Introduction
Acquired laryngotracheal stenosis (ALTS) is the pathological narrowing of the upper airway that may develop often after lifesaving airway interventions. It occurs in approximately 1–10 percent of patients who are intubated or undergo tracheotomy.Reference Halum, Ting, Plowman, Belafsky, Harbarger and Postma1–Reference Wain3 The biggest risk factors for its development are prolonged intubation, large endotracheal tube (ETT) size, and elevated cuff pressure. There may also be a genetic and racial predilection.Reference Anis, Zhao, Khurana, Krynetskiy and Soliman4, Reference Anis, Krynetskaia, Zhao, Krynetskiy and Soliman5 Both endotracheal intubation and tracheostomy can cause direct mechanical injury to the epithelium of the upper airway, resulting in ulceration, localized bacterial infection and subsequent inflammation of the underlying cartilage. Over time, aberrant wound healing mechanisms lead to hypertrophic scar formation which contract and result in airway stenosis.Reference Singh, Sandulache, Otteson, Barsic, Klein and Dohar6–Reference Papla, Dyduch, Frasik and Olechnowicz10
Specific medical comorbidities have been found to increase the risk of developing ALTS due to their effect on wound healing.Reference Plocienniczak, Sambhu, Noordzij and Tracy11, Reference Johnson, Bradshaw, Jaffal and Chorney12 Hillel et al. found that males with diabetes mellitus were associated with 888 percent increased odds of developing posterior glottic stenosis, although the study size was relatively small.Reference Hillel, Karatayli-Ozgursoy, Samad, Best, Pandian and Giraldez13 Higher body mass index (BMI) has also been associated with an increased risk of post-intubation tracheal stenosis.Reference Songu and Ozkul14–Reference Nicolli, Carey, Farquhar, Haft, Alfonso and Mirza16 Gastroesophageal reflux disease (GERD) additionally has been shown to be associated with tracheal stenosis as well as other respiratory diseases.Reference Cardoso, Minamoto, Bibas and Pego-Fernandes17, Reference Pappal, Burruss, Witt, Harryman, Ali and Bush18
Cardiovascular disease is a common medical condition that affects approximately 20.1 million adults and is the leading cause of death in the United States.Reference Virani, Newby, Arnold, Bittner, Brewer and Demeter19 It accounts for a significant percentage of hospitalizations and up to a third of ICU admissions.Reference Salah, Minhas, Khan, Khan, Ambrosy and Blumer20, Reference Illescas, Kache, Whittemore, Lucero, Quinn and Daskalakis21 Additionally, 7 percent to 20 percent of those with severe disease require endotracheal intubation, and approximately 6 percent require tracheotomy.Reference Di Marco, Tresoldi, Maggiolini, Bozzano, Bellani and Pesenti22, Reference Grammatico, Banna, Shahu, Gastanadui, Jimenez and Heck23 However, the risk of developing ALTS in this patient population has not been well studied. So et al. noted that patients with cardiovascular disease were more likely to develop posterior glottic injury after intubation for coronavirus disease 2019 (COVID-19).Reference So, Kavookjian, Ben Or, Snow, Akst and Hillel24 The senior author (A.S.) has also observed many patients presenting to his practice with ALTS after brief intubations for myocardial infarction. We sought to determine the incidence of ALTS in patients intubated for acute cardiac events and to characterize how these patients differ from those intubated for acute non-cardiac events.
Methods
Study design
After the study protocol was approved by the Temple University Institutional Review Board (Protocol Number 31717), we retrospectively reviewed all patients who presented to Temple University Health System (TUHS) hospitals between 1 January 2013 and 31 December 2023. The STROBE checklist was adhered to in reporting.
Patients
All patients greater than 18 years of age who were diagnosed with laryngotracheal stenosis (International Classification of Diseases 10th Revision [ICD-10] codes J38.6, J39.8, J95.5 and Ninth Revision [ICD-9] codes 478.74 and 519) after intubation for an acute cardiac event (ICD-10 codes I21, I22, I24, I46, I50.21, I50.23, I50.31, I50.33, I50.41, I50.811, I71.0, J81.0 and ICD-9 codes 410, 411, 427.5, 428.21, 428.23, 428.31, 428.33, 428.41, 441.03, 514) were included as the study group (Table 1). Similarly, all patients who developed ALTS after intubation for non-cardiac events during the study period served as the control group, which was derived using the same query without the cardiac ICD codes. At our institution, the diagnosis of ALTS was made in symptomatic patients by an otolaryngology attending after evaluation with laryngoscopy, bronchoscopy, or computed tomographic imaging (Figures 1 and 2). Non-cardiac events requiring intubation ranged from pulmonary embolism to trauma to stab wounds. Pregnant women, prisoners, and veterans were excluded, as were patients who underwent emergent cricothyroidotomy. The total number of adult patients intubated for acute cardiac or non-cardiac events during the study period were then each determined. Only intubations for acute events/illnesses were collected, not elective intubations for surgical procedures. The control group excluded any patient with ICD documentation of cardiac arrest, acute heart failure, myocardial infarction, or other cardiac events (Table 1).
Subglottic stenosis before and after endoscopic treatment.
Representative computed tomography neck images of tracheal stenosis. (A,B) Axial slices. (C) Sagittal section. (D) Coronal section.
Acute cardiac events included in the study group
Table 1 Long description
Acute cardiac event diagnoses are mapped to their corresponding ICD-10 and ICD-9 billing codes. Myocardial infarction categories include acute MI coded as I21 in ICD-10 and 410 and 411 in ICD-9, and STEMI and NSTEMI MI coded as I22 with no ICD-9 code listed. Other acute ischemic heart disease is coded as I24 in ICD-10 and 411 in ICD-9. Cardiac arrest is coded as I46 in ICD-10 and 427.5 in ICD-9. Several acute heart failure subtypes are listed with paired codes: acute systolic I50.21 and 428.21, acute on chronic systolic I50.23 and 428.23, acute diastolic I50.31 and 428.31, acute on chronic diastolic I50.33 and 428.33, and acute combined systolic and diastolic I50.41 and 428.41. Acute right heart failure appears only with ICD-10 code I50.811 and no ICD-9 code. Aortic dissection is coded as I71.0 in ICD-10 and 441.03 in ICD-9, and flash pulmonary edema is coded as J81.0 in ICD-10 and 514 in ICD-9. Dashes indicate diagnoses where an ICD-9 code is not provided in the table.
ICD = International Classification of Diseases; MI = myocardial infarction.
Statistical analyses
Patient demographics, comorbid medical conditions, medications given, length of intubation, endotracheal tube size, laryngoscopic findings, airway procedures performed, and length of hospitalization were collected. Post-hospitalization airway evaluations and subsequent procedures were also recorded. The distribution of continuous variables was examined by histogram. Normally distributed continuous variables were compared between groups using a two-tailed independent-samples t-test and reported as means (± SD). Non-normally distributed continuous variables were compared using the Wilcoxon rank-sum test and reported as medians (interquartile range [IQR]). The number of intubation attempts and ETT size are categorical variables and, therefore, were summarized as counts and proportions within each category and compared between groups using Fisher’s exact test. All other categorical variables were summarized as counts and percentages and compared using chi-square or Fisher’s exact tests, as appropriate. Statistical significance was set a priori at p less than 0.05. Statistical analyses were performed using SPSS (IBM, version 30, Armonk, NY).
Results
Baseline characteristics and incidence of ALTS
Of 3665 patients intubated for acute cardiac events during the study period, 10 (0.27 percent) developed ALTS. Four patients (40 percent) were male, 8 (80 percent) were African American, and 1 (10 percent) was Hispanic. The mean age was 55 (range, 34–72) years. Of these 3665 patients, 1630 (44.47 percent) died during their hospitalization. Of 1701 patients intubated for non-cardiac indications during the same period, 21 (1.23 percent) developed ALTS. Eleven (52.38 percent) were male, 10 (47.62 percent) were African American, four (19.05 percent) were Caucasian, and three (14.29 percent) were Hispanic. The mean age at admission was 51 (range, 23–70) years. Of these 1701 patients, 796 (46.79 percent) died during admission. All ALTS patients required ventilation with cuff inflation. No patients in either group had a documented prior history of airway stenosis.
Secondary characteristics
There was no difference between the study and control groups for age, sex, comorbid medical conditions, nicotine dependence, body mass index (BMI), or mortality, with 2035 (55.53 percent) and 905 (53.20 percent) patients, respectively, surviving long enough to undergo laryngeal assessment (Table 2). The absolute risk difference in incidence rates between ALTS patients intubated after non-cardiac and cardiac events was statistically significant (0.96 percent [1.23–0.27 percent], 95 percent confidence interval [CI] [0.43–1.56 percent]; p < 0.0001), corresponding to a relative risk of 4.53 (95 percent CI 2.14–9.59) for the non-cardiac group.
Demographics and medical comorbidities
Table 2 Long description
Comparison of demographics, body size, mortality, and selected comorbidities between ALTS patients intubated for a cardiac event (10 people) versus a non-cardiac indication (21 people), reporting p-values, effect sizes, and confidence intervals. Male sex was 40.00 percent in the cardiac group and 52.38 percent in the non-cardiac group, with overlapping confidence intervals. Mean age was 55.30 years in the cardiac group and 51.00 years in the non-cardiac group, and mean BMI was 25.14 versus 29.23, respectively. Diabetes occurred in 40.00 percent versus 47.62 percent; COPD in 30.00 percent versus 23.81 percent; and GERD in 20.00 percent versus 28.57 percent. Nicotine dependence was higher in the non-cardiac group (57.14 percent) than the cardiac group (30.00 percent), but the p-value still indicates no statistically significant difference. Alcohol-related disorders were absent in both groups. Overall, all reported p-values are above 0.10 and confidence intervals are wide, so differences should be interpreted cautiously given the small ALTS sample sizes.
1 3665 patients intubated for acute cardiac events.
2 1701 patients intubated for non-cardiac indications.
ALTS = acquired laryngotracheal stenosis; BMI = body mass index; CI = confidence interval; COPD = chronic obstructive pulmonary disease; GERD = gastroesophageal reflux disease.
Endotracheal tube size distribution, analyzed as a categorical variable by Fisher’s exact test, did differ significantly between groups (p = 0.023). In the cardiac event group, tube sizes were distributed as follows: 6.0 mm in one patient (10.0 percent), 7.0 mm in four (40.0 percent), 7.5 mm in four (40.0 percent) and 8.0 mm in one (10.0 percent). In the non-cardiac event group, no patient received a 6.0-mm tube; four patients received 7.0 mm (19.1 percent), seven received 7.5 mm (33.3 percent) and 10 received 8.0 mm (47.6 percent). The non-cardiac event group received statistically larger tubes (Table 3).
Airway management
Table 3 Long description
Airway management characteristics are compared between patients with acquired laryngotracheal stenosis who had a cardiac event (10 patients) versus a non-cardiac event (21 patients). Intubation attempts differed: 90 percent of the cardiac-event group were intubated in one attempt versus 66.7 percent in the non-cardiac group, while two attempts occurred in 0 percent versus 33.3 percent, and three attempts in 10 percent versus 0 percent; this comparison is statistically significant. Hospital stay medians were 20.5 days for the cardiac-event group and 16 days for the non-cardiac group, with no statistically significant difference. Length of intubation was similar, with medians of 3.5 versus 4 days and no statistically significant difference. Endotracheal tube size distribution also differed significantly: in the cardiac-event group, 7.0 mm and 7.5 mm were each 40 percent, with 6.0 mm and 8.0 mm each 10 percent; in the non-cardiac group, 8.0 mm was most common at 47.6 percent, followed by 7.5 mm at 33.3 percent and 7.0 mm at 19.1 percent, with no 6.0 mm tubes. P-values indicate statistical testing for group differences, but the small sample sizes mean results should be interpreted cautiously.
ALTS = acquired laryngotracheal stenosis; ETT = endotracheal tube; IQR = interquartile range.
† Fisher’s exact test.
‡ Wilcoxon rank-sum test.
The number of intubation attempts, analyzed as a categorical variable using Fisher’s exact test, differed significantly between groups (p = 0.04). A single intubation attempt was required in 90.0 percent of patients in the cardiac event group compared with 66.7 percent of the non-cardiac event group. No patients in the cardiac group required two attempts, whereas 33.3 percent in the non-cardiac group did. One patient (10.0 percent) in the cardiac group required three attempts; none in the non-cardiac group did (Table 3).
Length of hospital stay was numerically longer in the cardiac event group (median 20.5 days, IQR 14–70) compared with the non-cardiac event group (median 16 days, IQR 9–37), but this difference did not reach statistical significance (p = 0.35). Similarly, the duration of intubation was comparable between groups (median 3.5 vs. 4 days; IQR 2–10 vs. 2–13 days; p = 0.72) (Table 3).
Of patients intubated for acute cardiac events, six (60 percent) had posterior glottic stenosis, four (40 percent) had tracheal stenosis and four (40 percent) had subglottic stenosis. Four (40 percent) patients had multilevel stenosis. For the control group, eight (38.10 percent) patients had posterior glottic stenosis, nine (42.86 percent) patients had tracheal stenosis, eight (38.10 percent) patients had subglottic stenosis and one (4.76 percent) patient had supraglottic stenosis. Five (23.81 percent) patients had multilevel stenosis. There was no statistical difference in the types of stenoses seen or in the medications administered during hospitalization between the groups (Table 4).
Types of stenosis
Table 4 Long description
Comparison of stenosis patterns, number of operations, and decannulation outcomes between patients with acquired laryngotracheal stenosis who had a cardiac event (10 patients) versus a non-cardiac event (21 patients). Posterior glottic stenosis was more common in the cardiac-event group at 60 percent than in the non-cardiac group at 38.1 percent, but the p-value indicates no statistically significant difference. Tracheal stenosis occurred at similar rates, 40 percent with cardiac events and 42.86 percent with non-cardiac events, and subglottic stenosis was also similar at 40 percent versus 38.1 percent. Supraglottic stenosis was rare, with 0 percent in the cardiac-event group and 4.76 percent in the non-cardiac group. The average number of operations was higher in the cardiac-event group, 2.10 with a standard deviation of 2.13, compared with 1.57 with a standard deviation of 0.98, without statistical significance. Decannulation occurred less often with cardiac events, 30 percent versus 61.90 percent, and this difference trends toward but does not reach statistical significance. Confidence intervals are wide in several rows, especially in the smaller cardiac-event group, so estimates should be interpreted cautiously.
ALTS = acquired laryngotracheal stenosis; CI = confidence interval.
In the study and control groups, 519 (14.16 percent) and 332 (19.52 percent) patients, respectively, had a tracheostomy. There was no difference in tracheostomy size or time to decannulation. Five hundred sixty (15.28 percent) and 533 (31.33 percent), respectively, were never decannulated (Table 5).
Tracheostomy comparison
Table 5 Long description
Tracheostomy size and the length of time a tracheostomy was in place are compared for intubated patients who had a cardiac event versus a non-cardiac event. Average tracheostomy size was nearly identical between groups, about 6.9 in both, with overlapping confidence intervals and a non-significant p-value. The effect size for tracheostomy size was very small, suggesting little practical difference. Average tracheostomy duration was also similar, about 132 days for the cardiac-event group and about 119 days for the non-cardiac-event group. Confidence intervals for duration overlapped and the p-value was not significant, with a small effect size. The very large standard deviations for duration indicate wide variability, so group averages may not reflect typical individual experiences.
CI = confidence interval.
Discussion
ALTS is a well-recognized complication of endotracheal intubation, resulting from mucosal ischemia and subsequent fibrosis. Although cardiovascular disease is known to have systemic effects that may heighten airway vulnerability to stenosis, the risk of ALTS in patients specifically intubated for acute cardiac events has been poorly characterized. In this study, we sought to characterize ALTS occurring after intubation for acute cardiac events.
Endotracheal tube size is a known risk factor for ALTS as larger ETTs have a greater cross-sectional area, causing increased contact with the airway mucosa. A positive association between ETT size and posterior glottic and tracheal stenosis has also been observed, likely secondary to diminished mucosal perfusion leading to mucosal ischemia.Reference Schiff25 In our study, both ETT diameter and the overall incidence of ALTS were significantly lower in acute cardiac event patients. Although the lower incidence of ALTS in cardiac patients may be due to the smaller ETT size, given that the median difference in the size of the ETT was less than 0.5 mm (7.25 vs. 7.5 mm), there may be other factors at play. In our hospital system, 3665 cardiac event patients versus only 1701 non-cardiac event patients were intubated in the study time frame despite a similar number of admissions in each group. De Jong et al. found that patients who required rapid tracheal intubation after cardiac arrest had a significantly lower survival rate when compared to non-intubated patients, raising the possibility that some of the deceased patients may have had undiagnosed ALTS.Reference De Jong, Rolle, Molinari, Paugam-Burtz, Constantin and Lefrant26 In our study, however, there was no difference in mortality rate between the study and control groups. Despite the increased rate of intubation and similar mortality rate, acute cardiac patients were still less likely to develop ALTS, lending further support for possible additional causes besides the smaller ETT size, which warrants further study.
The number of intubation attempts also differed significantly between groups, with cardiac event patients more frequently intubated on the first attempt. No cardiac event patient required two attempts, whereas one-third of non-cardiac patients did. Acute cardiac arrest and similar emergencies often prompt immediate, coordinated airway management, circumstances that may inherently favor first-attempt intubation success. It is also possible that the hemodynamic instability of these patients and events leads providers to prioritize rapid airway control, ideally done in a single attempt. Repeated intubation attempts are more likely to be associated with mucosal trauma. Hasegawa et al. found repeated intubation attempts to be associated with increased odds of adverse events.Reference Hasegawa, Shigemitsu, Hagiwara, Chiba, Watase and CA III27 Thus, increased attempts can amplify airway damage that predisposes to ALTS development.
Increased length of intubation is also a known risk factor for the development of ALTS as prolonged pressure in the upper airway accelerates tissue injury through direct pressure or shearing forces.Reference Manica, Schweiger, Maróstica, Kuhl and Carvalho28 In our study, there was no significant difference in the length of intubation between our study and control groups, yet the acute cardiac event patients had a significantly lower incidence of ALTS. One possible explanation is that acute cardiac event patients often receive expedited interventions, such as revascularization or extracorporeal membrane oxygenation (ECMO), which reduce the need for prolonged vasopressor dependence with its associated mucosal hypoperfusion.Reference Russotto, Tassistro, Myatra, Parotto, Antolini and Bauer29 This may result in shorter effective ischemic time and potentially contribute to the lower incidence of ALTS in this group.
While the incidence of ALTS was lower after acute cardiac intubation, there was no difference in type of stenosis when compared to the control group. The incidence rates drew on thousands of intubated patients, but only 10 and 21 patients in the two groups developed stenosis, meaning this latter comparison was underpowered. The absence of significant difference in stenosis location should thus not be seen as equivalent and lends credence to the need for larger multi-site studies. Stenosis location was not impacted by clinical indication (cardiac vs. non-cardiac) for intubation and may be governed more by mucosal contact with the endotracheal tube itself.
Previous studies have suggested that females may be at increased risk for developing ALTS following intubation, which is usually attributed to anatomic differences, such as smaller tracheal diameters.Reference Barnett, Mudiyanselage, Haghbin, Banifadel, Adunse and Safi30, Reference Karmakar, Pate, Solowski, Postma and Weinberger31 In our study, however, we did not find any statistical difference between the sexes, with both the study and control groups having a roughly equal distribution between male and female patients.
Higher BMI and obesity can also lead to increased tracheotomy rates and, thus, a greater risk for developing ALTS, with previous studies showing that bariatric surgery reduced the number of endoscopic airway interventions.Reference Fehervari, Patel, Towning, Haire, Al Yaghchi and Sabharwal32 Obese patients additionally have decreased total lung capacity and elevated intrathoracic pressure, causing central airway narrowing.Reference Fehervari, Patel, Towning, Haire, Al Yaghchi and Sabharwal32 It is also thought that obesity is a chronic inflammatory state that can make these patients more susceptible to developing stenosis. In our study, there was no statistical difference in BMI between the study and control groups. Similarly, there were no statistically significant differences between the groups for diabetes mellitus or GERD, suggesting that these factors may not play a significant role in this specific patient cohort.
• Acquired laryngotracheal stenosis (ALTS) is a recognized complication of endotracheal intubation and tracheostomy. Established risk factors include prolonged intubation, larger endotracheal tube size, and obesity
• The impact of an acute cardiovascular event on the development of ALTS has not been well characterized
• Our study found that patients intubated after an acute cardiac event had a significantly lower incidence of ALTS compared to those intubated after a non-cardiac event. Cardiac patients, however, were more likely to be intubated with a smaller endotracheal tube and required fewer intubation attempts, which could potentially explain these results, although additional studies are needed
• Our findings suggest there may be other factors that influence development of ALTS
Limitations and strengths
This study has several limitations including those associated with a retrospective review, such as incomplete documentation and coding errors. ALTS was identified in symptomatic patients and, thus, may have been missed in asymptomatic cases, potentially underestimating incidence rates. The small number of ALTS patients in each group limits statistical power and increases the risk of type II errors, while also limiting analyses, and, thus, should be regarded as hypothesis-generating. By contrast, this series represents data from more than 5300 patients intubated over a 10-year period at a large tertiary Academic Medical Center, yielding a precise estimate of ALTS incidence after cardiac and non-cardiac intubation and as such significantly adds to the published literature. In addition, peri-intubation cardiovascular instability/collapse is common in critically ill intubations, and our institution is a referral center with a high-volume cardiac and transplant program that can skew the number of acute intubations for cardiac presentations, which may impact generalizability. However, both groups were drawn from a single institution during the same time period. While this means patients who were diagnosed elsewhere could have been missed, it also preserves internal validity. Finally, although no ALTS patients had documented prior airway stenosis, pre-existing subclinical airway pathology cannot be excluded in either cohort. In the future, a multi-institutional study with larger cohorts and prospective data collection potentially separating site of stenosis would be helpful to further investigate the mechanisms underlying ALTS in patients intubated for acute cardiac events.
Conclusion
Patients intubated for acute cardiac events have a lower incidence of ALTS than those intubated for non-cardiac events. Cardiac event patients were also intubated with a slightly smaller ETT and required fewer intubation attempts. Although these factors may partly account for the reduced ALTS incidence, other factors and mechanisms may also contribute, which warrant further investigation.
Data availability statement
The data collected and analyzed during this study are available from the corresponding author upon reasonable request.
Funding statement
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
The author(s) declare none.
Acknowledgements
None.
Ethical considerations
None.
Consent to participate
None. There were no patient participants.
Consent for publication
None.
