An AIDS-Associated Cause of the Difficult Airway: Supraglottic Kaposi’s Sarcoma

Abstract

Kaposi’s Sarcoma (KS) is a well known complication of human immunodeficiency virus (HIV) infection and the most common malignancy observed in patients with acquired immunodeficiency syndrome (AIDS). In 1989, as many as 40% of homosexual men with AIDS developed clinically evident KS over the course of the disease (1). Of these, approximately two thirds developed KS lesions involving the head and neck (1), and another 20% manifested respiratory tract involvement (2), which was almost always accompanied by cutaneous KS (3). Since 1996, the overall incidence of AIDS-associated KS in western countries reflects a well-documented downward trend, thanks in large part to the advent of highly active antiretroviral therapy (4–7). Although KS lesions of the upper airway remain a relatively common finding in patients with advanced AIDS, very rarely do they present as a life-threatening cause of airway obstruction, with only a handful of such cases reported in the literature (8–12). We report a case in which supraglottic KS in an AIDS patient was the cause of a difficult airway during the induction of general anesthesia and will discuss the clinical implications of such a case for the anesthesiologist. Case Report A 29-yr-old man with advanced AIDS presented to the emergency department complaining of fever, productive cough, and left-sided weakness lasting several days accompanied by increasing sore throat, dysphagia, and hoarseness of several weeks’ duration. The patient had recently been diagnosed with a new-onset seizure disorder and had been diagnosed on a magnetic resonance imaging scan (MRI) to have a 2.5-cm ring-enhancing lesion of the right frontoparietal lobe, for which he had been started on phenytoin and empiric antitoxoplasmosis therapy. He had also been evaluated by the otolaryngology service, whose examination by nasopharyngoscopy revealed KS lesions of the soft palate and supraglottis, the latter obscuring complete visualization of the vocal cords. The patient denied any symptoms of impending airway obstruction, including shortness of breath or stridor, and was subsequently referred to radiation oncology, where he underwent the first of three scheduled treatments just two days before his emergency department visit and ensuing admission. Shortly after admission for presumed pneumonia, the patient underwent a follow-up computed tomography scan (CT) of the head that showed slight enlargement of the ring-enhancing lesion, now 3.5 cm in diameter and seemingly unresponsive to 3 wk of empiric antitoxoplasmosis therapy. The neurosurgery service was consulted and recommended stereotactic brain biopsy to provide a definitive diagnosis and outline the treatment options. The preanesthetic evaluation of the patient by the anesthesia service was significant for mild respiratory compromise (decreased breath sounds bilaterally; an oxygen requirement of 3 L/min per nasal cannula to prevent dyspnea [Spo2 = 88% on room air]; pH = 7.52, Paco2 = 35 mm Hg, Pao2 = 97 mm Hg, SaO2 = 96%, HCO3 = 28 mEq/L) and multiple superficial KS lesions of the soft palate without evidence of clinically significant airway compromise. Vital signs (temperature = 40oC, blood pressure = 120/76 mm Hg, pulse = 98 bpm, respirations = 20 breaths/min) were stable and pertinent laboratory values documented (hematocrit = 34%, platelets = 200 th/μL, creatinine = 1.3 mg/dL); the patient’s ASA physical status was IV. In view of these findings and after consultation with the otolaryngology service, the decision was made to proceed with general anesthesia and simple direct laryngoscopy as a means of securing an airway and providing anesthesia for the procedure. After giving IV midazolam, 2 mg, anesthesia was induced with 75 μg of fentanyl and 150 mg of propofol IV (the patient weighed 65.5 kg). The ability to provide adequate ventilation via a mask was verified before administering 40 mg of rocuronium IV (succinylcholine was thought to be contraindicated because of the patient’s hemiparesis). Direct laryngoscopy was then performed with a curved (MacIntosh #3) blade, but tracheal intubation was not attempted because of an inability to visualize the vocal cords. During ventilation via a mask, a fiberoptic scope was passed into the pharynx without difficulty. It revealed a friable, mucocutaneous, much larger-than-anticipated supraglottic mass that obscured a complete view of the vocal cords and precluded the safe placement of an endotracheal tube for fear of trauma that might result in hemorrhage or introduction of tumor into the airway. The decision was made to postpone the operation pending further review of other anesthetic and surgical options. Postoperatively, the patient displayed an unusually prolonged response to the rocuronium-induced neuromuscular blockade. It was necessary to control ventilation via a mask for 90 min before it became possible to reverse the blockade (as determined by the recovery of two visible twitches on peripheral nerve stimulation) with neostigmine and glycopyrrolate and confirm adequate spontaneous ventilation. The following day, the otolaryngology and radiation oncology services were consulted, and a repeat examination of the mass with the aid of a nasopharyngoscope was attempted but not tolerated by the patient, although, once again, no stridor, paradoxical chest motion, nasal flaring, or other evidence of impending upper airway obstruction was observed. Thereafter, the patient received two additional radiation therapy treatments and a single treatment of chemotherapy with vincristine and bleomycin in an effort to shrink the tumor’s bulk enough to allow for safe tracheal intubation. A CT of the patient’s neck, taken 1 wk after the aborted operation, confirmed the obstructive nature of the lesion despite the continuing radiation treatment and chemotherapy (see Figure 1). After a follow-up MRI of his brain revealed a second lesion, the patient, with the support of his primary care physician, elected to forego any further invasive diagnostic or therapeutic efforts in favor of more conservative, strictly supportive management of his disease. He was discharged a few days later after 2 wk in the hospital and died several months later of complications related to his advanced AIDS.Figure 1: A computed tomography scan of the pharynx. The top image reveals a large pharyngeal mucosal space mass involving the left faucial tonsil, entire epiglottis, and extending into the left pyriform sinus, preepiglottic and paralaryngeal fat and well across the midline of the intraluminal space on the right. The bottom image is a tomographic cut just above the vocal cords and 9 mm below the level of the cut shown in the top image.Discussion KS of the larynx may be regarded as a relative newcomer to the list of potential causes of the difficult airway. The earliest report of KS of the larynx was provided in 1965 by Reynolds et al. (13), who described a case of classic, or non-AIDS-related, KS in which emergent tracheotomy was required to alleviate symptoms of impending airway obstruction. Since the emergence of AIDS, a number of similar cases have been reported. Greenberg et al. (8), in 1985, were the first to report a case of upper airway obstruction secondary to KS in a known AIDS patient. In 1991. Roy et al. (9) detailed a similar case. A 10-year retrospective review, conducted by Mochloulis et al. (11), identified 17 AIDS patients with laryngeal KS, all of whom presented with symptoms of upper airway obstruction, and found that the most common site of laryngeal involvement was the supraglottis (11 of 17 patients or 65%), followed by the glottis itself (8 of 17 or 47%), and then the subglottis (3 of 17 or 18%). As might be expected, the clinical implications of laryngeal KS in the AIDS patient depend greatly on the location of the lesion, associated risks, and the severity of airway compromise. In our case, the location and extent of the tumor were key and ultimately dictated the patient’s clinical course. The original decision to perform this procedure under general anesthesia after securing the airway by direct laryngoscopy instead of conscious sedation (i.e., monitored anesthesia care) with an unsecured airway was based on several factors: the patient’s existing respiratory compromise (i.e., pneumonia with oxygen requirement); the inability to perform direct laryngoscopy once the neurosurgical stereotactic frame is in place; and the patient’s relative lack of signs or symptoms to suggest significant upper airway obstruction, as determined by both otolaryngology and anesthesia personnel on separate occasions. The last of these factors proved to be a surprising and rather unique feature of this case. Life-threatening hemorrhage secondary to manipulation or biopsy is a significant risk and well documented cause of morbidity and mortality in patients with laryngeal KS, particularly those in acute respiratory distress. Greenberg et al. (8) and Roy et al. (9) report cases in which an AIDS patient with laryngeal KS and acute respiratory distress underwent emergent tracheostomy that resulted in uncontrolled hemorrhage, cardiorespiratory arrest, and death. Mochloulis et al. (11) and Pitchenick et al. (14) report brisk hemorrhage after simple biopsy of suspected laryngeal and tracheobronchial KS in patients who, as a result, suffered increased morbidity but not death. Again, in our case, the decision to forego tracheal intubation and abort the operation altogether was based on concerns about airway hemorrhage and the introduction of tumor into the trachea. Because obstructive KS lesions of the larynx in AIDS patients are rare, experience in how best to manage such cases clinically is understandably lacking. The world’s literature is limited to a small number of case reports and a single retrospective study of 17 such patients. Still, a number of conclusions seem self-evident. The preanesthetic evaluation of any AIDS patient with cutaneous KS should include a detailed historical and physical assessment of the airway. Any AIDS patient who complains of sore throat, cough, dysphagia, hoarseness, expectoration of blood or small tissue fragments, or other airway-related symptoms should be evaluated for the possibility of KS in the airway. In addition to a routine airway examination, the nasopharynx and oral cavity should be carefully inspected for evidence of KS. Any historical or physical evidence of oropharyngeal KS is an indication for otolaryngologic consultation to better define the airway management implications of the lesion. If KS in the airway is confirmed, the severity of the airway compromise and the appropriateness of various therapeutic interventions must be considered. Indirect or fiberoptic laryngoscopy, preferably in a conscious patient, and/or radiologic studies (CT, MRI) are indicated in the preoperative setting to define the anatomy accurately (and would have been especially helpful in the case described). In cases of impending airway obstruction, the patient will typically present with marked respiratory distress as evidenced by stridor, tachypnea, cyanosis, anxiety, sternal retractions, diaphoresis, and tachycardia. Once an obstructive KS lesion of the larynx has been established and endotracheal intubation by simple direct laryngoscopy is eliminated as a safe means of securing the airway, the literature speculates that cricothyroidotomy may be preferable to tracheostomy because of the latter’s association with significant bleeding in this setting (9). Unfortunately, experience with cricothyroidotomy in this setting is also lacking, and successful tracheostomy has been reported (11), suggesting that the exact location of the lesion, its proximity to laryngeal and tracheal structures, and the possibility of multiple lesions are complicating factors to consider. Here again, the appropriate radiologic study in conjunction with fiberoptic laryngoscopy is indicated to help delineate options for securing the airway. As for clinical management of the laryngeal KS lesion itself, once any life-threatening airway compromise has been addressed, there are a number of options: Conservative treatment would include radiation and chemotherapy, either separate or in combination, and has met with reasonable success for tumor regression and palliation of symptoms (11). More aggressive measures include tracheotomy, surgical or laser excision, cryotherapy, and photodynamic therapy. Another interesting nuance of this case relates to the relatively prolonged response to rocuronium. The patient did not have sufficient spontaneous recovery to permit pharmacologic reversal of his neuromuscular blockade until approximately 90 minutes after the rocuronium administration. This phenomenon has been reported previously in association with vecuronium (15). Although the etiology of the prolonged response is unclear, it may be related to an AIDS-associated myopathy or peripheral neuropathy (16,17). In summary, we report a case in which supraglottic KS in a patient with AIDS prevented safe endotracheal intubation by simple direct laryngoscopy before a brain biopsy procedure. The extent of the KS airway obstruction was not apparent preoperatively despite otolaryngologic consultation. This is the first report of an AIDS-associated difficult airway in the anesthesia literature. Although this lesion is undoubtedly rare and likely to become even more so with advances in and the increasing availability of HIV and AIDS therapies, AIDS patients presenting for anesthesia and surgery who are known to have oropharyngeal KS or any symptoms suggestive of possible airway compromise should be carefully screened for occult, near obstructing glottic or supraglottic lesions.