Endoscopic Management of Idiopathic Subglottic Stenosis

Introduction

Laryngotracheal stenosis (LTS) is a life-threatening condition that describes narrowing of the airway at different sites such as the larynx, subglottis, trachea, and carina.^1,2 It can often affect more than 1 site, with the subglottis being the most common, involving approximately 50% of cases. ³ Etiologies of subglottic stenosis include traumatic/prolonged intubation, external trauma, granulomatosis with polyangiitis (Wegener’s disease), and idiopathic.

Idiopathic subglottic stenosis (ISS) is a rare cause of subglottic stenosis and is a progressive fibro-inflammatory condition affecting the lamina propria of the subglottis and proximal trachea. ⁴ There are multiple hypotheses regarding the cause of ISS. As ISS is more prevalent in females, it is thought that estrogen may play a role. ⁵ With repetitive movement of the trachea, there may be recurrent micro-trauma to the subglottis, and estrogen could hinder wound healing. ⁴ However, lack of estrogen receptors on tissue specimens from affected patients decreases support for this hypothesis. ⁶ Alternatively, gastroesophageal reflux disease (GERD) may lead to upper airway inflammation and cause scarring. ⁵

The incidence of ISS is not well known. Some refer to it as the second most common cause of subglottic stenosis ¹ while others believe that it only accounts for 20% of cases. ⁷ Diagnostic delay is common in this population as patients present with nonspecific symptoms such as wheezing, progressive exertional dyspnea, stridor, or dysphonia.^4,7 Idiopathic subglottic stenosis is often misdiagnosed as adult-onset asthma, and patients are placed on inhaled corticosteroids and bronchodilators. ⁴ Establishing a natural history of the disease is challenging as patients present to multiple providers, including interventional pulmonologists, otolaryngologists, and allergists. ²

Currently, there are 2 general approaches to treatment: open and endoscopic. Open approaches have a curative intent and remove the stenotic segment. Endoscopic approaches mitigate symptoms, knowing that future interventions will likely be needed when the stenosis recurs. Endoscopic management may be appropriate for patients with ongoing inflammation and lack of cartilage involvement. ⁸ Adjunctive therapies such as steroids or mitomycin C have been used in an attempt to delay recurrence. ⁹

Gelbard et al ¹⁰ recently reported a multi-institutional study evaluating characteristics of patients with ISS and comparing outcomes with endoscopic versus open treatment approaches. Patients were typically middle-aged females, and 80% could be managed with endoscopic approaches alone. Additionally, as expected, disease recurrence was higher after endoscopic treatment than open. This study represents a key contribution to the literature on a rare disease that has traditionally been difficult to describe. While this enhances our knowledge on patient demographics and expected rates of recurrence after open versus endoscopic approaches, we still do not know what modifiable and non-modifiable factors affect time between endoscopic interventions. Therefore, we describe a homogeneous group of patients with ISS being treated primarily using endoscopic balloon dilation, evaluate the time between dilations and factors affecting this interval, and determine if concomitant steroid injection can prolong the time between successive dilations. We hypothesized that patients with ISS may have increased incidence of GERD compared to the general population, that body mass index (BMI) may correlate with inter-dilation interval (IDI), and that concomitant steroid injection would prolong the IDI.

Materials and Methods

This study was approved by the Health Sciences Institutional Review Board of the University of Wisconsin School of Medicine and Public Health. A retrospective chart review was performed on patients with LTS from 2003 to 2015. Patients 18 years of age or older with ISS who were treated with endoscopic balloon dilation between this period of time were identified. Patients were considered to have ISS if it was explicitly mentioned as the diagnosis in the chart or if it was not mentioned, the patient had a negative rheumatological workup and no history of severe GERD, external neck trauma, traumatic intubation, or prolonged intubation with post-extubation dyspnea. Out of the 88 LTS patients whose charts were reviewed, the aforementioned criteria resulted in exclusion of 51 patients and selection of 37 patients with ISS treated with endoscopic balloon dilation.

After identification of patients with ISS treated with endoscopic balloon dilation, electronic medical records were reviewed in detail, and the following information was gathered: demographics, co-morbidities, age at diagnosis, age at first balloon dilation, BMI, degree of stenosis pre- and post-endoscopic balloon dilation, number of balloon dilations, IDI, and use of concomitant steroid injection.

Results

Study Population

All patients (37/37) were female, with average age of 46.1 ± 14.9 years (range, 22-78 years) at time of diagnosis. Average age at first dilation was 47.9 ± 13.9 years (range, 27-80 years). Table 1 summarizes the age distribution.

OPEN IN VIEWER

Table 1.

Age at Time of Diagnosis and First Dilation. ^a

	No. of Patients
Age	Diagnosis	First Dilation
26-35	10	8
36-45	10	9
46-55	7	10
56-65	5	4
66-75	3	4
76-85	2	2

a

Mean ages were 45.9 ± 15.4 years and 48.0 ± 13.9 years, respectively.

Co-morbidities

Gastrointestinal co-morbidities were most common (64.9%), followed by cardiovascular (35.1%). The prevalence of pulmonary and psychiatric co-morbidities were 8.1% and 35.1%, respectively. Thirteen patients had psychiatric co-morbidities, which included anxiety (n = 10; 27.0%), depression (n = 7; 18.9%), and bipolar affective disorder (n = 1). The prevalence of GERD in premeopausal, perimenopauasal, and menopausal groups in our sample population were 50.0%, 70.6%, and 75.0%, respectively (Table 2). This is compared to prevalence rates of 25.5%, 41.7%, and 46.9%, respectively, in the general population. ¹¹ Prevalence of GERD in the total study population was higher than a weighted average in the general population (P = .036).

OPEN IN VIEWER

Table 2.

Prevalence of Gastroesophageal Reflux Disease (GERD) in Our Sample Compared to the General Population. ^a

Menopausal Status	Patients With GERD (%)	Prevalence in General Population (%) 10	P Value
Premenopause (age: 23-34)	4/8 (50.0)	25.5	.3017
Perimenopause (age: 35-50)	12/17 (70.6)	41.7	.084
Menopause (age: 51+)	8/12 (75.0)	46.9	.408
Total	24/37 (64.9)	40.0	.036

a

Age groups are based on age at time of first dilation. Values are presented as number of patients with GERD/number of patients in age category (% of patients in category with GERD). Estimated total prevalence in general population represents a weighted average based on ages of patients in this study. P values represent results of chi-square testing.

Body Mass Index

Subjects were stratified into BMI groups according to the BMI classification provided by Kulie et al ¹³ (Table 3). Average BMI at the first dilation was 29.8 ± 6.7 (range, 19.4-43.5). Average BMI at the last dilation was 30.9 ± 7.1 (range, 19.0-42.2). The difference between the BMI at first dilation and last dilation was not statistically significant (P = .2).

OPEN IN VIEWER

Table 3.

Number of Patients in Each Body Mass Index (BMI) Category.

BMI	Classification	No. of Patients
<18.5	Underweight	0
18.5-24.9	Normal weight	6
25.0-29.9	Overweight	9
30.0-34.9	Class I obesity	7
35.0-39.9	Class II obesity	4
≥40.0	Class III obesity	2

Inter-balloon Dilation Interval

Average IDI for all subjects was 635 ± 615 days (range, 49 to 3130 days) (Figure 1). Median IDI was 402 days.

OPEN IN VIEWER

Figure 1.

Histogram demonstrating variability in inter-dilation interval (IDI) for all patients. Average IDI was 635 ± 615 days, and median IDI was 402 days (range, 49-3130 days).

Number of Dilations

Six of 37 patients underwent only 1 dilation, and 31 of 37 patients underwent more than 1 dilation. Patients underwent an average of 3.8 ± 1.8 dilations (range, 1 to 11; mode: 3) (Figure 2).

OPEN IN VIEWER

Figure 2.

Histogram demonstrating number of patients undergoing a given number of dilations during the period evaluated. Patients underwent between 1 and 11 dilations.

Adjunctive Therapy

In total, 144 dilation procedures were done, and 98 included an adjunctive therapy (Table 4). The effect of concomitant steroid injection with balloon dilation on IDI was analyzed. Thirteen subjects had only concomitant steroid injection (either dexamethasone or triamcinolone) for all their dilations, and 4 patients had no adjunctive therapy. Patients not receiving steroids underwent 7.0 ± 3.7 dilations with an average IDI of 283 ± 36 days. Patients receiving steroids underwent 3.9 ± 1.6 dilations with an average IDI of 556 ± 397 days (Table 5). Based on a rank sum test, the difference in average IDI was not statistically significant (P = .079).

OPEN IN VIEWER

Table 4.

Number of Procedures Performed With Each Adjunctive Therapy.

Type of Adjunctive Therapy	No. of Procedures
Triamcinolone only	19
Dexamethasone only	74
Mitomycin C only	2
Mitomycin C + triamcinolone	1
Mitomycin C + dexamethasone	2

OPEN IN VIEWER

Table 5.

Difference in number of dilations and inter-dilation interval (IDI) for patients receiving and not receiving concomitant steroid injection (p=0.079). ^a

Group	No.	Average No. of Dilations	Average IDI (Days)
Steroid only	13	3.9 ± 1.6	556 ± 397
No steroid	4	7.0 ± 3.7	283 ± 36

a

Ninety-three procedures performed with steroid injection, and 48 were performed without.

Discussion

The low prevalence and incidence of ISS combined with its management by various specialties has limited the ability to define the natural history and clinical characteristics of this rare disease. As a result, management has also varied and included endoscopic approaches, open surgeries, and tracheotomy, with no fixed algorithm.^9,10

Endoscopic balloon dilation is not a new technique and has been used both in children and adults for various causes of subglottic stenosis, either benign or malignant. ³ However, its usage has not been well studied in a homogeneous ISS patient population. This study provides a 12-year retrospective review of patients with ISS treated with endoscopic balloon dilation aiming to: (1) provide detailed characteristics of this patient population, (2) evaluate the time between dilations and factors affecting this interval factors, and (3) determine the effect of concomitant steroid injection on IDI.

The mean age at diagnosis was 45.9 ± 15.4 years, similar to the average age of 47.8 ± 12.1 years found by Nouraei and Sandhu. ⁴ All our patients were female, which supports the finding in the literature that ISS typically affects females. In a study by Parker et al, ⁸ 94% of their ISS subjects were female. Why females are so disproportionately affected is not yet known.

Gastrointestinal co-morbidities, especially GERD, are common in ISS patients with 64.9% of our subjects having GERD. In a study by Kocdor et al, ³ 70% of their patients with either idiopathic or non-idiopathic LTS had GERD. Moreover, 78% of patients with LTS have abnormal findings on pH probe assessment. ¹⁴ Pepsin is also commonly found in the larynx or trachea of this patient population. ¹⁵ The prevalence of GERD in ISS patients is greater than in the general population. ⁵ The prevalence of GERD premeopausal, perimenopausal, and menopausal females is 25.5%, 41.7%, and 46.9%, respectively. ¹¹ These rates are lower than the rates for our ISS patients observed in this study. Other retrospective studies have suggested an increased GERD prevalence in ISS patients compared to the general population. ⁵ Excess acid exposure may play a role in the pathophysiology of ISS, and accordingly, treatment of GERD remains a key component of the management of this disorder. ⁵

Differences were observed in the prevalence of psychiatric disorders in our population with ISS compared to the general population. The rate of depression in the general population is approximately 10%, ¹⁶ lower than the 18.9% rate observed in this study. Likewise, the reported 16% prevalence of anxiety disorders in the general population ¹⁷ is lower than that reported in our sample (27.0%). Further investigation into the relationship between anxiety and stenosis severity may be of interest.

Our sample is borderline obese with an average BMI of 29.8 ± 6.7. Average BMI for females in the United States is 26.5. ¹⁸ We hypothesized that BMI may change over disease course, with potential for decrease in BMI over time as patients gain better control over disease and potentially increase exercise tolerance. However, there was no difference in BMI between the first and most recent dilation (P = .2). This may reflect a sustained level of relative exercise intolerance favoring a more sedentary lifestyle or a complex balance between decreased caloric expenditure due to decreased exercise combined with increased caloric expenditure due to increased work of breathing. Further evaluation into the impact of ISS on weight is warranted.

Average degree of stenosis at diagnosis was 53% ± 14%, slightly less than that observed in a recent multi-institutional study, which reported a mean stenosis of 63%. ¹⁰ This may reflect subtle differences in patient population at initial presentation or simply differences in method of stenosis estimation. It is currently difficult to quantify the degree of stenosis accurately, ⁹ and tools designed to improve the reliability of these estimates are worthy of future investigation.

The average IDI was 635 ± 615 days (range, 49 to 3130 days), which is equivalent to approximately 90.7 weeks (range, 7 to 447.1 weeks). Prior studies have reported average intervals of 38.4 and 54.8 weeks.^3,9 Differences in interval could be related to the type of procedure performed. In our study, all patients underwent endoscopic balloon dilation with some also receiving concomitant injection, while others have reported on patients undergoing not only balloon dilation but also scar lysis, open surgical interventions, or a combination thereof.^3,9 Due to the higher rate of disease recurrence with endoscopic procedures compared to open, ¹⁰ some may be hesitant to perform repeated dilations based on the premise that such repetitive trauma may cause more scarring and worsen the stenosis. However, repetitive dilations have not been shown to cause progression of stenosis. ³

In our study, the mean number of dilations was 3.8 ± 1.8, similar to that reported by Kocdor et al. ³ The maximum number of dilations was 11, which occurred in 1 patient. Notably, there are reports of patients undergoing more than 40 dilations, with many ultimately requiring tracheostomy. ¹⁹ It is difficult to say whether the repeated dilations resulted in expansion and progression of the stenosis, thus requiring tracheostomy, or that persons with severe disease are more likely to need repeated dilations and also more likely to require tracheostomy at any given time in their management. With endoscopic dilation, the rate of recurrence is about 40% to 70%, ⁹ with patients with Myer-Cotton grade 1-2 stenosis at initial presentation needing to undergo fewer interventions than those with grade 3-4.

Steroid injection is commonly performed at the time of dilation in an effort to delay scar recurrence. In our sample, 13 patients had steroid injection for all dilation procedures, and 4 did not have any adjunctive therapy. With the obvious limitations of small and disparate sample sizes, a trend was apparent for increased IDI with the concomitant steroid injection. This change did not reach statistical significance (P = .079). Given the difference in means, the minimal added risks associated with steroid injection at the time of balloon dilation, and the high inherent and potentially life-threatening risk of suboptimally managed subglottic stenosis, we favor steroid injection and now make it part of our routine practice.

Four patients underwent tracheotomy on a planned or emergent basis. The need for tracheotomy has been associated with cricatricial stenosis, stenosis greater than 1 cm in vertical length, tracheomalacia, loss of cartilage, and posterior laryngeal inlet stenosis with arytenoid fixation. ²⁰ Patients requiring frequent balloon dilations may ultimately require tracheostomy. ¹⁹ Notably, the 1 patient undergoing emergent tracheotomy had gone several years without evaluation, emphasizing the importance of close monitoring of this patient population and potential for the disease process to cause critical stenosis. Four patients underwent laryngotracheal reconstruction. Importantly, the patients undergoing reconstruction due to stenosis refractory to endoscopic interventions were early in the series, with prior endoscopic interventions including CO2 laser-based scar lysis and mitomycin C injection. Since the implementation of balloon dilation and steroid injection as routine therapy, 2 patients underwent reconstruction out of desire for a longer lasting solution rather than necessity secondary to failed endoscopic management.

Conclusion

Idiopathic subglottic stenosis is a rare disease that has recently been receiving more attention in the literature. This study adds to our knowledge regarding the typical characteristics of patients with ISS. Gastroesophageal reflux disease and obesity are common co-morbidities, making treatment with anti-reflux medications and counseling on weight loss important components of treatment. Further, in patients undergoing endoscopic management, injecting steroids at the time of dilation is likely to prolong the IDI, and regularly scheduled follow-up can help prevent the need for emergent intubation or tracheotomy. Determining when to perform the next dilation remains a challenge that must balance the desire to avoid unnecessary procedures with the need to avoid emergent airway situations. To this end, developing objective criteria for assessing airway functional status and reliable methods of measuring airway diameter in the office will be beneficial.