Different patterns of the survival probabilities of the time to revision by laryngotracheal cleft type emerged across time. Circles indicate censored.
Alessandro de Alarcón, Alexander J. Osborn, Meredith E. Tabangin, Aliza P. Cohen, Catherine K. Hart, Robin T. Cotton, Michael J. Rutter. Laryngotracheal Cleft Repair in Children With Complex Airway Anomalies. JAMA Otolaryngol Head Neck Surg. 2015;141(9):828–833. doi:10.1001/jamaoto.2015.1419
This study provides clinicians with relevant information regarding the surgical outcomes of patients with laryngotracheal cleft in the context of additional airway anomalies.
To determine the rates of surgical success in patients who underwent laryngotracheal cleft repair in the context of additional airway anomalies, determine the revision rate for cleft repair, characterize the time to revision among patients who underwent cleft repair, and assess the functional swallowing outcomes after cleft repair.
Design, Setting, and Participants
A retrospective study was conducted at a quaternary pediatric center of 81 children diagnosed as having laryngotracheal cleft with or without concomitant airway anomalies who underwent laryngotracheal cleft repair between February 1, 2000, and February 28, 2013. Analysis was conducted from October 1, 2012, through March 30, 2013.
Surgical repair of laryngotracheal cleft.
Main Outcomes and Measures
Surgery-specific success, overall surgical success, and revision rate. We defined surgery-specific success as a repair not requiring revision and overall surgical success as absence of a cleft or TEF at the last operative examination, regardless of the number of revisions required.
Of 81 patients with laryngotracheal cleft who underwent surgical repair, 48 (59%) had at least 1 additional airway finding; 24 (30%) had tracheomalacia and 21 (26%) had subglottic stenosis. Seventeen patients required a revision of laryngotracheal cleft repair, with a median time to revision of 8.3 months (interquartile range, 4.3-25.1 months). Six patients required a second revision, with a median interval between revisions of 44.2 months (interquartile range, 28.6-53.6 months). The surgery-specific success rate was 77% (37 of 48) in patients with additional airway anomalies and 82% (27 of 33) in those with laryngotracheal clefts alone. The overall surgical success rate was 92% (44 of 48) in patients with additional airway anomalies and 97% (32 of 33) in those with clefts alone. Fourteen (17%) patients demonstrated swallowing dysfunction postoperatively despite closure of the cleft.
Conclusions and Relevance
Although additional airway findings were common in our cohort of patients with laryngotracheal clefts, these anomalies did not affect surgery-specific or overall surgical success. The overall surgical success rate for those with and without additional airway anomalies was 92% and 97%, respectively. In view of the fact that cleft breakdown after surgical repair is not uncommon and may occur years after the initial repair, we strongly advocate long-term follow-up. Despite successful closure, a significant percentage of children with laryngotracheal cleft will have persistent swallowing dysfunction.
Laryngotracheal clefts (LTCs) (also called laryngotracheoesophageal clefts) comprise a spectrum of rare congenital anomalies of the upper aerodigestive tract. First described by Richter in the late 1700s, LTCs are a persistent connection between the posterior laryngotracheal airway and the esophagus.1 Symptoms vary greatly. Many patients present with mild or nonspecific symptoms that may mimic other conditions,2 whereas others present with recurrent pneumonia or symptoms of aspiration.3,4
Most children with an LTC have other congenital anomalies or syndromes, many of which compromise the airway (eg, tracheoesophageal fistula [TEF] and Opitz, Pallister-Hall, and VATER [vertebral anomalies, anal atresia, cardiac defects, tracheoesophageal fistula and/or esophageal atresia, renal and radial anomalies, and limb defects] syndromes).5- 10 Although these associations have been reported, to our knowledge, the effect of additional airway anomalies on LTC repair and the outcomes of LTC repair in patients with these complex conditions have not been amply explored.
The purpose of the present study was to report our surgical experience with a large cohort of children who underwent LTC repair. Our primary objectives were to determine the rates of surgical success in the context of additional airway anomalies, determine the revision rate for LTC repair, characterize the time to revision among patients who underwent LTC repair, and assess the functional swallowing outcomes after LTC repair.
We searched our clinical airway database to identify all patients diagnosed with LTC who underwent LTC repair from February 1, 2000, to February 28, 2013. Analysis was conducted from October 1, 2012, through March 30, 2013. We reviewed the medical records to obtain demographic data and information pertaining to comorbidities, cleft type, surgical management, revision surgery, complications, and swallowing outcomes. Patients who underwent LTC repair but had no follow-up data were excluded from the analysis. Laryngotracheal clefts were classified according to the grading system proposed by Benjamin and Inglis.3 Type 1 LTCs and deep interarytenoid notches were grouped together for analysis.10 We defined airway finding as an abnormality or significant clinical finding noted on endoscopic evaluation of the laryngotracheal airway. We defined surgery-specific success as a repair not requiring revision and overall surgical success as absence of a cleft or TEF at the last operative examination, regardless of the number of revisions required.
Functional swallowing outcomes were determined from a video swallow study and/or fiberoptic endoscopic evaluation of swallowing. Study data were collected and managed using REDCap (Research Electronic Data Capture) electronic data capture tools hosted at Cincinnati Children’s Hospital Medical Center.11 Patient data were deidentified as part of the study. We received approval for this study from the Cincinnati Children’s Hospital Medical Center Institutional Review Board.
Descriptive statistics, including medians and interquartile ranges (IQRs), are presented for nonnormally distributed continuous data. Categorical data are presented as percentages. The χ2 or Fisher exact test was used to test for group differences in proportions of patient characteristics, symptoms, and surgical success.
We characterized the time to revision among those with and without additional airway anomalies, as well as the time to revision by cleft grade using Kaplan-Meier survival curves. In survival analysis, censoring accounts for characteristics about the time to an event that may not otherwise be addressed with conventional parametric statistical methods. Each individual contributes information to the time to the end point (revision) or end of follow-up, and the resulting curve displays this survival probability against time. Since there were no revisions among type 4 LTCs, tests comparing probabilities of the time to revision by cleft grade were limited to cleft types: deep arytenoid notch and types 1, 2, and 3. The log-rank test was used to statistically compare survival curves among groups. P < .05 was considered statistically significant. We performed all analyses using SAS, version 9.3 (SAS Institute).
Of the 110 patients diagnosed with LTC during our study period, 88 met our inclusion criteria. Seven patients without follow-up data were excluded from the analysis. Forty-eight of these 81 patients had additional airway anomalies. Characteristics of the 2 groups are presented in Table 1.
With the exception of age at LTC repair, the presence of a prior TEF, tracheotomy placement, and duration of follow-up, characteristics of patients with and without concomitant airway anomalies were not significantly different (Table 1). Children with no additional airway anomalies had their LTC repaired at an earlier age (mean [SD], 2.6 [3.2] vs 4.6 [4.4] years; P = .03). Also, fewer children with no additional airway anomalies had TEF repair (3 [9%] vs 18 [38%]; P = .004) and underwent tracheotomy placement (11 [33%] vs 28 [58%]; P = .03). In addition, duration of follow-up among children with no additional airway anomalies was significantly shorter (median [IQR], 6.7 [2.3-16.6] vs 29.9 [12.3-50.7] months; P < .001). The distribution of cleft type among children with and without airway anomalies was somewhat different. Types 3 and 4 LTCs were less commonly associated with concomitant airway anomalies (type 3: 10 patients [21%] vs 14 [42%]; type 4: 0 vs 2 [6%]; P = .04).
As shown in Table 1, symptoms among the 2 groups were not significantly different. Nonetheless, frequent upper respiratory tract infections were more common in patients with additional airway anomalies (7 [15%] vs 0 [0%]).
Additional airway anomalies in our cohort are presented in Table 2. Tracheomalacia and subglottic stenosis (SGS) were the 2 most common anomalies (24 patients [30%] and 21 patients [26%], respectively). There was a significant association between cleft type and the presence of SGS (P = .006), with the proportion of patients with SGS decreasing with severity in patients with cleft anomalies. More specifically, 12 of 28 (42.9%), 8 of 27 (29.6%), and 1 of 24 (4.2%) patients with types 1, 2, and 3 LTCs, respectively, had SGS. The presence of the other individual concomitant airway anomalies was less than 10%.
Surgical management of all patients was based on cleft type; similar management approaches were used by the 3 otolaryngologists who performed surgical repairs (A.D.A., M.J.R., and R.T.C.). Of 28 patients with deep notch and type 1 LTCs, 19 (68%) were managed endoscopically, 5 (18%) underwent open surgery, 3 (11%) underwent open surgery with a graft, and 1 (4%) received laser surgery with no suturing. Of 27 patients with type 2 LTCs, 4 (15%) underwent endoscopic surgery, 14 (52%) underwent open surgery, and 9 (33%) underwent open surgery with a graft. Of 24 patients with type 3 LTCs, 17 (71%) underwent open surgery and the remaining 7 (29%) underwent open surgery with a graft. The 2 patients with type 4 LTCs underwent open surgery. Endoscopic LTC repair was more frequently performed during the last several years of the study period and more often performed in patients with type 1 or type 2 LTC.12 In addition, we did not encounter patients with type 3 LTCs who we thought were appropriate candidates for endoscopic repair (eg, the cleft was difficult to expose endoscopically or extended beyond the second tracheal ring).13,14 Open cleft repair was frequently performed with a complete laryngofissure and a 2-layered closure of the cleft with interrupted absorbable sutures.15 In 19 patients, an interposition graft was placed that consisted of periosteum (sternal or tibial) or a rib graft to reconstitute a deficient posterior cricoid plate.16
Surgery-specific success and overall surgical success are described in Table 3. Surgery-specific success was 77% (37 of 48) for those with and 82% (27 of 33) for those without additional airway anomalies. Overall success was 92% (44 of 48) for those with and 97% (32 of 33) for those without additional airway anomalies. There was no statistically significant difference in either success-related measure between those with and without additional airway anomalies.
The overall complication rate in our cohort was 15% (12 of 81 patients). Nine of 48 patients (19%) with additional airway anomalies had complications, whereas 3 of 33 (9%) with no additional findings had complications (P = .34). Specifically, 2 patients experienced failure of initial extubation attempts; 2 required supportive oxygen therapy for frequent desaturations; 1 experienced postoperative fever, requiring chest percussion therapy; 1 developed an infrastomal abscess; 1 experienced persistent emesis, requiring extended hospitalization; 1 had a persistent TEF within the cleft, requiring prolonged hospitalization; 1 developed pneumonia; 1 had a pneumothorax; 1 developed postoperative aspiration pneumonia; and 1 required tracheotomy for difficult extubation.
The proportions of patients undergoing initial revision surgery by cleft type are described in Table 4. Overall, 17 of 81 (21%) patients underwent revision surgery, with a median time to revision of 8.3 months (IQR, 4.3-25.1 months). An additional 6 patients (7%) required a second revision surgery, with a median interval between revisions of 44.2 months (IQR, 28.6-53.6 months). In some patients, revisions occurred more than 3 years after the initial repair. For example, the median time to revision after initial surgery was 44.2 months (IQR, 25.1-50.2 months) for patients with type 3 LTC. Similarly, the interval between the first and second revision for patients with type 3 LTC was 50.7 months (IQR, 37.7-107.1 months). There was no association between having an additional airway anomaly and the need for revision surgery.
Survival probabilities of the time to revision were similar between patients with or without an additional airway anomaly (P = .47), but different patterns by cleft type emerged across time (P = .06) (Figure). Type 2 LTCs were repaired earlier, although the probability of not needing revision was greater than 60% at 12 months. For type 3 LTCs, the probability of not needing revision at 24 months was greater than 84% and was 70% at 44 months.
Of the 81 patients in our cohort, 59 had swallowing evaluations available for review. Of these patients, 14 (24%) demonstrated aspiration. Swallowing recommendations were based on the following findings from the video swallow study or fiberoptic endoscopic evaluation of swallowing: 6 patients (10%) continued to receive no food or drink by mouth, 11 (19%) required thickened fluids, 11 (19%) required minor modifications of all food and drink consistencies, and 31 (53%) required no restrictions.
In view of previously reported findings,5- 9,17 it was not surprising that 48 patients (59%) had additional airway findings. Contrary to our expectations, however, our results demonstrated that regardless of cleft type, additional airway anomalies did not have a significant effect on primary repair or overall repair. We suspect that both the rigor of our clinical practice with regard to preoperative assessment and optimization and the relative consistency in surgical management within our clinical practice have resulted in improved surgery-specific and overall surgical outcomes. These findings do not reflect whether children who were severely ill required other surgical procedures or tracheotomy placement, as these issues were not the focus of the study. Although the technical aspects of surgery should not be underestimated, we believe that achieving optimal outcomes requires patients to be in an optimal state at the time of surgery. To this end, we perform a thorough assessment of the patient’s conditions that are likely to contribute to a higher failure rate. Specifically, the possible presence of gastroesophageal reflux disease, eosinophilic esophagitis, active larynx, and pseudomonal or methicillin-resistant Staphylococcus aureus airway colonization is explored and addressed by an interdisciplinary team of pediatric subspecialists before undertaking LTC repair.
An explanation of our unanticipated surgical outcomes must also consider that our 2 most common additional airway anomalies were tracheomalacia and SGS. For patients with LTC and concomitant SGS, we typically address both concurrently. Staging may be preferable in these patients and in patients with LTC and severe concomitant tracheomalacia. Patients with SGS or tracheomalacia are optimized before LTC repair, with optimization including an individualized pulmonary plan for postoperative care. Our findings demonstrate the efficacy of this approach.
Seventeen patients (21%) required revision surgery. Nonetheless, all clefts were amenable to revision surgery, with a high degree of success. The survival curves by cleft type depict the occurrence of late surgical failure, particularly among type 3 LTCs (Figure). Across all types, the probability of survival without revision remained constant after 50 months, suggesting that these cleft types may require a minimum of 50 months of follow-up to assess breakdown.
A strength of this study is the long duration of follow-up. Children who have undergone LTC repair routinely undergo annual surveillance endoscopy subsequent to successful initial repair. This management approach has enabled us to identify patients with late cleft breakdown. This is particularly important, as breakdown detection prevents the long-term possibility of aspiration and chronic lung damage.
Although the primary focus of the current study was on the relationship between airway findings and LTC repair, we also examined swallowing outcomes; this information is pertinent in that swallowing difficulties may persist despite successful cleft repair. Because we are a quaternary pediatric care center with a far-stretching referral pattern, not all patients had follow-up swallowing studies performed at our institution and available for review. Given that we did not have swallowing evaluations on 22 patients, we may have overestimated or underestimated the prevalence of swallowing dysfunction. That being said, we found that a large proportion of patients had swallowing difficulties despite a successful cleft repair (6 patients [10%] continued to receive no food or drink by mouth and 11 [19%] required a thickened liquid diet). Swallowing outcomes were examined in an earlier study at our institution.18
Our study has the design limitations inherent in all retrospective studies, which includes missing information. The later age of repair in our cohort may be at least partially attributed to the possibility that LTC was masked by other airway anomalies in patients referred from outside institutions. As an example, a child with a severe SGS and an LTC may have few symptoms from the LTC; therefore, the LTC may have been missed in the outside evaluation and not documented until the patient underwent a complete airway evaluation at our institution. Alternatively, there could be communication gaps between caregivers, referring physicians, and our institution that resulted in incomplete documentation. In addition to this limitation, the high rates of associated airway anomalies in our cohort could theoretically reflect the referral bias (ie, more complex patients) typically associated with a quaternary center. Nevertheless, as mentioned above, our findings regarding additional airway anomalies are consistent with those in the published literature.
Although additional airway anomalies were common in our cohort of patients with LTC, these anomalies did not affect surgery-specific or overall surgical success. The overall surgical success rate was 92% (44 of 48) in patients with additional airway anomalies and 97% (32 of 33) in those without additional airway anomalies. In view of the fact that cleft breakdown after surgical repair is not uncommon and may occur years after the initial repair, we strongly advocate long-term follow-up. Despite successful closure of the cleft, a significant percentage of children will have persistent swallowing dysfunction.
Accepted for Publication: June 6, 2015.
Corresponding Author: Alessandro de Alarcón, MD, MPH, Division of Pediatric Otolaryngology–Head and Neck Surgery, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave, Mail Location C-2018, Cincinnati, OH 45229 (email@example.com).
Published Online: August 6, 2015. doi:10.1001/jamaoto.2015.1419.
Author Contributions: Dr de Alarcón had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: de Alarcón, Tabangin, Hart, Cotton, Rutter.
Acquisition, analysis, or interpretation of data: de Alarcón, Osborn, Tabangin, Cohen, Hart.
Drafting of the manuscript: de Alarcón, Tabangin.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: de Alarcón, Osborn, Tabangin.
Administrative, technical, or material support: de Alarcón, Rutter.
Study supervision: de Alarcón, Rutter.
Conflict of Interest Disclosures: None reported.
Previous Presentation: This study was presented at the American Society of Pediatric Otolaryngology Annual Meeting; April 26, 2013; Washington, DC.