Surgeons who make airway reconstruction a major part of their practice inevitably are faced with children in whom initial surgical management fails. Searching for the possible causes of failure and determining how best to minimize the risk of repeated failure is often an exhaustive process. Establishing a framework and guidelines for approaching and managing these failures can improve the chances of success significantly. The aim of this article is to provide such a framework.
What are the potential causes of failure?
It is estimated that surgical management will be unsuccessful in 10% to 20% of pediatric patients who undergo airway reconstruction . In some cases, failure occurs despite sound clinical judgment and excellent surgical skill, and there is no identifiable cause. In many cases, however, careful review may suggest the cause for initial failure. A fundamental consideration is the goal of the original surgery and determining if this goal was achieved. Was the goal to decannulate the patient, to restore voice, or to create a safer airway? Other factors to be considered include the possibility of insufficient preoperative evaluation, inappropriate patient selection, technical errors or misjudgments, failure to optimize the patient’s status before surgery, and factors inherent to the child .
Insufficient preoperative evaluation
Insufficient diagnostic information can lead to a misdiagnosis, failure to detect another airway lesion, or a missed non-airway diagnosis, all of which can result in a failed procedure. An example of an insufficient preoperative evaluation is a child who has both subglottic stenosis and vocal fold immobility who undergoes surgery to repair the subglottic stenosis without the vocal fold immobility being addressed. If the presence of the vocal fold immobility was missed in the diagnostic evaluation, decannulation may be unsuccessful because of persistent glottic obstruction.
Inappropriate patient selection
Patient selection can affect overall clinical outcome significantly. Although the goal of creating an anatomically normal airway at the site of reconstruction may be achieved from a technical perspective, if a child remains dependent on a tracheotomy because of oxygen requirements, ventilation requirements, or other levels of airway obstruction that were not addressed during the surgery, then, in essence, the operation has failed. Similarly, if decannulation was achieved, but the operation has resulted in a child suffering from chronic aspiration, then, in a more global sense, the operation still has failed.
Technical causes of failure
Technical causes of surgical failure that warrant assessment include
- •
Using a technique that may not have been the best surgical option for managing a particular lesion
- •
Performing a single-stage operation rather than a two-stage procedure (In a single-stage procedure, the tracheotomy is removed, and pathology is addressed concurrently. In a two-stage procedure, the tracheotomy is maintained during airway reconstruction, and a suprastomal stent is placed to support the reconstruction as healing occurs.)
- •
Choosing a stent that may have been inappropriate for the procedure performed or a stenting period that may not have been optimal
Optimization
The authors have found that optimizing patient status before surgery (whether an initial or a revision procedure) plays a central role in airway management and is absolutely crucial to the success of surgery. They believe this point cannot be overemphasized. Many children have multiple comorbidities that complicate the treatment of an isolated airway problem. Lung disease caused by prematurity, a history of long-term ventilation, gastrointestinal disease, feeding disorders, and neurologic impairment all can have a negative impact on a well-conceived and well-executed surgical plan.
What are the operative and perioperative risks that can lead to surgical failure?
Although an appropriate surgical plan may have been in place, there are elements of particular operations that can precipitate failure.
Laryngotracheoplasty
Laryngotracheoplasty (LTP) with cartilage grafting can fail as a consequence of graft necrosis. In the senior author’s (MJR) experience, approximately 1 in 50 rib grafts are lost. Graft loss is caused most commonly by infection, steroid usage, or inadequate muscle coverage.
Infection
Any type of bacterial infection can predispose to graft loss; however, the most common causative organisms are methicillin-resistant Staphylococcus aureus (MRSA) and Pseudomonas aeruginosa . The authors have found that children who have colonized bacteria in the nares or the trachea are at higher risk for graft loss from infection. It is prudent to screen for these bacteria, because they may result in graft infection following LTP. In the authors’ experience, preoperative antibiotic prophylaxis can decrease the risk of graft loss. The protocols used for screening and antibiotic prophylaxis are described in Box 1 and Table 1 , respectively.
Tracheal aspirate
Bacterial culture and sensitivities
MRSA culture and sensitivities
Nares
MRSA culture and sensitivities
Perianal
MRSA culture and sensitivities
Abbreviations: MRSA, methicillin-resistant Staphylococcus aureus .
a Performed within 3 months of anticipated surgery (at least 10 days before surgery).
| Timing | MRSA protocol | Pseudomonas aeruginosa protocol |
|---|---|---|
| Preoperative | ▵ tracheostomy tube. Topical or nebulized antipseudomonal (tobramycin or Ciprodex) | |
| Perioperative | Intravenous vancomycin 1 hour before incision | Intravenous piperacillin and tazobactam 1 hour before incision |
| Postoperative | Intravenous vancomycin every 6 to 8 hours for 24 to 48 hours; then change to oral TMP/SMX for total of 14 days’ treatment | Intravenous piperacillin and tazobactam every 6 hours for 24 to 48 hours |
a Children who have MRSA-positive cultures from the nares are treated with mupirocin ointment twice daily for 72 hours before surgery.
b MRSA resistant to TMP is treated with an alternate oral antibiotic based on sensitivities.
Steroids
The effect of steroids on wound healing is well known. When administered at high dosages and over long periods of time, steroids can affect wound healing negatively . Steroids can impede the vascularization, mucosalization, and integration of the graft and thus predispose to graft loss. In single-stage reconstruction, the use of high-dose steroids for a brief period may facilitate extubation, with negligible risk of graft loss.
Inadequate muscle coverage
As a cartilage graft incorporates into the airway, the early blood supply is of paramount importance. The blood supply to the graft is provided by overlying muscle, usually the strap musculature. The risk of graft loss is increased when this early blood supply is compromised. Although a Penrose drain often is placed between the strap muscles and the graft, the authors have found that it can prevent the early blood supply from vascularizing the graft, especially if the drain is left in place for more than 48 hours. If an anterior graft is not sutured properly and there is a significant air leak, the resultant aerocele or mucocele can lift the muscle off the graft and lead to devascularization ( Fig. 1 ). Pre-emptive procedures that can help decrease the risk of graft loss include testing for an air leak once the anterior graft is sutured in place; sealing small leaks with Tisseel or a muscle plug; and closing the strap musculature loosely with a superficially placed Penrose drain that can allow egress of some air while maintaining the contact of the graft to its blood supply.
Cricotracheal resection
As with LTP with cartilage graft reconstruction, cricotracheal resection (CTR) can fail because of the inherent risks of the operation. Dehiscence is a dreaded cause of failure and can be life threatening. The risk of dehiscence is increased by steroid usage and by bacterial infection (MRSA and Pseudomonas aeruginosa ) . Preoperative screening for bacterial colonization and subsequent perioperative antibiotic prophylaxis can decrease the risk of dehiscence.
Children who have Down syndrome are at an increased risk of dehiscence. Many of these children have hyperflexible neck mobility that, if not taken into account, can cause added strain on an anastomosis and potentially lead to dehiscence. In the authors’ experience, this group of children is particularly problematic, and the surgeon must be extremely cautious when performing a CTR; the authors often leave the chin-to-chest sutures in place for at least 2 weeks rather than the more common period of 7 to 10 days.
The risk of dehiscence increases as the tension across the anastomosis increases. This may be seen in children with long-segment stenosis or in those who have had previous distal tracheal surgery. It also is seen in patients in whom the trachea is difficult to mobilize. The surgeon should exercise particular caution in these patients.
The recurrent laryngeal nerves can be injured unilaterally or bilaterally during CTR, causing a temporary vocal fold paresis or permanent paralysis . When such injury occurs bilaterally, an iatrogenic glottic obstruction can prevent decannulation. The relative risk of damage to the recurrent laryngeal nerves is much higher with CTR than with LTP because of the technical aspects of the operation. With CTR, the recurrent laryngeal nerves can be injured during skeletonization of the trachea or during the resection of the anterior cricoid ring near the cricothyroid joint.
Hidden airway lesions
Children who require laryngotracheal reconstruction can have an additional lesion of the airway that may have been undetected before the initial surgery. Such lesions typically present as failure to extubate following a single-stage procedure or failure to decannulate following a two-stage procedure. Hidden lesions (eg, laryngomalacia, vocal cord dysfunction, and tracheomalacia) classically are dynamic and may not be seen on rigid bronchoscopy . If extubation or decannulation is unsuccessful and there is no obvious stenosis, a dynamic lesion should be suspected. An endoscopic evaluation during spontaneous respiration may reveal the cause of obstruction.
Perioperative care
The perioperative care of children who have undergone airway reconstruction can have an important effect on the end result of surgery. Scarring, granulation tissue, and mild restenosis can be managed endoscopically in the early postoperative phase, but failure to address such problems promptly can lead to restenosis. The authors typically use a vigilant management approach, which is discussed later in this article.
What are the operative and perioperative risks that can lead to surgical failure?
Although an appropriate surgical plan may have been in place, there are elements of particular operations that can precipitate failure.
Laryngotracheoplasty
Laryngotracheoplasty (LTP) with cartilage grafting can fail as a consequence of graft necrosis. In the senior author’s (MJR) experience, approximately 1 in 50 rib grafts are lost. Graft loss is caused most commonly by infection, steroid usage, or inadequate muscle coverage.
Infection
Any type of bacterial infection can predispose to graft loss; however, the most common causative organisms are methicillin-resistant Staphylococcus aureus (MRSA) and Pseudomonas aeruginosa . The authors have found that children who have colonized bacteria in the nares or the trachea are at higher risk for graft loss from infection. It is prudent to screen for these bacteria, because they may result in graft infection following LTP. In the authors’ experience, preoperative antibiotic prophylaxis can decrease the risk of graft loss. The protocols used for screening and antibiotic prophylaxis are described in Box 1 and Table 1 , respectively.
Tracheal aspirate
Bacterial culture and sensitivities
MRSA culture and sensitivities
Nares
MRSA culture and sensitivities
Perianal
MRSA culture and sensitivities
Abbreviations: MRSA, methicillin-resistant Staphylococcus aureus .
a Performed within 3 months of anticipated surgery (at least 10 days before surgery).
| Timing | MRSA protocol | Pseudomonas aeruginosa protocol |
|---|---|---|
| Preoperative | ▵ tracheostomy tube. Topical or nebulized antipseudomonal (tobramycin or Ciprodex) | |
| Perioperative | Intravenous vancomycin 1 hour before incision | Intravenous piperacillin and tazobactam 1 hour before incision |
| Postoperative | Intravenous vancomycin every 6 to 8 hours for 24 to 48 hours; then change to oral TMP/SMX for total of 14 days’ treatment | Intravenous piperacillin and tazobactam every 6 hours for 24 to 48 hours |
a Children who have MRSA-positive cultures from the nares are treated with mupirocin ointment twice daily for 72 hours before surgery.
b MRSA resistant to TMP is treated with an alternate oral antibiotic based on sensitivities.
Steroids
The effect of steroids on wound healing is well known. When administered at high dosages and over long periods of time, steroids can affect wound healing negatively . Steroids can impede the vascularization, mucosalization, and integration of the graft and thus predispose to graft loss. In single-stage reconstruction, the use of high-dose steroids for a brief period may facilitate extubation, with negligible risk of graft loss.
Inadequate muscle coverage
As a cartilage graft incorporates into the airway, the early blood supply is of paramount importance. The blood supply to the graft is provided by overlying muscle, usually the strap musculature. The risk of graft loss is increased when this early blood supply is compromised. Although a Penrose drain often is placed between the strap muscles and the graft, the authors have found that it can prevent the early blood supply from vascularizing the graft, especially if the drain is left in place for more than 48 hours. If an anterior graft is not sutured properly and there is a significant air leak, the resultant aerocele or mucocele can lift the muscle off the graft and lead to devascularization ( Fig. 1 ). Pre-emptive procedures that can help decrease the risk of graft loss include testing for an air leak once the anterior graft is sutured in place; sealing small leaks with Tisseel or a muscle plug; and closing the strap musculature loosely with a superficially placed Penrose drain that can allow egress of some air while maintaining the contact of the graft to its blood supply.
Cricotracheal resection
As with LTP with cartilage graft reconstruction, cricotracheal resection (CTR) can fail because of the inherent risks of the operation. Dehiscence is a dreaded cause of failure and can be life threatening. The risk of dehiscence is increased by steroid usage and by bacterial infection (MRSA and Pseudomonas aeruginosa ) . Preoperative screening for bacterial colonization and subsequent perioperative antibiotic prophylaxis can decrease the risk of dehiscence.
Children who have Down syndrome are at an increased risk of dehiscence. Many of these children have hyperflexible neck mobility that, if not taken into account, can cause added strain on an anastomosis and potentially lead to dehiscence. In the authors’ experience, this group of children is particularly problematic, and the surgeon must be extremely cautious when performing a CTR; the authors often leave the chin-to-chest sutures in place for at least 2 weeks rather than the more common period of 7 to 10 days.
The risk of dehiscence increases as the tension across the anastomosis increases. This may be seen in children with long-segment stenosis or in those who have had previous distal tracheal surgery. It also is seen in patients in whom the trachea is difficult to mobilize. The surgeon should exercise particular caution in these patients.
The recurrent laryngeal nerves can be injured unilaterally or bilaterally during CTR, causing a temporary vocal fold paresis or permanent paralysis . When such injury occurs bilaterally, an iatrogenic glottic obstruction can prevent decannulation. The relative risk of damage to the recurrent laryngeal nerves is much higher with CTR than with LTP because of the technical aspects of the operation. With CTR, the recurrent laryngeal nerves can be injured during skeletonization of the trachea or during the resection of the anterior cricoid ring near the cricothyroid joint.
Hidden airway lesions
Children who require laryngotracheal reconstruction can have an additional lesion of the airway that may have been undetected before the initial surgery. Such lesions typically present as failure to extubate following a single-stage procedure or failure to decannulate following a two-stage procedure. Hidden lesions (eg, laryngomalacia, vocal cord dysfunction, and tracheomalacia) classically are dynamic and may not be seen on rigid bronchoscopy . If extubation or decannulation is unsuccessful and there is no obvious stenosis, a dynamic lesion should be suspected. An endoscopic evaluation during spontaneous respiration may reveal the cause of obstruction.
Perioperative care
The perioperative care of children who have undergone airway reconstruction can have an important effect on the end result of surgery. Scarring, granulation tissue, and mild restenosis can be managed endoscopically in the early postoperative phase, but failure to address such problems promptly can lead to restenosis. The authors typically use a vigilant management approach, which is discussed later in this article.
What are disease-specific causes of failure?
There are a number of disease-specific risk factors for laryngotracheal reconstruction failure, and surgeons should be particularly mindful of these factors when considering possible causes of failure. As emphasized earlier, each of these clinical circumstances mandates stabilization or optimization before a revision operation is undertaken.
Eosinophilic esophagitis
Eosinophilic esophagitis (EE) is an uncommon disorder that, if left untreated, may have a significant effect on the aerodigestive tract. Many patients who have EE have esophageal, laryngotracheal, and sinonasal complaints; however, some patients are asymptomatic . The diagnosis of EE is made by histologic examination of biopsies taken from the esophagus at the time of esophageal endoscopy. In patients who have active EE, the laryngotracheal complex often is inflamed. Revision surgery in the presence of active EE often elicits a brisk inflammatory response that can lead to graft failure and or restenosis. If a child requires revision LTP, the authors recommend that the preoperative evaluation include esophagoscopy with biopsies specifically to exclude EE. If EE is present, the authors recommend medical management followed by repeat endoscopy with biopsies. Once biopsies demonstrate no active EE, revision surgery may be performed. For more in-depth information on EE and its treatment, the reader is referred to the recently published guidelines by Furuta and colleagues .
Gastroesophageal reflux disease
Because of the potential impact of gastroesophageal reflux disease (GERD) on postoperative healing, the authors routinely administer prophylactic pre- and postoperative therapy to patients undergoing airway reconstruction . Most patients are managed with a daily proton pump inhibitor and nighttime H 2 -blocker therapy. Patients continue the antireflux regimen for up to 1 year following successful reconstruction. Given the minimal side effects and the significant potential benefit of preventing restenosis and failure, the authors consider this approach prudent. The evaluation for GERD can include esophagoscopy with biopsies, esophagram, impedance probe, and/or dual pH probe testing. Although medical management of GERD often is sufficient, non-acidic reflux can play a role also. The authors believe that in some cases non-acidic reflux can cause damage in the reconstructed airway and potentially lead to failure. When medical treatment fails or non-acid reflux is suspected, a Nissen fundoplication should be considered before airway reconstruction.
Obstructive sleep apnea
Obstructive sleep apnea (OSA) can be difficult to diagnose and treat in children. It can cause failure in an otherwise well-executed operative plan. In children who have a tracheotomy and a known fixed airway lesion above the tracheotomy, OSA may be difficult to identify because the tracheotomy cannot be capped during a sleep study. If a single-stage procedure was performed, failure may occur because of the inability to determine the presence of clinically significant OSA. In two-stage procedures, an inability to decannulate may have been a function of OSA that was unrecognized.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
