eTable. Patients Undergoing Open Airway Operation
Lennon CJ, Gelbard A, Bartow C, Garrett CG, Netterville JL, Wootten CT. Dysphagia Following Airway Reconstruction in Adults. JAMA Otolaryngol Head Neck Surg. 2016;142(1):20-24. doi:10.1001/jamaoto.2015.2562
Patients who undergo open airway reconstruction procedures are likely to experience some degree of postoperative dysphagia symptoms and delayed return to oral intake.
To review the duration of postoperative dysphagia symptoms and outcomes in a group of adult patients.
Design, Setting, and Participants
Retrospective review of the medical records of adult patients undergoing laryngotracheoplasty, posterior cricoid split laryngoplasty, tracheal resection, and cricotracheal resection in a tertiary hospital between July 2009 and September 2014.
Laryngotracheoplasty, posterior cricoid split laryngoplasty, tracheal resection, and cricotracheal resection.
Main Outcomes and Measures
Demographic characteristics, etiology of airway stenosis, surgical procedure, stent type, and duration of dysphagia symptoms.
Thirty-eight patients (14 men, 24 women; mean [SD; range] age, 48 [14.4; 20-80] years) fitting the inclusion criteria were identified. Twenty-four (63%) patients had laryngotracheal stenosis secondary to prolonged intubation, with 3 (8%), 5 (13%), and 6 (16%) cases being due to autoimmune, idiopathic, or other etiology, respectively. Twenty-five (66%) patients underwent tracheal or cricotracheal resection, and 13 (34%) underwent laryngotracheoplasty or posterior cricoid split laryngoplasty. Of the 17 patients with stents placed, 6 (35%) patients had a suprastomal stent sewn at the top with a polypropylene suture using a horizontal mattress technique, 6 (35%) patients had a suprastomal stent capped with an extended Silastic thoracic T-tube segment, and 5 (29%) patients had either a T-tube or hood bronchial stent. Eight of 17 patients used a nasogastric feeding tube while the stent was in place (up to 5 weeks). All patients returned to their preoperative diet. The mean (SD) duration of dysphagia symptoms in all patients (both those without a stent and following stent removal) was 8 (27.2) days (median, 1.5 days). The mean (SD) duration of dysphagia symptoms in patients who did not have a stent placed was 4.8 (5.3) days (median, 4 days).
Conclusions and Relevance
In this study of adults who underwent open airway reconstruction, all returned to their preoperative diet, but those without stents had a shorter duration of dysphagia symptoms than those with stents. Approximately half as many patients with a stent had a prolonged course with dysphagia symptoms compared with those without a stent.
Laryngotracheal stenosis (LTS) is a common end point to multiple pathophysiologic processes. In modern practice, the most common cause of LTS in adults is tracheal intubation, followed by tracheostomy and laryngeal trauma. Additionally, airway compromise can occur as a result of collagen vascular disorders, idiopathic fibroinflammatory disease, disorders of epithelial barrier function, physical and thermal trauma, and extrinsic compression or invasion by benign and malignant disease.1- 6
Operative interventions to correct LTS range from endoscopic procedures (scar incision or excision, endoluminal dilation) to major open airway reconstruction.7 If the degree of stenosis warrants major airway reconstruction, expansion framework procedures or resection procedures may be necessary. Expansion framework procedures involve the use of an anterior and/or posterior graft(s) (from the thyroid, conchal, or rib cartilages) placed through either the cricoid or tracheal cartilage, categorized as laryngotracheoplasty or posterior cricoid split laryngoplasty, respectively. Resection procedures are used most commonly for severe stenosis, ideally at least 3 mm below the vocal folds.8 These procedures involve removal of the stenotic segments with reapproximation of the healthy tracheal segment to either another tracheal segment or the posterior cricoid remnant and thyroid lamina.
The primary aims of open airway reconstruction are enlargement of the airway to support adequate ventilation and/or allow for decannulation. This is accomplished through physical manipulation of the cartilaginous laryngotracheal complex. Swallowing dysfunction with a frequent, albeit temporary, increase in aspiration risk has been established as a complication of open airway reconstruction in the pediatric population.9,10 The transient disruption of normal cartilaginous support, the mobilization of suprahyoid musculature, and the temporary placement of endoluminal airway stents following reconstructive procedures have been hypothesized to affect normal muscular and sensory mechanisms of deglutition in adults. Furthermore, placement of posterior grafts may result in destabilization of the cricoarytenoid joints, resulting in difficult maintenance of airway closure during swallowing.11
At this time, both the degree and duration of dysphagia symptoms following open airway reconstruction in adults are unclear. One study by Herrington et al12 examined toleration of regular diet as well as placement of gastrostomy tube following major airway reconstruction; however, this study did not look at mean time to resolution of dysphagia symptoms. The purpose of the present study was to report the mean amount of time that adults experience dysphagia symptoms following open airway reconstruction, as well as to identify factors associated with dysphagia symptoms following open airway reconstruction in the adult population.
Following approval from the Vanderbilt University Medical Center Institutional Review Board, records were extracted by querying the Department of Otolaryngology–Head and Neck Surgery procedure database for patients older than 18 years between September 2009 and September 2014. Consent was waived due to the retrospective nature of the study. Records were selected on the basis of surgical coding for rib cartilage graft, laryngoplasty (thyroplasty), laryngoplasty (cricoid split), tracheoplasty (cervical), and excision tracheal stenosis and anastomosis (cervical).
A medical record review was conducted and surgical procedures were categorized as tracheal resection, cricotracheal resection, laryngotracheoplasty, or posterior cricoid split laryngoplasty (whether performed in an open or endoscopic fashion). Tracheal or cricotracheal resections were defined as resection of stenosis involving either the trachea, cricoid cartilage, or both with subsequent reanastomosis. Laryngotracheoplasty involved either an anterior or both anterior and posterior grafts, whereas posterior cricoid split laryngoplasty used only a posterior graft for expansion. Procedures were either single stage, involving reconstruction with intraoperative closure or resection of the tracheostoma; or multistage, in which various stenting options were used to support the laryngeal repair site while a tracheostomy remained in place. The most commonly used stenting option involved a suprastomal stent fashioned from a Silastic T-tube either sewn at the top with a polypropylene suture using a horizontal mattress technique or capped. These closure techniques were used to minimize aspiration while the stent was in place.
Demographic characteristics, comorbidities, primary underlying etiology of LTS, preoperative diet, surgical procedure, presence of stent, stent type, and duration of dysphagia symptoms following surgery were reviewed and recorded. Dysphagia symptoms were defined as aspiration on swallow study, coughing on bedside swallow evaluation, and/or the sensation of difficulty swallowing reported by the patient. Postoperative swallow study reports were reviewed, when available. Video fluoroscopic swallow studies predominated, although bedside swallow evaluations or fiberoptic endoscopic evaluation of swallow were also performed when clinically indicated. The dates of clinical resolution of dysphagia symptoms, or the length of time to return to a preoperative diet, were recorded from the electronic medical record. When a stent was placed, success or failure with swallowing while a stent was in place was recorded. Additionally, duration of dysphagia symptoms was recorded from the time of stent removal.
The data were entered into the REDCap database available through Vanderbilt University. Patients who underwent tracheal and cricotracheal resection were compared with those who underwent laryngotracheoplasty and posterior cricoid split laryngoplasty based on the similarities between these respective procedures. Statistical analyses were performed to examine the relationship between type of laryngotracheal reconstruction and duration of dysphagia symptoms experienced. We also compared outcomes between patients who had a stent placed vs those who had none. Demographic data for the 2 groups were summarized. Differences in continuous variables were tested using the t test. Differences in categorical variables (sex) were tested using the Fisher exact test. Statistical significance was declared at α = .05. Statistical analyses were conducted using GraphPad software.
Thirty-eight patients underwent tracheal or cricotracheal resection, laryngotracheoplasty, or posterior cricoid split laryngoplasty between September 2009 and September 2014. The ages of participants in the study ranged from 20 to 80 years (Table 1). The mean (SD) age was 48 (14.4) years. The population included 14 men and 24 women.
Twenty-four (63%) patients had LTS secondary to prolonged intubation, with 3 (8%), 5 (13%), and 6 (16%) of cases being due to autoimmune, idiopathic, or other etiology, respectively (Table 1). Twenty-five (66%) patients underwent resections, with 5 (13%) undergoing cricotracheal resections and 20 (53%) undergoing tracheal resection. Thirteen (34%) patients underwent luminal expansion procedures with 6 (16%) undergoing laryngotracheoplasty and 7 (18%) undergoing posterior cricoid split laryngoplasty. Seventeen (45%) patients had a stent placed during their surgery (Table 2). Of these patients, 12 (71%) had concomitant graft placement as well. Of those with stents placed, 6 (35%) patients had a suprastomal stent sewn at the top with a polypropylene suture using a horizontal mattress technique, 6 (35%) patients had a suprastomal stent capped with an extended Silastic thoracic T-tube segment, and 5 (29%) patients had either a T-tube or hood bronchial stent. Participants’ surgical characteristics are summarized in the eTable in the Supplement.
There was no significant difference in the mean (SD) age of those with or without grafts (52.3 [16.6] vs 46.3 [13.1] years; P = .23). Similarly, there was no significant difference in the mean age of those with or without stents placed (51.5 [16.1] vs 45.8 [12.8] years; P = .23). The sex differences between those with and without grafts and stents were similarly nonsignificant (P > .99 and P = .49, respectively).
All patients returned to their preoperative diet. Only 1 patient had dysphagia with resulting feeding tube dependence recorded prior to surgery, which was the result of a traumatic brain injury. The mean (SD) duration of dysphagia symptoms in all patients (both those without a stent and following stent removal) was 8 (27.2) days (median, 1.5 days) (Figure, A). The mean (SD) duration of dysphagia symptoms in patients who did not have a stent placed was 4.8 (5.3) days (median, 4 days) (Figure, B). The mean (SD) duration of dysphagia symptoms among those with grafts was 2.2 (5.7) days (although a majority had no dysphagia symptoms) whereas the mean (SD) among those without grafts was 10.8 (33.1) days (median, 3 days) (P = .35).
Among patients who had a stent placed, 16 of 17 received a video fluoroscopic swallow study with the stent in place. Six patients were not cleared for oral intake because they demonstrated moderate or moderate to severe dysphagia symptoms with aspiration. Five patients demonstrated either mild or silent aspiration but adequate compensatory strategies were taught and the patient was cleared for oral intake. Five patients demonstrated trace penetration without aspiration and were cleared for oral intake. Eight of 17 patients either required or opted for a feeding tube while the stent was present. Six of these patients had the feeding tube removed at the time of stent removal. The mean (SD) duration of dysphagia symptoms while the stent was in place in those not requiring a feeding tube was 3.7 (1) days (median, 4 days). In those who required a feeding tube, the mean (SD) total duration of dysphagia symptoms, both with the stent in place and after its removal, was 50.8 (53.6) days (median, 35 days). In this same group, the mean (SD) amount of time the stent was in place was 27.2 (9.1) days (median, 25.5 days). In all patients with a stent, the mean (SD) duration of dysphagia symptoms following stent removal was 11.1 (40.7) days (although a majority had no dysphagia symptoms). Of note, 1 patient with a stent (and a feeding tube) had continued dysphagia symptoms for nearly 6 months following stent removal due to postoperative complications including presumed cerebrovascular accident. This is reflected in the standard deviations of the means.
Open airway reconstruction allows decannulation in a majority of patients.13 However, the therapeutic tradeoffs in swallowing function after major open reconstruction have not been defined in adults. The aim of this study was to examine swallowing outcomes among a group of adult patients following major airway reconstructive surgery and provide normative data on the presence, degree, and duration of postoperative dysphagia symptoms.
Overall, the mean duration of dysphagia symptoms among our patients was approximately 8 days. However, the median duration of dysphagia symptoms was 1.5 days. There was a difference between the 2 values due to an outlier who experienced a postoperative cerebrovascular accident. These values were measured from date of stent removal, where applicable. In those patients who did not have a stent placed, the mean duration of dysphagia symptoms was 4.8 days, with a median of 4 days. A similar duration of dysphagia symptoms was seen in those who had a stent placed but did not require a feeding tube in the long term (mean dysphagia symptom duration, 3.7 days; median, 4 days). Importantly, in those who required a feeding tube, the mean (SD) total duration of dysphagia symptoms, both with the stent in place and after its removal, was 50.8 (53.6) days (median, 35 days). Forty-seven percent of patients with a stent required a feeding tube while the stent was in place. Thus, approximately half as many patients with a stent will have a significantly prolonged course with dysphagia symptoms compared with those without a stent—amplified by the duration of stenting.
Open airway reconstructive procedures by definition involve alteration of laryngeal anatomy and have the potential to affect laryngeal closure as well as overall swallowing coordination at the glottic and supraglottic level. It has been hypothesized that expansion procedures might affect laryngeal closure more than resection procedures. Although the limited size of our cohort did not lend itself to direct comparisons between laryngotracheoplasty and posterior split cricoid laryngoplasty, or between tracheal resection and cricotracheal resection, the data did not support a difference in outcome among procedure types.
Although anterior and posterior grafting with in-dwelling stents has been associated with increased duration of postsurgical dysphagia symptoms in children,14 we show that this did not occur in adults. We attribute this difference to the different characteristics of populations of adults and children undergoing airway reconstruction. Many children develop their airway injury in the setting of prematurity and prolonged intubation. Impaired or delayed feeding skills, lingering pulmonary dysfunction, and neurocognitive concerns are important comorbid conditions. Because the operation-specific decannulation rate is significantly lower than the overall decannulation rate across the pediatric population,15 complex revision laryngotracheal operations are sometimes indicated, requiring prolonged stenting. Smith et al10,16 described dysphagia symptoms associated with swallowing during the postoperative period in pediatric patients with stents following airway reconstruction. They noted in this study that dysphagia was frequently a short-term perioperative problem while a stent was in place following open reconstruction. Other studies have noted a residual deficiency in swallowing ability following open reconstruction in children due to ongoing laryngeal structural concerns. Miller et al9 described laryngeal closure patterns present in patients undergoing airway reconstruction and demonstrated that patients who experienced aspiration were more likely to present with structural problems including glottis scar, true vocal fold height mismatch, or prolapsed arytenoids. They also observed that difficulty with coordination seemed to exacerbate patients’ ability to compensate effectively.
There are many potential factors that may facilitate the rapid recovery of swallowing function in adults after open airway reconstruction. In adults the majority of airway injuries occur as a result of intubation (although collagen vascular disorders, idiopathic fibroinflammatory fibrosis, and trauma do play sizeable roles). This population is typically in their mid- to late 50s, with substantial comorbid disease,2 but without a history of dysphagia symptoms. Adults may also be more receptive to instruction on compensatory strategies by speech pathologists. Additionally, adults generally have more coordinated and practiced swallowing function preoperatively, which may facilitate accelerated postoperative compensation. Adults also typically do not require the multiple days of sedation (given their extubation), which is known to potentiate global motor weakness.17
We acknowledge several limitations of the present study. First, the retrospective nature of our investigation allows for variability in the data collected. For example, formal swallow studies were not conducted postoperatively in all patients. The small number of patients included in this study is also an important limitation. Although descriptive statistics were provided, the power of our comparisons is limited. Finally, because not all patients underwent a preoperative swallowing evaluation, there may be unknown underlying swallowing impairments in some patients that were not accounted for.
Future directions include comprehensive prospective swallowing assessments of patients both before and after major airway reconstruction in our multispecialty airway program (AeroVU) within the Vanderbilt Bill Wilkerson Center. Building on the insights generated from this initial work, we hope that future prospective data on dysphagia symptoms in patients with LTS will better elucidate the risk factors for postoperative dysphagia symptoms following open airway reconstruction in adults.
Major airway reconstruction in adults is an effective treatment modality for LTS, often providing definitive treatment for patients. Adults who undergo open airway reconstruction experience recovery of previous swallowing habits. Adults generally adapt well and return to their preoperative diet following these procedures.
Submitted for Publication: May 14, 2015; final revision received September 22, 2015; accepted September 25, 2015.
Corresponding Author: Christopher T. Wootten, MD, Department of Otolaryngology, Vanderbilt University Medical Center, 7209 Medical Center East-South Tower, 1215 21st Ave, South, Nashville, TN 37232-8605 (email@example.com).
Published Online: November 12, 2015. doi:10.1001/jamaoto.2015.2562.
Author Contributions: Dr Lennon had full access to all of 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: Lennon, Gelbard, Garrett, Wootten.
Acquisition, analysis, or interpretation of data: Gelbard, Bartow, Netterville, Wootten.
Drafting of the manuscript: Lennon.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Lennon, Gelbard, Wootten.
Administrative, technical, or material support: Garrett.
Study supervision: Gelbard, Netterville, Wootten.
Conflict of Interest Disclosures: None reported.
Previous Presentation: This work was presented in a poster presentation at the American Laryngological Spring Meeting; April 22, 2015; Boston, Massachusetts.