Objective assessment of supraglottoplasty outcomes using polysomnography

Results From a total of 46 patients, 10 were suitable for inclusion into the study. Mean age at first presentation was 2 months and 19 days (range 30–134 days). The surgical procedure performed was bilateral division of aryepiglottic folds alone in 1 patient (1/10), bilateral division of aryepiglottic folds and bilateral trimming of arytenoid mucosa in 5 patients (5/10) and bilateral aryepiglottic fold division, bilateral arytenoid mucosal trimming and epiglottic trimming or epiglottopexy in 4 patients (4/10). Statistically significant improvements occurred in mean values for Total Sleep Time (TST) ( P = 0.049), Lowest Oxygen Saturation Levels (SpO2 nadir) ( P = 0.006), Obstructive Apnoea Hypopnoea Index (OAHI) ( P = 0.009) and Respiratory Disturbance Index (RDI) ( P = 0.002), following supraglottoplasty. An improvement in mean Transcutaneous Carbon Dioxide (TcCO2) value occurred, but did not achieve statistical significance (57.1 vs. 52.8) ( P = 0.259). The mean age at which post-operative polysomnography confirmed a reversal of abnormal respiratory parameters following surgery was 5 months and 18 days. Conclusions Polysomnography is an effective method for objectively assessing the efficacy of supraglottoplasty for laryngomalacia. Supraglottoplasty effectively reverses the abnormal respiratory parameters occurring in moderate to severe laryngomalacia. Keywords Laryngomalacia Polysomnography Supraglottoplasty Obstructive sleep apnea 1 Introduction Laryngomalacia accounts for 75% of neonatal stridor [1] . Presentation typically occurs in the first few weeks of life with most infants presenting simply with intermittent inspiratory stridor. Most cases are entirely self-limiting with a reported 70% rate of spontaneous resolution in the first year of life [2] . In approximately 10% of cases however, infants will have a more severe form of laryngomalacia associated with significant respiratory difficulties, poor feeding and failure to thrive. Pathologically laryngomalacia occurs due to collapse of the structures of the supraglottis towards the glottic airway during inspiration. Supraglottoplasty is the principal surgical treatment option for moderate to severe laryngomalacia and involves surgically modifying the supraglottis in an effort to reduce its collapse during respiration. To-date several studies have assessed the efficacy of supraglottoplasty for treating laryngomalacia, however most studies have assessed subjective outcomes, such as reduced stridor, improved feeding and reduced respiratory difficulties. Few studies however have measured the outcomes of supraglottoplasty objectively; in particular its effect on reversing measurable respiratory parameters affected by laryngomalacia. Obstructive sleep apnoea syndrome is estimated to have a prevalence of 3% among children aged 6 months to 3 years of age, with no figures estimating the incidence below the age of 6 months [3] . The effects of untreated obstructive sleep apnoea among children are well documented and include fatigue, developmental delay and behavioural difficulties. More serious medical consequences include cor-pulmonale, pulmonary hypertension and an association with sudden infant death syndrome (SIDS) [4] . Obstructive sleep apnoea remains an overlooked component of laryngomalacia and the consideration of laryngomalacia in the context of obstructive sleep apnoea syndrome would seem obvious. Polysomnography is the gold standard method of assessment of patients with obstructive sleep apnoea and therefore its use in the assessment of patients with laryngomalacia would also seem logical. The use of polysomnography in laryngomalacia allows the opportunity to objectively assess the severity of airway obstruction prior to surgery and equally to measure outcomes following supraglottoplasty. The aim of this study was to objectively assess, using both pre- and post-operative polysomnography, the efficacy of supraglottoplasty in treating moderate to severe laryngomalacia. 2 Methods 2.1 Study design Approval was obtained from the Institutional Review Board of Princess Margaret Hospital prior to the commencement of this retrospective study. Clinical notes were reviewed for all patients undergoing supraglottoplasty for laryngomalacia while under the care of the senior author of this article, from March 2006 to February 2009. 2.2 Patient selection Inclusion criteria were a clinical diagnosis of moderate to severe laryngomalacia requiring surgical intervention, with the diagnosis being confirmed by history and clinical findings, including fibreoptic nasopharyngoscopy. Furthermore, for inclusion in our study, patients were required to have been assessed by full overnight polysomnography, both before and after supraglottoplasty. 2.3 Exclusion criteria In order to avoid potential delays in the treatment of more severe cases, patients whose clinical presentation with laryngomalacia included recurrent episodes of cyanosis and oxygen desaturations, were not referred for polysomnography prior to surgery. Further exclusion criteria included patients undergoing revision supraglottoplasty surgery, cases with second airway lesions and cases where supraglottoplasty was performed in combination with other procedures such as adenoidectomy or adenotonsillectomy. 2.4 Surgical procedure At our institution supraglottoplasty is performed under general anaesthesia, with induction and maintenance of anaesthesia occurring by either inhalational or intravenous techniques. Spontaneous ventilation is maintained throughout the procedure, which is generally performed without endotracheal intubation occurring. An initial full airway assessment is performed using a rigid zero degree Hopkins telescope to assess the larynx, trachea and distal airways. The purpose of this initial assessment is threefold: (1) to observe and confirm the diagnosis of laryngomalacia during spontaneous ventilation, (2) to assess in each case the particular structural abnormalities causing laryngomalacia and (3) to exclude a second airway lesion. A suspension laryngoscopy is then performed to expose the supraglottic structures and an operating microscope is positioned. A single dose of dexamethasone (0.6 mg/kg – max 8 mg) is administered intravenously. Supraglottoplasty surgery then involves treatment specific to the abnormalities identified on the initial airway assessment and involves varying combinations of: (1) division of both aryepiglottic folds, (2) resection bilaterally of redundant arytenoid mucosa and (3) trimming of the epiglottis or epiglottopexy; with the procedure being performed with cold instrumentation throughout. Patients are generally is transferred to the intensive care unit for overnight observation following the procedure. Omeprazole (1 mg/kg) is prescribed orally for a period of 3 months following surgery with clinical follow-up arranged for approximately 6 weeks post-operatively. 2.5 Polysomnography Pre-operative polysomnography was scheduled for all patients prior to supraglottoplasty at the time of booking for surgery. Following surgery a post-operative polysomnogram was scheduled at the first post-operative appointment. Polysomnography at our institution involves overnight admission to the paediatric sleep laboratory where formal polysomnographic data is collected. Respiratory parameters measured include, Total Sleep Time (TST), Central Apnoea Index (CAI), Obstructive Apnoea Index (OAI), Total Apnoea Index (TAI), Obstructive Apnoea Hypopnoea Index (OAHI), Respiratory Disturbance Index (RDI), Lowest Oxygen Saturation Measurement (SpO2 nadir) and Highest Carbon Dioxide reading by Transcutaneous Measurement (TcCO2). All polysomnographic studies were reviewed and formally reported-on by a consultant paediatric respiratory physician. 2.6 Statistical methods Pre- and post-operative polysomnographic data for all patients was compared using a paired Student's t -test. A probability value ( P -value) of less than 0.05 was considered statistically significant. 3 Results 3.1 Demographics Fort-six patients underwent supraglottoplasty surgery during the 35-month period of the study. Of these 10 patients had undergone both pre- and post-operative polysomnography and were therefore suitable for inclusion in our study. Clinical notes were available for review in all 10 cases. These patients comprised of seven males and three females. Patient ages at first presentation ranged from 30 days to 134 days with a mean of 2 months and 19 days while ages at post-operative polysomnography ranged from 67 days to 241 days with a mean of 5 months 18 days. 3.2 Co-morbidities Two patients were known to have Downs’ Syndrome (Patients 4 and 6). 3.3 Waiting times The mean waiting time from (1) first clinical appointment to pre-operative polysomnography was 10.6 days; (2) from pre-operative polysomnography to supraglottoplasty was 12.1 days; (3) from supraglottoplasty to post-operative polysomnography was 83.2 days; and (4) from first appointment to post-operative polysomnography was 100.7 days. 3.4 Surgery 3.4.1 Surgical findings The most commonly observed anatomical abnormalities causing laryngomalacia were short aryepiglottic folds (10/10 patients), prolapsing arytenoid mucosa (9/10) and prolapsing or omega-shaped epiglottis (4/10) ( Table 1 ). 3.4.2 Procedure performed In all cases the surgical procedure performed for each patient was dictated by the findings at the time of surgery. For example: patient 4 had short aryepiglottic folds, prolapsing arytenoid mucosa and a prolapsing epiglottis and therefore underwent, bilateral aryepiglottic fold division, bilateral arytenoid mucosal trimming and trimming of the epiglottis. One patient (1/10) had division of both aryepiglottic folds alone, while five patients (5/10) underwent division of both aryepiglottic folds with bilateral trimming of their arytenoid mucosa. Four patients (4/10) had bilateral division of aryepiglottic folds, bilateral trimming of arytenoid mucosa and epiglottic trimming or epiglottopexy ( Table 1 ). 3.5 Polysomnography Polysomnographic data obtained before and after supraglottoplasty was compared and the statistical difference between the results was calculated using a paired Student's t -test. Improvements occurred in all respiratory parameter mean values following surgery. Changes in Total Sleep Time (TST), Lowest Oxygen Saturation Levels (SpO2 nadir) ( Fig. 1 ), Obstructive Apnoea Hypopnoea Index (OAHI) ( Fig. 2 ) and Respiratory Disturbance Index (RDI) ( Fig. 3 ) were all deemed statistically significant ( P < 0.05), while an improvement in the mean Highest Carbon Dioxide Measurement occurred (TcCO2) (57.1 vs. 52.8) which was not statistically significant ( Table 2 ). 3.6 Complications One patient was intubated post-operatively for a period of 24 h and developed a partial left lower lobe collapse due to inadvertent movement of the endotracheal tube into the right main bronchus. 4 Discussion Our study demonstrates the efficacy of supraglottoplasty in reversing abnormal respiratory parameters in laryngomalacia detected on polysomnography. Statistically significant improvements in mean values for Total Sleep Time, Lowest Oxygen Saturation Level, Obstructive Apnoea Hypopnoea Index and Respiratory Disturbance Index occurred following supraglottoplasty. An improvement in mean Highest Carbon Dioxide levels occurred, which did not achieve statistical significance. Three previous studies have assessed supraglottoplasty outcomes using pre- and post-operative polysomnography with equivalent or smaller sample sizes. A 1990 study by Marcus et al. assessed the outcomes of supraglottoplasty on 6 patients using pre- and post-operative daytime nap polysomnography. They observed improvements in oxygen saturation levels and expired carbon dioxide levels for all patients following surgery. The polysomnographic data did not assess Obstructive Apnoea Hypopnoea Index (OAHI) or Respiratory Disturbance Index (RDI) [5] . In 2006 Valera et al. studied 7 patients with severe laryngomalacia; 2 patients underwent surgical tracheostomy for pharyngolaryngomalacia; with the remaining 5 patients undergoing supraglottoplasty with pre- and post-operative polysomnography. Statistically significant improvements occurred in Respiratory Disturbance Index (RDI) in all 5 patients however a similar improvement in lowest Oxygen Saturation levels failed to occur [6] . A 2008 study by Zafereo et al. obtained statistically significant improvements in mean SpO2 nadir, Obstructive Apnoea Hypopnoea Index (OAHI), Respiratory Disturbance Index (RDI) for all 10 patients undergoing supraglottoplasty for laryngomalacia [7] . Similarly to our study no statistically significant improvement was observed in mean Peak End-Tidal Carbon Dioxide levels following surgery. In contrast to our study, the authors performed unilateral trimming of arytenoid mucosa with bilateral division of aryepiglottic folds, where we perform bilateral arytenoid mucosal trimming with bilateral division of aryepiglottic folds. Traditionally supraglottoplasty outcomes are assessed primarily through subjective assessment. Following surgery typically a reduction in stridor occurs, along with improved feeding and a resumption of weight-gain. Few methods exist however to objectively assess the efficacy of supraglottoplasty in the treatment of laryngomalacia. Polysomnography is a reliable and accurate method for collecting data for the assessment of any upper airway abnormality. Indeed at our institution, polysomnography is widely used in assessing the degree of paediatric upper airway obstruction, not only in laryngomalacia but also in many other paediatric airway disorders. This practice has been well documented by previous authors [8] . We realise however that the routine use of paediatric polysomnography may not be a realistic expectation in all institutions. Criticisms of polysomnography include cost, requirement for an overnight hospital stay and questionable night-to-night variability in polysomnographic data. A 2002 study by Katz et al. dispels the criticism of night-to-night variability of polysomnography, with a single night's shown to provide sufficient evidence to detect Obstructive Sleep Apnoea [9] . Larygomalacia is known to occur due to prolapse of the supraglottic laryngeal structures towards the glottic lumen during inspiration, leading typically to inspiratory stridor and in more severe cases airway obstruction, feeding difficulties and failure to thrive. Supraglottoplasty has been widely documented as being successful in modifying the dynamics of the supraglottis in laryngomalacia, resulting in a reduction in stridor, airway obstruction and improved feeding [10,11] . Many authors have proposed modifications of supraglottoplasty with differing combinations of the following being practised: unilateral/bilateral division of aryepiglottic folds, unilateral/bilateral excision of redundant arytenoid mucosa and unilateral or bilateral epiglottic trimming and epiglottopexy. The procedure may be performed with cold instrumentation, Carbon Dioxide Laser, using microdebrider assisted techniques or using combination techniques. It is our policy to perform surgery directed at the dynamic abnormalities observed at initial diagnostic fibreoptic nasendoscopy and subsequently confirmed during spontaneous ventilation at microlaryngoscopy. Furthermore, it is our practice to perform bilateral incision of the aryepiglottic folds with bilateral excision of redundant arytenoid mucosa where indicated. We are mindful equally of the concerns regarding supraglottic stenosis where bilateral arytenoid mucosal trimming occurs, however to-date we have had no cases of supraglottic stenosis in our series of 46 patients surgically treated for laryngomalacia. In our current study, we observed a mean age at initial presentation to our department with laryngomalacia, of 2 months and 19 days (2.6 months) (range 30–134 days). When compared with previously published series where surgical intervention was required for laryngomalacia, our patient group were significantly younger than previous studies have reported. Other authors have recorded mean ages at presentation of wide variation: from 38.3 months to 27 months in one study and 10.3 months in another series [12,5] . More recently however authors have reported younger mean ages at presentation of: 6.8 months and 4 months, respectively [6,7] . Our figures however would appear to indicate that we performed supraglottoplasty at a younger mean age than all previously published series. Perhaps increasing primary care and general paediatric awareness of laryngomalacia, its’ disease-related morbidity and treatment options has led progressively to earlier detection of cases. We acknowledge however, due to the nature of our study our small sample size may limit this observation when compared with larger series of patients. Previous studies have documented poor outcomes in treating obstructive sleep apnoea in children with co-morbidities associated with generalised hypotonia; these include Downs’ Syndrome and Cerebral Palsy [13,14] . These observations have been primarily in children undergoing adenotonsillectomy for obstructive sleep apnoea syndrome. Fraga et al. observed a poorer outcome of supraglottoplasty for laryngomalacia in a group of 8 patients; 4 of which had underlying neurological impairment [15] . Two patients in our study group (Patients 4 and 6) had a co-morbidity associated with hypotonia; both patients had Downs’ Syndrome. In both cases we observed significant improvements in SpO2 nadir, Obstructive Apnoea Hypopnoea Index and Respiratory Disturbance Index in following supraglottoplasty. Limitations of the current study include a retrospective method of data collection. While the overall sample size was small, when taken in the context of similar previous studies it has an equivalent sample size to the largest previously published study on this topic [7] . Furthermore despite the small sample size statistical significance was achieved for all but one of the observed respiratory parameters (TcCO2). Limiting factors to increasing study numbers include the availability of both pre- and post-operative polysomnography and an observed parent reluctance to undergo post-operative polysomnography, where obvious subjective clinical improvement has occurred following supraglottoplasty. A potential selection bias exists in our method of patient selection, with patients with obvious recurrent cyanosis and desaturations not being included in the study group to prevent potential delays in their treatment. Our study group therefore may underestimate the severity of laryngomalacia which may occur in severe cases and therefore our results may indeed underestimate the efficacy of supraglottoplasty in the treatment of laryngomalacia. There may also be subtle differences in surgical technique as the surgery was performed by several surgeons over the study period; however in all cases the senior author either performed the procedure or directly supervised the procedure. Finally, it could be argued that the improvements which occurred in the polysomnographic data in our patients were the result of the natural spontaneous resolution which occurs in laryngomalacia. Studies quote the mean age at which spontaneous resolution occurs in laryngomalacia as 9 months of age [2] . In our study at a mean age of 5 months and 18 days, post-operative polysomnography confirmed the reversal of previously abnormal pre-operative respiratory parameters. This strongly suggests a causal relationship between the resolution of laryngomalacia in our study group and surgical intervention. However to conclusively evaluate whether the resolution of polysomnography findings occurred as a result of surgical intervention, a randomised controlled trial would be necessary. This would be unethical as some children may be potentially exposed to untreated hypoxemia and failure to thrive. 5 Conclusion Polysomnography is an effective method for objectively assessing the efficacy of supraglottoplasty for laryngomalacia. Supraglottoplasty effectively reverses the abnormal respiratory parameters occurring in moderate to severe laryngomalacia. Conflict of interest None. Acknowledgements The authors wish to acknowledge Dr. Andrew Wilson and Dr. Dave Mullane of the Paediatric Respiratory Department at Princess Margaret Hospital for Children, for their assistance in the timely processing of patients requiring polysomnographic assessment during our study. References [1] G.F. Tucker Laryngeal development and congenital lesions Ann. Otol. Rhinol. Laryngol. Suppl. 89 1980 142 145 [2] D.R. Olney J.H. Greinwald R.J. Smith N.M. Bauman Laryngomalacia and its treatment Laryngoscope 109 1999 1770 1775 [3] T. Gislason B. Benediktsdottir Snoring apneic episodes and nocturnal hypoxemia among children 6 months to 6 years old. An epidemiologic study of lower limit of prevalence Chest 107 1995 963 966 [4] A.C. Engelberts The role of obstructive sleep apnea in sudden infant death syndrome and apparent life threatening event Int. J. Pediatr. Otorhinolaryngol. 32 1995 59 62 [5] C.L. Marcus D.M. Crockett S.L. Ward Evaluation of epiglottoplasty for as treatment for severe laryngomalacia J. Pediatr. 117 1990 706 710 [6] F.C. Valera E. Tamashiro M.M. de Araujo H.H. Sander D.S. Kupper Evaluation of the efficacy of supraglottoplasty in obstructive sleep apnea syndrome associated with severe laryngomalacia Arch. Otolaryngol. Head Neck Surg. 132 2006 489 493 [7] M.E. Zafereo R.J. Taylor K.D. Pereira Supraglottoplasty for laryngomalacia with obstructive sleep apnea Laryngoscope 118 2008 1873 1877 [8] M.T. Goncalves J. Sato M.A. Avelino G.U. Pizarro G.A. Moreira M.P. Hallinan Polisomnographic findings on children with laryngopathies Rev. Braz. Otorhinolaryngol. 72 2006 187 192 [9] E.S. Katz M.G. Greene K.A. Carson P. Galster G.M. Loughlin J. Carroll Night-to-night variability of polysomnography in children with suspected obstructive sleep apnea J. Pediatr. 140 2002 589 594 [10] G. Roger F. Denoyelle J.M. Triglia E.N. Garabedian Severe laryngomalacia: surgical indications and results in 115 patients Laryngoscope 105 1995 1111 1117 [11] N.B. Solomons C.A. Prescott Laryngomalacia. A review and the surgical management for severe cases Int. J. Pediatr. Otorhinolaryngol. 13 1987 31 39 [12] D.K. Reddy B.H. Matt Unilateral vs. bilateral supraglottoplasty for severe laryngomalacia in children Arch. Otolaryngol. Head Neck Surg. 127 2001 694 699 [13] M. Strome Obstructive sleep apnea in Down syndrome children: a surgical approach Laryngoscope 96 1986 1340 1342 [14] E. Hultcrantz H. Svanholm Down syndrome and sleep apnea—A therapeutic challenge Int. J. Pediatr. Otorhinolaryngol. 21 1991 263 268 [15] J.C. Fraga L. Schopf V. Volker S. Canani Endoscopic supraglottoplasty in children with severe laryngomalacia with and without neurological impairment J. Pediatr. (Rio J) 77 2001 420 424