this group. Some children with persistent OSA were scheduled for additional treatment based on DISE findings (orthodontics, medication, transoral robotic surgery [TORS]); however these results were not yet available. 5. Conclusion In this study, most of the children with DS presented with severe OSA and a multilevel UA collapse on DISE. Adenotonsillectomy significantly decreased postoperative oAHI; however, more than half

70

M. Maris et al. / Sleep Medicine 24 (2016) 63–70

of the patients had persistent OSA, probably due to multilevel UA collapse. Conflict of interest The ICMJE Uniform Disclosure Form for Potential Conflicts of Interest associated with this article can be viewed by clicking on the following link: http://dx.doi.org/10.1016/j.sleep.2016.06.018. References [1] Bull MJ, Committee on Genetics. Health supervision for children with Down syndrome. Pediatrics 2011;128:393–406. [2] Chin CJ, Khami MM, Husein M. A general review of the otolaryngologic manifestations of Down syndrome. Int J Pediatr Otorhinolaryngol 2014;78:899– 904. [3] Shott SR. Down syndrome: common otolaryngologic manifestations. Am J Med Genet C Semin Med Genet 2006;142C:131–40. [4] Lal C, White DR, Joseph JE, et al. Sleep-disordered breathing in down syndrome. Chest 2015;147:570–9. [5] Bertrand P, Navarro H, Caussade S, et al. Airway anomalies in children with Down syndrome: endoscopic findings. Pediatr Pulmonol 2003;36:137–41. [6] Ye J, Liu H, Zhang GH, et al. Outcome of adenotonsillectomy for obstructive sleep apnea syndrome in children. Ann Otol Rhinol Laryngol 2010;119:506–13. [7] Mitchell RB. Adenotonsillectomy for obstructive sleep apnea in children: outcome evaluated by pre- and postoperative polysomnography. Laryngoscope 2007;117:1844–54. [8] Fung E, Witmans M, Ghosh M, et al. Upper airway findings in children with down syndrome on sleep nasopharyngoscopy: case-control study. J Otolaryngol Head Neck Surg 2012;41:138–44. [9] Myatt HM, Beckenham EJ. The use of diagnostic sleep nasendoscopy in the management of children with complex upper airway obstruction. Clin Otolaryngol Allied Sci 2000;25:200–8. [10] Shete MM, Stocks RM, Sebelik ME, et al. Effects of adeno-tonsillectomy on polysomnography patterns in down syndrome children with obstructive sleep apnea: a comparative study with children without down syndrome. Int J Pediatr Otorhinolaryngol 2010;74:241–4. [11] Shott SR, Donnelly LF. Cine magnetic resonance imaging: evaluation of persistent airway obstruction after tonsil and adenoidectomy in children with down syndrome. Laryngoscope 2004;114:1724–9. [12] Manickam PV, Shott SR, Boss EF, et al. Systematic review of site of obstruction identification and non-CPAP treatment options for children with persistent pediatric obstructive sleep apnea. Laryngoscope 2015;[published Online First: Epub Date]. [13] Durr ML, Meyer AK, Kezirian EJ, et al. Drug-induced sleep endoscopy in persistent pediatric sleep-disordered breathing after adenotonsillectomy. Arch Otolaryngol Head Neck Surg 2012;138:638–43.

[14] Lin AC, Koltai PJ. Sleep endoscopy in the evaluation of pediatric obstructive sleep apnea. Int J Pediatr Otorhinolaryngol 2012;2012:576719. [15] Ulualp SO, Szmuk P. Drug-induced sleep endoscopy for upper airway evaluation in children with obstructive sleep apnea. Laryngoscope 2013;123:292–7. [16] Croft CB, Thomson HG, Samuels MP, et al. Endoscopic evaluation and treatment of sleep-associated upper airway obstruction in infants and young children. Clin Otolaryngol Allied Sci 1990;15:209–16. [17] Truong MT, Woo VG, Koltai PJ. Sleep endoscopy as a diagnostic tool in pediatric obstructive sleep apnea. Int J Pediatr Otorhinolaryngol 2012;76:722–7. [18] Brodsky L. Modern assessment of tonsils and adenoids. Pediatr Clin North Am 1989;36:1551–69. [19] Berry RB, Budhiraja R, Gottlieb DJ, et al. Rules for scoring respiratory events in sleep: update of the 2007 AASM manual for the scoring of sleep and associated events. Deliberations of the sleep apnea definitions task force of the American Academy of Sleep Medicine. J Clin Sleep Med 2012;8:597–619. [20] Marcus CL, Moore RH, Rosen CL, et al. A randomized trial of adenotonsillectomy for childhood sleep apnea. N Engl J Med 2013;368:2366–76. [21] Boudewyns A, Verhulst S, Maris M, et al. Drug-induced sedation endoscopy in pediatric obstructive sleep apnea syndrome. Sleep Med 2014;15:1526–31. [22] Guimaraes CV, Donnelly LF, Shott SR, et al. Relative rather than absolute macroglossia in patients with down syndrome: implications for treatment of obstructive sleep apnea. Pediatr Radiol 2008;38:1062–7. [23] Mitchell RB, Call E, Kelly J. Diagnosis and therapy for airway obstruction in children with down syndrome. Arch Otolaryngol Head Neck Surg 2003; 129:642–5. [24] Revell SM, Clark WD. Late-onset laryngomalacia: a cause of pediatric obstructive sleep apnea. Int J Pediatr Otorhinolaryngol 2011;75:231–8. [25] Mitchell RB, Garetz S, Moore RH, et al. The use of clinical parameters to predict obstructive sleep apnea syndrome severity in children: the Childhood Adenotonsillectomy (CHAT) study randomized clinical trial. JAMA Otolaryngol Head Neck Surg 2015;141:130–6. [26] Fricke BL, Donnelly LF, Shott SR, et al. Comparison of lingual tonsil size as depicted on MR imaging between children with obstructive sleep apnea despite previous tonsillectomy and adenoidectomy and normal controls. Pediatr Radiol 2006;36:518–23. [27] Donnelly LF, Shott SR, LaRose CR, et al. Causes of persistent obstructive sleep apnea despite previous tonsillectomy and adenoidectomy in children with down syndrome as depicted on static and dynamic cine MRI. AJR Am J Roentgenol 2004;183:175–81. [28] Sedaghat AR, Flax-Goldenberg RB, Gayler BW, et al. A case-control comparison of lingual tonsillar size in children with and without down syndrome. Laryngoscope 2012;122:1165–9. [29] Goldberg S, Shatz A, Picard E, et al. Endoscopic findings in children with obstructive sleep apnea: effects of age and hypotonia. Pediatr Pulmonol 2005;40:205–10. [30] Ehsan Z, Mahmoud M, Shott SR, et al. The effects of anesthesia and opioids on the upper airway: a systematic review. Laryngoscope 2016;126:270–84. [31] Mahmoud M, Jung D, Salisbury S, et al. Effect of increasing depth of dexmedetomidine and propofol anesthesia on upper airway morphology in children and adolescents with obstructive sleep apnea. J Clin Anesth 2013;25:529–41.