Tracheal and cricotracheal resection for laryngotracheal stenosis: experience in 54 consecutive cases

Abstract

Objective: Partial tracheal resection (Küster operation (KO)) and cricotracheal resection (Pearson operation (PO)) are currently the standard operative techniques in the curative treatment of tracheal and cricotracheal stenosis, respectively. This study aims to analyze the outcomes of tracheal and cricotracheal resection when a specific protocol is applied. Methods: Between 1990 and 2004 we treated 54 patients with laryngotracheal stenosis. The mean age was 44.9 years with a sex ratio of 1:1. All patients were treated according to the random protocol “Lesions of the main airway (MA) protocol,” which considers the following stenosis variables: stage of development (S), caliber (C), and length (L). We performed 38 Küster operations, 14 Pearson operations, and 2 combined Pearson–Küster–Rethi operations (ROs). Results: Overall mortality of the series was 1.85%, with a specific morbidity of 27.7%. A total of 96.2% of patients were cured (85.6% of Pearson operation and 100% of Küster operation). We performed 3.7% re-interventions (14.2% of Pearson operation and 0% of Küster operation), and the failure rate was 3.7% (14.4% of Pearson operation and 0% of Küster operation). We had 27.5% who had postoperative complications (28.5% of Pearson operation and 26.3% of Küster operation). The most frequent complications were restenosis (14.2%), granulation tissue (13.1%), edema (10.5%), anastomotic dehiscence (7.1%), and tracheoesophageal fistula (7.1%). In terms of the SCL variables, significant differences were only observed with respect to morbidity between the S4 group and the other cases without tracheoesophageal fistula in the Küster operation group; we found no differences in Pearson operation. Conclusions: Application of the Main Airway protocol allowed development of a strategy for the surgical treatment of main airway stenosis. This, in turn, enabled a strict selection of cases and meticulous preoperative preparation that, coupled with a highly effective surgical technique, led to excellent outcomes with minimal sequel. The presence of tracheoesophageal fistula could increase the complications.

Laryngotracheal stenosis, Tracheal resection, Cricotracheal resection, Tracheoesophageal fistula

1 Introduction

Within the field of main airway (MA) surgery, controversy remains as to which technique should be used: conservative treatments [1] (laser and endotracheal prosthesis) or full-scale surgery. The usual pathologies requiring this type of treatment are inflammatory stenosis of the trachea and/or tracheoesophageal fistulas in critical patients on artificial ventilation [2], and less frequently in the case of traumatic lesions, tracheal and/or mediastinum tumors.

Currently, the standard techniques used in the curative treatment of tracheal and cricotracheal stenosis are the segmental resection and anastomoses of the trachea, respectively, by means of the Küster operation (KO) [3] and cricotracheal resection through the Pearson operation (PO) [4]; the Rethi operation (RO) [5] is reserved for cases of glottic and subglottic stenosis. Most of the research in this area has taken the form of retrospective studies [6–13], thus illustrating the difficulty of standard therapeutic indications.

The aim of the present study was to analyze the outcomes of tracheal and cricotracheal resection when applying a specific protocol.

2 Methods

The sample comprised 54 patients operated on consecutively between January 1990 and October 2004. Paper information about the technique, results, and complications were given during the first consultation. This study is a review of a prospectively gathered database. It was approved by our institutional committee on human research, and all patients provided written informed consent.

The etiology of the benign stenosis (whether or not associated with a tracheoesophageal fistula) was prolonged intubation or post-tracheostomy for cardiorespiratory resuscitation; no cases of idiopathic stenosis were observed. Table 1 shows the specific baseline pathology of the patients. Systematic bronchoscopy was performed in all cases; neck and chest computerized tomography (CT) and/or nuclear magnetic resonance (NMR) were carried out in 49 patients (90.74%); and the flow/volume (F/V) curve was applied in non-tracheostomy cases. Follow-up was completed in all patients from 2 month to 12 years, with a median follow-up of 36 months; the evaluation included bronchoscopy, neck and chest CT scan, and F/V curve monthly for 3 months.

2.1 Surgical technique

To perform tracheal or cricotracheal cuff resections, the following procedures were carried out in all the cases: (a) general anesthesia with orotracheal intubation using tubes of caliber 5–6 (90.74% of cases), which were placed in a prestenotic position in nine cases (16.66%); in tracheostomized patients, an intubation tube was left in the larynx; (b) after the patient was placed in the position of hyperextension, transversal cervicotomy was performed, with partial sternotomy of the manubrium in two cases and associated with Grillo's right thoracotomy in one case; and (c) dissection of the visceral compartment of the neck with transverse section of the infrahyoid muscles, the anterior surface of the trachea was exposed and circumferential dissection of the trachea sparing the inferior laryngeal nerves, and circumferential tracheal and/or cricotracheal resection was performed. Tracheal anastomosis or end-to-end tracheal tugging was performed with sutures of 3-0 polyglactin (Vicryl^®, Ethicon Inc.). In cases of tracheoesophageal fistula, an esophagorrhaphy was performed with interrupted sutures of 3-0 polyglactin and esophageal suture was buttressed with sternocleidomastoid muscular flap to isolate it from trachel anastomosis.

During exeresis and anastomosis, the surgical field was kept ventilated by means of a distal intubation tube. In the cricoid resection (PO), the anterior and lateral portions of the cartilage were removed, sparing the perichondrium and mucosa; with the aim of sparing the caudal laryngeal nerves, rongeur forceps were used until reaching the thyrocricoid joint, described by Couraud et al. [8].

On one occasion, the tracheobronchial bifurcation was released to reduce the tension in the anastomosis line, which was not necessary to carry out the Dedo [14] maneuver. In seven cases of PO in which the suture was placed very close to the vocal chords, a Montgomery [15] T-tube was inserted for greater safety and remained in place for 18 months. The patients were extubated in the operating room and were kept under observation for 48 h the intensive care unit.

2.2 “Lesions of the main airway (MA)” protocol

In 1990 our hospital began to apply the “Diagnostic and Therapeutic MA protocol” (Tables 2 and 3³). This protocol classified the inflamed laryngotracheal stenosis according to a topographic and lesional criteria, incorporating three independent variables: stage of development (S), caliber (C) and length (L). This protocol was approved by the Institutional Ethics Committee.

From the very beginning of applying MA protocol, all patients admitted and diagnosed with main airway pathology in our department were evaluated using bronchoscopy, CT, and/or NMR. The results were classified according to Table 2 and the composition of three variables (S, C, and L) was used to determine the stage of lesion and treatment option, as described in Table 3. Importantly, each case and the results of diagnosing procedures were thoroughly evaluated by a multidisciplinary committee of surgeons, endoscopy specialists, intensive care doctors and otorhinolaryngologists. The committee discussed the cases and assigned a stage of injury (SCL), and the best therapeutic option was chosen and performed following Table 3. The distributions for the three variables are listed in Table 4.

2.3 Tracheal (KO) and cricotracheal (PO) resection

We performed 38 tracheal resections (KO), 14 cricotracheal resections (PO) and two PO + KO + RO (Table 5).

Inflammatory lesions were staged according to the SCL variables, while neoplastic tumors were tabulated according to the variables C and L (Table 4). Twenty-eight cases with inflammatory stenosis receiving laser treatment were given a maximum of three sessions over a period of 6 weeks, and, in the event of failure, 3 months were allowed to elapse before performing the surgery.

The mean age was 44.9 years (range, 18–79) with a sex ratio of 1:1. The etiology of the stenosis was prolonged intubation and/or tracheostomy in 47 cases, with a mean intubation time of 9 days (range, 2–26); seven patients had neoplastic stenosis with tracheal invasion. In 43 cases, surgery was performed without applying a prosthesis; in seven patients a Montgomery T-tube was inserted; and in four cases an endotracheal prosthesis was fitted.

3 Results

The overall mortality of the series was 1.85%, with a specific morbidity of 27.7%. A total of 96.2% of patients were cured; the figures for re-intervention and minor procedures were 3.7% and 24%, respectively, with a failure rate of 3.7%. Below, the results are analyzed according to the type of intervention.

3.1 Küster operation

This operation was performed in 38 cases (17 stenosis, 14 tracheoesophageal fistulas, and 7 neoplastic tumors). One case (2.63%) died 5 days after surgery due to a post-traumatic aneurysm of the internal carotid artery.

The specific morbidity was 26.3%: one case showed suture dehiscence in the anterior face of the anastomosis (2.6%) with subsequent formation of granulomas, requiring the insertion of a Montgomery T-tube; four cases (10.5%) presented with edema in the anastomosis, which was resolved by endotracheal prosthesis; and five cases (13.1%) presented granuloma tissue that was treated by laser. There were no re-interventions.

In terms of the SCL variables, significant differences were only observed with respect to morbidity between the S4 group and the other cases without tracheoesophageal fistula.

In the subgroup receiving radical treatment for tracheoesophageal fistula (Table 5), the S4 variable presented a specific morbidity of 50% compared with 25% for the variables S1, S2, and S3. The S4 variable was also associated with 28.5% granulomas, 14.2% vocal chord paralysis, and 7.1% tracheal restenosis. Five minor procedures were carried out, with the evolution of one case remaining pending.

The cure rate for this type of intervention was 100%, and minor procedures were performed in 26.3% of cases but there was no re-intervention (Table 6).

3.2 Pearson operation

This operation was performed in 14 cases, and no mortality was recorded. The specific morbidity rate was 28.5%: two cases of restenosis (14.2%) in the same patient, who also required a temporary Montgomery T-tube (a complex case because of inhalation of inflammable gases); one case of tracheoesophageal fistula (7.1%); and one case of anastomotic dehiscence treated successfully with surgical debridement and temporary Montgomery T-tube.

Complications were not related to the SCL variables. The cure rate for this subgroup was 85.7%, and the failure rate was 14.2% (Table 7).

3.3 Combined Küster and Pearson operations

This procedure was carried out in two cases that showed a good evolution in terms of clinical and endoscopic parameters. A Rethi laryngoplasty was also performed in both cases: one case presented granulomas that were treated by laser while the other showed a good evolution without incident (Table 7).

4 Discussion

The present study concerns a group of patients who, after application of the MA protocol, were assigned either to surgery or conservative treatment, thus enabling selection of surgical stages. In a previous report [1], we described the benefits of conservative treatment with laser and/or endotracheal prostheses.

When consulting surgical reports on this issue [6–13,16], we noticed the lack of prospective protocols despite the extensive experience of the authors in question. Mortality was found to range from 0% to 24% (Table 8). For example, in an extensive series [13] of 503 patients, the figure was 2.4% with major complication rates of 13.9% (trachea–trachea anastomosis, KO), 15.4% (trachea–cricoid anastomosis), and 12.9% (trachea–thyroid anastomosis, PO), respectively. For the same series, the therapeutic failure rates were 2.2%, 6.0%, and 8.1% of the above-mentioned levels of anastomosis, respectively.

In our series of 54 cases, mortality was 1.85% for both tracheal (KO) and cricotracheal (PO) resection; although overall specific morbidity was higher at 27.7% (26.3% for KO, 28.5% for PO, and 50% for KO + PO + RO), the rate of major complications (dehiscence, restenosis, and tracheoesophageal fistula) was only 9.2%. In terms of morbidity, a higher rate was observed in the tracheoesophageal fistula (S4) subgroup, while the C and L variables showed no differences. Unlike previous reports [17], there was no need for re-intervention, and neither was it necessary, as some authors propose [18,19], to apply prophylactic mitomycin-C to avoid restenosis.

In the extensive series reported by Grillo et al. [13], outcomes were good in 84.9% of cases and satisfactory in 5.3%, with 3.6% having re-interventions, and 3.8% failures. Mortality in this series was 2.6% in cases of stenosis without tracheoesophageal fistula and 5% in those with such a fistula. In our study of 54 patients with stenosis (both benign and malignant), outcomes were good overall in 96.2%, with a failure rate of 3.7% (0% in KO and 13.7% in PO). In terms of the 38 cases of tracheal resection (KO), 14 presented with a fistula and the following results were found:

• In the tracheal resection (KO) without fistula group (variables S1, S2, and S3), the specific morbidity was 25% (major complications 4.1%, minor complications 20.9%), and no mortality was recorded. The cure rate for this group was 100%, and 25% required endoscopic and/or prosthetic procedures.

• In cases of tracheal resection (KO) for tracheoesophageal fistula (S4), the specific morbidity was 50.0% (major complications 21.3%, minor complications 28.5%), and no mortality was recorded. The cure rate for this group was 100% without re-intervention and 26.3% required endoscopic procedures.

We believe that the highly satisfactory nature of the outcomes was due to protocol-based selection, which enabled the specialists involved (intensive care staff, surgeons, otorhinolaringologists, and endoscopy technicians) to agree on the best approach and thus undoubtedly improve the quality of life of these patients.

Application of the MA protocol in the surgical treatment of tracheal and cricotracheal stenosis, with or without tracheoesophageal fistulas, has enabled a more objective selection of cases that may be cured at the first attempt. The outcomes obtained were excellent or good in 96.2% of cases, with an overall failure rate of 3.7% (0% in tracheal resection and 13.7% in cricotracheal resection). These highly satisfactory outcomes may be related to the protocol-based selection of cases; surgery was reserved for potentially curable states, and we chose the most suitable moment in the evolution of the process and the optimum conditions for favoring good scar-tissue formation.

Our prospective study demonstrated that evaluating and cataloguing the SCL-independent variables enabled the development of stage reference that subsequently was applied to determine and perform the most appropriate intervention for each specific case of laryngotracheal stenosis. Application of the MA protocol has resulted in excellent outcomes of surgical treatment, highlighting the benefit of standardization of the protocol for daily use in the clinic.

Table 1

Number of cases/year for every hospital

Table 2

Groups of hospitals according to number of cases/year

Table 3

Descriptive data of the groups

Table 4

General mortality in the extreme groups

Table 5

Adjusted mortality for different risk factors

Table 6

General morbidity and mortality in pneumonectomies and in operations in patients above the age of 75

Table 7

Prognosis and number of operations/year

Table 8

Prognosis and number of operations/year