The Use of Heliox and the Laryngeal Mask Airway in a Child with an Anterior Mediastinal Mass

The patient with an anterior mediastinal mass and respiratory distress presents one of the most difficult problems in anesthetic management. When the mass is a malignancy, patients require an accurate tissue diagnosis to determine the optimal therapy, which may limit the preoperative administration of chemotherapy, steroids, or radiotherapy. Nonmalignant masses, such as teratomas and cystic hygromas, are not amenable to such therapy. Numerous reports in the anesthesia literature have described both the pitfalls and the optimal management of these patients. We report the use of heliox, a gas containing 20% oxygen and 80% helium, and the laryngeal mask airway (LMA) for the administration of anesthesia to a 3-yr-old boy with asthma and a large anterior mediastinal mass. Case Report A 16-kg, 3 10/12-yr-old boy presented to the emergency room in acute severe respiratory distress. He had a 3-yr history of asthma, treated with daily nebulized albuterol, cromalyn sodium, and intermittent steroids. On examination, he was sitting upright with biphasic stridor. Breath sounds were nearly inaudible over the entire left chest. He had no plethora of the head and neck, suggesting the absence of superior vena cava syndrome, and had a palpable right supraclavicular lymph node. A chest radiograph showed a massive mediastinal mass Figure 1, and computed tomography demonstrated compression of the carina and left mainstem bronchus Figure 2. He was given heliox and oxygen (FIO2 = 0.30) and responded with both symptomatic improvement and an increase of his oxygen saturation from 76% to 96%. Because of his history of steroid-dependent asthma, and because of the significant symptoms from the mediastinal mass, he was given 1 mg/kg of intravenous prednisone.Figure 1: Anteroposterior chest radiograph demonstrating the large mediastinal mass.Figure 2: Computed tomography scan of the thorax, just inferior to the level of the carina. There is compression of both mainstem bronchi, with complete obliteration of the lumen of the left and partial occlusion of the right.After premedication with a total of 1.5 mg of midazolam (titrated in 0.25- to 0.5-mg increments to achieve a calm yet awake state), he was taken to the operating room where anesthesia was induced in a sitting position with slowly increasing concentrations of halothane in heliox. Spontaneous ventilation was sustained throughout the anesthesia, and 3-5 cm H2 O of continuous positive end-expiratory pressure (CPAP) was administered by adjusting the pressure relief valve and maintaining a tight mask seal. At an end-expired concentration of 2%, a No. 2 LMA was inserted, and 5 cm H2 O of CPAP continued. The anesthetic was maintained with halothane (0.7%-1.0% end-expired) in heliox, during which time all vital signs remained stable. Lidocaine 1% was infiltrated in the operative field and the supraclavicular node was excised. During the wait for the biopsy results, an implantable central venous port was placed through a cutdown of the external jugular vein that had been exposed in the same operative field. The patient was placed on his left side without alteration in ventilation, and a lumbar puncture and bone marrow aspirate and biopsy were performed. At the end of the procedures he was awakened, the LMA was removed, CPAP was discontinued, and he was returned to the intensive care unit breathing heliox, still in a semisitting position. No bronchospasm or other respiratory or cardiac complications were observed during the anesthesia or emergence. His respiratory symptoms diminished rapidly over the next 48 h after beginning chemotherapy. The tissue diagnosis was acute lymphoblastic leukemia. Discussion Our patient presented with significant symptoms from his large anterior mediastinal mass that put him at considerable risk of developing life-threatening complications during anesthetic induction, maintenance, and emergence. His history of asthma further increased these risks. (In retrospect, we speculate that some of the reduced wheezing after steroid therapy in the months prior to diagnosis was perhaps related to shrinkage of the yet-to-be-discovered mass.) Because of these two problems, we wanted to maintain spontaneous ventilation and avoid the use of muscle relaxants, maintain as light a plane of anesthesia as possible, and minimize stimulation of the airways, especially below the glottis, to avoid precipitating bronchospasm. In addition, we sought to decrease the resistance to airflow that was severely limited through a critically narrowed subglottic airway. Two major complications may develop in patients with large anterior mediastinal masses: inability to ventilate, often despite the presence of an endotracheal tube, and circulatory collapse. Respiratory failure is usually thought to be secondary to airway compression [1,2] and appears to occur most often as tracheal and chest wall tone diminishes with the onset of neuromuscular block. Most authors conclude that the maintenance of spontaneous ventilation and avoidance of muscle relaxants appear to be crucial in lessening the risk of intractable respiratory collapse [1-3]. Cardiovascular complications are thought to be secondary to impairment of venous return, right heart output [4], or direct cardiac compression and tamponade [5,6]. Maintaining a semisitting posture appears to decrease the compression of the heart and great vessels [4,5]. Ferrari and Bedford [7] concluded that, in the absence of life-threatening symptoms, general anesthesia could be safely performed in patients without prior radiation or chemotherapy to shrink the mass's size. Despite the successful outcomes using optimal management strategies, these authors described potentially fatal complications in 44% of the nine patients in their series who presented with preoperative symptoms and were unable to extubate one third of their patients at the conclusion of surgery because of airway insufficiency. Although the use of heliox in other disease states with impairment of airflow has been previously described [8-11], none of the literature on the anesthetic management of anterior mediastinal masses mentions this potentially beneficial technique. Heliox is less than one third the density of oxygen and almost one fifth that of 30% oxygen/70% nitrous oxide, permitting more laminar gas flow and decreasing the resistance in the conducting airways [12]. The physics of laminar gas flow through a tube is described by the Hagen-Poiseuille equation: Equation 1 where L is the length of the tube, r is its radius, and mu is the viscosity of the gas. During laminar flow, which normally occurs in the large conducting airways, the fourth power of the radius and length are the primary determinants of resistance, since the viscosities of air, oxygen, and heliox are nearly the same. However, when an obstruction occurs in these larger airways, where the linear velocity of gas is high, turbulence develops. It is this turbulent flow through the obstruction, as well as the decrease in the cross-sectional area of the conducting passage, that results in the greatly increased work of breathing. Where flow is turbulent, the driving pressure required becomes proportional to the density of the gas and inversely proportional to the fifth power of the radius, and resistance is less well described by the Hagen-Poiseuille equation. The characteristics of turbulent flow were described more than 100 yr ago by Reynolds, who showed that turbulence increases as the density, linear velocity of flow, and tube diameter increase. The Reynolds number (Re) predicts whether flow will be linear or turbulent: Equation 2 where rho is the density, U is the velocity, d is the diameter of the tube, and mu is the viscosity. When Re in a straight, unbranched tube is greater than 1500-2000, turbulent flow occurs. In a branching tube, or, as in our case, around an obstruction, turbulent flow occurs above a even smaller Re [13]. It is primarily through an obstructed region where both turbulence and gas velocity are high that heliox, with its lower density, can be effective in decreasing Re, resulting in more linear gas flow, and decreasing resistance to flow [12]. Note that were the density of heliox to be substituted for that of oxygen in the Equation andall other values held unchanged, the magnitude of Re would decrease by 60%. Heliox is stored in tanks with the concentration of helium varying from 20% to 30%, depending on the supplier. The gas can be attached to the anesthesia machine's air inlet with a standard threaded connector, but the rotameter will read inaccurately, and one must read the actual FIO2 with an oxygen analyzer [14]. One then adjusts the oxygen and air rotameters to deliver the desired heliox concentration. Because the beneficial clinical effects from decreased density are directly related to the amount of helium delivered, the lowest FIO2 (and conversely, the highest concentration of helium) that will permit adequate oxygen saturation should be administered. The LMA was chosen because a conventional mask would have been awkward to use so close to the operative field. It allowed the maintenance of a relatively secure airway with spontaneous ventilation and a light plane of anesthesia and minimized lower airway irritation that could have been precipitated by endotracheal intubation, thus decreasing the risk of bronchospasm in a susceptible patient. Should we not have been able to insert the LMA, anesthesia by conventional face mask would have been attempted in preference to an endotracheal tube, because of (a) the ability to provide an anesthetic with less depth than needed for the insertion of an endotracheal tube and (b) the patient's history of asthma, which also had the potential to require greater anesthetic depth to blunt any bronchoconstriction that may have been precipitated in response to the tracheal stimulation of intubation. We believe that the combination of the management strategies of sitting position, spontaneous ventilation with inhalation induction, minimization of airway stimulation by use of the LMA, and use of heliox contributed to the successful outcome in this patient. One must, of course, always be prepared to intervene with either manipulations of patient position in the event of airway compromise (including upright, lateral, and prone) or more aggressive strategies, such as rigid bronchoscopy and even median sternotomy (in the case of intractable cardiovascular collapse), or to allow the patient to awaken if critical airway or cardiovascular compromise becomes evident at any time during the course of the anesthesia. The author wishes to thank Michael Sborov, MD, Chief of Anesthesia, for his helpful review and comments; Ted Carter, MD, Pediatric Pulmonology, for assistance with heliox administration; and Randall Holland, MD, Chief of Pediatric Surgery, and Bonnie Heyer, MD, staff anesthesiologist, all of Madigan Army Medical Center, for their assistance in the clinical management of the patient.