Given reduced rates of laryngospasm, sore throat, hoarseness, and coughing,1–3 supraglottic airway devices (SADs) have replaced endotracheal tubes (ETTs) for numerous indications, yet endotracheal intubation has firmly remained the standard of anesthetic care in tracheal resection and reconstruction.4 Regarding the challenges of airway management in tracheal surgery, however, SADs may be beneficial, especially in patients with severe subglottic stenosis.5 After induction of general anesthesia, SAD insertion obviates the need for endotracheal intubation, which may be impossible in cases of extreme airway narrowing. Furthermore, endotracheal intubation could result in mechanical trauma, swelling, bleeding, tumor fragmentation, and subsequent airway occlusion. Where the stenosis is in the subglottic area, ETT placement and cuff inflation above the level of stenosis may be difficult and might be associated with a high risk for dislocation. Once the anastomosis is completed, SADs do not cause mechanical stress to the freshly sutured airway and will not impair its blood supply. In addition, the risk for coughing at anesthesia emergence, another potential hazard to the anastomosis, is decreased. Furthermore, laryngeal nerve function can be easily evaluated by inserting a flexible bronchoscope over the SAD for direct visualization of the vocal cords and the larynx in spontaneously breathing patients.
The actual use of SADs for tracheal resection has been documented in 5 isolated case reports.6–10 Clearly, no conclusion regarding safety, efficacy, and ease of use can be drawn from isolated reports. In our institution, SADs have become an established alternative airway technique for cervical tracheal resection. This report reviews our experience with SADs for this application over a 5-year period since our first use. Its purpose was to document the feasibility of this new concept of airway management in tracheal surgery.
The study was performed in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement. After ethical approval by our institutional review board, all cervical tracheal resections, including cricotracheal resections, performed in our university hospital from 2010 to 2015 were screened retrospectively for airway management techniques. The requirement for written informed consent was waived by the institutional review board.
We included all adult patients (≥18 years of age) managed with SADs after induction of general anesthesia, regardless whether SAD insertion was planned as the primary/elective method for airway management, or if they were used as a rescue device after endotracheal intubation failed. Patients in which SADs failed to provide adequate ventilation would also be included (intention-to-treat analysis). The use of SADs for tracheal resection and reconstruction was considered feasible/successful if positive pressure ventilation (PPV) was unimpaired (median peak airway pressures ≤20 cm H2O and isolated maximum peak airway pressures of ≤25 cm H2O) and the peripheral oxygen saturation remained ≥95% with an end-tidal carbon dioxide level of ≤50 mm Hg. SADs failed if the device was exchanged for an ETT due to ventilation difficulties.
Medical records were obtained from our institution’s electronic database and reviewed for demographic parameters, comorbidities, periprocedural specifics, and postoperative outcomes. We analyzed preoperative diagnostic procedures (computed tomography, bronchoscopy, blood gas analysis, lung function testing), perioperative anesthetic, and surgical records, as well as the postoperative course documented by the intensive/intermediate care unit and general wards. The retrospective observation period ended with hospital discharge.
We used IBM SPSS Statistics 24 (IBM Corp, Armonk, NY) for descriptive statistical analysis. Results are presented as median and range (minimum–maximum) for continuous variables. Furthermore, both absolute (n) and relative (%) values are provided.
Regarding the total number of cervical tracheal resections performed (n = 81) during the 6-year screening period (2010–2015), SADs were used in 12% (n = 10) of all patients (Figure 1).
The first use of a SAD for cervical tracheal resection was an emergency situation in 2010, when it successfully served as a rescue device after multiple failed attempts at endotracheal intubation. The remaining cases (n = 9; 90%), however, were scheduled for elective SAD use. After extensive preparations and team training, this was done with increasing frequency from 2013 (5.6% of cases) to 2015 (46.1% of cases) (Figure 1). LMA® masks were the only SADs used in this study. All patients that met the inclusion criteria were included in this study and completed the retrospective observation period until hospital discharge.
Patients and Methods
Patients managed with SADs were mostly middle-aged women with normal body weight and advanced systemic diseases (Table 1). Tracheal stenoses were sequelae of prolonged mechanical ventilation (n = 6; 60%) or of idiopathic origin (n = 4; 40%). The stenoses were located a median distance of 20.0 mm (10.0–40.0 mm) from the vocal chords and had a median length of 15.0 mm (10.0–33.0 mm). The median residual airway lumen of the stenoses was 5.0 mm (4.0–7.0 mm) or 20.0% (10.0%–35.0%) of the normal tracheal diameter. Four patients (40%) underwent interventional treatments (eg, dilation, laser resection) before presenting for surgery. Preoperative medical history and physical examination revealed a high prevalence of severely compromised respiratory systems. Stridor at rest was present in almost all patients (n = 8; 80%). Multiple patients (n = 4; 40%) showed signs of respiratory insufficiency and required preoperative oxygen therapy. Preexisting chronic pulmonary disease, such as chronic obstructive pulmonary disease, was noted in 3 cases (30%). There was 1 anticipated difficult airway, caused by advanced ankylosing spondylitis. Assessment of American Society of Anesthesiologists and revised cardiac risk index scores indicated the presence of advanced disease of multiple organ systems in 7 patients (70%).
Standard preoperative evaluations included arterial blood gas analysis, lung function testing, and bronchoscopy. Although these tests were performed in all patients, either by the referring medical center or by our institution, 5 examination reports (blood gas analysis: 1 report missing; bronchoscopy: 1 report missing; lung function testing: 3 reports missing) were missing from our electronic patient records and paper chart archive and could not be reproduced by the referring hospitals. Preoperative testing demonstrated that most patients had extensive, scarred airway stenosis shortly distal to the vocal cords that severely impaired transstenotic air flow and respiratory volume (Figure 2).
Several patients (n = 6; 60%) received oral midazolam (3.75–7.5mg) for preoperative anxiolysis. After establishment of standard anesthesia monitoring (electrocardiography, oscillatory blood pressure, peripheral oxygen saturation) and preoxygenation, general anesthesia was induced intravenously using sufentanil (median dose: 0.33 µg/kg [0.21–0.58]) and propofol (median dose: 1.9 mg/kg [0.5–3.1]). The majority (n = 9; 90%) received muscle relaxants, such as rocuronium (n = 7; 70%), succinylcholine (n = 1; 10%), or mivacurium (n = 1; 10%). Repetitive paralytic doses were given in 7 of 9 patients (78%). Invasive hemodynamic monitoring (radial arterial line) was instituted in 6 patients (60%). SADs were primarily selected for airway management in 9 cases (90%). One patient was initially scheduled for conventional endotracheal intubation, however, it was not possible to advance a 5.0-mm (internal diameter) cuffed ETT across a stenotic segment with a residual tracheal lumen of 6.0 mm located 18 mm below the vocal cords. After multiple attempts with differently sized ETTs, a SAD was inserted without any difficulties. There were no other complications related to airway management after induction of anesthesia. There were no cases of SAD failure.
During the procedure, PPV was not impaired by increased transstenotic airway resistance or surgical manipulation, although 1 American Society of Anesthesiologists physical status IV patient with preoperative respiratory failure and end-stage COPD had persistent hypercarbia during PPV. Oxygenation was unimpaired at all times (Table 2). Once surgical preparation and mobilization of the cervical trachea were completed, all patients were preoxygenated with an fraction of inspired oxygen of 1.0, as a short period of apnea was required during the excision of the stenotic segment. After resection of the stenosis, the surgeon carefully inserted the proximal tip of a jet ventilation catheter (11 or 14 Fr Cook Airway Exchange Catheter; Cook Medical, Bloomington, IL) retrogradely into the incised trachea and advanced it without resistance toward the larynx, until it was accessed either inside the SAD lumen or in the oropharynx after removal of the SAD, for connection to the ventilator. Afterwards, the distal tip of the jet catheter was placed in the distal trachea by the surgeon. In 1 instance, retrograde catheter insertion failed, most likely due to catheter impingement on the vocal cords. To minimize the risk for vocal cord trauma, the SAD was exchanged for an ETT to facilitate anterograde jet catheter placement. The driving pressure of high frequency jet ventilation was set to 1.5 bar with a median frequency of 225/min (200–250) and an fraction of inspired oxygen of 1.0. Except for acute hypercarbia (median peak Pco2 = 98.0 mm Hg [70.0–116.0], lowest pH = 7.21 [7.09–7.31]) detected in 3 of 6 patients with arterial lines placed, there were no complications observed resulting from HFJV. During HFJV, the lowest peripheral oxygen saturation recorded was 91%. Whenever necessary and always before completion of the anastomosis, the distal airways were suctioned by the surgeon to reduce the risk for aspiration of blood or secretions.
Once the tracheal walls were reconnected, the jet catheter was removed by the anesthesiologist and PPV was resumed via SAD in 9 patients (90%), except for the 1 patient who was intubated endotracheally to facilitate jet catheter placement. In this case, the ETT was left in situ for the remainder of the procedure. Before extubation, a bronchoscope was inserted through the SAD orifice to check recurrent laryngeal nerve function, evaluate the anastomosis, and suction the distal airways if necessary. There were no cases of vocal cord paralysis or unsatisfactory surgical results. All patients were extubated before leaving the operating theatre. There were no anesthesia-related complications. Specifics on ventilator settings, timing, and other anesthetic interventions are outlined in Table 2.
Surgically, all tracheal resections were performed via cervical incisions. Sternotomies were not necessary. The cricoid cartilage was part of the resected tissue in 70% of cases (n = 7) (ie, cricotracheal resections). The median length of the resected tracheal segment was 31.0 mm (21.0–70.0). Except for 2 cases of minor bleeding (median blood loss: 300.0 mL [300.0–500.0]), there were no intraoperative surgical complications.
Postoperatively all patients were breathing spontaneously with minor supplemental oxygen requirements. Most were transferred to the intensive care unit (ICU) for postoperative monitoring (n = 9; 90%). One patient (10%) was monitored on an intermediate care unit.
The majority (n = 6; 60.0%) had uneventful postoperative courses. In these patients, the median ICU length of stay was 1.0 days (0.0–1.0). After bronchoscopic confirmation of regular anastomotic healing, they were discharged from hospital 7.5 days (4.0–19.0) after the operation.
Postoperative complications occurred in 4 patients (40%) and led to prolonged lengths of stay in the ICU (median days: 5.5 [2.0–15.0]) and in the hospital (median days: 17.0 [5.0–30.0]). In 1 instance, a COPD patient required short-term noninvasive ventilatory support due to hypercarbia. Another patient required a surgical revision after postoperative hemorrhage and developed refractory glottis edema, which finally led to a tracheotomy. Despite maximum therapy (including extracorporeal membrane oxygenation), there were 2 in-hospital deaths resulting from severe pneumonia occurring in 2 patients with preexisting advanced pulmonary and vascular disease.
We examined the feasibility of using SADs (LMA® masks) for elective and emergency airway management in 10 adult patients scheduled for cervical tracheal resection and reconstruction. There were no intra- or postoperative complications related to anesthetic management, apart from transient hypercarbia during jet ventilation. Regarding the prevalence of advanced systemic diseases, such as COPD with signs of preoperative respiratory failure, coronary and peripheral artery disease, these procedures were performed in patients at high risk for postoperative complications. Even though the majority (n = 6; 60%) had uneventful postoperative courses, 2 patients (20%) developed pneumonia and died during their hospital stay. There were no indicators, however, that these deaths were in any way associated with anesthetic airway management. Both patients were extubated at the end of surgery and were breathing spontaneously when they were transferred to the ICU.
Currently, there are 5 isolated case reports of previous SAD use for cervical tracheal resections in patients with severe subglottic stenosis.6–10 In contrast to choosing HFJV during the phase of resection and reconstruction, as we did exclusively in our case series, several authors used cross-field intubation of the distal tracheal stump and subsequent PPV.8–10 While this is a safe and well-established technique in tracheal surgery,4 cross-field intubation may significantly obstruct surgical access to the posterior tracheal wall and may impair suturing conditions. There are no data, however, comparing HFJV to cross-field intubation with regards to postoperative surgical results (eg, quality/healing of anastomosis). Hypercarbia is a common sequel of HFJV and was noted in 3 of our patients (30.0%). However, concerns of negative hemodynamic effects of rapidly induced (permissive) hypercarbia, such as impaired cardiac contractility or increased arrhythmic potential, have mostly been discarded.11 None of the patients in our series exhibited any signs of cardiovascular instability apart from minor vasopressor requirements, despite peak CO2 levels of up to 116 mm Hg.
While most authors of previous reports chose conventional ETT for the last phase of the operation, to us, SADs may be the preferable choice, as they do not cause any mechanical stress to the anastomosis or its blood supply and are associated with reduced rates of coughing or airway irritation at the emergence from anesthesia.3 Before extubation, a bronchoscope can be inserted over the SAD to visualize the vocal cords and to evaluate laryngeal nerve function in spontaneously breathing patients.
Looking solely at previously published data, it was not possible to reliably assess the safety of SADs for airway management in cervical tracheal resection and reconstruction, especially with regards to patient selection and the risk for ventilation difficulties. There are reports of fatalities after ventilation across SADs failed in patients with tracheal stenosis during resuscitations, because of increased airway resistance.12 Although this could be interpreted as a potential hazard of SAD use in patients with airway stenosis, detailed assessment of the peak airway pressures applied in our case series and previous reports demonstrate that there are no indicators that PPV via SAD is significantly impaired in the setting of elective surgery, even in cases of severe airway stenosis. Nevertheless, rescue airway techniques must be available in case of SAD failure. Conventional endotracheal intubation with a small-caliber ETT is probably the least traumatic option in patients with soft and/or less extensive tracheal stenosis. In patients with scarred and severe airway narrowing, endotracheal intubation is likely to fail. Hence, the surgical staff should be present at anesthesia induction and be prepared to perform an emergency surgical airway in case of a “cannot ventilate, cannot intubate” scenario.
“Awake” SAD insertion after topicalization of the oropharynx with local anesthetics under minor sedation could be a suitable approach to evaluate the impact of the airway stenosis on PPV. Should PPV not be possible, the patient could easily return to spontaneous breathing.
There are potential weaknesses to our study. First, its retrospective nature does not allow an unrestricted evaluation of potential benefits of SAD use in tracheal surgery, as rates of coughing at anesthesia emergence, for example, are not usually documented. There were 5 preoperative examination reports (blood gas analysis, lung function testing, bronchoscopy) missing, which were performed in referring medical centers and could not reproduced. The impact of these missing data on our results is expected to be minimal. Although this is a retrospective analysis, we applied a standardized set of criteria identifying suitable patients for SAD use during the study period. The indications and contraindications for SAD use were based on our department’s procedural guidelines in tracheal surgery (Table 3).
As in any retrospective study, however, the existence of unidentified confounders regarding patient selection cannot be excluded. Although this is the largest case series on SAD use in tracheal resections, a sample of 10 patients will not allow a comprehensive safety evaluation, especially with regards to rare periprocedural complications (eg, aspiration risk). Large patient cohorts, however, are difficult to evaluate, as cervical tracheal resections are relatively rare, even in specialized centers.4 An international registry database could be a potential option to evaluate the risks and benefits in a large multicenter cohort.
This trial focused mostly on high-risk patients, considered unsuitable for endotracheal intubation. The benefits of SADs, however, are expected to apply equally to most patients undergoing cervical tracheal resection. Currently, we are conducting a prospective observational trial on SAD use in all patients scheduled for cervical tracheal resections, with a specific focus on ventilation mechanics, intraoperative gas exchange, and potential risks and benefits (WHO International Clinical Trials Registry Platform/German Clinical Trials Register ID: DRKS00010826).
This retrospective case series demonstrates the feasibility of using supraglottic airways alongside high-frequency jet ventilation for airway management in 10 cases of cervical tracheal resection and reconstruction. However, given the small number of cases examined limitations apply regarding indications, contraindications, and periprocedural safety.
Name: Mark Schieren, MD.
Contribution: This author helped with study design, data acquisition, data analysis and interpretation, drafting of manuscript, and final approval of the version to be published.
Name: Enikö Egyed, MD.
Contribution: This author helped with data interpretation, critical manuscript revision, and final approval of the version to be published.
Name: Burkhard Hartmann, MD.
Contribution: This author helped with data interpretation, critical manuscript revision, and final approval of the version to be published.
Name: Armen Aleksanyan, MD.
Contribution: This author helped with data acquisition, critical manuscript revision, and final approval of the version to be published.
Name: Erich Stoelben, PhD, MD.
Contribution: This author helped with critical manuscript revision and final approval of the version to be published.
Name: Frank Wappler, PhD, MD.
Contribution: This author helped with drafting the manuscript, critical manuscript revision, and final approval of the version to be published.
Name: Jerome M. Defosse, MD.
Contribution: This author helped with study design, data acquisition and interpretation, drafting the manuscript, and final approval of the version to be published.
This manuscript was handled by: David Hillman, MD.