- Email: [email protected]
Subglottic Stenosis in Granulomatosis With Polyangiitis: The Role of Laryngotracheal Resection
Subglottic Stenosis in Granulomatosis With Polyangiitis: The Role of Laryngotracheal Resection Christina L. Costantino, MD, John L. Niles, MD, Cameron D. Wright, MD, Douglas J. Mathisen, MD, and Ashok Muniappan, MD Division of Thoracic Surgery, Massachusetts General Hospital, Harvard Medical School, Boston; and Division of Nephrology, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts
Background. Granulomatosis with polyangiitis (GPA) is associated with development of subglottic stenosis in about one-fourth of all patients. Although endoscopic management is the primary treatment method for tracheobronchial stenosis, some patients have refractory disease, and tracheostomy is required. It is unclear if laryngotracheal resection and reconstruction (LTRR) can be safely performed in patients with GPA. Methods. A retrospective review was performed of 11 patients with GPA undergoing LTRR. Results. Eleven female patients with GPA and a median age of 47 years underwent LTRR. Six patients were diagnosed with GPA after LTRR and had not received any induction immunosuppression regimen. Five patients had received induction immunosuppression regimen and were in clinical remission before LTRR. LTRR was performed with a protective tracheostomy in 3 patients, which was eventually removed in all. There
were no major complications and no postoperative deaths. One patient (9%) failed surgical management and had replacement of a permanent tracheostomy 4 months after LTRR. Six patients (55%) required additional tracheal dilations after LTRR. Ten patients (91%) had durable control of symptoms and freedom from tracheostomy with a median follow-up of 9.7 years. Two patients (18%) experienced subsequent lower airway stenoses. Conclusions. Surgical treatment of subglottic stenosis in highly selected patients with GPA is effective and associated with minimal morbidity. Although long-term outcomes are encouraging, additional procedures may be necessary, and patients are at risk of experiencing lower airway disease.
G
GPA was almost always fatal before the advent of immunosuppression regimens that lead to remission in most patients. Different regimens are chosen for inducing and maintaining remission [3]. Airway complications such as symptomatic SGS are typically treated with bronchoscopic techniques, and they include dilation, laser debridement, and corticosteroid injection. Patients with refractory SGS require tracheostomy placement. Surgical resection, generally laryngotracheal resection and reconstruction (LTRR), of SGS in patients with GPA is rarely performed because of the concern for anastomotic complications and questionable long-term efficacy. The aim of this study is to characterize patient selection, immunosuppressive management, and outcomes associated with LTRR in patients with GPA.
ranulomatosis with polyangiitis (GPA, formerly Wegener’s granulomatosis) is a systemic disorder characterized by necrotizing vasculitis of small arteries, necrotizing granulomatous inflammation of the upper and lower respiratory tracts, and necrotizing glomerulonephritis. GPA is a member of a family of small vessel vasculitides referred to as antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis [1]. Limited GPA is found in approximately one-fourth of GPA patients and refers to disease that is restricted to the airways, but it can also be associated with ocular and sinus involvement. The term “limited” is a misnomer because disease severity can be considerable. Approximately 15% to 25% of all GPA patients will experience subglottic stenosis (SGS) [2]. Pathologic assessment of airway disease often reveals nonspecific fibrosis and infiltration by neutrophils and epitheliod histiocytes without overt vasculitis.
(Ann Thorac Surg 2017;-:-–-) Ó 2017 by The Society of Thoracic Surgeons
Patients and Methods Patients Accepted for publication July 17, 2017. Presented at the Poster Session of the Fifty-third Annual Meeting of The Society of Thoracic Surgeons, Houston, TX, Jan 21–25, 2017. Address correspondence to Dr Muniappan, 55 Fruit St, Blake 1570, Boston, MA 02114; email: [email protected].
Ó 2017 by The Society of Thoracic Surgeons Published by Elsevier Inc.
This study was approved by the Institutional Review Board of the Massachusetts General Hospital (MGH), who determined that patient consent was unnecessary. The study population consisted of all patients undergoing tracheal operation at MGH from January 1988 through 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2017.07.026
2
COSTANTINO ET AL LARYNGOTRACHEAL RESECTION AND GPA
Ann Thorac Surg 2017;-:-–-
December 2016. Patients were included in this study if they underwent tracheal resection for SGS and had a diagnosis of GPA. Six patients, all from 1988 to 2001, were diagnosed with GPA when ANCA testing was found to be positive after LTRR was performed. Five patients, all from 2002 and onward, had a preoperative diagnosis of GPA.
complications and efficacy of surgical management. Long-term follow-up was achieved by reviewing the most recent clinical note or contacting the patient to determine subsequent need for immunosuppression regimen, additional airway interventions, and quality of breathing.
GPA Diagnosis
Results
All patients met the Chapel Hill Consensus Guidelines for ANCA-associated vasculitis based on ANCA serologic or clinicopathologic findings [1]. Serologic testing of ANCA antibody consisted of both indirect immunofluorescence detection of cytoplasmic staining (C-ANCA) and enzyme-linked immunosorbent assay (ELISA) quantification of the titers of anti-proteinase 3 (PR3) and antimyeloperoxidase (MPO) antibodies [4].
Patient Characteristics
Surgical Management All patients underwent a single-stage LTRR using a technique previously described by our group for management of idiopathic laryngotracheal stenosis [5]. The standard procedure involved removal of the anterior and about one-half of the lateral cricoid cartilage in addition to resection of the involved proximal trachea to enlarge the subglottic lumen. Lateral heavy polyglactin 910 (2-0) stay sutures were placed, followed by interrupted circumferential fine polyglactin 910 (4-0) sutures, after which time the airway was approximated and the sutures were sequentially tied. The anastomosis was buttressed with strap muscle or thyroid gland. A protective tracheostomy was performed at the time of LTRR in a minority of patients (3 of 11 patients) when there was concern about edema at the level of the anastomosis or glottis. All patients underwent routine surveillance bronchoscopy approximately 1 week after the operation.
Outcomes and Follow-Up A retrospective electronic and paper chart review of all patients was performed to determine postoperative
There were 11 patients with GPA who underwent LTRR at our institution in the period from 1988 to 2016 (Table 1). The median age at onset of airway symptoms was 35 years (range: 17 to 47 years), and the median age at operation was 47 (range: 27 to 63 years). All patients were women. The primary symptoms related to upper airway disease were dyspnea on exertion (n ¼ 6, 55%) and stridor (n ¼ 5, 45%). Six patients (55%) were diagnosed with GPA after undergoing LTRR for a presumed diagnosis of idiopathic laryngotracheal stenosis. Five of these patients had positive ANCA antibody levels identified after operation. All tracheal resection specimens exhibited nonspecific fibrosis or inflammation and were free of granulomatous or vasculitic changes. A sixth patient did not have any ANCA testing available but underwent sleeve resection for left main-stem stenosis 1 year after LTRR. The pathologic specimen revealed changes consistent with the healed phase of GPA. Five other patients (45%) had a diagnosis of GPA before undergoing LTRR. They were all considered to be in clinical remission before LTRR and had been weaned completely from immunosuppressive drugs before operation. All patients with previous history of GPA had negative ANCA levels just before operation. All but one patient in this series had limited GPA with a primary clinical feature of SGS. The exception was a patient who had undergone kidney transplantation for GPA glomerulonephritis and then experienced SGS. Of the 10 patients with limited GPA, 5 patients had additional involvement of sinus, eye, or lung.
Table 1. Characteristics of 11 Patients With Granulomatous Polyangiitis Undergoing Surgical Treatment of Subglottic Stenosis at Massachusetts General Hospital Preoperative Diagnosis Patient of GPA 1 2 3 4 5 6 7 8 9 10 11
No No No No No No Yes Yes Yes Yes Yes
ANCA
Other Organs
Positive Positive Positive Positive Unknown Positive Unknown Unknown Negative Positive Positive
Nasal septum, eye, joints Lung None None None Sinus, joints None Lung, sinus, nasal septum Kidney, sinus, joints Eye None
ANCA ¼ anti-neutrophil cytoplasmic antibody;
Prior Interventions
Age at Onset of Airway Symptoms, Years
Age at Operation, Years
22 Unknown 43 46 29 47 35 17 27 35 39
46 63 45 49 36 49 50 27 47 42 49
Dilations, steroid injection Tracheal dilations None None Laser debridement, tracheostomy Dilations Dilations Dilations, T-tube, stents Laser debridement, T-tube Dilations, tracheostomy Dilations
GPA ¼ granulomatous polyangiitis;
T-tube ¼ tracheal T-tube.
Ann Thorac Surg 2017;-:-–-
COSTANTINO ET AL LARYNGOTRACHEAL RESECTION AND GPA
Most patients (n ¼ 9, 82%) had undergone bronchoscopic tracheal interventions before LTRR. Interventions included tracheal dilation (n ¼ 7 patients, 64%), steroid injection (n ¼ 1, 9%), laser debridement (n ¼ 2, 18%), and tracheal stenting (n ¼ 1, 9%). Most patients had multiple interventions, and 4 patients (36%) had undergone more than 15 dilations before LTRR. Three patients (27%) had prior tracheostomies, one of which was in place at the time of LTRR.
Operative Management All patients underwent LTRR to correct SGS and any stricture related to prior tracheal interventions. A singlestage laryngotracheal resection with primary anastomosis was used in all patients, and 3 patients (27%) had a protective tracheostomy placed distal to the anastomosis to manage anticipated airway and laryngeal edema (Table 2). All patients were decannulated before discharge, one with a tracheal button in place that was subsequently removed as an outpatient.
Complications There were no major postoperative complications and no postoperative deaths (Table 2). Two patients were reintubated to manage transient airway swelling and were promptly extubated. One patient required tube thoracostomy for pneumothorax. One patient required bronchoscopy and debridement of anastomotic granulation tissue. Another patient was re-admitted for a wound infection that required incision and drainage.
Long-Term Outcomes The median follow-up was 10.9 years (range: 4 months to 28 years) after LTRR (Table 2). Ten of 11 patients (91%) were considered to have a good airway outcome, with control of symptoms related to their prior SGS and avoidance of tracheostomy. One patient (9%) was considered an operative failure, because a tracheostomy tube was replaced 2 months after LTRR. She remained cannulated 8 months later. This patient had relatively
3
severe GPA (eye enucleation, SGS leading to tracheostomy) that was refractory to induction immunosuppression regimen with cyclophosphamide and abatacept but ultimately responded to rituximab. She was decannulated before operation, and the ANCA antibody titer resolved after rituximab treatment. She was the only patient with a history of PR3 antibody, whereas all other patients had MPO antibodies. Her SGS recurred without an obvious anastomotic complication and despite perioperative immunosuppression regimen, including rituximab. Additional tracheal dilations were necessary for 6 patients (55%) after LTRR. The median number of dilations in these 6 patients was two (range: 1 to 8 dilations). One patient presented with left main-stem bronchial obstruction 1 year after LTRR and underwent successful sleeve reconstruction. This patient was free from recurrent tracheal stenosis. Four patients (36%) had no further airway interventions after LTRR. Two of these 4 patients had a positive ANCA antibody test after LTRR and did not receive any immunosuppressive drugs. One patient died 16 years after LTRR with progressive pulmonary and lower airway GPA. This patient required multiple tracheal dilations for 2 years after LTRR but did not have recurrent SGS before death. Seven patients (64%) received immunosuppression regimen after LTRR. Immunosuppression regimens included prednisone, adalimumab, azathioprine, tacrolimus, and rituximab. Rituximab was used to manage suspected recrudescence of GPA in 2 patients and to prevent reactivation of GPA in 1 patient. One patient received tacrolimus after undergoing a second renal transplantation.
Comment In an earlier review of our experience with LTRR, we concluded that open surgical treatment of SGS in patients with rheumatologic disease such as GPA should be avoided [5]. The primary concerns were disease relapse,
Table 2. Outcomes Associated With Laryngotracheal Resection in Patients With Granulomatous Polyangiitis
Patient Operation 1 2 3 4 5 6 7 8 9 10 11 a
LTRR LTRR LTRR LTRRb LTRRb LTRR LTRR LTRR LTRR LTRRb LTRR
Complications
Follow-Up Period (Years)
Delayed Interventions
Reintubation None Wound infection Anastomotic granulation Pneumothorax Reintubation None Vocal cord paresis None None None
27.9 16 18.5 21.8 1 15.9 10.9 9.7 8.8 0.9 0.3
Dilations (n ¼ 8) Dilations (n ¼ 4) None None LMSB sleeve resection Dilations (n ¼ 2) Dilations (unknown number) None Dilation (n ¼ 1) Dilation (n ¼ 1) None
Deceased with left main-stem bronchial stenosis and pulmonary disease.
LMSB ¼ left main-stem bronchus;
b
Protective tracheostomy.
LTRR ¼ laryngotracheal resection and reconstruction.
Status of Airway Good Gooda Good Good Good Good Good Good Good Failure Good
Postoperative Immunosuppression Regimen Adalimumab, methotrexate Prednisone None None None Rituximab Unknown Azathioprine Tacrolimus, prednisone Rituximab, azathioprine Rituximab
4
COSTANTINO ET AL LARYNGOTRACHEAL RESECTION AND GPA
anastomotic complications, and the need for immunosuppressive drugs that could impair healing. The cornerstone of management of SGS developing in patients with GPA remains bronchoscopic intervention, whereas refractory disease might require tracheostomy. Although clinical remission rates and overall outcomes are markedly improved with newer immunosuppression regimens, SGS may prove refractory to pharmacologic management alone [3, 6, 7]. In a highly selected group of patients with GPA and SGS, 91% of patients achieved a good airway outcome after LTRR and avoided tracheostomy during long-term follow-up. Four patients (36%) required no further dilation or airway intervention and had an excellent outcome. Seven patients (64%) required additional bronchoscopic airway intervention, only one (9%) of which did not resolve or stabilize, necessitating a tracheostomy. These outcomes were achieved without major complications and with no postoperative deaths. One of the earliest reports suggesting that LTRR could be safely performed in patients with GPA was by Herridge and colleagues [8], who described their experience with 3 patients. Although long-term follow-up was only available for one of the patients, the investigators concluded that the operation could be performed safely as long as patients were in remission and received perioperative immunosuppression regimen. A more recent report by Wester and colleagues [9], who had a series of 8 GPA patients undergoing open surgical management, observed that 75% of these patients required additional tracheal dilation and only 1 patient (13%) required permanent tracheostomy, which is similar to our observed rates of 55% and 9%, respectively. These rates are clearly higher than observed for patients undergoing LTRR for idiopathic laryngotracheal stenosis and suggests that even with GPA remission and maintenance immunosuppression regimen, there is likely a chronic phase of the disease that affects long-term outcomes. In addition to concern about recurrent SGS, a small proportion of GPA patients appear to be also susceptible to lower airway stenosis. In our experience, 2 patients (18%) went on to experience left main-stem stenosis 1 and 9 years after LTRR, respectively. Guardiani and colleagues [10] also noted that 18% of their patients with GPA and SGS experienced lower airway disease. The risk of developing lower airway disease in GPA patients after LTRR mandates long-term clinical follow-up to ensure early diagnosis and effective management. Although bronchoscopic management of lower airway stenosis is the mainstay of treatment, on occasion sleeve resection may be successfully performed, as in our experience with 1 patient and the report by Soo and colleagues [11]. The precise role of ANCA antibody testing in management of SGS in GPA has yet to be defined. Patients who experience SGS seem to represent a distinct subset of all patients with GPA and are known to have a higher incidence of negative ANCA testing [12]. A high index of suspicion needs to be maintained in patients undergoing evaluation for SGS, and clinical manifestations of rheumatologic disease, such as sinus, ocular, pulmonary, or
Ann Thorac Surg 2017;-:-–-
renal involvement, must be thoroughly investigated. Two patients in our early experience with LTRR had positive ANCA tests shortly after operation and have had excellent long-term outcomes (18 and 21 years of follow-up) without any immunosuppression regimen. We consider these patients to be the exception to the rule and have adhered to the standard that all patients with GPA undergoing LTRR must have negative ANCA testing at the time of operation. There is likely a role for serial monitoring of ANCA antibody levels with PR3 and MPO ELISA to guide maintenance immunosuppression regimen, as explained by Han and colleagues [13]. It remains to be seen if this contemporary strategy for managing postoperative immunosuppression regimen will improve long-term outcomes. Because ANCA levels do not absolutely correlate with disease activity in GPA, bronchoscopic examination to rule out active inflammation is recommended before LTRR. In our experience, all patients were free of bronchoscopic signs of inflammation such as airway edema and hyperemia (Fig 1). In our 1 patient who was an operative failure, there was an early recurrence of SGS that necessitated a tracheostomy and was likely related to reactivation of her GPA. Her ANCA/PR3 levels rose several weeks after LTRR and did not decline with resumption of rituximab. PR3 antibody was only detected in this 1 patient, whereas all other patients had MPO antibody. Han and colleagues [13] noted that the presence of PR3 antibody is associated with increased risk of relapse and disease severity. An intriguing report on the utility of magnetic resonance imaging in assessing for granulomatous involvement of the airway and
Fig 1. Bronchoscopic image of subglottic stenosis in patient with granulomatosis with polyangiitis before laryngotracheal resection. The stenosis is mature and free of inflammation.
Ann Thorac Surg 2017;-:-–-
inflammation suggests that imaging might also have a role in assessing a patient’s suitability for LTRR [14]. Although this is the largest report of GPA patients undergoing LTRR to date, the relatively small number of patients and large time span in which care was delivered prevent more rigorous assessment of patient variables that may affect outcomes. The disparate immunosuppression regimens used in the study period also makes it difficult to draw conclusions about ideal medical management of GPA patients with SGS. Nonetheless, we conclude that there are some patients who might benefit from LTRR when bronchoscopic interventions become more frequent or when tracheostomy is threatened. The ideal patient has limited GPA with airway disease confined to a short segment of the subglottic trachea, evidence of clinical remission for at least 6 months after induction immunosuppression regimen, negative ANCA antibody testing, minimal or no steroid requirement just before operation, and a bronchoscopic examination revealing no active airway inflammation. Patients should be fully informed about the risk of recurrence, need for future airway dilations, and the possibility of subsequent lower airway disease. Although there is no difference in LTRR technique performed for patients with GPA and idiopathic SGS, the need for protective tracheostomy is more likely in GPA patients. Long-term control of the disease requires careful surveillance of clinical variables and ANCA levels as well as a lower threshold to escalate immunosuppression regimen. We have demonstrated that there are reasonable long-term outcomes, including freedom from tracheostomy, in a highly selected group of GPA patients undergoing LTRR.
References 1. Jennette JC, Falk RJ, Bacon PA, et al. 2012 revised International Chapel Hill Consensus Conference Nomenclature of Vasculitides. Arthritis Rheum 2013;65:1–11.
COSTANTINO ET AL LARYNGOTRACHEAL RESECTION AND GPA
5
2. Langford CA, Sneller MC, Hallahan CW, et al. Clinical features and therapeutic management of subglottic stenosis in patients with Wegener’s granulomatosis. Arthritis Rheum 1996;39:1754–60. 3. Wojciechowska J, Krajewski W, Krajewski P, Krecicki T. Granulomatosis with polyangiitis in otolaryngologist practice: a review of current knowledge. Clin Exp Otorhinolaryngol 2016;9:8–13. 4. Niles JL. Antineutrophil cytoplasmic antibodies in the classification of vasculitis. Annu Rev Med 1996;47:303–13. 5. Ashiku SK, Kuzucu A, Grillo HC, et al. Idiopathic laryngotracheal stenosis: effective definitive treatment with laryngotracheal resection. J Thorac Cardiovasc Surg 2004;127: 99–107. 6. Girard C, Charles P, Terrier B, et al. Tracheobronchial stenoses in granulomatosis with polyangiitis (Wegener’s): a report on 26 cases. Medicine (Baltimore) 2015; 94:e1088. 7. Rhee EP, Laliberte KA, Niles JL. Rituximab as maintenance therapy for anti-neutrophil cytoplasmic antibody-associated vasculitis. Clin J Am Soc Nephrol 2010;5:1394–400. 8. Herridge MS, Pearson FG, Downey GP. Subglottic stenosis complicating Wegener’s granulomatosis: surgical repair as a viable treatment option. J Thorac Cardiovasc Surg 1996;111: 961–6. 9. Wester JL, Clayburgh DR, Stott WJ, Schindler JS, Andersen PE, Gross ND. Airway Reconstruction in Wegener’s granulomatosis-associated laryngotracheal stenosis. Laryngoscope 2011;121:2566–71. 10. Guardiani E, Moghaddas HS, Lesser J, et al. Multilevel airway stenosis in patients with granulomatosis with polyangiitis (Wegener’s). Am J Otolaryngol 2015;36:361–3. 11. Soo A, Aziz R, Buckley M, Young V. Bronchoplastic procedure for an unusual indication–Wegener’s granulomatosis. Interact Cardiovasc Thorac Surg 2009;9:530–1. 12. Hoffman GS, Kerr GS, Leavitt RY, et al. Wegener granulomatosis: an analysis of 158 patients. Ann Intern Med 1992;116:488–98. 13. Han WK, Choi HK, Roth RM, McCluskey RT, Niles JL. Serial ANCA titers: useful tool for prevention of relapses in ANCAassociated vasculitis. Kidney Int 2003;63:1079–85. 14. Klink T, Holle J, Laudien M, et al. Magnetic resonance imaging in patients with granulomatosis with polyangiitis (Wegener’s) and subglottic stenosis. MAGMA 2013;26: 281–90.
Background. Granulomatosis with polyangiitis (GPA) is associated with development of subglottic stenosis in about one-fourth of all patients. Although endoscopic management is the primary treatment method for tracheobronchial stenosis, some patients have refractory disease, and tracheostomy is required. It is unclear if laryngotracheal resection and reconstruction (LTRR) can be safely performed in patients with GPA. Methods. A retrospective review was performed of 11 patients with GPA undergoing LTRR. Results. Eleven female patients with GPA and a median age of 47 years underwent LTRR. Six patients were diagnosed with GPA after LTRR and had not received any induction immunosuppression regimen. Five patients had received induction immunosuppression regimen and were in clinical remission before LTRR. LTRR was performed with a protective tracheostomy in 3 patients, which was eventually removed in all. There
were no major complications and no postoperative deaths. One patient (9%) failed surgical management and had replacement of a permanent tracheostomy 4 months after LTRR. Six patients (55%) required additional tracheal dilations after LTRR. Ten patients (91%) had durable control of symptoms and freedom from tracheostomy with a median follow-up of 9.7 years. Two patients (18%) experienced subsequent lower airway stenoses. Conclusions. Surgical treatment of subglottic stenosis in highly selected patients with GPA is effective and associated with minimal morbidity. Although long-term outcomes are encouraging, additional procedures may be necessary, and patients are at risk of experiencing lower airway disease.
G
GPA was almost always fatal before the advent of immunosuppression regimens that lead to remission in most patients. Different regimens are chosen for inducing and maintaining remission [3]. Airway complications such as symptomatic SGS are typically treated with bronchoscopic techniques, and they include dilation, laser debridement, and corticosteroid injection. Patients with refractory SGS require tracheostomy placement. Surgical resection, generally laryngotracheal resection and reconstruction (LTRR), of SGS in patients with GPA is rarely performed because of the concern for anastomotic complications and questionable long-term efficacy. The aim of this study is to characterize patient selection, immunosuppressive management, and outcomes associated with LTRR in patients with GPA.
ranulomatosis with polyangiitis (GPA, formerly Wegener’s granulomatosis) is a systemic disorder characterized by necrotizing vasculitis of small arteries, necrotizing granulomatous inflammation of the upper and lower respiratory tracts, and necrotizing glomerulonephritis. GPA is a member of a family of small vessel vasculitides referred to as antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis [1]. Limited GPA is found in approximately one-fourth of GPA patients and refers to disease that is restricted to the airways, but it can also be associated with ocular and sinus involvement. The term “limited” is a misnomer because disease severity can be considerable. Approximately 15% to 25% of all GPA patients will experience subglottic stenosis (SGS) [2]. Pathologic assessment of airway disease often reveals nonspecific fibrosis and infiltration by neutrophils and epitheliod histiocytes without overt vasculitis.
(Ann Thorac Surg 2017;-:-–-) Ó 2017 by The Society of Thoracic Surgeons
Patients and Methods Patients Accepted for publication July 17, 2017. Presented at the Poster Session of the Fifty-third Annual Meeting of The Society of Thoracic Surgeons, Houston, TX, Jan 21–25, 2017. Address correspondence to Dr Muniappan, 55 Fruit St, Blake 1570, Boston, MA 02114; email: [email protected].
Ó 2017 by The Society of Thoracic Surgeons Published by Elsevier Inc.
This study was approved by the Institutional Review Board of the Massachusetts General Hospital (MGH), who determined that patient consent was unnecessary. The study population consisted of all patients undergoing tracheal operation at MGH from January 1988 through 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2017.07.026
2
COSTANTINO ET AL LARYNGOTRACHEAL RESECTION AND GPA
Ann Thorac Surg 2017;-:-–-
December 2016. Patients were included in this study if they underwent tracheal resection for SGS and had a diagnosis of GPA. Six patients, all from 1988 to 2001, were diagnosed with GPA when ANCA testing was found to be positive after LTRR was performed. Five patients, all from 2002 and onward, had a preoperative diagnosis of GPA.
complications and efficacy of surgical management. Long-term follow-up was achieved by reviewing the most recent clinical note or contacting the patient to determine subsequent need for immunosuppression regimen, additional airway interventions, and quality of breathing.
GPA Diagnosis
Results
All patients met the Chapel Hill Consensus Guidelines for ANCA-associated vasculitis based on ANCA serologic or clinicopathologic findings [1]. Serologic testing of ANCA antibody consisted of both indirect immunofluorescence detection of cytoplasmic staining (C-ANCA) and enzyme-linked immunosorbent assay (ELISA) quantification of the titers of anti-proteinase 3 (PR3) and antimyeloperoxidase (MPO) antibodies [4].
Patient Characteristics
Surgical Management All patients underwent a single-stage LTRR using a technique previously described by our group for management of idiopathic laryngotracheal stenosis [5]. The standard procedure involved removal of the anterior and about one-half of the lateral cricoid cartilage in addition to resection of the involved proximal trachea to enlarge the subglottic lumen. Lateral heavy polyglactin 910 (2-0) stay sutures were placed, followed by interrupted circumferential fine polyglactin 910 (4-0) sutures, after which time the airway was approximated and the sutures were sequentially tied. The anastomosis was buttressed with strap muscle or thyroid gland. A protective tracheostomy was performed at the time of LTRR in a minority of patients (3 of 11 patients) when there was concern about edema at the level of the anastomosis or glottis. All patients underwent routine surveillance bronchoscopy approximately 1 week after the operation.
Outcomes and Follow-Up A retrospective electronic and paper chart review of all patients was performed to determine postoperative
There were 11 patients with GPA who underwent LTRR at our institution in the period from 1988 to 2016 (Table 1). The median age at onset of airway symptoms was 35 years (range: 17 to 47 years), and the median age at operation was 47 (range: 27 to 63 years). All patients were women. The primary symptoms related to upper airway disease were dyspnea on exertion (n ¼ 6, 55%) and stridor (n ¼ 5, 45%). Six patients (55%) were diagnosed with GPA after undergoing LTRR for a presumed diagnosis of idiopathic laryngotracheal stenosis. Five of these patients had positive ANCA antibody levels identified after operation. All tracheal resection specimens exhibited nonspecific fibrosis or inflammation and were free of granulomatous or vasculitic changes. A sixth patient did not have any ANCA testing available but underwent sleeve resection for left main-stem stenosis 1 year after LTRR. The pathologic specimen revealed changes consistent with the healed phase of GPA. Five other patients (45%) had a diagnosis of GPA before undergoing LTRR. They were all considered to be in clinical remission before LTRR and had been weaned completely from immunosuppressive drugs before operation. All patients with previous history of GPA had negative ANCA levels just before operation. All but one patient in this series had limited GPA with a primary clinical feature of SGS. The exception was a patient who had undergone kidney transplantation for GPA glomerulonephritis and then experienced SGS. Of the 10 patients with limited GPA, 5 patients had additional involvement of sinus, eye, or lung.
Table 1. Characteristics of 11 Patients With Granulomatous Polyangiitis Undergoing Surgical Treatment of Subglottic Stenosis at Massachusetts General Hospital Preoperative Diagnosis Patient of GPA 1 2 3 4 5 6 7 8 9 10 11
No No No No No No Yes Yes Yes Yes Yes
ANCA
Other Organs
Positive Positive Positive Positive Unknown Positive Unknown Unknown Negative Positive Positive
Nasal septum, eye, joints Lung None None None Sinus, joints None Lung, sinus, nasal septum Kidney, sinus, joints Eye None
ANCA ¼ anti-neutrophil cytoplasmic antibody;
Prior Interventions
Age at Onset of Airway Symptoms, Years
Age at Operation, Years
22 Unknown 43 46 29 47 35 17 27 35 39
46 63 45 49 36 49 50 27 47 42 49
Dilations, steroid injection Tracheal dilations None None Laser debridement, tracheostomy Dilations Dilations Dilations, T-tube, stents Laser debridement, T-tube Dilations, tracheostomy Dilations
GPA ¼ granulomatous polyangiitis;
T-tube ¼ tracheal T-tube.
Ann Thorac Surg 2017;-:-–-
COSTANTINO ET AL LARYNGOTRACHEAL RESECTION AND GPA
Most patients (n ¼ 9, 82%) had undergone bronchoscopic tracheal interventions before LTRR. Interventions included tracheal dilation (n ¼ 7 patients, 64%), steroid injection (n ¼ 1, 9%), laser debridement (n ¼ 2, 18%), and tracheal stenting (n ¼ 1, 9%). Most patients had multiple interventions, and 4 patients (36%) had undergone more than 15 dilations before LTRR. Three patients (27%) had prior tracheostomies, one of which was in place at the time of LTRR.
Operative Management All patients underwent LTRR to correct SGS and any stricture related to prior tracheal interventions. A singlestage laryngotracheal resection with primary anastomosis was used in all patients, and 3 patients (27%) had a protective tracheostomy placed distal to the anastomosis to manage anticipated airway and laryngeal edema (Table 2). All patients were decannulated before discharge, one with a tracheal button in place that was subsequently removed as an outpatient.
Complications There were no major postoperative complications and no postoperative deaths (Table 2). Two patients were reintubated to manage transient airway swelling and were promptly extubated. One patient required tube thoracostomy for pneumothorax. One patient required bronchoscopy and debridement of anastomotic granulation tissue. Another patient was re-admitted for a wound infection that required incision and drainage.
Long-Term Outcomes The median follow-up was 10.9 years (range: 4 months to 28 years) after LTRR (Table 2). Ten of 11 patients (91%) were considered to have a good airway outcome, with control of symptoms related to their prior SGS and avoidance of tracheostomy. One patient (9%) was considered an operative failure, because a tracheostomy tube was replaced 2 months after LTRR. She remained cannulated 8 months later. This patient had relatively
3
severe GPA (eye enucleation, SGS leading to tracheostomy) that was refractory to induction immunosuppression regimen with cyclophosphamide and abatacept but ultimately responded to rituximab. She was decannulated before operation, and the ANCA antibody titer resolved after rituximab treatment. She was the only patient with a history of PR3 antibody, whereas all other patients had MPO antibodies. Her SGS recurred without an obvious anastomotic complication and despite perioperative immunosuppression regimen, including rituximab. Additional tracheal dilations were necessary for 6 patients (55%) after LTRR. The median number of dilations in these 6 patients was two (range: 1 to 8 dilations). One patient presented with left main-stem bronchial obstruction 1 year after LTRR and underwent successful sleeve reconstruction. This patient was free from recurrent tracheal stenosis. Four patients (36%) had no further airway interventions after LTRR. Two of these 4 patients had a positive ANCA antibody test after LTRR and did not receive any immunosuppressive drugs. One patient died 16 years after LTRR with progressive pulmonary and lower airway GPA. This patient required multiple tracheal dilations for 2 years after LTRR but did not have recurrent SGS before death. Seven patients (64%) received immunosuppression regimen after LTRR. Immunosuppression regimens included prednisone, adalimumab, azathioprine, tacrolimus, and rituximab. Rituximab was used to manage suspected recrudescence of GPA in 2 patients and to prevent reactivation of GPA in 1 patient. One patient received tacrolimus after undergoing a second renal transplantation.
Comment In an earlier review of our experience with LTRR, we concluded that open surgical treatment of SGS in patients with rheumatologic disease such as GPA should be avoided [5]. The primary concerns were disease relapse,
Table 2. Outcomes Associated With Laryngotracheal Resection in Patients With Granulomatous Polyangiitis
Patient Operation 1 2 3 4 5 6 7 8 9 10 11 a
LTRR LTRR LTRR LTRRb LTRRb LTRR LTRR LTRR LTRR LTRRb LTRR
Complications
Follow-Up Period (Years)
Delayed Interventions
Reintubation None Wound infection Anastomotic granulation Pneumothorax Reintubation None Vocal cord paresis None None None
27.9 16 18.5 21.8 1 15.9 10.9 9.7 8.8 0.9 0.3
Dilations (n ¼ 8) Dilations (n ¼ 4) None None LMSB sleeve resection Dilations (n ¼ 2) Dilations (unknown number) None Dilation (n ¼ 1) Dilation (n ¼ 1) None
Deceased with left main-stem bronchial stenosis and pulmonary disease.
LMSB ¼ left main-stem bronchus;
b
Protective tracheostomy.
LTRR ¼ laryngotracheal resection and reconstruction.
Status of Airway Good Gooda Good Good Good Good Good Good Good Failure Good
Postoperative Immunosuppression Regimen Adalimumab, methotrexate Prednisone None None None Rituximab Unknown Azathioprine Tacrolimus, prednisone Rituximab, azathioprine Rituximab
4
COSTANTINO ET AL LARYNGOTRACHEAL RESECTION AND GPA
anastomotic complications, and the need for immunosuppressive drugs that could impair healing. The cornerstone of management of SGS developing in patients with GPA remains bronchoscopic intervention, whereas refractory disease might require tracheostomy. Although clinical remission rates and overall outcomes are markedly improved with newer immunosuppression regimens, SGS may prove refractory to pharmacologic management alone [3, 6, 7]. In a highly selected group of patients with GPA and SGS, 91% of patients achieved a good airway outcome after LTRR and avoided tracheostomy during long-term follow-up. Four patients (36%) required no further dilation or airway intervention and had an excellent outcome. Seven patients (64%) required additional bronchoscopic airway intervention, only one (9%) of which did not resolve or stabilize, necessitating a tracheostomy. These outcomes were achieved without major complications and with no postoperative deaths. One of the earliest reports suggesting that LTRR could be safely performed in patients with GPA was by Herridge and colleagues [8], who described their experience with 3 patients. Although long-term follow-up was only available for one of the patients, the investigators concluded that the operation could be performed safely as long as patients were in remission and received perioperative immunosuppression regimen. A more recent report by Wester and colleagues [9], who had a series of 8 GPA patients undergoing open surgical management, observed that 75% of these patients required additional tracheal dilation and only 1 patient (13%) required permanent tracheostomy, which is similar to our observed rates of 55% and 9%, respectively. These rates are clearly higher than observed for patients undergoing LTRR for idiopathic laryngotracheal stenosis and suggests that even with GPA remission and maintenance immunosuppression regimen, there is likely a chronic phase of the disease that affects long-term outcomes. In addition to concern about recurrent SGS, a small proportion of GPA patients appear to be also susceptible to lower airway stenosis. In our experience, 2 patients (18%) went on to experience left main-stem stenosis 1 and 9 years after LTRR, respectively. Guardiani and colleagues [10] also noted that 18% of their patients with GPA and SGS experienced lower airway disease. The risk of developing lower airway disease in GPA patients after LTRR mandates long-term clinical follow-up to ensure early diagnosis and effective management. Although bronchoscopic management of lower airway stenosis is the mainstay of treatment, on occasion sleeve resection may be successfully performed, as in our experience with 1 patient and the report by Soo and colleagues [11]. The precise role of ANCA antibody testing in management of SGS in GPA has yet to be defined. Patients who experience SGS seem to represent a distinct subset of all patients with GPA and are known to have a higher incidence of negative ANCA testing [12]. A high index of suspicion needs to be maintained in patients undergoing evaluation for SGS, and clinical manifestations of rheumatologic disease, such as sinus, ocular, pulmonary, or
Ann Thorac Surg 2017;-:-–-
renal involvement, must be thoroughly investigated. Two patients in our early experience with LTRR had positive ANCA tests shortly after operation and have had excellent long-term outcomes (18 and 21 years of follow-up) without any immunosuppression regimen. We consider these patients to be the exception to the rule and have adhered to the standard that all patients with GPA undergoing LTRR must have negative ANCA testing at the time of operation. There is likely a role for serial monitoring of ANCA antibody levels with PR3 and MPO ELISA to guide maintenance immunosuppression regimen, as explained by Han and colleagues [13]. It remains to be seen if this contemporary strategy for managing postoperative immunosuppression regimen will improve long-term outcomes. Because ANCA levels do not absolutely correlate with disease activity in GPA, bronchoscopic examination to rule out active inflammation is recommended before LTRR. In our experience, all patients were free of bronchoscopic signs of inflammation such as airway edema and hyperemia (Fig 1). In our 1 patient who was an operative failure, there was an early recurrence of SGS that necessitated a tracheostomy and was likely related to reactivation of her GPA. Her ANCA/PR3 levels rose several weeks after LTRR and did not decline with resumption of rituximab. PR3 antibody was only detected in this 1 patient, whereas all other patients had MPO antibody. Han and colleagues [13] noted that the presence of PR3 antibody is associated with increased risk of relapse and disease severity. An intriguing report on the utility of magnetic resonance imaging in assessing for granulomatous involvement of the airway and
Fig 1. Bronchoscopic image of subglottic stenosis in patient with granulomatosis with polyangiitis before laryngotracheal resection. The stenosis is mature and free of inflammation.
Ann Thorac Surg 2017;-:-–-
inflammation suggests that imaging might also have a role in assessing a patient’s suitability for LTRR [14]. Although this is the largest report of GPA patients undergoing LTRR to date, the relatively small number of patients and large time span in which care was delivered prevent more rigorous assessment of patient variables that may affect outcomes. The disparate immunosuppression regimens used in the study period also makes it difficult to draw conclusions about ideal medical management of GPA patients with SGS. Nonetheless, we conclude that there are some patients who might benefit from LTRR when bronchoscopic interventions become more frequent or when tracheostomy is threatened. The ideal patient has limited GPA with airway disease confined to a short segment of the subglottic trachea, evidence of clinical remission for at least 6 months after induction immunosuppression regimen, negative ANCA antibody testing, minimal or no steroid requirement just before operation, and a bronchoscopic examination revealing no active airway inflammation. Patients should be fully informed about the risk of recurrence, need for future airway dilations, and the possibility of subsequent lower airway disease. Although there is no difference in LTRR technique performed for patients with GPA and idiopathic SGS, the need for protective tracheostomy is more likely in GPA patients. Long-term control of the disease requires careful surveillance of clinical variables and ANCA levels as well as a lower threshold to escalate immunosuppression regimen. We have demonstrated that there are reasonable long-term outcomes, including freedom from tracheostomy, in a highly selected group of GPA patients undergoing LTRR.
References 1. Jennette JC, Falk RJ, Bacon PA, et al. 2012 revised International Chapel Hill Consensus Conference Nomenclature of Vasculitides. Arthritis Rheum 2013;65:1–11.
COSTANTINO ET AL LARYNGOTRACHEAL RESECTION AND GPA
5
2. Langford CA, Sneller MC, Hallahan CW, et al. Clinical features and therapeutic management of subglottic stenosis in patients with Wegener’s granulomatosis. Arthritis Rheum 1996;39:1754–60. 3. Wojciechowska J, Krajewski W, Krajewski P, Krecicki T. Granulomatosis with polyangiitis in otolaryngologist practice: a review of current knowledge. Clin Exp Otorhinolaryngol 2016;9:8–13. 4. Niles JL. Antineutrophil cytoplasmic antibodies in the classification of vasculitis. Annu Rev Med 1996;47:303–13. 5. Ashiku SK, Kuzucu A, Grillo HC, et al. Idiopathic laryngotracheal stenosis: effective definitive treatment with laryngotracheal resection. J Thorac Cardiovasc Surg 2004;127: 99–107. 6. Girard C, Charles P, Terrier B, et al. Tracheobronchial stenoses in granulomatosis with polyangiitis (Wegener’s): a report on 26 cases. Medicine (Baltimore) 2015; 94:e1088. 7. Rhee EP, Laliberte KA, Niles JL. Rituximab as maintenance therapy for anti-neutrophil cytoplasmic antibody-associated vasculitis. Clin J Am Soc Nephrol 2010;5:1394–400. 8. Herridge MS, Pearson FG, Downey GP. Subglottic stenosis complicating Wegener’s granulomatosis: surgical repair as a viable treatment option. J Thorac Cardiovasc Surg 1996;111: 961–6. 9. Wester JL, Clayburgh DR, Stott WJ, Schindler JS, Andersen PE, Gross ND. Airway Reconstruction in Wegener’s granulomatosis-associated laryngotracheal stenosis. Laryngoscope 2011;121:2566–71. 10. Guardiani E, Moghaddas HS, Lesser J, et al. Multilevel airway stenosis in patients with granulomatosis with polyangiitis (Wegener’s). Am J Otolaryngol 2015;36:361–3. 11. Soo A, Aziz R, Buckley M, Young V. Bronchoplastic procedure for an unusual indication–Wegener’s granulomatosis. Interact Cardiovasc Thorac Surg 2009;9:530–1. 12. Hoffman GS, Kerr GS, Leavitt RY, et al. Wegener granulomatosis: an analysis of 158 patients. Ann Intern Med 1992;116:488–98. 13. Han WK, Choi HK, Roth RM, McCluskey RT, Niles JL. Serial ANCA titers: useful tool for prevention of relapses in ANCAassociated vasculitis. Kidney Int 2003;63:1079–85. 14. Klink T, Holle J, Laudien M, et al. Magnetic resonance imaging in patients with granulomatosis with polyangiitis (Wegener’s) and subglottic stenosis. MAGMA 2013;26: 281–90.