Advertisement

Surgical Management of Subglottic Stenosis

  • Cameron D. Wright
    Correspondence
    Address reprint requests to Cameron D. Wright, MD, Associate Professor of Surgery, Harvard Medical School, Division of Thoracic Surgery, Massachusetts General Hospital, Boston, MA 02114.
    Affiliations
    Harvard Medical School, Division of Thoracic Surgery, Massachusetts General Hospital, Boston, Massachusetts.
    Search for articles by this author
      Subglottic stenosis is usually due to either endotracheal tube ischemic necrosis or idiopathic larygotracheal stenosis. The ischemic area involves varying depths of injury, causing a spectrum of injury from mucosal to full thickness, including the cartilage. The narrowest portion of the airway is the cricoid in the lower larynx. Selection of an endotracheal tube that is too large relative to the size of the cricoid can lead to circumferential injury to the cricoid mucosa, which may lead to subglottic stenosis. Idiopathic laryngotracheal stenosis is an unusual disease that almost always just affects women with a peculiar nonspecific scarring that only affects the lower larynx and upper few tracheal rings. Autoimmune disease and Wegner’s granulomatosis must be eliminated in the differential diagnosis by appropriate blood tests. Stenosis involving the larynx is more complex to repair and the results are less certain than simple tracheal resection. Stenosis also may involve the glottis from endotracheal tube damage to the vocal cords and arytenoids. An abnormal glottis must be corrected before correction of any tracheal problem as the glottis must provide an adequate airway at the conclusion of a laryngotracheal resection.
      Resection should be deferred until the patient fully recovers from the original illness that precipitated mechanical ventilation. All necessary operations that require general anesthesia (ie, burn patients that require multiple excisions and scar releases) should be completed before embarking on tracheal resection. Often steroids are mistakenly administered to treat new-onset “asthma” after prolonged intubation. Steroids should be rapidly weaned off as they play no role in ameliorating postintubation stenosis and interfere with healing of the anastomosis. Glottic pathologic problems should be corrected before treating the tracheal pathologic problem. Mucosal inflammation identified by bronchoscopy should be allowed to resolve before resection to reduce the chance of an anastomotic problem. If the patient is in otherwise good shape, definitive treatment by resection may be performed. The airway can be temporarily opened by rigid bronchoscopic dilation if only a short period (ie, weeks) of time is needed before definitive treatment. Laser bronchoscopy further injures the airway and should be avoided when one wishes to temporize before proceeding to definitive resection. Self-expanding metallic stents should not be used as they effectively lengthen the airway injury and often convert a simple short resection to a very long resection, which has a much greater failure rate.
      In general, any patient with obstructive airway symptoms and airway stenosis after ventilation should have definitive laryngotracheal resection and reconstruction if they have favorable anatomy. Patients with idiopathic stenosis often have significant inflammation and scarring coming very close to the vocal cords, leading to an inadequate subglottic vestibule to perform an anastomosis (Figure 1, Figure 2). These patients must be selected with care as there is little margin for error. Absolute contraindications are few and include a nonreconstructable airway (usually due to excessive length of the damaged airway), severe comorbidities, and a continued need for ventilation. Relative contraindications include a history of radiation to the trachea, continued mucosal inflammation beyond the area of resection, and active steroid use.
      Figure thumbnail gr1
      Figure 1Bronchoscopic view of the vocal cords and subglottic larynx in a woman with idiopathic laryngotracheal stenosis. There is maximal circumferential scar at the level of the cricoid. There is relative preservation of the immediate subglottic vestibule, indicating that this is a favorable case for resection.
      Figure thumbnail gr2
      Figure 2Bronchoscopic view just at the level of the vocal cords in a patient with postintubation stenosis with complete subglottic stenosis at the level of the top of the cricoid. There is mild narrowing of the subglottic vestibule but there is a reasonable opening to perform an anastomosis.
      Although everyone has a trachea, anatomic and demographic factors influence how much trachea can be safely resected and reconstructed. Probably the most important factor is body habitus; young, tall, thin patients with long necks have an abundance of trachea and are the most straightforward candidates for surgery. Alternatively, elderly, kyphotic, short, obese patients with bull necks and a cricoid that is at the sternal notch are the most challenging patients. Recent high-dose steroid use and insulin-dependent diabetes seem to be risk factors for anastomotic problems, so special attention needs to be paid to the details and amount of tension on the anastomosis. Pediatric patients tolerate anastomotic tension less well than adults; ideally resections only under 30% should be done in children, whereas some adults can tolerate resections up to 50% of the length of the trachea. Previous tracheal surgery causes relative fixation of the remaining trachea and thus limits any possible re-resection to less than normal. Previous radiation therapy to the trachea limits tracheal mobility and impairs microvascular blood supply, both of which place constraints on how much trachea can be safely removed.
      Patients who present with critical airway obstruction almost always can be managed with bronchoscopic dilation. It is very rare that a tracheostomy is needed to establish an airway in postintubation tracheal stenosis or idiopathic laryngotracheal stenosis. Patients should be promptly taken to the operating room for urgent brochoscopic dilation. If the airway is judged to be critical, muscle relaxants are avoided and induction of anesthesia is done by an inhalational technique, which allows the patient to breathe spontaneously until the airway is secured by bronchoscopy. The surgeon must be present during induction of anesthesia with all equipment immediately available and absolute competence to establish an airway. When the patient is deep enough to allow airway manipulation, rigid bronchoscopy is commenced with an adult no. 7 or 8 rigid bronchoscope to visualize the airway and stenosis. Dilation is begun with plastic small-diameter Jackson dilators through the bronchoscope until the maximal size is reached (about 6 mm). At this point, the airway is open enough to allow suctioning of retained distal secretions and to allow the anesthesiologist to ventilate the patient enough to return the end-tidal CO2 to normal. The airway can be further dilated by passage of serial-sized rigid bronchoscopes from a pediatric no. 3 up to an adult no. 9. The surgeon can appreciate the degree of resistance with the passage of each size scope to decide exactly how far to dilate so as to avoid tearing the airway. The surgeon should select the proper sized bronchoscope that is just a little too small to fit through the stenosis to begin with dilation. This allows for a graded stretching of the stenosis. If too large a size is selected, the lip of the bronchoscope can instead tear the mucosa, leading to development of a flap, which can occlude the end of the bronchoscope and lead to loss of the airway. If dilation is planned as a temporizing measure, in general, the dilation should be to the largest size possible so it will last for as long as possible. An alternate technique is to use the new noncompliant balloon dilators with either a rigid or a flexible bronchoscope to achieve the same result. If dilation is being done just before a resection, usually one dilates just enough to accept a no. 6 endotracheal tube.

      Operative Technique

      Figure thumbnail gr3
      Figure 3The patient is supine on the operating room table with an inflatable bag underneath the shoulders to extend the neck. A small transverse incision is placed below the palpable cricoid cartilage, usually from the edge of each sternocleidomastoid muscle. If it is a typical idiopathic stenosis where just the first two rings of trachea are removed, the incision is usually about 2 cm below the bottom of the cricoid. The incision is placed progressively lower if more trachea must be removed as the anastomosis will always be brought closer to the sternum by neck flexion. Subplatysmal flaps are created from the top of the thyroid cartilage and to the top of the sternal notch. Gelpi retractors are placed on each side. The strap muscles are separated in the midline and the anterior aspect of the airway is exposed. The thyroid isthmus is dissected off the front of the trachea, clamped, ligated, and divided. The pretracheal fascia at the level of the sternal notch is divided transversely and a finger is swept under the fascia, on top of the trachea and descending down into the mediastinum to free the anterior trachea.
      Figure thumbnail gr4
      Figure 4A typical short subglottic stenosis, usually from either postintubation stenosis or idiopathic stenosis. In this case, which is favorable for repair, the laryngeal stenosis involves the anterolateral cricoid but spares the posterior cricoid and the immediate subglottic space. The small dashed lines indicate the limits of the stenosis. The bold dashed lines indicate the lines of division. The trachea is dissected circumferentially to the proposed distal transection point staying directly on the trachea to avoid injury to the recurrent laryngeal nerve. The recurrent nerve is never dissected out and in fact is almost never seen as long as one dissects right on the wall of the trachea. Typically one dissects about two rings distal to the true division point so there is enough room to suture the anastomosis. One should not dissect too far distal as that will devascularize the trachea. If the distal extent of stenosis is unclear externally, it is best to bronchoscope the patient with the endotracheal tube pulled back to mark the exact internal location of the end of the stenosis. One can either just transilluminate the trachea with the light or pass a 25-gauge needle through the anterior tracheal wall to define the true bottom of the stenosis. Before the cricoid is cut, the cricothyroid muscle must be dissected off the front aspect. I use the cautery set on very low to peel it off the perichondrium to just beyond the proposed line of division. The cricoid is divided obliquely to preserve the recurrent nerve and to maximally open the subglottic airway. The distal trachea is also divided obliquely (usually two rings worth) to match the upper division line. Ventilation is performed by a sterile armored endotracheal tube in the distal trachea while the trachea is divided. The native endotracheal tube is usually just removed and then replaced at the end of the suture placement. n = nerve.
      Figure thumbnail gr5
      Figure 5Stay sutures of 2-0 Vicryl are placed at 3 and 9 o’clock at both the tracheal side and the laryngeal side of the anastomosis. On the tracheal side they are placed at least one ring away from the edge and at least around one whole ring. They are placed so that the knots are tied on the outside. The cricoid stay sutures are placed in a deep bite of cricoid but are not through and through the mucosa. They are extramucosal here to try to minimize edema and a possible nidus for granulation tissue in the more fragile larynx. They will be tied first when the anastomosis is approximated to take the tension off the true anastomotic sutures. m = muscle; n = nerve.
      Figure thumbnail gr6a
      Figure 6Anastomotic sutures are placed starting posteriorly with knots to be tied on the outside. Each successive suture is placed so that the loose ends of the suture are in front of the previous suture. This is done so that when the sutures are tied, they are tied in reverse order (front to back) and the suture being tied does not interfere with the more posterior suture. Sutures are placed around one ring of cartilage (or the equivalent) and about 3 to 4 mm apart. Usually 4-0 Vicryl sutures are used for the anastomosis in an adult. All the sutures are placed before tying any. The ends of the sutures are snapped with a hemostat and then the hemostat is snapped to the surrounding drapes in a circumferential fashion to organize the sutures. Before tying the sutures, the inflatable bag beneath the shoulders is deflated and the head is mildly flexed to reduce tension on the anastomosis. In straightforward cases where a short resection is done, the lateral stay sutures are approximated and tied first, followed by the true anastomotic sutures. When there is a substantial resection, a preliminary trial approximation is done to assess the tension on the anastomosis. If this seems excessive, additional release maneuvers are warranted. The distal airway release is again maximized by loosening all anterior attachments of the airway to the mediastinum as far as a finger will reach into the mediastinum. If this is not enough, then a suprahyoid laryngeal release will be necessary. In general, if you are thinking that the tension is a little too much, a hyoid release should be done.
      Figure thumbnail gr6b
      Figure 6Anastomotic sutures are placed starting posteriorly with knots to be tied on the outside. Each successive suture is placed so that the loose ends of the suture are in front of the previous suture. This is done so that when the sutures are tied, they are tied in reverse order (front to back) and the suture being tied does not interfere with the more posterior suture. Sutures are placed around one ring of cartilage (or the equivalent) and about 3 to 4 mm apart. Usually 4-0 Vicryl sutures are used for the anastomosis in an adult. All the sutures are placed before tying any. The ends of the sutures are snapped with a hemostat and then the hemostat is snapped to the surrounding drapes in a circumferential fashion to organize the sutures. Before tying the sutures, the inflatable bag beneath the shoulders is deflated and the head is mildly flexed to reduce tension on the anastomosis. In straightforward cases where a short resection is done, the lateral stay sutures are approximated and tied first, followed by the true anastomotic sutures. When there is a substantial resection, a preliminary trial approximation is done to assess the tension on the anastomosis. If this seems excessive, additional release maneuvers are warranted. The distal airway release is again maximized by loosening all anterior attachments of the airway to the mediastinum as far as a finger will reach into the mediastinum. If this is not enough, then a suprahyoid laryngeal release will be necessary. In general, if you are thinking that the tension is a little too much, a hyoid release should be done.
      Figure thumbnail gr7
      Figure 7The anastomosis is now completed. The anastomosis should be checked for airtightness by having the anesthesiologist give 20 to 30 cm of positive airway pressure (with the endotracheal cuff deflated) while the anastomosis is submerged under saline. If there are any small leaks, an additional suture should be placed to correct this. In addition, a bronchoscopy is done (with withdrawal of the endotracheal tube under direct bronchoscopic control) to visually inspect the anastomosis from the inside for technical problems, such as loose sutures, inadequate approximation, or an inadequate airway (luminal diameter) reconstruction. m = muscle.
      Figure thumbnail gr8
      Figure 8The cricothyroid muscles are reattached over the anastomosis with two fine sutures which typically include a bite of muscle, the airway in the midline, and then the opposite muscle. This helps cover the anastomosis and facilitates reattachment of the muscle, which tenses the vocal cord and is important in voice pitch.
      Figure thumbnail gr9
      Figure 9When the stenosis extends higher and with more posterior involvement, a slight modification of the anastomosis is called for. The transection of the inferior larynx is about the same as before and this will open up the anterior subglottic larynx nicely. In this case however there is a thick layer of posterior scar that limits the airway that must be dealt with. The posterior cricoid cannot be removed entirely because of the recurrent nerve and the esophagus behind the cricoid plate. The posterior mucosa and submucosa can be removed to within several millimeters of the inferior vocal cords to open up the airway. If need be, a portion of the posterior cricoid plate can be removed to gain another couple of millimeters of airway space. This raw surface then needs to be resurfaced to allow satisfactory healing. This is done by cutting the distal trachea in such a fashion that preserves a portion (it should be less than 1 cm so it is well vascularized) of membranous wall that will cover the posterior cricoid plate.
      Figure thumbnail gr10
      Figure 10The anastomosis is much the same as the standard case except for the posterior mucosal flap. The same lateral stay sutures are placed. The sutures for the superior aspect of the flap are placed first and will need to be tied from within the airway lumen (as opposed to the usual where the posterior sutures are tied on the outside from the lateral aspect of the anastomosis). The sutures are again placed from deep to superficial so that the knots will tend to lie down. The lateral sutures are placed in the usual fashion from the corner posteriorly up to the stay sutures (usually no more than three sutures). The anterior sutures will not be placed like usual to allow the posterior sutures to be tied easily without any obstructions. In some cases the laryngeal opening is wide enough such that the posterior flap sutures will be able to be placed and tied from the anterior airway defect after the two lateral stay sutures are tied. However in many cases the opening is too small. In these cases a laryngofissure should be done to further open up the anterior larynx to allow gentle suture tying. Once the posterior and posterolateral sutures are tied, the anterior sutures are placed and tied in the standard fashion. The anastomosis is again checked externally and internally. These high, complex anastomoses are more prone to incite laryngeal edema, which can temporarily obstruct the airway. Certainly if significant edema of the vocal cords is seen at the end of the procedure, strong consideration should be given to a temporary tracheostomy until the airway edema resolves. If performed, a tracheostomy should be done well distal to the anastomosis; the anastomosis should be separated from the tracheostomy with viable good tissue such as a strap muscle, and the anastomosis should be covered as well with viable tissue to keep it away from tracheostomy secretions.
      Figure thumbnail gr11
      Figure 11If needed, a laryngofissure is performed by vertically splitting the thyroid cartilage exactly in the midline. This will widely open up the immediate subglottic larynx to allow good access. It is closed with several 4-0 Vicryl sutures and heals readily. Care must be taken when closing the laryngofissure that the sutures do not come too close to the vocal cords such that motion/function would be impaired. A laryngofissure does tend to increase the chances of significant vocal cord edema.

      Results

      In a recent series review of idiopathic laryngotracheal stenosis, we reported 91% of patients had good to excellent results with only 1% needing long-term dilation.
      • Ashiku S.K.
      • Kuzucu A.
      • Grillo H.C.
      • et al.
      Idiopathic laryngotracheal stenosis: effective definitive treatment by laryngotracheal resection.
      In our unit’s most recent series of patients with postintubation stenosis, 82% had a good result, 8% had a satisfactory result, and 8% had a failed result after laryngotracheal resection and reconstruction.
      • Grillo H.C.
      • Mathisen D.J.
      • Wain J.C.
      Laryngotracheal resection and reconstruction for subglottic stenosis.
      Figure 12 illustrates an excellent anastomotic result after laryngotracheal reconstruction.
      Figure thumbnail gr12
      Figure 12Bronchoscopic view just below the vocal cords of the patient in one month after laryngotracheal resection, demonstrating a satisfactory anastomosis of the trachea to the thyroid cartilage in an oblique fashion.

      References

        • Ashiku S.K.
        • Kuzucu A.
        • Grillo H.C.
        • et al.
        Idiopathic laryngotracheal stenosis: effective definitive treatment by laryngotracheal resection.
        J Thorac Cardiovasc Surg. 2004; 127: 99-107
        • Grillo H.C.
        • Mathisen D.J.
        • Wain J.C.
        Laryngotracheal resection and reconstruction for subglottic stenosis.
        Ann Thorac Surg. 1992; 53: 54-63