Pediatric Endoscopic Airway Management With Posterior Cricoid Rib Grafting

Transcription

1 The Laryngoscope VC 2011 The American Laryngological, Rhinological and Otological Society, Inc. Pediatric Endoscopic Airway Management With Posterior Cricoid Rib Grafting Matthew J. Provenzano, MD; Stephanie L. Hulstein, BA; Donald H. Solomon, MD; Nancy M. Bauman, MD; Jose M. Manaligod, MD; Deborah S. F. Kacmarynski, MD; Richard J. H. Smith, MD Objectives/Hypothesis: To confirm and extend reported successful treatment of posterior glottic stenosis in pediatric patients using endoscopic laser division of the posterior cricoid plate with augmentation using costal cartilage. Study Design: A retrospective chart review and case series. Methods: Medical records were examined to determine the surgical indications, outcomes, and postoperative complications of this procedure. Results: Twelve patients underwent the procedure, six females and six males, with an average age of 7 years (range, 2 26 years). There were 8/12 (67%) patients successfully decannulated after being tracheostomy dependent. There were no consistent anatomic abnormalities or surgical findings predictive of failure to decannulate. Average hospital stay was 3.6 days (range, 2 9 days). There were no deaths or other major complications; one patient had extrusion. Conclusions: Endoscopic posterior cricoid grafting is a valuable surgical option for patients with posterior glottic stenosis. The procedure is associated with low morbidity and permits decannulation in the majority of patients. Key Words: posterior glottic expansion, costal cartilage, posterior glottic stenosis. Level of Evidence: 4. Laryngoscope, 121: , 2011 INTRODUCTION The primary goal of pediatric airway surgery is to provide the young patient with an airway that allows for tracheotomy-free respiration, appropriate phonation, and normal swallowing. Treatment for glottic and subglottic stenosis has traditionally involved open procedures such as anterior or posterior cricoid expansion. Although these procedures have high rates of decannulation, the patients often have significant postoperative morbidity. Postoperative vocal cord malposition resulting from the laryngofissure necessary to access the posterior cricoid can leave the patient with phonation difficulties. There is now emerging evidence demonstrating phonation changes in pediatric patients who have undergone open airway reconstruction. 1 Open procedures also leave external scaring and likely have longer hospitalizations compared to endoscopic alternatives. We have encountered a number of these postoperative difficulties in our experience with open airway procedures in children. Determining the correct vocal From the Departmentof Otolaryngology Head and Neck Surgery (M.J.P., S.L.H., D.H.S., J.M.M., D.S.F.K., R.J.H.S.), University of Iowa, Hospital and Clinic, Iowa City, Iowa; and the Department of Otolaryngology (N.M.B.), Children s National Medical Center, Washington, District of Columbia, U.S.A. Editor s Note: This Manuscript was accepted for publication December 7, The authors have no funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Richard J. H. Smith, MD, Department of Otolaryngology Head and Neck Surgery, University of Iowa Hospitals and Clinics, 200 Hawkins Drive, Iowa City, IA richard-smith@uiowa.edu DOI: /lary cord position after laryngofissure is often difficult and results in unpredictable voice outcomes. Some patients have significant hospital length of stays, including initial observation in the intensive care unit. 2 Given the postoperative morbidity of open procedures, innovative surgical procedures have been developed. In 2003, Inglis and colleagues introduced laser-assisted, endoscopic posterior cricoid expansion for patients with bilateral vocal cord paralysis and posterior glottic-subglottic stenosis. 3 Their work showed that the procedure carried minimum morbidity while allowing for decannulation and improved speech and swallowing. 3 The technique, which built on a procedure introduced by Gray et al. in 1994, 4 has now been cited in a case series including the treatment of a 3-week-old neonate. 5 In this study, we present the largest series of patients who have undergone laser-assisted, endoscopic posterior cricoid expansion with chondral cartilage. The decannulation rate was high, with minimal intensive care unit time. Postoperative phonation and deglutition were excellent. Results were consistent across all airway abnormalities, confirming the Inglis study and demonstrating the utility of this procedure in the treatment of pediatric airway abnormalities. MATERIALS AND METHODS After institutional review board approval, the medical charts of patients undergoing endoscopic posterior cricoid grafting from 2003 to 2010 were examined. Patient demographics, procedure information, and outcomes as presented in the Results section were collected and analyzed with Microsoft Excel (Microsoft Corp., Redmond, WA).

2 To harvest costal cartilage, a 3-cm incision is made medial to the synchondrosis approximately one finger breadth above and parallel to the inferior margin of the rib cage. The soft tissues are retracted to expose the underlying cartilage, which is marked and harvested with its external perichondrium. Elevating the internal perichondrium from the cartilage as the graft is harvested minimizes the possibility of a pneumothorax. After taking the required amount of cartilage, the wound is irrigated under positive pressure to confirm the absence of a pneumothorax and then closed, carefully approximating the muscle fascia to minimize postoperative pain. At the conclusion of the entire procedure, local anesthetic is infiltrated around the incision. The patient is then placed in suspension using a Lindholm laryngoscope. Draping with moistened towels and minimizing the FiO 2 are standard laser precautions used in all laser procedures at our institution. To optimize access to the pediatric larynx, we have customized a number of instruments for use in this procedure. We use a modified laryngeal spreader to visualize the entire depth of the cricoid (Fig. 1). The intra-arytenoid musculature is pushed posteriorly with the modified spatula, thereby exposing the superior aspect of the cricoid. The CO 2 laser is used to divide the overlying mucosa and the cricoid cartilage, working slowly from superior to inferior to divide the Fig. 1. A number of instruments were modified for use in pediatric endoscopic airway procedures. The smaller retractors allow for greater access for the laser and other instruments. (A) Viewed from above, the adult and laryngeal spreaders have the same profile. The adult spreaders are on the right. (B) The adult spreaders are on are bottom and the modified pediatric spreaders on top. The length of the spreader blades measures approximately 1 cm in both. (C) The pediatric spreader is on top. The blade height measures approximately 3 mm. The blade on the adult spreader (bottom) measures approximately 5 mm in height. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] Our technique followed the procedure as described by Inglis et al., 3 with some modifications. In brief, the diagnosis of posterior glottic stenosis or vocal cord paralysis/fixation is confirmed by a functional evaluation of the larynx in the clinic and operating room. On the day of endoscopic posterior cricoid expansion, a tracheotomy is performed if not already present. Autologous rib harvest is the preferred grafting material, although on occasion, we use cadaveric cartilage. Both options work without difficulty and provide sufficient material for carving multiple grafts. Fig. 2. The modified pediatric spatula is smaller than its adult counterpart and aids in distraction of the cricoid and placement of the cartilage graft. (A) The two spatulas have the same length. (B) The length of the working end of the two spatulas is approximately the same (3 cm). (C) The pediatric spatula (left) has a height of 3 mm, whereas the height of the adult spatula (right) is 5 mm. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] 1063

3 Fig. 3. Preoperative (A) and postoperative (B) photographs demonstrate the significant posterior expansion after graft placement. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] cricoid plate. Progress is assessed by palpation using the modified pediatric spatula. Once the superior/posterior aspect of the cartilage is divided, a laser platform can be used to protect the esophageal mucosa. We also use a right-angle hook on the superior aspect of the cricoid to rotate the cricoid plate and provide better exposure for the laser. Once the cricoid plate is divided, the pediatric spatula is used to distract the halves (Fig. 2). The costal cartilage is then carved. Prior to placement, a safety suture is placed through the graft to aid in its retrieval if repeated attempts at placement are required. Depending on the size of the harvested cartilage, I-beam or keystone shapes can be fashioned, although we prefer the I-beam shape because it firmly locks in position. The pediatric spatula is used to push the graft into place, a maneuver that often takes considerable force. The safety suture is then cut and photographs are taken (Fig. 3). Postoperatively, the majority of patients are observed in the pediatric intensive care unit. Decannulation is performed at a subsequent procedure after a corking trial with a fenestrated tracheostomy tube. RESULTS Twelve patients underwent this procedure (six females, six males) (Table I). The average age at surgery was 7 years (range, 2 26 years), with follow-up ranging from 1 to 5 years (Table II). Included was one adult patient who underwent the same procedure using the TABLE I. Patient Demographics, Diagnosis, and Outcomes. Gender Age, yr B-O* Type VC Paralysis Subglottic Stenosis Grade Previous Anterior Graft Additional Findings Decannulated ICU (d) Length of Hospitalization (d) F 8 4 No III Yes Yes 0 3 M 2 2 Yes III Yes Laryngeal web Yes 1 2 F 2 2 Yes II Yes Bilateral furrows and No 4 9 granulation tissue F 17 4 Yes IV Yes Yes 0 4 M 26 4 Yes None Yes No 0 2 M 3 2 Yes III No Yes 1 2 F 5 1 Yes I No Type 3 laryngomalacia Yes 2 3 M 6 4 Yes III Yes Yes 6 8 M 3 4 Yes None No Tracheomalacia No 1 2 M 9 4 Yes III No Tracheomalacia Yes 1 2 F 6 3 Yes None No Tracheomalacia Yes 1 2 F 2 2 No III No No 0 4 *Bogdasarian and Olson classification of posterior glottic stenosis. New tracheostomy at time of cricoid expansion. VC ¼ vocal cord; ICU ¼ intensive care unit; F ¼ female; M ¼ male. 1064

4 TABLE II. Outcomes Data. Average 6 SD (Range) Gender (M/F), % 50/50 Age, yr (2 26) Follow-up, mo (2 80) Decannulation rate (n¼8), % 67 Length of ICU stay, d Including new tracheostomies (1 6) Excluding new tracheostomies (1 2) Length of hospitalization, d Including new tracheostomies (2 9) Excluding new tracheostomies (2 4) SD ¼ standard deviation; M ¼ male; F ¼ female; ICU ¼ intensive care unit. same technique as in the pediatric patients. Most patients had multiple upper airway abnormalities. Ten patients had a pre-existing tracheostomy, and in the remaining two patients, a tracheostomy was placed in conjunction with the procedure. Eight patients (67%) were successfully decannulated. One half of patients had type 4 Bogdasarian-Olson posterior glottic stenosis. 6 Three of the patients yet to be decannulated have bilateral vocal cord paralysis and varying degrees of posterior glottic stenosis. Repeat direct laryngoscopic evaluation of their airways has confirmed posterior cricoid expansion, and decannulation is planned. One patient who has not undergone decannulation had a recent posterior graft. Decannulation is expected. None has yet undergone a corking trial with a fenestrated tracheotomy tube, which is our practice prior to decannulation. 7 One patient who was decannulated subsequently developed additional scar tissue distal to the posterior graft site and required a second tracheostomy. A number of patients also had open anterior laryngotracheal reconstruction prior to endoscopic cricoid expansion. Six patients had balloon dilations prior to posterior cricoid expansion (average number, 6.2; range, 1 19). Hospitalization after graft placement averaged 3.6 days (range, 2 9 days) including eight patients observed in the intensive care unit for an average of 2 days (range, 1 6 days). Excluding the two patients who underwent tracheostomy in conjunction with grafting, the average hospitalization was 2.6 days (range, 2 4 days), with 1 day in the intensive care unit (Table II). The first patient to undergo this procedure had extrusion of the graft 7 days after the procedure and subsequently underwent a repeat procedure. There were no deaths or major complications from the procedure. DISCUSSION As described by Inglis and colleagues, endoscopic posterior cricoid expansion offers a number of advantages over an open procedure. Most importantly, it obviates the need for an anterior laryngofissure and the associated potential morbidities of true vocal cord malalignment, anterior glottic webbing, and an anterior cervical scar. Our series of 12 patients is the largest yet published and confirms and extends previous findings. Included in our series are an adult (age 26 years) and an older pediatric patient. Inclusion of these two patients demonstrates the applicability of this technique to a wide patient population. This series, together with the reports of Inglis and Thakkar, provides data on 24 patients. Our experience has shown the procedure to be safe and associated with few complications. Although initially concerned about potential donor site morbidity, esophageal injury, changes to speech and swallowing, or infection, we encountered no major or minor complications aside from graft extrusion after the first procedure. Postoperative flexible and direct laryngoscopy demonstrated rapid graft mucosalization. Our experience compares favorably to open procedures that have reported complications including dysphonia and arytenoid prolapse. 8 Initially, we used adult laryngeal instruments, but the size of these instruments limited access, visualization, and manipulation of the cricoid cartilage. We now use modified instruments including a smaller laryngeal spreader (Fig. 1A and 1B). The reduced profile of the pediatric laryngeal spreader increases visualization and minimizes trauma to surrounding tissue. Rotation of the cricoid cartilage is facilitated with a right angle hook. The pediatric spatula aids in distraction of the cricoid halves, a step we found difficult with the adult-sized instruments. Our modified instruments have the requisite strength to manipulate the cricoid cartilage and to place the graft into position with greater ease. Two patients received a posterior graft fashioned from cadaveric cartilage for confounding medical reasons. Although our experience is limited, carving, placement, and postoperative outcomes appear to be no different when comparing patients with cadaveric and autologous grafts. Nevertheless, based on our long-term experience with anterior costal grafting to expand the trachea and subglottis, we believe that autologous cartilage is preferable. A tracheostomy was present in 10 of 12 patients prior to the procedure, with the remaining two undergoing tracheostomy during graft placement. Eight (67%) patients have been decannulated; the four remaining patients are expected to be decannulated after a corking trial using a fenestrated tracheostomy tube. This result is comparable to that reported by Inglis and colleagues, who decannulated five of seven patients (71%) in their series. Similar to our experience, the remaining patients in their series demonstrate an improved airway. Taken together, 13 of 19 (68%) patients have been decannulated after posterior cricoid expansion. This decannulation rate is lower than the 97% reported in some open procedure series; however, we anticipate that all of our patients will eventually be decannulated. 8 A longer follow-up period will allow for accurate comparison between open and endoscopic decannulation rates. One patient in our series was successfully decannulated but later developed stenosis distal to the posterior 1065

5 cricoid graft site and required a second tracheotomy. This patient had a subsequent distal airway expansion procedure and has now been decannulated for a second time. Our practice of requiring a tracheotomy for endoscopic posterior cricoid grafting is in contrast to both Inglis and Thakkar, who have performed posterior cricoid expansion in patients without a tracheostomy. Thakkar and Gerber report expansion in a 3-week-old infant who weighed 3.4 kg at birth. We believe that cricoid expansion without tracheostomy should be reserved for select patients without additional airway abnormalities and after gaining experience with this procedure. Inglis and colleagues have commented on the difficulty of decannulating patients with high grade III and grade IV subglottic stenosis. Only one of our patients had a grade IV stenosis, whereas six had a grade III stenosis. We did not find that grade III or IV stenosis precludes successful decannulation. Our single grade IV stenosis was in a 17-year-old female who required a posterior cricoid graft and additional anterior airway procedures prior to decannulation. This patient illustrates the importance of using posterior cricoid expansion as a technique to address a specific portion of the airway. It should be used in conjunction with additional airway procedures focused on the other airway abnormalities that are frequently present in these patients. Neither the severity of the posterior glottic pathology as graded using the Bogdasarian-Olson classification nor the dynamic status of the vocal folds (paralyzed or not paralyzed) correlated with decannulation outcome, although the number of patients in all series is too small for statistical analysis. We documented no phonation or swallowing difficulties, with some parents reporting improved phonation in their children. Inglis and colleagues also reported that many patients had improved phonation, feeding, and exercise tolerance. Future studies should examine quality-of-life measures, pre- and postoperative voice quality, and feeding outcome. CONCLUSION Advancements in pediatric airway management have seen the development of a number of endoscopic techniques, one of which is laser-assisted posterior cricoid expansion. Using modified pediatric laryngeal instruments, this technique is a reliable and safe method to address posterior glottic pathology. Morbidity is minimal, and in conjunction with other airway procedures, decannulation is possible in the majority of tracheotomydependent children. BIBLIOGRAPHY 1. Kelchner LN, Weinrich B, Brehm SB, Tabangin ME, de Alarcon A. Characterization of supraglottic phonation in children after airway reconstruction. Ann Otol Rhinol Laryngol 2010;119: Cable BB, Manaligod JM, Bauman NM, Smith RJ. Pediatric airway reconstruction: principles, decision-making, and outcomes at the University of Iowa hospitals and clinics. Ann Otol Rhinol Laryngol 2004;113: Inglis AF Jr, Perkins JA, Manning SC, Mouzakes J. Endoscopic posterior cricoid split and rib grafting in 10 children. Laryngoscope 2003;113: Gray SD, Kelly SM, Dove H. Arytenoid separation for impaired pediatric vocal fold mobility. Ann Otol Rhinol Laryngol 1994;103: Thakkar K, Gerber ME. Endoscopic posterior costal cartilage graft placement for acute management of pediatric bilateral vocal fold paralysis without tracheostomy. Int J Pediatr Otorhinolaryngol 2008;72: Bogdasarian RS, Olson NR. Posterior glottic laryngeal stenosis. Otolaryngol Head Neck Surg 1980;88: Merritt RM, Bent JP, Smith RJ. Suprastomal granulation tissue and pediatric tracheotomy decannulation. Laryngoscope 1997;107: Rutter MJ, Cotton RT. The use of posterior cricoid grafting in managing isolated posterior glottic stenosis in children. Arch Otolaryngol Head Neck Surg 2004;130: