Obstructive sleep apnea (OSA) is caused by obstacles to the free flow of air through the upper airway from the lips or nasal tip to the larynx and therefore is within the otolaryngologist’s sphere of responsibility. Management can involve weight loss, treatment of nasal obstruction, surgery of the pharynx, or the use of various ventilatory assist devices such as continuous positive airway pressure (CPAP). This condition has serious health risks and consequences and should be thoroughly researched and investigated to be treated effectively. Furthermore, it is often multifactorial, requiring a multidisciplinary approach to identify the etiology and optimal treatment for each patient.
Tracheotomy (or tracheostomy) provides the most efficacious bypass of upper airway obstruction and is rightfully considered among the treatments of OSA. Although they can singularly cure OSA, tracheotomy/tracheostomy are regarded as extreme modes of treatment due to associated surgical and postoperative morbidities. In various instances these procedures are reserved for emergencies such as difficult intubation or are considered as last resorts in patients with severe, refractory, or life-threatening apnea. Other less frequent indications include morbid obesity, complicated oropharyngeal obstruction with severe hypoxia, failed management attempts, and disabling daytime somnolence.
To understand the role of tracheostomy in the management of OSA, one must distinguish between tracheotomy and tracheo s tomy . These terms are frequently interchanged, mixed, or misused. In this chapter, tracheotomy refers to an operation in which an opening is made in the trachea for the purposes of short-term cannulation, which closes spontaneously when the cannula is removed. On the other hand, tracheostomy is derived from the Greek term stoma and is reserved for procedures that are performed with the intent of establishing a patent, long-term, or permanent opening between the trachea and the overlying skin.
When this surgically established stoma safely retains its patency, it obviates the dependency on tubes or other devices for support and is designated as a long-term tube-free tracheostomy .
With this semantic clarification in mind, one can understand that tracheotomy and tracheostomy have two distinct roles in the management of OSA. For instance, tracheotomy can be performed ad hoc in urgent conditions to temporarily secure problematic high-risk airways, making it possible and safe to manage the disease through other techniques. By contrast, tracheostomy may be applied as a primary and definitive treatment of OSA whenever long-term or permanent bypass of the obstructed upper airway is indicated. Once established, tracheostomy is intended to provide safe, complication-free, and readily accessible control of the airway during which extended management of a patient’s disease with adjunct procedures may be undertaken with minimal risk to the patient’s safety. Reversal of both procedures is possible, although with tracheostomy the closure is an operative procedure, whereas a tracheotomy most often closes spontaneously after cannula removal. This chapter discusses both tracheotomy and tracheostomy, yet the true focus will be upon the use of tracheostomy as a definitive therapy for OSA.
Historically, the documented performance of tracheotomy (semantically misrepresented as tracheostomy) for Pickwickian syndrome was first described in 1969. Since then, various modifications of the procedure have evolved into the “gold standard” for management of OSA because, when successful, it bypasses upper airway obstruction in 100% of the cases in which it is used. Unfortunately, as mentioned earlier, patients with OSA are often difficult candidates for performance and maintenance of routine tube-dependent tracheotomy; their unfavorable anatomy may be compounded by obesity, diabetes, infection, and cardiopulmonary comorbidity. With each negative experience, lessons were learned that led to modifications of tracheotomy as it related to the treatment of OSA. Through this process, tracheotomy evolved into increasingly refined techniques such as tube-dependent tracheostomy and subsequently into the tube-free tracheostomy, which has been researched, promoted, and formulated by the authors. Further classification of these terms is provided next.
Tracheotomy (incisional or percutaneous)
Tracheostomy, tube or stent dependent
Tracheostomy, long-term tube-free (LTTFT)
As a treatment option for patients with OSA, tracheotomy/tracheostomy can be performed either as definitive therapy or as a temporary means to control the airway while another upper airway surgery is performed. For instance, the risks of difficult intubation or postoperative airway obstruction due to perioperative edema and hemorrhage may necessitate short-term tracheotomy before, during, or after other surgeries. For most patients, the perioperative risks of such other surgery can be obviated through CPAP or remain low enough that temporary tracheotomy is not necessary. When tracheotomy is performed for this indication, the technique may be in some instances similar to that for any other short-term, tube-dependent procedure. However, as will be further discussed, the degree of obesity and the difficult anatomy that frequently causes OSA may lead to higher rates of intraoperative and/or postsurgical complications that mandate modification of traditional tracheotomy. Furthermore, as patients may fail to lose weight or may not benefit from other surgeries, tracheotomy may not always be as short term as originally planned. For these reasons, traditional tracheotomy has given way to both tube-dependent and tube-free tracheostomy techniques in the management of OSA.
When permanent tracheostomy is considered, it is generally indicated for one of three scenarios. The first scenario is dictated by the patient’s abnormal anatomy that cannot be managed under the limitations of traditional tracheotomy and therefore requires tracheostomy. The second scenario may be encountered in patients with OSA who have failed more conservative medical and/or surgical therapies. As the rest of this book demonstrates, CPAP may not always be applicable, effective, or well tolerated, and other surgical techniques lack the 100% success rate of bypassing obstruction, which is the hallmark of tracheostomy. For patients who fail at weight loss or continue to have medically significant OSA despite conservative medical and/or surgical management, tracheostomy may be considered the final measure of what is a graduated, stepwise approach to treatment.
In the last scenario, tracheotomy/tracheostomy is performed as the primary procedure for control of severe disease and/or life-threatening cardiopulmonary manifestations of OSA. Polysomnographic indications for tracheostomy as a primary treatment modality for OSA include Respiratory Disturbance Index >50 and oxygen desaturation to <60%. Additionally, if OSA of any severity causes significant cor pulmonale or cardiac dysrhythmias in association with the apneic episodes, tracheostomy might be warranted as the initial management of choice. For instance, severe bradycardia, asystole, multiple premature ventricular contractions, and even runs of ventricular tachycardia in association with apneas may all serve as indications for primary tracheostomy. Lastly, other comorbid states such as impending or recent myocardial infarction, unmanageable hypertension, or diabetes mellitus might serve as indications for immediate and conclusive management through urgent tracheostomy. For example, an epileptic who experiences seizures while apneic or a patient with coronary artery disease who develops nocturnal angina in association with OSA might be promptly and effectively managed with tracheostomy. When indicated, such patients may benefit from the use of modified transtracheal ventilation (e.g. CPAP) at all stages of their treatment (see later).
Tracheotomy and Tube-Dependent Tracheostomy
“Standard tracheotomy” might be performed in situations in which the duration of cannulation is not expected to exceed several days. These tracheotomies might be performed in OSA if they are used to temporarily secure the airway while other upper airway surgery is performed. Even then, the technique may require modifications such as lipectomy or resort to specially designed and constructed tubes to be safe and effective. Once postoperative edema is resolved and the risk of perioperative hemorrhage and any other problems have been overcome, the tracheotomy tube may be withdrawn and the tracheotomy site allowed to heal by secondary intent. If prolonged cannulation is expected, such “standard tracheotomy” (even when modified) is not the procedure of choice. Patients with severe OSA are prone to developing more serious complications as they often have short, thick necks and high incidence of anatomic deformations. Chondritis, granulation tissue, infection, and stenosis make long-term maintenance of a “temporary standard tracheotomy” undesirable.
Techniques for “permanent” tracheostomy were developed to minimize these complications whenever cannulation is expected to be prolonged. Among the several indications for establishment of a “permanent” skin-lined tracheostoma are severe laryngeal or laryngotracheal stenosis, bilateral vocal cord fixation (conditions that may cause OSA in and of themselves), neurologic conditions such as myasthenia gravis and amyotrophic lateral sclerosis, and, of course, severe OSA. The first report of permanent tracheostomy for the management of pulmonary disease was by Penta and Mayer in 1960; the technique was subsequently popularized for OSA by Fee and Ward in 1977. These techniques aim to decrease the length of the tracheocutaneous tract and bring the margins of the tracheal fenestration into direct contact and continuity with the cervical skin. Shortening the skin-to-trachea tract limits granulation tissue within the wound bed, promotes primary healing of the mucocutaneous junction, inhibits stomal stenosis, and aims to limit other complications associated with prolonged tracheotomy.
There are several techniques for creating a “permanent” tracheostoma depending upon the degree of continuity established between the trachea and the cervical skin. One early method for creating a permanent tracheostomy was the Bjork flap, in which an inferiorly based tracheal flap (generally created from the anterior portion of the second or third tracheal ring) is sutured to the inferior skin margin. Use of this flap relative to standard tracheotomy reduced rates of accidental decannulation and facilitated reinsertion of a displaced tracheotomy tube. Later methods for tracheostomy attempted to further improve the primary attachments between the trachea and the skin. An example is the “H-flap” technique described by Mickelson. By mobilizing both cervical skin and tracheal flaps, a fully circumferential stoma is created through this technique.
More extensive surgical approaches were researched with the intent to achieve maximal shortening of the tracheocutaneous tract, at the same time producing a tension-free, fully circumferential, permanently patent mucocutaneous junction. These modifications are designed to promote primary healing, ease wound care, and facilitate recannulation more readily than the Bjork flap or the H-flap. Protracted healing by secondary intent, which is often accompanied by stenosis or scar formation, is thus minimized. In past experiences, even these permanent tracheostomies depended upon prolonged placement of irritating foreign bodies in the form of tracheotomy tubes or designated stoma stents. As will be discussed later, these tube- or stent-dependent tracheostomy techniques, although better tolerated than standard tracheotomy, still develop inevitable tube-related complications, which have led to the evolution of the tube-free concepts advocated by the authors.
Review of the Literature on the Efficacy of Tracheotomy and Tracheostomy in OSA
The reader should be aware that these two terms may be inappropriately interchanged in the following references.
Once successfully performed, tracheotomy or tracheostomy should provide secure and complete bypass of any ventilatory obstruction proximal to the tracheotomy’s point of entry into the trachea. This achievement is unmatched by any other form of therapy for OSA. In one study, for instance, tracheostomy successfully relieved OSA in 24 patients; 22 of these patients had already failed other treatment modalities. Nor does the efficacy of tracheostomy fade with time, provided that local tracheal complications are avoided. A retrospective review of 79 patients with tracheostomies for OSA reveals that, even with a mean follow-up of over 8 years, tracheostomy eliminated the obstructive component in OSA in all cases. Long-term follow-up suggests that tracheostomy improves parameters such as Apnea/Hypopnea Index, snoring, and excessive somnolence with efficacy unmatched by any other therapy.
Another benefit of tracheostomy is the immediacy with which it cures OSA. Obstruction is relieved soon after the tracheostomy is established. Polysomnographic measures of sleep architecture reveal that delta-wave and rapid eye movement sleep are greatly increased soon after tracheotomy, in a “rebound” phenomenon in which the body attempts to replenish those stages of deep sleep lost to apneic episodes. Within 1 month after tracheostomy, normal sleep architecture is returned; within weeks after tracheostomy, patients experience resolution of daytime sleepiness and snoring. In association with these improved sleep parameters, Guilleminault et al. also found that tracheostomy resolved the personality changes, erratic behavior, enuresis, morning headache, sleepwalking, and other symptoms related to OSA.
Other studies demonstrate the effectiveness of tracheostomy in reducing the mortality associated with OSA. He et al. found that in patients with apnea indices >20, the 38% 8-year mortality of untreated patients was reduced to 0% in 33 patients with tracheostomy. Partinen et al. report similar findings in their cohort of 198 patients, 71 treated with tracheostomy and 127 treated conservatively with weight loss. All deaths at 5 years were in the weight loss group, with a 5-year mortality rate of 11% compared with 0% for tracheostomy. These differences in mortality are even more remarkable given that the tracheostomy group had, on average, higher apnea and body mass indices than the comparable weight loss group.
One early study on tracheostomy and hemodynamic changes in OSA was published by Motta et al. in 1978. Comparing patients with OSA before and after tracheostomy, these investigators found that during sleep, mean pulmonary and femoral artery pressures were significantly reduced after tracheostomy, whereas arterial oxygen levels were increased. Although the underlying mechanisms of increased cardiovascular morbidity and mortality are not entirely understood, one postulated etiology is that repeated hypoxic events in OSA initiate oxidative stress, cause endothelial cell dysfunction, and exacerbate atherogenic injury.
Unfortunately, there are some limits to the effectiveness of tracheostomy for treatment of sleep apnea. First, tracheostomy can successfully treat only obstructive sleep apnea; central apnea may continue to be a problem for these patients. Indeed, one case report even describes the obstructive apneas corrected by tracheostomy being replaced by central events of similar duration and severity. Second, tube-dependent tracheostomy can only provide effective bypass of upper airway obstruction as long as the tube remains patent. In a patient with a large body habitus, excess soft tissue or skin folds over the chest and under the chin may actually obstruct the tube’s proximal end. It is for that reason that some authors recommend specially designed tubes in addition to adequate lipectomy in combination with tracheostomy for obese patients. Misalignment between the tube and the trachea may result in impaction of the tube’s caudal end against the tracheal wall, causing local ridge formation or stenosis, both of which may be compounded by traumatic suctioning. Unfortunately, patients with OSA compounded by cardiopulmonary decompensation may be more prone to complications, including infections and even persistent apnea, after tracheostomy (from central events, stenosis, or tube obstruction) than patients with “uncomplicated” OSA.
Complications of Tube-Dependent Techniques (Tracheotomy or Tube-Dependent Tracheostomy)
The complications of tube-dependent tracheotomy/tracheostomy can generally be divided into short-term and long-term categories. Short-term complications include possible intraoperative consequences such as damage to the great vessels, injury of the tracheoesophageal party wall, pneumothorax, and pneumomediastinum. Early postoperative consequences of tracheotomy such as tube obstruction, tube displacement, and infection can also be considered short-term complications. Among the long-term complications are such problems as deep tissue infection, granulation tissue, and laryngeal and/or tracheal stenosis at the stomal level or at the caudal end of the tube with predisposition to tracheo–innominate artery fistula. In addition to these physical complications, the often-persistent morbidity associated with tube-dependent tracheostomy may have emotional consequences (such as depression) as well, for both the patient and the patient’s family.
Several studies have documented the incidence of short-term complications after tracheotomy/tracheostomy for OSA. Thatcher and Maisel’s review of 79 patients counts four instances of peristomal wound infection, four instances of skin flap necrosis, and one fatal episode of cardiac arrest among early complications. Harmon et al. report a 15% incidence of peristomal infection after tracheostomy for OSA, whereas Guilleminault et al. suggest that the incidence of low-grade infection causing peristomal granulation may be closer to 42%. This same group also reports that a similar percentage (42%) of patients experienced early problems with poorly fitted tracheotomy tubes, such that the tube would become obstructed when the patient’s head was flexed, hyperextended, or turned to the side.
Postobstructive pulmonary edema has been commonly reported after relief of short-term upper airway obstruction, such as that found with postoperative laryngospasm. However, postobstructive pulmonary edema has also been described as a consequence of OSA, perhaps secondary to effects of OSA on the heart and pulmonary vasculature. Given that tracheostomy for severe OSA provides immediate relief of prolonged upper airway obstruction, it should not be surprising that postobstructive pulmonary edema has been described in this setting as well. Comparing 45 OSA patients to 25 non-OSA patients, Burke et al. found an incidence of pulmonary edema of 67% in the OSA group after tracheotomy compared with a control group incidence of 20%. Although most OSA patients with posttracheotomy pulmonary edema were graded as “mild,” 8/30 patients with pulmonary edema were graded as “moderate” or “severe,” and 2 patients (7% of the overall OSA group) died of complications related to cor pulmonale in the posttracheotomy period.
Among long-term complications of tracheostomy, tracheo–innominate artery fistula is the most feared. Though rare (only 1 of 79 patients in Thatcher and Maisel’s study ), it carries a mortality rate of 73%. One controllable factor that might predispose toward fistula is a “too low” tracheotomy, and for this reason, tracheotomy is rarely performed beneath the third tracheal ring. Unfortunately, emphysematous, barrel-chested, or overweight OSA patients may have a high innominate artery, mandating modifications in the surgical procedure that may promote additional damage, to the larynx in particular.
A much more common long-term complication of tracheostomy is progressive, tube-related deformation of the trachea itself, including the development of laryngotracheal stenosis. Estimates of long-term tracheal damage after tracheostomy for OSA vary and depend on the surgical technique and end point used in each study. For instance, Conway et al. reported on eight patients with standard tracheotomy and found that seven (87.5%) experienced tracheal granuloma or stomal stenosis. Among eight patients (five revision, three primary) with skin-lined permanent tracheostomy, the incidence of tracheal complications decreased to 25%. Thatcher and Maisel do not record the overall proportion of patients who developed granulation tissue, but document that 8 of 79 (10%) of their cohort needed to return to the operating room for management of this problem.
Such retrospective studies might underestimate the degree of structural laryngotracheal deformities, infections, chondritis, granulation tissue, and stenosis compared with prospective studies. Law et al. reviewed several studies and found that retrospective reviews report obstructive tracheal lesions at rates of 1% to 11%, whereas prospective studies found obstructive lesions in 20% to 64% of patients with long-term tracheostomies. Their own study of 81 patients with long-term tracheotomy (mean duration 4.9 months) utilizing fiber-optic bronchoscopy before decannulation discovered obstructive lesions in 54 patients (67%). Of these 54 patients with tracheal obstruction secondary to prolonged tube-dependent tracheotomy, almost half (23/54, 43%) had more than one type of obstructing lesion. All of the lesions were located above the level of the stoma and included 45 obstructing tracheal granulomas (all from the anterior tracheal wall), 19 patients with tracheomalacia, and 10 episodes of tracheal stenosis.
Unfortunately, damage to the trachea and larynx above the level of the stoma is too commonly an unavoidable consequence of prolonged cannulation with a curved tracheotomy tube. The introduction of a mismatched curved foreign body (the routinely used tracheotomy tube) into an otherwise straight airway displaces the suprastomal portion of the anterior tracheal wall inwards ( Fig. 61.1 ). This pathologic buckling of the trachea is an unrelenting dynamic process perpetuated by the movements of the tube relative to the trachea during coughing, breathing, swallowing, and manipulation of the tube for cleaning. In this fashion, progressive deformation of the trachea can erode the anterosuperior tracheal wall and lead to suprastomal inflammation, infection, granulation, perichondritis, chondritis, and eventually necrosis, potentially extending to the cricoid arch and conus elasticus. When the tube is withdrawn, various degrees of tracheomalacia and tracheal stenosis may occur. Variations of standard tracheotomy tubes (i.e. tubes that are narrower, smoother, softer, more flexible, shorter, etc.) can help reduce but cannot eliminate the laryngotracheal morbidity associated with prolonged cannulation.