Cervical Tracheal Resection: New Lessons Learned

Transcription

1 Cervical Tracheal Resection: New Lessons Learned Christopher J. Mutrie, MD, Shady M. Eldaif, MD, Caleb W. Rutledge, MS, Seth D. Force, MD, William J. Grist, MD, Kamal A. Mansour, MD, and Daniel L. Miller, MD Department of Surgery, Section of General Thoracic Surgery, and Department of Otolaryngology, Emory University School of Medicine, Atlanta, Georgia GENERAL THORACIC Background. Cervical tracheal stenosis can be a difficult condition to manage. Depending on the etiology, location, and extent of the stenosis, tracheal or cricotracheal resection may be required. Intraoperative decisions may predict outcome. Methods. We performed a retrospective chart review of all patients undergoing cervical tracheal or cricotracheal resection from April 2000 through March Results. One hundred and five patients underwent 108 tracheal or cricotracheal resections. Median age was 65 years (range, 15 to 78); 68% were women. Indication for operation included postintubation tracheal stenosis (38), idiopathic (31), tracheostomy stenosis (19), invasive thyroid cancer (9), and other (8). Median length of trachea resected was 2.7 cm (range, 1.5 to 6.0 cm); 48 patients (46%) underwent extended cricotracheal resections. Twenty-six patients (25%) had an intraoperative chin stitch placed. Hospital stay was a median of 4 days (range, 2 to 33). Operative mortality was (1%); 1 patient died of myocardial infarction on postoperative day 3. Four patients (4%) had hoarseness or vocal cord immobility. Median follow-up was 36 months (range, 1 to 79). Eighteen patients (17%) required dilation postoperatively. Seven patients (7%) required tracheostomy; 2 (2%) are tracheostomy dependent. Three patients (3%) underwent a re-resection for recurrent stenosis. Multivariate analysis of indication for resection, type of resection, length of resection, anastomotic technique, and use of chin stitch did not predict the need for postoperative dilation, tracheostomy, or reoperation. Conclusions. Cervical tracheal resection can be performed safely with low morbidity and mortality. Only 5% of patients required a long-term tracheostomy or re-resection for recurrent tracheal stenosis. Specific intraoperative decisions did not predict long-term success. (Ann Thorac Surg 2011;91:1101 6) 2011 by The Society of Thoracic Surgeons Accepted for publication Nov 8, Presented at the Fifty-sixth Annual Meeting of the Southern Thoracic Surgical Association, Marco Island, FL, Nov 4 7, Address correspondence to Dr Miller, General Thoracic Surgery, Emory University Clinic, 1365 Clifton Rd NE, Atlanta, GA 30322; daniel.miller@emoryhealthcare.org. Cervical tracheal stenosis usually occurs as a result of injury to the trachea secondary to prolonged intubation or a prior tracheostomy, neoplasm, trauma, infection, irradiation, or idiopathic inflammatory causes [1]. As many as 10% of intubated patients will have cricotracheal stenosis, and although the risk appears to rise with prolonged intubation, subglottic injury has been demonstrated within hours of intubation [2 4]. Therefore, a large population of patients exists who are at risk for tracheal stenosis. Subglottic tracheal stenosis presents the greatest challenge for tracheal resection and reconstruction because it is the narrowest segment of airway and is in proximity to the vocal cords and recurrent laryngeal nerve insertions [5, 6]. Single-stage resection and reconstruction, as described by Grillo and coworkers [2], has achieved excellent results. A wide array of operative interventions exists for treatment of cervical tracheal stenosis from outpatient procedures such as sequential airway dilation to major airway reconstruction such as laryngotracheoplasty, tracheal resection, or cricotracheal resection. Tracheal dilation may be successful when the tracheal injury is identified early and the stricture is limited in distance. Major airway reconstruction is a definitive treatment and is usually performed as a single-stage procedure. Detailed reports on the management and results of treatment are infrequent, except from Grillo s institution at Massachusetts General Hospital [1, 2, 6 10]. In 1994, our senior author reported on our first series of tracheal resections performed at Emory University Hospital. Over the last 15 years, the management of cervical tracheal stenosis has changed at our institution. To determine if that change has been advantageous, we evaluated our more recent experience utilizing a multivariate analysis of indication for resection, type of resection, length of resection, anastomotic technique, and use of chin stitch as predictors of a satisfactory outcome based on the need for postoperative dilation, tracheostomy, or reoperation. Material and Methods We retrospectively reviewed the charts of all patients who underwent surgical treatment for cervical tracheal disease from January 2000 through March 2008 at Emory University Hospital. Excluded were patients who had tracheal lesions below the aortic arch or who required a thoracotomy as the approach. One hundred five patients were identified who 2011 by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 GENERAL THORACIC 1102 MUTRIE ET AL Ann Thorac Surg CERVICAL TRACHEAL RESECTION 2011;91: underwent tracheal or cricotracheal resection and reconstruction by querying the Section of General Thoracic Surgery and the Department of Otolaryngology databases. After obtaining approval and patient consent waiver from our Institutional Review Board, the medical records were reviewed for the patient s demographics, etiology of stenosis, previous treatments, type of resection (tracheal versus cricotracheal), length of trachea resected, duration of hospital stay, operative morbidity and mortality, and the need for postoperative intervention. We reviewed all operative reports to determine the technique of reconstruction and if a postoperative chin stitch was used and why. We also reviewed all postoperative bronchoscopy reports for the presence of recurrent stenosis. Statistical analysis was performed by one-way analysis of variance and multivariate analysis for the indication for resection, type of resection, length of resection, anastomotic technique, and use of a chin stitch as a need for postoperative dilation, tracheostomy, or reoperation. Univariate analysis was performed using a 2 test to establish any relationship between these factors. The statistical significance of the differences in surgical outcome was evaluated by the log rank test. The statistical significance was considered to be significant if the p value was below Data were analyzed with the use of the commercially available MedCalc statistical program (MedCalc Software, Mariakerke, Belgium). Results There were 105 patients who underwent tracheal or cricotracheal resection and reconstruction; 71 of the patients (68%) were women. The median age of the patients was 65 years old (range, 15 to 78). The etiology of the stenosis was an iatrogenic injury in the majority of patients; 57 patients (54%) had tracheal stenosis secondary to prolonged endotracheal intubation (n 38) or tracheostomy complication (n 19). Seventeen patients (16%) had tracheal stenosis secondary to neoplasms; 9 of the neoplasms (53%) were locally invasive thyroid cancers. Of the remaining 31 patients (30%), the cause was inflammatory in nature; 2 patients (1%) had tracheal stenosis secondary to Wegener s granulomatosis and 29 patients (18%) had an idiopathic etiology (Table 1); 11 patients (10%) had associated gastroesophageal reflux disease (GERD). Prior treatments included tracheal dilation in 80 patients (76%), laser treatment in 30 patients (28%), stent placement in 5 patients (5%), and prior tracheal resection in 5 patients (5%) (Table 2). Fifty-seven patients (54%) underwent cervical tracheal resection, and 48 patients (46%) underwent cricotracheal resection. The median length of resected trachea was 2.7 cm (range, 1.5 to 6.0 cm). No patients required laryngeal release or sternal division. A chin-to-sternum stitch was performed in 26 patients (25%). In the operating room, 101 patients (92%) were extubated successfully without airway compromise. Operative mortality was 1%; 1 patient died of a myocardial infarction on postoperative day 3. Complications occurred in 18 patients (17%) (Table 3). The median hospital of stay was 4 days (range, 2 to 33). Table 1. Etiology of Cervical Tracheal Disease Postoperative tracheotomy was required in 7 patients (7%); 2 (2%) of these patients are tracheotomy dependent. A T-tube was placed in 1 patient (1%). Four patients (4%) experienced hoarseness or vocal cord immobility. Median follow-up was 36 months (range, 1 to 79). Ten patients (10%) were lost to follow-up. Postoperative dilatation was required in 18 patients (17%), all within 12 weeks after the resection; 12 patients (67%) had undergone cricotracheal resection. Five patients required more than one dilatation. Three patients underwent tracheal re-resection at 6, 12, and 24 months after the initial resection. Multivariate analysis of indication for resection, type of resection, length of resection, anastomotic technique, and use of chin stitch did not predict the need for postoperative dilation, tracheostomy, or reoperation. Patient s overall outcome after the final intervention were recorded as good for 88 patients, satisfactory for 10, fair for 5, and poor for 2; 93% of patients had either a good or satisfactory result after their tracheal surgery. Comment Etiology Number of Patients (%) Iatrogenic injury 57 (54) Postintubation 38 Tracheostomy 19 Neoplasms 17 (16) Thyroid 9 ACC 4 Other 4 Inflammatory 31 (30) Wegner s 2 Idiopathic 29 ACC adenoid cystic carcinoma. Because of the rarity of tracheal tumors, the surgical procedures of the trachea has developed more slowly than most other areas of thoracic surgery [2]. The most common conditions of the trachea that can be treated with surgical resection are a result of traumatic, neoplastic, and inflammatory processes [1]. Traumatic lesions most commonly arise from postintubation tracheal stenosis after prolonged intubation where overinflated low pressure cuffs causes circumferential pressure necrosis or posttracheostomy stenosis caused by triangular tracheal scarring [5, 6]. Primary tracheal tumors have an annual incidence of only 2.7 cases per million people. The two most common primary malignancies are squamous cell carcinoma, mostly found in smokers, and adenoid cystic carcinoma, as seen in 7 of our patients [7]. Secondary neoplasms involve the trachea through direct extension, most commonly thyroid carcinomas, as in 8 of our patients, by invading the trachea usually at the second or third cartilaginous ring [8]. Idiopathic tracheal stenosis most commonly involves the upper trachea and lower larynx. It is more common among women and is characterized by circumferential keloidal fibrosis with thickening of the lamina propria

3 Ann Thorac Surg MUTRIE ET AL 2011;91: CERVICAL TRACHEAL RESECTION Table 2. Prior Treatments Intervention Number of Patients (%) Dilatation 80 (76) Rigid 52 Balloon 28 Laser treatment 30 (29) Tracheal resection 5 (5) Cancer 2 Postintubation 2 Idiopathic 1 Stent placement 5 (5) Combination 32 (30) 1103 Fig 1. Cross-sectional computed tomography scan demonstrating a severe tracheal stenosis of the cervical trachea secondary to granulation tissue obstructing a metallic stent. GENERAL THORACIC without cartilaginous destruction [9]. If the etiology of the stenosis is unknown, then further investigation needs to be performed to determine if the cause could be related to GERD. Gastroesophageal reflux disease is thought to result in chronic irritation and subsequent subglottic stenosis. Several studies have found an association between subglottic stenosis and GERD [10 13]. However, most patients with cricotracheal stenosis do not have GERD. To note, 67% of our patients who underwent cricotracheal resection for inflammatory causes had GERD. The role of GERD as the mechanism for their tracheal stenosis is unknown. However, if GERD is discovered, then treatment should consist of either prolonged proton-pump inhibitor therapy or a laparoscopic reflux procedure to prevent potential recurrence of the tracheal stenosis, which had been initially treated successfully with resection and reconstruction. Further study is warranted in this unique population of idiopathic inflammatory cricotracheal stenotic patients. Computed tomography (CT) scanning has proved to be the most useful modality for evaluation of cricotracheal and tracheal stenosis by revealing the extent and luminal diameter of the stenosis (Figs 1 and 2). Three-dimensional CT scan has been also been utilized more lately to help facilitate and plan for resection for complex tracheal problems such as removal and reconstruction after inappropriate stent placement (Figs 3 and 4). Positron emission tomography scanning is also useful in patients with cervical tracheal malignancies, either primary or metastatic, to help in staging of the patients in regards to distant disease and to access local disease such as lymphadenopathy. Bronchoscopic evaluation is essential in the preoperative workup of tracheal stenosis. Bronchoscopic measurement, preferably by a rigid scope, of the amount of normal trachea proximal and distal to the lesion and the relation to the vocal cords and carina allows for operative planning as to approach. The majority of tracheal resections can be performed through a cervical incision with or without upper sternal split. All 105 of our patients underwent resection through a cervical incision. As many as 50% of the trachea can be resected and anastomosed primarily. In addition, confirmation of the diagnosis and treatment of impending airway obstruction can be performed during bronchoscopic evaluation. Rigid bronchoscopy enables the surgeon to dilate and control the airway as necessary, and is especially important in evaluating patients with critical airway stenosis (more than 4 mm). Inadvertent induction of secretions, edema, and bleeding can be fatal in such fragile patients. A thorough knowledge of the anatomy of the trachea and its associated structures is essential to perform the correct surgical procedures for long-term relief of tracheal stenosis. Several features of the trachea make it particularly difficult for surgery. Its structural rigidity, short length, lack of Table 3. Complications Complication Number of Patients (%) Tracheostomy 7 (7) Hoarseness 4 (4) Wound infection 3 (3) Pneumonia 2 (2) Atrial fibrillation 1 (1) Tracheal dehiscence 1 (1) Total 18 (17) Fig 2. Coronal computed tomography scan showing severe trachea stenosis caused by intraluminal scarring secondary to a metallic stent in the cervical trachea.

4 GENERAL THORACIC 1104 MUTRIE ET AL Ann Thorac Surg CERVICAL TRACHEAL RESECTION 2011;91: Fig 3. Three-dimensional computed tomography scan of the cervical trachea showing a metallic stent. longitudinal elasticity, proximity to major vascular structures, and segmental blood supply make surgical management difficult. The trachea averages 10 to 11 cm in length from the lower border of the cricoid cartilage to the carinal spur, with an additional 1.5 to 2.0 cm of subglottic intralaryngeal airway. There are 18 to 22 cartilaginous U-shaped rings in the trachea providing structural support [14]. The cricoid is the only complete cartilaginous ring in the normal airway. Cervical hyperextension elevates almost half of the trachea into the neck, and flexion devolves the trachea almost entirely into the mediastinum [15]. The recurrent laryngeal nerves ascend in the tracheoesophageal grooves bilaterally and pass medial to the inferior cornua of the thyroid cartilage. Meticulous operative technique with great attention to the above outlined anatomical detail is the basis for the governing principles of tracheal surgery. Operative dissection is performed directly on the trachea and associated scar tissue to avoid injury to the lateral blood supply and recurrent laryngeal nerves. The recurrent laryngeal nerves are not mobilized or identified at any time during the procedure. Circumferential dissection of the trachea is confined to the area of stenosis and no more than 1 to 2 cm of normal trachea above and below the stenosis. This method of detailed dissection preserves the lateral segmental blood supply of the trachea. The majority of cervical trachea disease can be approached through a low collar incision or a collar incision combined with an upper partial sternotomy. If a stoma is present, then it is usually incorporated into the collar incision; however, if it is higher than normal, then it can be excised and closed separately. Subplatysmal flaps are elevated to the thyroid notch superiorly and the sternal notch inferiorly. The strap muscles are saved and retracted laterally. The thyroid isthmus is divided and reflected laterally. Mobilization of the trachea is performed only on the anterior surface before resection from the cricoid to the carina. The level of the stenosis is determined by extrinsic anatomical deformity of the trachea or by intraoperative bronchoscopy with transillumination of the trachea. Circumferential dissection is carried out directly at the stenosis to minimize injury to the nerves and maintain the segmental blood supply. If the cricoid is involved, then the trachea is first divided below the stenosis in an area of normal trachea, and dissection is carried proximally to free up the esophagus posterior to the trachea. If the cricoid cartilage is involved, the cricoid is transected laterally in an oblique fashion to prevent injury to the recurrent laryngeal nerves. As for lower lesions, the trachea is divided proximally. After the stenotic segment is removed, cross-field ventilation is achieved in the distal trachea. Stay sutures are placed laterally two rings above and below the planned resected segment, and with neck flexion of the patient, assessment of the tension on the completed anastomosis is carried out. Cervical neck flexion and anterior mobilization to the carina will allow, in the majority of cases, a tensionfree anastomosis of the cervical trachea. If additional length is required, a suprahyoid laryngeal release can be performed by the Montgomery technique [16]. A tension-free anastomosis with preserved blood supply ensures the best chance of healing without restenosis or dehiscence. Surgical resection and reconstruction for cervical tracheal stenosis can yield excellent results. The largest reported series of 503 patients demonstrated good to satisfactory results in more than 93% of patients, with an overall operative mortality of 2.4% [17]. Our results were identical, with 93% of our patients having either a good or satisfactory outcome and an operative mortality of 1%. In our 1994 series, our senior colleagues advocated Fig 4. Three-dimensional computed tomography scan of the trachea demonstrating a lower cervical trachea stenosis caused from prolonged postintubation (arrow).

5 Ann Thorac Surg MUTRIE ET AL 2011;91: CERVICAL TRACHEAL RESECTION several lessons that they learned from their experience over a 19-year period of performing tracheal surgery for a wide variety of disease processes [4]. First, symptoms related to tracheal stenosis are usually vague, occur late, and are usually thought to be caused by other disease processes, tracheal asthma rather than tracheal stenosis. In our series, patients with injury to the trachea secondary to prolonged intubation or prior tracheostomy presented earlier than patients with stenosis related to inflammatory conditions or neoplasms, which was a similar finding in the 1994 series. Second, tomograms and CT scanning was the modality of choice to evaluate the patient in the earlier series. Today, CT scanning has replaced tomograms entirely as the modality of choice to evaluate the trachea, especially utilizing a three-dimensional reconstruction program. Also, positron emission tomography scanning has been used more frequently to help stage patients with a tracheal malignancy. Positron emission tomography was not available during the period of our first series. Like the first series, bronchoscopy, especially a rigid bronchoscope, is still the preferred tool to facilitate the diagnosis and treatment of tracheal stenosis, especially in improving airway quality during the assessment phase of treatment for tracheal stenosis. Third, in the earlier series one of the most important lessons was that the reconstruction should not be performed with stents, using cardiopulmonary bypass, or with lasers. None of these devices was used in any of our patients, except that in 1 patient, a Montgomery T tube was placed after an extended cricotracheal resection for severe tracheal stenosis secondary to multiple metallic stents. Montgomery T tubes have proved to be an enviable adjunct for the successful treatment of complex cervical tracheal stenosis. The final lesson learned in the original series was that when tracheal resection was performed for cancer, either a primary lesion or direct invasion of a secondary tumor, that the long-term survival was satisfactory. We found a similar survival advantage (more than 80% 5-year survival) in our series. Also, among patients who required adjuvant radiotherapy, there was no associated sequelae, such as recurrent stenosis or airway compromise, related to the radiation. Technically, there have been several changes over the 15 years since the first Emory series was published. All of our anastomoses now are performed with polydioxanone suture (PDS) and not Vicryl. In 2002, we modified our suture technique from all interrupted sutures to a running 4-0 PDS in the mucosal layer and interrupted 3-0 PDS in the cartilaginous layer, similar to our anastomotic technique for our lung transplant patients. To prevent early postoperative neck extension in our first series, all patients underwent a chin stitch, whereas only 25% did in the later series. The chin stitch had a significant impact on hospital stay because the chin stitch was left in place for an average of 7 to 10 days, compared with 5 days when used in the later series. The median hospital stay in the first series was 17 days (range, 4 to 38) and only 4 days (range, 2 to 33) in the later group. Today, we only do a chin stitch if the resection is greater than 4 cm or if the procedure is a re-resection. Other methods for prevention of neck extension is mechanical ventilation, neck brace (360 degrees), posterior plaster splint, or upper body cast, all of which can prolong hospital stay and increase morbidity. Reinforcement of the tracheal anastomosis with viable tissue was performed on a routine basis in the earlier series and not at all in the later series. Also, now, before initiation of oral intake, a modified barium swallow is performed to prevent the risk of aspiration, which could be a fatal complication. This has also proven beneficial in our lung transplant and lung resection patients to decrease postoperative complications such as aspiration pneumonia, readmission to the intensive care unit, or requirement for bronchoscopy [18]. The need for further intervention after resection was evaluated. Postoperative dilation was required in 18 patients (17%); 12 patients (67%) had undergone cricotracheal resection. All of these cricotracheal resections had complex histories before resection at our institution; 4 were re-resections, 4 had a metallic stent removed at the time of resection, 3 were the required re-resection, and 1 with Wegener s granulomatosis. Three patients underwent tracheal re-resection at 6, 12, and 24 months after the initial resection. The 2 patients who are tracheostomy dependent had metallic stents in place for more than 4 years before resection and had significant peritracheal scar tissue at the time of cricotracheal resection. In this series, multivariate analysis of indication for resection, type of resection, length of resection, anastomotic technique, and use of chin stitch did not predict the need for postoperative dilation, tracheostomy, or reoperation. In conclusion, cervical tracheal or cricotracheal resection can be performed safely with low morbidity and mortality utilizing a multidisciplinary team approach. In all cases, resection was performed as a one-stage surgical resection and reconstruction, and only 5% of patients required a long-term tracheostomy or re-resection for recurrent tracheal stenosis. The lessons learned in this series and in our earlier series have significantly improved the care and outcome of our patients with tracheal stenosis, and we hope they will be the basis for continued improvements in the future. References Grillo HC. Primary reconstruction of airway after resection of subglottic laryngeal and upper tracheal stenosis. Ann Thorac Surg 1982;33: Grillo HC. The history of tracheal surgery. Chest Surg Clin North Am 2003;13: Gould SJ, Young M. Subglottic ulceration and healing following endotracheal intubation in the neonate: a morphometric study. Ann Otol Rhinol Laryngol 1992;101: Mansour KA, Lee RB, Miller JI. Tracheal resection: lesson learned. Ann Thorac Surg 1994:57; Klainer AS, Turndorf H, Wu WH, Maewal H, Allender P. Surface alterations due to endotracheal intubation. Am J Med 1975;58: Cooper JD, Grillo HC. Experimental production and prevention of injury due to cuffed tracheal tubes. Surg Gynecol Obstet 1969;129: Grillo HC, Mathisen DJ. Primary tracheal tumors: treatment and results. Ann Thorac Surg 1990;49: GENERAL THORACIC

6 GENERAL THORACIC 1106 MUTRIE ET AL Ann Thorac Surg CERVICAL TRACHEAL RESECTION 2011;91: Grillo HC, Suen HC, Mathisen DJ, Wain JC. Resectional management of thyroid carcinoma invading the airway. Ann Thorac Surg 1992;54: Grillo HC, Mark EJ, Mathisen DJ, Wain JC. Idiopathic laryngotracheal stenosis and its management. Ann Thorac Surg 1993;56: Bain WM, Harrington JW, Thomas LE, et al. Head and neck manifestations of gastroesophageal reflux disease. Laryngoscope 1983;93: Koufman JA. The otolaryngologic manifestations of gastroesophageal reflux disease (GERD). Laryngoscope 1991;101: Walner DL, Stern Y, Gerber ME, et al. Gastroesophageal reflux in patients with subglottic stenosis. Arch Otolaryngol Head Neck Surg 1998;124: Jindal JR, Milbrath MM, Shaker R, et al. Gastroesophageal reflux disease as a likely cause of idiopathic subglottic stenosis. Ann Otol Rhinol Laryngol 1994;103: Grillo HC. Surgical treatment of postintubation tracheal stenosis. J Thorac Cardiovasc Surg 1979;78: Grillo HC, Dignan EF, Miura T. Extensive resection and reconstruction of mediastinal trachea without prosthesis or graft: an anatomical study in man. J Thorac Cardiovasc Surg 1964;48: Mulliken JB, Grillo HC. The limits of tracheal resection with primary anastomosis: further anatomical studies in man. J Thorac Cardiovasc Surg 1968;55: Young-Beyer P, Wilson RS. Anesthetic management for tracheal resection and reconstruction. J Cardiothorac Anesth 1988;2: Keeling B, Hernandez, Lewis V, et al. Increased age is an independent risk factor for radiographic aspiration and laryngeal penetration after thoracotomy for pulmonary resection. J Thorac Cardiovascular Surg 2010;140: DISCUSSION DR DAVID R. JONES (Charlottesville, VA): Dr Mutrie, you did a wonderful job presenting the paper and represented your coauthors very well. Many of you may know that Chris is a general surgery resident, and he is going to continue his interest in tracheal surgery as a resident at MGH in thoracic surgery. As I said, I enjoyed your paper. The notable findings that I took away were that you had no indication or chose not to do any kind of suprahyoid release, yet you had a team approach with ear, nose, and throat (ENT) not infrequently. I thought the modified barium swallow at the end of the hospitalization was a great idea, and we certainly are doing that as well. I have just a few questions for you. What technique do you currently use for the true subglottic stenosis? And second, for the GERD patients, which were maybe 11 or 12 in your series, how did you sort out that it was truly related to GERD? What was the time interval that you may have used after dealing with the tracheal problem to then deal with the reflux problem with an antireflux procedure, for instance? Do you ever perform tracheoscopy before they leave the hospital or do you just send them out on postoperative day 4 as long as they are asymptomatic, in other words, have no crepitus in their neck, and so forth? And finally, given that this was a retrospective review, what was the impetus, if you will, to make some of the changes to your technique and to your postoperative care? Things seemed to be going pretty good from the first series, so why did you change things? Again, I enjoyed the paper. DR MUTRIE: Thank you very much, Dr Jones. I appreciate your comments. To address these questions in reverse order, very interestingly, the impetus to change the technique I think was addressed in our first series in 1993 when our average hospital stay ranged from 10 to 14 days, with all of the patients having a chin stitch, as was written by Dr Grillo in their early series. In the interim with the new multimodality approach, we have utilized worked with our ENT colleagues who rarely use the chin stitch, and we found a tremendous improvement in the hospital stay as well as an equally satisfactory result in our patients. And the need for a tracheoscopy, I can t say to be 100% sure, but I do not believe we performed tracheoscopy in any of our patients routinely. And to evaluate our patients with GERD for a possible reflux procedure, that factor also was not analyzed, and Dr Miller may have a bit to add to that. And the technique for our subglottic stenosis, we did a running 4-0 PDS in the posterior layer with an interrupted 3-0 PDS in the anterior layer. DR MILLER: To answer your questions, our ENT colleague, Dr Bill Grist, and ourselves do these together. We do an oblique cut on the cricoid and a lot of cephalad mobilization anteriorly. We have not had to do a hyoid release. And then we use either interrupted 3-0 PDS or 2-0 PDS for the cricoid. We cut obliquely posteriorly to not injure the insertion of the recurrent laryngeal nerve. We do not do a bronchoscopy on the patients before they go home. In regard to the idiopathic patients, all patients get extensively evaluated. First, we perform a rigid bronchoscopy with inspection, measurements, and dilation. After the rigid bronchoscopy, we evaluate them with a 24-hour ph and motility study. Of the 18 patients in our series, 10 of those went on within a 6-month period to have laparoscopic Nissen fundoplications. The remaining 8 patients were treated with proton-pump inhibitors. DR ROBERT J. CERFOLIO (Birmingham, AL): I have been cutting the distal trachea so there is a tongue-like projection in its most anterior part to slide that into the upper aspect of the cricoid. I thought you were doing something similar to that, am I wrong? DR MILLER: We bevel the distal trachea slightly, to prevent ischemia. Working with our ENT colleagues is very helpful. DR RODNEY LANDRENEAU (Pittsburgh, PA): One question I had is what were the criteria for doing a resection? We have been really aggressive with dilating these things and trying to get these people through it with that instead of resecting whenever. What were the criteria for a resection? DR MUTRIE: That is also an excellent question. Approximately 76% of our patients required a dilation before our operative intervention, and of these, 60% had multiple dilations. So the two biggest factors were a history of repeated dilations and severity of shortness of breath. DR MILLER: A lot of these were postintubation subglottic stenosis. We usually dilated them once. If you had to dilate them within a 3- to 4-week period, we would proceed with resection unless they had other comorbidities. Morbid obesity was also an issue. Two patients underwent a bariatric procedure before resection. One weighed more than 400 pounds, lost almost 180 over about a 14-month period, and then we did their resection.