Main Bronchial Sleeve Resection With Pulmonarv Conservation J
Joseph R. Newton, Jr, MD, Hermes C. Grillo, MD, and Douglas J. Mathisen, MD General Thoracic Surgical Unit, Massachusetts General Hospital, and Department of Surgery, Harvard Medical School, Boston, Massachusetts
Twenty-seven main bronchial resections (19 left, 8 right) were performed without pulmonary resection between 1975 and 1991. The patients were 17 men and 9 women with an average age of 35 years (range, 20 to 65 years). Tumors comprised 55% of the lesions, including 9 carcinoid tumors (33%), 2 mucoepidermoid tumors, 2 fibrous histiocytomas, 1 hemangiopericytoma, and 1 large cell carcinoma. Scarring and stenosis secondary to multiple causes occurred in 10 patients (37%). Two patients had miscellaneous lesions. Presenting symptoms included dyspnea (52%), wheezing or stridor (44%),cough (41%), hemoptysis (37%), and pneumonia (18%). Preoperative chest roentgenogram was abnormal in 60% of patients, whereas tomograms delineated the lesion in 94%. All patients had bronchoscopy for lesion evaluation. Anesthesia was accomplished through a long single-lumen endotracheal tube in 19 cases and a double-lumen tube in
T
racheal, carinal, and pulmonary sleeve resections have been described in various reports [l-31, but isolated main bronchial lesions are unusual, and scant information is available about their clinical course. Over the past 16 years, 26 patients at the Massachusetts General Hospital underwent 27 main bronchial resections with complete pulmonary preservation. The presenting signs and symptoms, preoperative evaluation, resectional technique, pathology, perioperative management, and follow-up are presented. The specific operative techniques that allow main bronchial resection and avoidan'ce of pneumonectomy are described in detail. Furthermore, pathologic comparison is made with lesions in other segments of the airway.
Material a n d Methods
Patients and Pathology Twenty-six patients comprise this 16-year series (1975 to 1991) of 27 main bronchial resections. Seventeen patients were male and 9 female. Data were collected from hospital and office records with follow-up data obtained from questionnaires or by direct contact. Pathology is shown in Table 1. Tumors represented the Presented at the Twenty-seventh Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Feb 18-20, 1991. Address reprint requests to Dr Grillo, Thoracic Surgical Unit, Warren 1101, Massachusetts General Hospital, Boston, MA 02114.
0 1991 by The Society of Thoracic Surgeons
8 cases. Mobilization and exposure techniques to create a
tension-free anastomosis were critical for left main bronchial resections and included pretracheal mobilization (loo%),neck flexion (loo%),tracheal and main bronchial retraction (85%), aortic and pulmonary artery retraction (44%),and intrapericardial hilar release (33%).All resections were for cure; there was no operative mortality. Morbidity in 4 patients (15%) included an anastomotic stenosis (successfully reresected), prolonged air leak and pneumonia, transient recurrent nerve palsy, and atelectasis. Median 5-year follow-up revealed 92% of patients alive, with only one of two late deaths being diseaserelated. Main bronchial resection is an ideal technique for selected benign and malignant lesions, allowing complete pulmonary parenchymal preservation. (Ann Thorac Surg 1991;52:1272-80)
single largest category of main bronchial lesions (55%). Nine patients (33%; 7 men and 2 women) had typical carcinoid tumors. Six were in the left main bronchus and three were in the right main bronchus. No patient had carcinoid syndrome. Two patients (7.5%) had left main bronchial mucoepidermoid tumors. Fibrous histiocytoma was present in 2 patients (7.5%)--one right and one left. One patient had a left main bronchial hemangiopericytoma. An additional patient had biopsy performed on a discrete right main bronchial nodule at the time of right upper lobe sleeve resection for a peripheral large cell carcinoma. Despite negative intraoperative frozen section, the nodule proved to be a separate focus of large cell carcinoma on permanent pathological review. One month after lobectomy the patient underwent right main bronchial resection. All tumors resected had negative histological margins and negative lymph nodes. Ten lesions (37%) were benign stenoses of varying cause and occurred in 5 male and 5 female patients. Stenosis developed in 4 of these patients after a previous chest operation. Two patients had undergone right upper lobectomy for carcinoma and 1 patient, left lower lobectomy for carcinoid before referral. Main bronchial stenosis developed in these patients at the site of closure of the lobar bronchus. Another patient in this series was operated on twice owing to anastomotic distraction and stenosis after resection of a left main bronchial mucoepidermoid tumor. Five months after initial resection, her stenosis was repaired, and she remains well 6 years later. 0003-4975/91/$3.50
NEWTON ET AL MAIN BRONCHIAL RESECTION
Ann Thorac Surg 1991;52:1272-80
Table 1. Pathology of Main Bronchial Lesions Pathology Tumors (55%) Carcinoid (33%) Mucoepidermoid Fibrous histiocytoma Hemangiopericytoma Large cell carcinoma Stenosis (37%) Postoperative (15%) Posttraumatic rupture Secondary to laser use Idiopathic Tuberculosis Other (8%) Inflammatory pseudotumor Lymphoid aggregate Total
Left
Right
1273
peared in the trachea and right main bronchus. These 2 patients were categorized as having idiopathic stenosis, as no other cause was identified (ie, mediastinal fibrosis, histiocytosis, tuberculosis or other infection, specific inflammatory disease, trauma, chemical or thermal exposure). One final patient with severe left main stenosis secondary to tuberculosis underwent total left main resection. Two other patients had miscellaneous pathology. One patient had a lymphoid aggregate (endobronchial lymph node) in the left main bronchus. Another patient had an inflammatory pseudotumor.
Operative Technique
1 1 19
Stenosis developed in 2 patients after blunt traumatic rupture of the bronchus. One was seen 4 months after rupture of the right main bronchus without having undergone definitive initial treatment, and one 8 months after failed primary repair of the left main bronchus. Left main bronchial stenosis developed in another patient after ill-advised multiple photodynamic laser treatments for bronchial metaplasia found on bronchoscopy done to evaluate a chronic cough. After resection, the specimen showed chronic scarring and lymph nodes consistent with sarcoid. Two patients with benign stenosis were women, both of whom had histories of asthma and bronchitis since childhood and were found subsequently to have discrete fibrotic left main bronchial lesions. Although both recovered well initially after left main bronchial resections, additional fibrotic stenoses later ap-
Initially, bronchoscopy (rigid or flexible) was performed to precisely identify the location and extent of the lesion. A biopsy with frozen section examination was performed when it appeared the operative strategy would be affected by the results. Biopsy was avoided in vascular-appearing carcinoid tumors. Occasionally, in patients with obstructive pneumonia it was necessary to suction away secretions in the distal airway. After bronchoscopy, continued intraoperative anesthesia was provided by reintubating the patient with either a Wilson tube into the opposite bronchus (19 patients) or a double-lumen tube (8 patients). The Wilson tube (Fig 1)is a long, flexible, single-lumen tube constructed by attaching an extension onto a flexible armored tube. There is also an extender on the balloon cuff inflation port. The end is square, not beveled, for better fit into the main bronchi, and the balloon cuff is smaller to minimize herniation. The tube is advanced into the opposite bronchus before thoracotomy. Tube position is verified by examination or by use of a flexible pediatric bronchoscope. When double-lumen tubes were used, right double-lumen tubes were placed for left main lesions ( 5 patients) and left double-lumen tubes were placed for right main lesions (3 patients). The relative rigidity of the double-lumen tube compared with the Wilson tube can
Fig I . The Wilson tube constructed by attaching an extension tube onto a square-ended, flexible armored tube.
1274
NEWTON ET AL MAIN BRONCHIAL RESECTION
Ann Thorac Surg 1991;52:1272-80
Fig 2 . Exposure for left main bronchial resection. Note the tracheal, right main bronchial, pulmonary artery, and aortic retraction that provides improved exposure.
Pulmonary artery I
f
Left main bi
I
4 limit mobility of the carina and is a disadvantage of the double-lumen tube in extensive left main bronchial resection. The operative approach was through the bed of the resected fifth rib or in the fourth interspace on the appropriate side. Intraoperative exploration revealed lobar collapse in 3 patients and hyperinflation in 1. The lesion could usually be palpated, and nodes were occasionally enlarged by inflammation. After assessment of resectability, mobilization techniques were initiated and varied with extent of resection and whether right or left main bronchial resection was done. The inferior pulmonary ligament was freed in all cases. For right main lesions, tension was not a problem during reconstruction, given the short length of the right bronchus; therefore, additional mobilization and exposure techniques were not required. For left main bronchial resections, the length of the left main bronchus resected, the prominent aortic arch, and left heart border made the use of tensionreducing maneuvers important. Pretracheal mobilization from the carina cephalad was initiated by sharp dissection and extended by blunt finger dissection. This dissection can be done in patients after prior mediastinoscopy although less mobility may be gained. Temporary neck flexion was used in all left-sided cases after resection to facilitate approximation of the anastomosis. A guardian suture from the underside of the chin to the premanubrial skin was placed after the thoracotomy was closed but before extubation to maintain the neck in 45-degree flexion for 7 postoperative days only when extensive left main resection was carried out. Both pretracheal dissection and neck flexion provide added proximal main bronchial length by allowing the trachea, carina, and proximal main bronchus to descend.
recurrent nerves
I
Right main bronchus
In most patients (85%)before resection, traction tapes (6.4-mm [Ih-inch] Penrose drains) were placed encircling the trachea for right main bronchial resections and encircling the trachea and right main bronchus for left main bronchial resections. Tracheal and bronchial traction is a key exposure technique in the successful completion of the operation. Additional dissection to retract the aortic arch (without division of intercostal vessels) and pulmonary artery was required for exposure in 8 of the 19 left main bronchial resections (44%). Smith and Nigam [4] have previously described aortic mobilization with intercostal vessel division to expose the left main bronchus. An example of the extent of exposure obtained by careful dissection for left main resection is shown in Figure 2. The trachea and right main bronchus are pulled up into the field by the traction tapes to expose the origin of the left main bronchus. The recurrent laryngeal nerve must be carefully identified and preserved for left main bronchial resections. If needed, added distal airway mobility can be obtained by the intrapericardial hilar release technique, which allows the distal end of the bronchus and hilum to ascend approximately 1.0 to 2.0 cm. In 6 of the 19 patients undergoing left main bronchial resection (33%), intrapericardial hilar release was performed to assist in providing a tension-free anastomosis. It was especially helpful for proximal, subtotal, or total left main resections. Hilar release (Fig 3 ) involved a U-shaped incision in the pericardium beneath the inferior pulmonary vein to allow upward movement of the hilum. The pericardial incision may be extended superiorly or completely circumscribe the hilar vessels if additional mobility is required for reconstruction, but this is rarely needed. If the extended pericardial incision is used, the subvenous frenulum
NEWTON ET AL MAIN BRONCHIAL RESECTION
Ann Thorac Surg 1991:52:1272-80
1275
Aorta
/ ’
Fig 3 . The left side intrapericardial hilur release technique, showing the U shaped pericardial incision allowing 2 to 2 cm of upward hilar mobility to facilitate the creation of a tension-free anastomosis.
should be divided and the posterior pedicle of lymph nodes preserved to avoid lymphatic obstruction in the early postoperative days. Resection was performed by circumferential complete division of the airway proximal and distal to the lesion. A typical pathology specimen is shown in Figure 4. Resection margins were negative by frozen section for malignant lesions and at grossly normal mucosa for benign
Fig 4 . The pathology specimen of a right main bronchial carcinoid tumor. i n this patient, only a short segment of the airway needed to be removed.
stenoses. Routine airway anastomosis reconstruction was carried out as described elsewhere [5] using two lateral traction sutures (2-0 Vicryl; Ethicon, Somerville, NJ) for orientation control and assistance in creating a tensionfree anastomosis. Interrupted sutures (4-0 Vicryl) placed 3 to 4 mm apart and 3 to 4 mm from the cut edge with knots tied outside the airway completed the reconstruction. All patients had their anastomosis tested for an airtight seal. In 96% the anastomosis was buttressed with a flap of viable pleura or pericardial fat pad. Pulmonary artery kinking was not observed. The average length of resected right main bronchus was 1.5 cm (range, 1.0 tos 2.1 cm), and for the left main bronchus the average resected specimen was 2.8 cm (range, 1.0 to 4.0 cm). Right main bronchial resections tended to be the total length of the bronchus, as it is naturally short, whereas left main bronchial resections were more variable. The entire left main bronchus was removed in 3 patients (2 for benign stenosis [l idiopathic, 1 postoperative] and 1 for stenosis secondary to tuberculosis). The intrapericardial hilar release was an especially critical tension-reducing maneuver during reconstruction in these patients. An additional patient with a carcinoid tumor had the carina between the left upper and lower lobes resected with anastomosis of the upper and lower lobe bronchi together before anastomosis to the proximal left main bronchus. Perioperative management began with immediate extubation in the operating room and intensive care observation, usually for 24 hours and occasionally 48 hours. All patients received aggressive chest physiotherapy and perioperative antibiotics, and 4 patients required postoperative bronchoscopy for clearance of secretions. The
1276
NEWTON ET AL MAIN BRONCHIAL RESECTION
A
Ann Thorac Surg 1991;52:1272-80
B
Fig 5. ( A ) Preoperative tracheabronchial tomograins demonstrating a right carcinoid tumor, identified by the arrow. ( B ) Postoperative tracheobronchial tomograms demonstrating a patent airway uiith an arrow marking the anastomosis.
average total hospital stay was 9 days (range, 6 to 36 days). Adjuvant therapy was not required for any patient.
Results
Symptoms and Preoperative Evaluation All patients were symptomatic and had an average of 2.5 symptoms. The presenting signs and symptoms included dyspnea (52%), wheeze or stridor (44%), cough (41%), hemoptysis (37%), pneumonia (18%), fever (ll%),and chest pain (7%).Left or right main bronchial location or pathology of the lesion produced no difference in symptoms. The average age at presentation was 35.0 years (range, 20 to 65 years), although generally patients with left main bronchial lesions were younger (33.1years) than those with right-sided lesions (42.6 years). Associated medical problems were uncommon (23%) and included coronary artery disease in 2 patients (1 of whom also had diabetes and diverticulosis) and chronic obstructive pulmonary disease, hypertension, gout, and Crohn’s disease in 1 patient each. Fifteen of the 26 patients (58%) had a history of smoking. Preoperative evaluation included a chest roentgenogram in all patients, which was normal in 41%. A specific
lesion was visible in the main bronchus in 18%, and secondary effects of atelectasis, pneumonia, or air trapping were noted in 41%. Standard tracheobronchial tomograms (Figs 5, 6) were especially useful in delineating the lesion in 16 of 17 patients (94%). Twelve patients had computed axial tomography before referral, detecting the lesion in 9 (Fig 7 ) . One patient, before referral, had a magnetic resonance scan which detected the lesion. Pulmonary function tests revealed an average forced expiratory volume in 1 second of 2.5 L (range, 0.8 to 5.4 L). Ventilation and perfusion scanning usually showed decreased function on the affected side, although depending on the size of the lesion, values were occasionally near normal. No patient had bone or brain scans, pulmonary angiography, or mediastinoscopy. Nineteen patients had undergone bronchoscopy before referral, which occasionally established the correct diagnosis, and all patients underwent bronchoscopy immediately before thoracotomy. The respective bronchus was on average stenosed to approximately 80% of normal caliber by visual inspection, with 15 patients showing nearly total obstruction. Despite this degree of obstruction, only 2 patients required preoperative therapeutic bronchoscopy to clear secretions of a postobstructive pneumonia. One patient
NEWTON ET AL MAIN BRONCHIAL RESECTION
Ann Thorac Surg 1991;52:1272-80
A
1277
B
Fig 6 . ( A ) Preoperative tracheobronchial totnograms demonstrating total left main stenosis secondany to tziberc-ulosis as outlined by the arrows. ( B ) Postoperative tracheobronchial tomagrams demonstrating left upper and lower division bronchi anastoniosed to the lateral trachea at the site of the arrow.
with a left main bronchial fibrous histiocytoma underwent a therapeutic bronchoscopy 48 hours before thoracotomy, and 1 patient with a right main bronchial carcinoid underwent therapeutic bronchoscopy 24 hours before thoracotomy. Both needed prompt relief of obstruction and drainage of postobstructive pneumonia in conjunction with appropriate preoperative antibiotics and chest physiotherapy to reduce sepsis in the involved lung.
Morbidity and Mortality Complications (Table 2) occurred in 4 of the 26 patients (15%). A 29-year-old woman with a 1-cm mucoepidermoid tumor of the left main bronchus had an initial 2.2-cm resection performed. Postoperatively, shortness of breath and wheezing prompted bronchoscopic examination, which revealed partial dehiscence due to anastomotic tension, which progressed to stenosis. She was managed with five rigid bronchoscopic dilations over the ensuing 5 months before undergoing subtotal resection of the remainder of her left main bronchus with intrapericardial hilar release. She recovered without difficulty; 6 years postoperatively she is working without respiratory symptoms. Transient hoarseness and temporary left vocal cord
Fig 7 . Computed tomographic S C ~ I Iof tumor (arrow).
a
/eft main bronc.hia/ rarcinoid
1278
NEWTON ET AL MAIN BRONCHIAL RESECTION
Ann Thorac Surg 1991;52:1272-80
Table 2. Morbidity of Main Bronchial Resections" Morbidity Anastomotic stenosis Transient recurrent nerve palsy Pneumonia Prolonged air leak Atelectasis a
Right
Left
0 0 1 1 0
1 1 0 0 1
There was no operative mortality
palsy developed in 1 patient with left main bronchial resection for a lymphoid aggregate but resolved without treatment after 6 months. One patient undergoing resection for stenosis after right upper lobectomy for adenocarcinoma had development of a prolonged (30-day) air leak which healed spontaneously. He also had a postobstructive Hemophilus pneumonia on presentation; postoperatively a new left lower lobe infiltrate developed despite clearing of the right lung. This responded to appropriate antibiotics. Minimal left lower lobe atelectasis, which responded to chest physiotherapy and did not prolong hospital stay, occurred in 1 patient after left main bronchial resection. There were no problems with anastomotic granulations, pulmonary torsion, or fistula. No patient died before hospital discharge or within 30 days of operation. Long-term follow-up showed that 92% of the patients (24 of 26) were alive a median of 5.0 years after operation (range, 1 month to 13 years). One patient, a 55-year-old woman who underwent left main resection for carcinoid, died 22 months postoperatively of a severe head injury after a motor vehicle accident. The second death occurred in a 35-year-old woman who underwent total left main bronchial resection for idiopathic stenosis. She originally was seen 5 years before operation with diffuse tracheobronchial inflammation. She required periodic dilations. During a period of decreased inflammation the persistent isolated left main bronchial stenosis was resected. Initially her postoperative course was uncomplicated, but recurrent dyspnea and wheezing 4 months postoperatively mandated four bronchoscopic dilations over a 14.5-month period before her death. Diffuse inflammation and stenosis of the distal trachea and both bronchi were noted. She suffered a respiratory arrest at home and died subsequently of anoxic cerebral injury at another hospital. One other patient also has suffered from diffuse idiopathic tracheobronchial stenosis. She is alive 12 years after left main bronchial resection with a patent anastomosis but requires dilation of her distal trachea and right main bronchus every 4 months.
Postoperative Evaluation Twenty-four patients underwent postoperative bronchoscopy (usually approximately 3 months after resection), and 21 (86%) were normal. Two of the abnormal examinations revealed diffuse inflammatory stenosis in the trachea and right main bronchus, both in patients with idiopathic stenosis and left main bronchial resection. One
abnormal examination was in the patient requiring reoperation for anastomotic stenosis. Her bronchoscopic examination is now normal. Eleven patients had postoperative tomograms; 10 were normal (see Figs 5, 6). The abnormal examination was in 1 of the patients with idiopathic stenosis. Twenty-three of the 24 living patients (96%)currently report no symptoms whatsoever and have returned to work. One patient with idiopathic stenosis has dyspnea. One patient who underwent resection of a right main bronchial stenosis after right upper lobectomy and irradiation (3,500 cGy) for lung carcinoma had a second primary lung tumor resected from his left lung 3 years after his main bronchial operation. He is the only patient in the series other than the 2 with idiopathic stenosis in whom any other chest pathology developed. Early postoperative pulmonary function testing was formally performed on 3 patients who also had such studies preoperatively. Performed an average of 7 days postoperatively (just before discharge), the studies had essentially the same results as preoperative studies. The lack of demonstrated improvement probably is due to persistent thoracotomy discomfort at 7 days.
Comment Main bronchial circumferential resection with end-to-end anastomosis has been described previously, mostly for traumatic disruptions and resultant stenosis and occasionally for primary tumors or tuberculosis. Although main bronchial traumatic injuries were described in 1873 [6]and 1947 [7], it was not until 1949 that Griffith [8] successfully performed a near circumferential resection of a posttraumatic stenosis of the left main bronchus with end-to-end reconstruction. Previous experimental work by Jackson and associates in 1948 [9] showed that main bronchial resection and primary anastomosis could be carried out successfully in the dog model. A variety of other authors have reported resection and reconstruction of main bronchial posttraumatic stenosis, with Deslaurier and colleagues' series of 7 in 1981 [lo] and Hood and Sloan's series of 4 in 1959 [ll] representing the largest experiences. Few reports of main bronchial resection and reconstruction for tumor, which form the bulk of this report, have appeared. Isolated cases of successful surgical treatment by circumferential resection and end-to-end anastomosis for carcinoid [12-141, fibrous histiocytoma (15), tuberculosis [16], and carcinoma [17]have been described. Frist and co-workers [3] noted some of the present cases in a report that dealt mostly with sleeve pulmonary resections. Other authors have reported bronchoplastic techniques either with lateral closure or with interposition or augmentation of the airway diameter with artificial or autologous materials [17-251. These techniques may result in narrowing of the airway due to either the initial repair or subsequent scarring. Avoidance of any narrowing is a primary concern in airway reconstruction. Circumferential sleeve resection with precise primary end-to-end
NEWTON ET AL MAIN BRONCHIAL RESECTION
Ann Thorac Surg 1991;52:1272-80
anastomosis as described here usually avoids this complication. The pathology of main bronchial lesions differs somewhat from lesions found in the remainder of the airway and generally permits their isolated resection. At the Massachusetts General Hospital, tracheal resection and reconstruction is undertaken most commonly (75% of cases) for postintubation stenosis [l-31. Stenosis was a common main bronchial lesion as well (37%), although none were secondary to intubation. Prior surgical manipulation, trauma, laser injury, and idiopathic stenosis were the pathologic lesions responsible for main bronchial stenosis. Previous surgical manipulation was usually pulmonary resection with tangential lobar bronchial division and resultant narrowing. Such stenoses would probably have been avoided if sleeve lobectomy had been carried out primarily. The most common tumors of the main bronchi were carcinoid (33%), mucoepidermoid (7.5%), and fibrous histiocytoma (7.5%). In the trachea, the most common tumors are adenoid cystic (approximately 40%) and squamous carcinoma (approximately 35%) [2], neither of which were found as isolated main bronchial lesions. Isolated carinal lesions, perhaps representing a transition zone between the trachea and main bronchus, reflect a broad mixture of lesions including adenoid cystic carcinoma (approximately 40%), carcinoid, stenosis, squamous cell carcinoma, and mucoepidermoid tumor [l]. Distal airway resections (sleeve lobectomy) were done at the Massachusetts General Hospital mainly for bronchogenic carcinoma (70%) and less commonly for carcinoid (21%) [3]. Based on the data presented, preoperative evaluation of patients with main bronchial lesions should include standard chest roentgenogram and tomography. Bronchoscopic examination of patients should be performed along with pulmonary function testing. If there is concern regarding potential pneumonectomy, ventilation-perfusion scanning should be performed. If the clinical course suggests an obstructive pneumonia, preoperative culture-specific antibiotics and bronchoscopic decompression should be used, followed in 24 to 48 hours by resection. Metastatic work-up and mediastinoscopic staging generally are not indicated as the majority of these lesions are benign or of low-grade malignancy (96% of cases). Patients with persistent tracheobronchial inflammation and focal areas of stenosis without identifiable causative factors should be handled cautiously as further airway problems may develop in these patients. This w a s exemplified by the 2 patients with idiopathic stenosis. Operative management is facilitated by use of the long, flexible Wilson tube. Tension-reducing maneuvers are critical for left main bronchial resection. Initially all patients should have the inferior pulmonary ligament divided and should have pretracheal dissection and neck flexion performed. If, upon approximation of the lateral traction sutures, tension still exists, then intrapericardial hilar release may yield 1 to 2 cm of added length, especially for subtotal or total left main bronchial resections. Tracheal and bronchial traction tapes, pulmonary
1279
artery retraction, and aortic mobilization are critical maneuvers for exposure. Clear resection margins and a precise anastomosis with absorbable sutures complete the operation. Postoperative tomograms and follow-up bronchoscopy either before discharge or several months postoperatively or both assist in ensuring an excellent result. A failed anastomosis can be successfully resected again as long as maximal tension reduction can be achieved. Postoperative stenosis, granulations, fistula, or other complications should be watched for but are rare. Main bronchial resection with pulmonary preservation can be carried out with minimal morbidity and mortality. The usually benign or low-grade malignant pathology permit such resections without pneumonectomy. Successful preservation of the normal lung allows virtually all patients to return to unrestricted activity.
References 1. Grillo HC. Carinal resection. Ann Thorac Surg 1982;34: 356-73. 2. Grillo HC, Mathisen DJ. Primary tracheal tumors: treatment and results. Ann Thorac Surg 1990;49:69-77. 3. Frist WH, Mathisen DJ, Hilgenberg AD, Grillo HC. Bronchial sleeve resection with and without pulmonary resection. J Thorac Cardiovasc Surg 1987;93:350-7. 4. Smith RA, Nigam BK. Resection of proximal left main bronchus carcinoma. Thorax 1979;34:61&20. 5. Grillo HC. Tracheal resection. In: Ravitch M, Steichen F, eds. Atlas of general thoracic surgery. Philadelphia: W.B. Saunders, 1988:296. 6. Seavre M. Crushing injury from wheel of an omnibus: rupture of the left main bronchus. Bull SOCAnat Paris 1873:48:680-2. 7. Kinsella TJ, Johnsrud LW. Traumatic rupture of the bronchus. J Thorac Surg 1947;16:571-83. 8. Griffith JL. Fracture of the bronchus. Thorax 1979;4:1059. 9. Jackson TL, Lefkin P, Tuttle W, Hampton F. An experimental study in bronchial anastomosis. J Thorac Surg 1949;18:63042. 10. Deslauriers J, Beaulieu M, Archambault G, LaForge J, Bernier R. Diagnosis and long-term follow-up of major bronchial disruptions due to nonpenetrating trauma. Ann Thorac Surg 1981;33:32-9. 11. Hood RM, Sloan HE. Injuries of the trachea and major bronchi. J Thorac Cardiovasc Surg 1959;38:458-80. 12. Spitzer SA, Segal 1, Lubin E, Nili M, Levy M. Unilateral increased transradiancy of the lung caused by bronchial carcinoid tumor. Thorax 1979;34:73944. 13. Grant JL, Naylor RW, Crandell WB. Bronchial adenoma resection with relief of hypoxic pulmonary vasoconstriction. Chest 1980;77:44&9. 14. DeLima R. Bronchial adenoma. Chest 1980;77814. 15. Yamazaki K, Kubo Y, Kawabata M. A case of fibrous histiocytoma in the left main bronchus. Jpn J Thorac Surg 1988;41: 34G2. 16. Caligiari PA, Banner AS, Jensik RJ. Tuberculous mainstem bronchial stenosis treated with sleeve resection. Arch Intern Med 1984;1244:1302-3. 17. Keszler P. Sleeve resection and other bronchoplasties in the surgery of bronchogenic tumors. lnt Surg 1986;71:229-32. 18. Lowe JE, Bridgman AH, Sabiston DC. The role of broncho-
1280
NEWTON ET AL MAIN BRONCHIAL RESECTION
plastic procedures in the surgical management of benign and malignant pulmonary lesions. J Thorac Cardiovasc Surg 1982;83:227-34. 19. Gebauer PW. Plastic reconstruction of tuberculous bronchostenosis with dermal grafts. J Thorac Surg 1950;19:604-28. 20. Thomas CP. Conservative resection of the bronchial tree. J R Coll Surg Edinb 1956;3:16&86. 21. Bjork VO, Rodriguez LE. Reconstruction of the trachea and its bifurcation. J Thorac Surg 1958;35:596-603.
Ann Thorac Surg 1991;52:1272-80
22. Nicks R. Restoration and reconstruction of the trachea and main bronchi. J Thorac Surg 1956;32:22645. 23. Mathey J, Galey JJ, Lemoire G. Tracheal and tracheobronchial resections. J Thorac Cardiovasc Surg 1966;51:1-13. 24. Keshishian JM, Blades B, Beattie EJ. Tracheal reconstruction. J Thorac Surg 1956;32:707-27. 25. Paulson DL, Urschel HC, McNamara JJ, Shaw RR. Bronchoplastic procedures for bronchogenic carcinoma. J Thorac Cardiovasc Surg 1970;59:38-48.
DISCUSSION DR HAROLD C. URSCHEL, JR (Dallas, TX): I congratulate Dr Newton and his colleagues on a landmark paper. I want to make two technical points. I talked to Hermes Grillo about a recent patient of mine. This 26-year-old patient had a totally occluded left main bronchus and underwent laser removal of a mucoepidermoid carcinoma of the left main bronchus. A stent was placed to support the bronchus, allowing time for drainage of the infected lung before resection. This is a valuable technique for total occlusions if an attempt is to be made to save the whole lung. The second point is that he described the inferior pericardial release, which we use. But in this case we performed a total circular pericardial release, which we use for lung transplants. The whole lung comes up easily without any tension at all on the tracheal suture line. I think the principles that you have illustrated technically are very valuable to all of us. DR NEWTON: Thank you very much.
DR ROBERT J. GINSBERG (New York, NY): I want to emphasize one technical point that Newton and associates mentioned briefly, and that is the type of tube used. Those who use a right-sided double-lumen tube will find that they cannot always mobilize the carina the way Newton and associates have demonstrated so nicely. It is very important to use something like the Roger Wilson adaptation, or a bronchial blocker [I], to allow the carina to be pulled down through the aortic arch. A doublelumen tube will fix the carina on the other side of the arch, making the anastomosis extremely difficult. I think this is a very important point in performing left main bronchial anastomoses. DR NEWTON: I agree with that point. We found the flexibility of the Wilson tube to be very important.
Reference 1. Ginsberg RJ. New technique for one-lung anesthesia using endobronchial blocker. J Thorac Cardiovasc Surg 1981;82: 542-6.
Joseph R. Newton, Jr, MD, Hermes C. Grillo, MD, and Douglas J. Mathisen, MD General Thoracic Surgical Unit, Massachusetts General Hospital, and Department of Surgery, Harvard Medical School, Boston, Massachusetts
Twenty-seven main bronchial resections (19 left, 8 right) were performed without pulmonary resection between 1975 and 1991. The patients were 17 men and 9 women with an average age of 35 years (range, 20 to 65 years). Tumors comprised 55% of the lesions, including 9 carcinoid tumors (33%), 2 mucoepidermoid tumors, 2 fibrous histiocytomas, 1 hemangiopericytoma, and 1 large cell carcinoma. Scarring and stenosis secondary to multiple causes occurred in 10 patients (37%). Two patients had miscellaneous lesions. Presenting symptoms included dyspnea (52%), wheezing or stridor (44%),cough (41%), hemoptysis (37%), and pneumonia (18%). Preoperative chest roentgenogram was abnormal in 60% of patients, whereas tomograms delineated the lesion in 94%. All patients had bronchoscopy for lesion evaluation. Anesthesia was accomplished through a long single-lumen endotracheal tube in 19 cases and a double-lumen tube in
T
racheal, carinal, and pulmonary sleeve resections have been described in various reports [l-31, but isolated main bronchial lesions are unusual, and scant information is available about their clinical course. Over the past 16 years, 26 patients at the Massachusetts General Hospital underwent 27 main bronchial resections with complete pulmonary preservation. The presenting signs and symptoms, preoperative evaluation, resectional technique, pathology, perioperative management, and follow-up are presented. The specific operative techniques that allow main bronchial resection and avoidan'ce of pneumonectomy are described in detail. Furthermore, pathologic comparison is made with lesions in other segments of the airway.
Material a n d Methods
Patients and Pathology Twenty-six patients comprise this 16-year series (1975 to 1991) of 27 main bronchial resections. Seventeen patients were male and 9 female. Data were collected from hospital and office records with follow-up data obtained from questionnaires or by direct contact. Pathology is shown in Table 1. Tumors represented the Presented at the Twenty-seventh Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Feb 18-20, 1991. Address reprint requests to Dr Grillo, Thoracic Surgical Unit, Warren 1101, Massachusetts General Hospital, Boston, MA 02114.
0 1991 by The Society of Thoracic Surgeons
8 cases. Mobilization and exposure techniques to create a
tension-free anastomosis were critical for left main bronchial resections and included pretracheal mobilization (loo%),neck flexion (loo%),tracheal and main bronchial retraction (85%), aortic and pulmonary artery retraction (44%),and intrapericardial hilar release (33%).All resections were for cure; there was no operative mortality. Morbidity in 4 patients (15%) included an anastomotic stenosis (successfully reresected), prolonged air leak and pneumonia, transient recurrent nerve palsy, and atelectasis. Median 5-year follow-up revealed 92% of patients alive, with only one of two late deaths being diseaserelated. Main bronchial resection is an ideal technique for selected benign and malignant lesions, allowing complete pulmonary parenchymal preservation. (Ann Thorac Surg 1991;52:1272-80)
single largest category of main bronchial lesions (55%). Nine patients (33%; 7 men and 2 women) had typical carcinoid tumors. Six were in the left main bronchus and three were in the right main bronchus. No patient had carcinoid syndrome. Two patients (7.5%) had left main bronchial mucoepidermoid tumors. Fibrous histiocytoma was present in 2 patients (7.5%)--one right and one left. One patient had a left main bronchial hemangiopericytoma. An additional patient had biopsy performed on a discrete right main bronchial nodule at the time of right upper lobe sleeve resection for a peripheral large cell carcinoma. Despite negative intraoperative frozen section, the nodule proved to be a separate focus of large cell carcinoma on permanent pathological review. One month after lobectomy the patient underwent right main bronchial resection. All tumors resected had negative histological margins and negative lymph nodes. Ten lesions (37%) were benign stenoses of varying cause and occurred in 5 male and 5 female patients. Stenosis developed in 4 of these patients after a previous chest operation. Two patients had undergone right upper lobectomy for carcinoma and 1 patient, left lower lobectomy for carcinoid before referral. Main bronchial stenosis developed in these patients at the site of closure of the lobar bronchus. Another patient in this series was operated on twice owing to anastomotic distraction and stenosis after resection of a left main bronchial mucoepidermoid tumor. Five months after initial resection, her stenosis was repaired, and she remains well 6 years later. 0003-4975/91/$3.50
NEWTON ET AL MAIN BRONCHIAL RESECTION
Ann Thorac Surg 1991;52:1272-80
Table 1. Pathology of Main Bronchial Lesions Pathology Tumors (55%) Carcinoid (33%) Mucoepidermoid Fibrous histiocytoma Hemangiopericytoma Large cell carcinoma Stenosis (37%) Postoperative (15%) Posttraumatic rupture Secondary to laser use Idiopathic Tuberculosis Other (8%) Inflammatory pseudotumor Lymphoid aggregate Total
Left
Right
1273
peared in the trachea and right main bronchus. These 2 patients were categorized as having idiopathic stenosis, as no other cause was identified (ie, mediastinal fibrosis, histiocytosis, tuberculosis or other infection, specific inflammatory disease, trauma, chemical or thermal exposure). One final patient with severe left main stenosis secondary to tuberculosis underwent total left main resection. Two other patients had miscellaneous pathology. One patient had a lymphoid aggregate (endobronchial lymph node) in the left main bronchus. Another patient had an inflammatory pseudotumor.
Operative Technique
1 1 19
Stenosis developed in 2 patients after blunt traumatic rupture of the bronchus. One was seen 4 months after rupture of the right main bronchus without having undergone definitive initial treatment, and one 8 months after failed primary repair of the left main bronchus. Left main bronchial stenosis developed in another patient after ill-advised multiple photodynamic laser treatments for bronchial metaplasia found on bronchoscopy done to evaluate a chronic cough. After resection, the specimen showed chronic scarring and lymph nodes consistent with sarcoid. Two patients with benign stenosis were women, both of whom had histories of asthma and bronchitis since childhood and were found subsequently to have discrete fibrotic left main bronchial lesions. Although both recovered well initially after left main bronchial resections, additional fibrotic stenoses later ap-
Initially, bronchoscopy (rigid or flexible) was performed to precisely identify the location and extent of the lesion. A biopsy with frozen section examination was performed when it appeared the operative strategy would be affected by the results. Biopsy was avoided in vascular-appearing carcinoid tumors. Occasionally, in patients with obstructive pneumonia it was necessary to suction away secretions in the distal airway. After bronchoscopy, continued intraoperative anesthesia was provided by reintubating the patient with either a Wilson tube into the opposite bronchus (19 patients) or a double-lumen tube (8 patients). The Wilson tube (Fig 1)is a long, flexible, single-lumen tube constructed by attaching an extension onto a flexible armored tube. There is also an extender on the balloon cuff inflation port. The end is square, not beveled, for better fit into the main bronchi, and the balloon cuff is smaller to minimize herniation. The tube is advanced into the opposite bronchus before thoracotomy. Tube position is verified by examination or by use of a flexible pediatric bronchoscope. When double-lumen tubes were used, right double-lumen tubes were placed for left main lesions ( 5 patients) and left double-lumen tubes were placed for right main lesions (3 patients). The relative rigidity of the double-lumen tube compared with the Wilson tube can
Fig I . The Wilson tube constructed by attaching an extension tube onto a square-ended, flexible armored tube.
1274
NEWTON ET AL MAIN BRONCHIAL RESECTION
Ann Thorac Surg 1991;52:1272-80
Fig 2 . Exposure for left main bronchial resection. Note the tracheal, right main bronchial, pulmonary artery, and aortic retraction that provides improved exposure.
Pulmonary artery I
f
Left main bi
I
4 limit mobility of the carina and is a disadvantage of the double-lumen tube in extensive left main bronchial resection. The operative approach was through the bed of the resected fifth rib or in the fourth interspace on the appropriate side. Intraoperative exploration revealed lobar collapse in 3 patients and hyperinflation in 1. The lesion could usually be palpated, and nodes were occasionally enlarged by inflammation. After assessment of resectability, mobilization techniques were initiated and varied with extent of resection and whether right or left main bronchial resection was done. The inferior pulmonary ligament was freed in all cases. For right main lesions, tension was not a problem during reconstruction, given the short length of the right bronchus; therefore, additional mobilization and exposure techniques were not required. For left main bronchial resections, the length of the left main bronchus resected, the prominent aortic arch, and left heart border made the use of tensionreducing maneuvers important. Pretracheal mobilization from the carina cephalad was initiated by sharp dissection and extended by blunt finger dissection. This dissection can be done in patients after prior mediastinoscopy although less mobility may be gained. Temporary neck flexion was used in all left-sided cases after resection to facilitate approximation of the anastomosis. A guardian suture from the underside of the chin to the premanubrial skin was placed after the thoracotomy was closed but before extubation to maintain the neck in 45-degree flexion for 7 postoperative days only when extensive left main resection was carried out. Both pretracheal dissection and neck flexion provide added proximal main bronchial length by allowing the trachea, carina, and proximal main bronchus to descend.
recurrent nerves
I
Right main bronchus
In most patients (85%)before resection, traction tapes (6.4-mm [Ih-inch] Penrose drains) were placed encircling the trachea for right main bronchial resections and encircling the trachea and right main bronchus for left main bronchial resections. Tracheal and bronchial traction is a key exposure technique in the successful completion of the operation. Additional dissection to retract the aortic arch (without division of intercostal vessels) and pulmonary artery was required for exposure in 8 of the 19 left main bronchial resections (44%). Smith and Nigam [4] have previously described aortic mobilization with intercostal vessel division to expose the left main bronchus. An example of the extent of exposure obtained by careful dissection for left main resection is shown in Figure 2. The trachea and right main bronchus are pulled up into the field by the traction tapes to expose the origin of the left main bronchus. The recurrent laryngeal nerve must be carefully identified and preserved for left main bronchial resections. If needed, added distal airway mobility can be obtained by the intrapericardial hilar release technique, which allows the distal end of the bronchus and hilum to ascend approximately 1.0 to 2.0 cm. In 6 of the 19 patients undergoing left main bronchial resection (33%), intrapericardial hilar release was performed to assist in providing a tension-free anastomosis. It was especially helpful for proximal, subtotal, or total left main resections. Hilar release (Fig 3 ) involved a U-shaped incision in the pericardium beneath the inferior pulmonary vein to allow upward movement of the hilum. The pericardial incision may be extended superiorly or completely circumscribe the hilar vessels if additional mobility is required for reconstruction, but this is rarely needed. If the extended pericardial incision is used, the subvenous frenulum
NEWTON ET AL MAIN BRONCHIAL RESECTION
Ann Thorac Surg 1991:52:1272-80
1275
Aorta
/ ’
Fig 3 . The left side intrapericardial hilur release technique, showing the U shaped pericardial incision allowing 2 to 2 cm of upward hilar mobility to facilitate the creation of a tension-free anastomosis.
should be divided and the posterior pedicle of lymph nodes preserved to avoid lymphatic obstruction in the early postoperative days. Resection was performed by circumferential complete division of the airway proximal and distal to the lesion. A typical pathology specimen is shown in Figure 4. Resection margins were negative by frozen section for malignant lesions and at grossly normal mucosa for benign
Fig 4 . The pathology specimen of a right main bronchial carcinoid tumor. i n this patient, only a short segment of the airway needed to be removed.
stenoses. Routine airway anastomosis reconstruction was carried out as described elsewhere [5] using two lateral traction sutures (2-0 Vicryl; Ethicon, Somerville, NJ) for orientation control and assistance in creating a tensionfree anastomosis. Interrupted sutures (4-0 Vicryl) placed 3 to 4 mm apart and 3 to 4 mm from the cut edge with knots tied outside the airway completed the reconstruction. All patients had their anastomosis tested for an airtight seal. In 96% the anastomosis was buttressed with a flap of viable pleura or pericardial fat pad. Pulmonary artery kinking was not observed. The average length of resected right main bronchus was 1.5 cm (range, 1.0 tos 2.1 cm), and for the left main bronchus the average resected specimen was 2.8 cm (range, 1.0 to 4.0 cm). Right main bronchial resections tended to be the total length of the bronchus, as it is naturally short, whereas left main bronchial resections were more variable. The entire left main bronchus was removed in 3 patients (2 for benign stenosis [l idiopathic, 1 postoperative] and 1 for stenosis secondary to tuberculosis). The intrapericardial hilar release was an especially critical tension-reducing maneuver during reconstruction in these patients. An additional patient with a carcinoid tumor had the carina between the left upper and lower lobes resected with anastomosis of the upper and lower lobe bronchi together before anastomosis to the proximal left main bronchus. Perioperative management began with immediate extubation in the operating room and intensive care observation, usually for 24 hours and occasionally 48 hours. All patients received aggressive chest physiotherapy and perioperative antibiotics, and 4 patients required postoperative bronchoscopy for clearance of secretions. The
1276
NEWTON ET AL MAIN BRONCHIAL RESECTION
A
Ann Thorac Surg 1991;52:1272-80
B
Fig 5. ( A ) Preoperative tracheabronchial tomograins demonstrating a right carcinoid tumor, identified by the arrow. ( B ) Postoperative tracheobronchial tomograms demonstrating a patent airway uiith an arrow marking the anastomosis.
average total hospital stay was 9 days (range, 6 to 36 days). Adjuvant therapy was not required for any patient.
Results
Symptoms and Preoperative Evaluation All patients were symptomatic and had an average of 2.5 symptoms. The presenting signs and symptoms included dyspnea (52%), wheeze or stridor (44%), cough (41%), hemoptysis (37%), pneumonia (18%), fever (ll%),and chest pain (7%).Left or right main bronchial location or pathology of the lesion produced no difference in symptoms. The average age at presentation was 35.0 years (range, 20 to 65 years), although generally patients with left main bronchial lesions were younger (33.1years) than those with right-sided lesions (42.6 years). Associated medical problems were uncommon (23%) and included coronary artery disease in 2 patients (1 of whom also had diabetes and diverticulosis) and chronic obstructive pulmonary disease, hypertension, gout, and Crohn’s disease in 1 patient each. Fifteen of the 26 patients (58%) had a history of smoking. Preoperative evaluation included a chest roentgenogram in all patients, which was normal in 41%. A specific
lesion was visible in the main bronchus in 18%, and secondary effects of atelectasis, pneumonia, or air trapping were noted in 41%. Standard tracheobronchial tomograms (Figs 5, 6) were especially useful in delineating the lesion in 16 of 17 patients (94%). Twelve patients had computed axial tomography before referral, detecting the lesion in 9 (Fig 7 ) . One patient, before referral, had a magnetic resonance scan which detected the lesion. Pulmonary function tests revealed an average forced expiratory volume in 1 second of 2.5 L (range, 0.8 to 5.4 L). Ventilation and perfusion scanning usually showed decreased function on the affected side, although depending on the size of the lesion, values were occasionally near normal. No patient had bone or brain scans, pulmonary angiography, or mediastinoscopy. Nineteen patients had undergone bronchoscopy before referral, which occasionally established the correct diagnosis, and all patients underwent bronchoscopy immediately before thoracotomy. The respective bronchus was on average stenosed to approximately 80% of normal caliber by visual inspection, with 15 patients showing nearly total obstruction. Despite this degree of obstruction, only 2 patients required preoperative therapeutic bronchoscopy to clear secretions of a postobstructive pneumonia. One patient
NEWTON ET AL MAIN BRONCHIAL RESECTION
Ann Thorac Surg 1991;52:1272-80
A
1277
B
Fig 6 . ( A ) Preoperative tracheobronchial totnograms demonstrating total left main stenosis secondany to tziberc-ulosis as outlined by the arrows. ( B ) Postoperative tracheobronchial tomagrams demonstrating left upper and lower division bronchi anastoniosed to the lateral trachea at the site of the arrow.
with a left main bronchial fibrous histiocytoma underwent a therapeutic bronchoscopy 48 hours before thoracotomy, and 1 patient with a right main bronchial carcinoid underwent therapeutic bronchoscopy 24 hours before thoracotomy. Both needed prompt relief of obstruction and drainage of postobstructive pneumonia in conjunction with appropriate preoperative antibiotics and chest physiotherapy to reduce sepsis in the involved lung.
Morbidity and Mortality Complications (Table 2) occurred in 4 of the 26 patients (15%). A 29-year-old woman with a 1-cm mucoepidermoid tumor of the left main bronchus had an initial 2.2-cm resection performed. Postoperatively, shortness of breath and wheezing prompted bronchoscopic examination, which revealed partial dehiscence due to anastomotic tension, which progressed to stenosis. She was managed with five rigid bronchoscopic dilations over the ensuing 5 months before undergoing subtotal resection of the remainder of her left main bronchus with intrapericardial hilar release. She recovered without difficulty; 6 years postoperatively she is working without respiratory symptoms. Transient hoarseness and temporary left vocal cord
Fig 7 . Computed tomographic S C ~ I Iof tumor (arrow).
a
/eft main bronc.hia/ rarcinoid
1278
NEWTON ET AL MAIN BRONCHIAL RESECTION
Ann Thorac Surg 1991;52:1272-80
Table 2. Morbidity of Main Bronchial Resections" Morbidity Anastomotic stenosis Transient recurrent nerve palsy Pneumonia Prolonged air leak Atelectasis a
Right
Left
0 0 1 1 0
1 1 0 0 1
There was no operative mortality
palsy developed in 1 patient with left main bronchial resection for a lymphoid aggregate but resolved without treatment after 6 months. One patient undergoing resection for stenosis after right upper lobectomy for adenocarcinoma had development of a prolonged (30-day) air leak which healed spontaneously. He also had a postobstructive Hemophilus pneumonia on presentation; postoperatively a new left lower lobe infiltrate developed despite clearing of the right lung. This responded to appropriate antibiotics. Minimal left lower lobe atelectasis, which responded to chest physiotherapy and did not prolong hospital stay, occurred in 1 patient after left main bronchial resection. There were no problems with anastomotic granulations, pulmonary torsion, or fistula. No patient died before hospital discharge or within 30 days of operation. Long-term follow-up showed that 92% of the patients (24 of 26) were alive a median of 5.0 years after operation (range, 1 month to 13 years). One patient, a 55-year-old woman who underwent left main resection for carcinoid, died 22 months postoperatively of a severe head injury after a motor vehicle accident. The second death occurred in a 35-year-old woman who underwent total left main bronchial resection for idiopathic stenosis. She originally was seen 5 years before operation with diffuse tracheobronchial inflammation. She required periodic dilations. During a period of decreased inflammation the persistent isolated left main bronchial stenosis was resected. Initially her postoperative course was uncomplicated, but recurrent dyspnea and wheezing 4 months postoperatively mandated four bronchoscopic dilations over a 14.5-month period before her death. Diffuse inflammation and stenosis of the distal trachea and both bronchi were noted. She suffered a respiratory arrest at home and died subsequently of anoxic cerebral injury at another hospital. One other patient also has suffered from diffuse idiopathic tracheobronchial stenosis. She is alive 12 years after left main bronchial resection with a patent anastomosis but requires dilation of her distal trachea and right main bronchus every 4 months.
Postoperative Evaluation Twenty-four patients underwent postoperative bronchoscopy (usually approximately 3 months after resection), and 21 (86%) were normal. Two of the abnormal examinations revealed diffuse inflammatory stenosis in the trachea and right main bronchus, both in patients with idiopathic stenosis and left main bronchial resection. One
abnormal examination was in the patient requiring reoperation for anastomotic stenosis. Her bronchoscopic examination is now normal. Eleven patients had postoperative tomograms; 10 were normal (see Figs 5, 6). The abnormal examination was in 1 of the patients with idiopathic stenosis. Twenty-three of the 24 living patients (96%)currently report no symptoms whatsoever and have returned to work. One patient with idiopathic stenosis has dyspnea. One patient who underwent resection of a right main bronchial stenosis after right upper lobectomy and irradiation (3,500 cGy) for lung carcinoma had a second primary lung tumor resected from his left lung 3 years after his main bronchial operation. He is the only patient in the series other than the 2 with idiopathic stenosis in whom any other chest pathology developed. Early postoperative pulmonary function testing was formally performed on 3 patients who also had such studies preoperatively. Performed an average of 7 days postoperatively (just before discharge), the studies had essentially the same results as preoperative studies. The lack of demonstrated improvement probably is due to persistent thoracotomy discomfort at 7 days.
Comment Main bronchial circumferential resection with end-to-end anastomosis has been described previously, mostly for traumatic disruptions and resultant stenosis and occasionally for primary tumors or tuberculosis. Although main bronchial traumatic injuries were described in 1873 [6]and 1947 [7], it was not until 1949 that Griffith [8] successfully performed a near circumferential resection of a posttraumatic stenosis of the left main bronchus with end-to-end reconstruction. Previous experimental work by Jackson and associates in 1948 [9] showed that main bronchial resection and primary anastomosis could be carried out successfully in the dog model. A variety of other authors have reported resection and reconstruction of main bronchial posttraumatic stenosis, with Deslaurier and colleagues' series of 7 in 1981 [lo] and Hood and Sloan's series of 4 in 1959 [ll] representing the largest experiences. Few reports of main bronchial resection and reconstruction for tumor, which form the bulk of this report, have appeared. Isolated cases of successful surgical treatment by circumferential resection and end-to-end anastomosis for carcinoid [12-141, fibrous histiocytoma (15), tuberculosis [16], and carcinoma [17]have been described. Frist and co-workers [3] noted some of the present cases in a report that dealt mostly with sleeve pulmonary resections. Other authors have reported bronchoplastic techniques either with lateral closure or with interposition or augmentation of the airway diameter with artificial or autologous materials [17-251. These techniques may result in narrowing of the airway due to either the initial repair or subsequent scarring. Avoidance of any narrowing is a primary concern in airway reconstruction. Circumferential sleeve resection with precise primary end-to-end
NEWTON ET AL MAIN BRONCHIAL RESECTION
Ann Thorac Surg 1991;52:1272-80
anastomosis as described here usually avoids this complication. The pathology of main bronchial lesions differs somewhat from lesions found in the remainder of the airway and generally permits their isolated resection. At the Massachusetts General Hospital, tracheal resection and reconstruction is undertaken most commonly (75% of cases) for postintubation stenosis [l-31. Stenosis was a common main bronchial lesion as well (37%), although none were secondary to intubation. Prior surgical manipulation, trauma, laser injury, and idiopathic stenosis were the pathologic lesions responsible for main bronchial stenosis. Previous surgical manipulation was usually pulmonary resection with tangential lobar bronchial division and resultant narrowing. Such stenoses would probably have been avoided if sleeve lobectomy had been carried out primarily. The most common tumors of the main bronchi were carcinoid (33%), mucoepidermoid (7.5%), and fibrous histiocytoma (7.5%). In the trachea, the most common tumors are adenoid cystic (approximately 40%) and squamous carcinoma (approximately 35%) [2], neither of which were found as isolated main bronchial lesions. Isolated carinal lesions, perhaps representing a transition zone between the trachea and main bronchus, reflect a broad mixture of lesions including adenoid cystic carcinoma (approximately 40%), carcinoid, stenosis, squamous cell carcinoma, and mucoepidermoid tumor [l]. Distal airway resections (sleeve lobectomy) were done at the Massachusetts General Hospital mainly for bronchogenic carcinoma (70%) and less commonly for carcinoid (21%) [3]. Based on the data presented, preoperative evaluation of patients with main bronchial lesions should include standard chest roentgenogram and tomography. Bronchoscopic examination of patients should be performed along with pulmonary function testing. If there is concern regarding potential pneumonectomy, ventilation-perfusion scanning should be performed. If the clinical course suggests an obstructive pneumonia, preoperative culture-specific antibiotics and bronchoscopic decompression should be used, followed in 24 to 48 hours by resection. Metastatic work-up and mediastinoscopic staging generally are not indicated as the majority of these lesions are benign or of low-grade malignancy (96% of cases). Patients with persistent tracheobronchial inflammation and focal areas of stenosis without identifiable causative factors should be handled cautiously as further airway problems may develop in these patients. This w a s exemplified by the 2 patients with idiopathic stenosis. Operative management is facilitated by use of the long, flexible Wilson tube. Tension-reducing maneuvers are critical for left main bronchial resection. Initially all patients should have the inferior pulmonary ligament divided and should have pretracheal dissection and neck flexion performed. If, upon approximation of the lateral traction sutures, tension still exists, then intrapericardial hilar release may yield 1 to 2 cm of added length, especially for subtotal or total left main bronchial resections. Tracheal and bronchial traction tapes, pulmonary
1279
artery retraction, and aortic mobilization are critical maneuvers for exposure. Clear resection margins and a precise anastomosis with absorbable sutures complete the operation. Postoperative tomograms and follow-up bronchoscopy either before discharge or several months postoperatively or both assist in ensuring an excellent result. A failed anastomosis can be successfully resected again as long as maximal tension reduction can be achieved. Postoperative stenosis, granulations, fistula, or other complications should be watched for but are rare. Main bronchial resection with pulmonary preservation can be carried out with minimal morbidity and mortality. The usually benign or low-grade malignant pathology permit such resections without pneumonectomy. Successful preservation of the normal lung allows virtually all patients to return to unrestricted activity.
References 1. Grillo HC. Carinal resection. Ann Thorac Surg 1982;34: 356-73. 2. Grillo HC, Mathisen DJ. Primary tracheal tumors: treatment and results. Ann Thorac Surg 1990;49:69-77. 3. Frist WH, Mathisen DJ, Hilgenberg AD, Grillo HC. Bronchial sleeve resection with and without pulmonary resection. J Thorac Cardiovasc Surg 1987;93:350-7. 4. Smith RA, Nigam BK. Resection of proximal left main bronchus carcinoma. Thorax 1979;34:61&20. 5. Grillo HC. Tracheal resection. In: Ravitch M, Steichen F, eds. Atlas of general thoracic surgery. Philadelphia: W.B. Saunders, 1988:296. 6. Seavre M. Crushing injury from wheel of an omnibus: rupture of the left main bronchus. Bull SOCAnat Paris 1873:48:680-2. 7. Kinsella TJ, Johnsrud LW. Traumatic rupture of the bronchus. J Thorac Surg 1947;16:571-83. 8. Griffith JL. Fracture of the bronchus. Thorax 1979;4:1059. 9. Jackson TL, Lefkin P, Tuttle W, Hampton F. An experimental study in bronchial anastomosis. J Thorac Surg 1949;18:63042. 10. Deslauriers J, Beaulieu M, Archambault G, LaForge J, Bernier R. Diagnosis and long-term follow-up of major bronchial disruptions due to nonpenetrating trauma. Ann Thorac Surg 1981;33:32-9. 11. Hood RM, Sloan HE. Injuries of the trachea and major bronchi. J Thorac Cardiovasc Surg 1959;38:458-80. 12. Spitzer SA, Segal 1, Lubin E, Nili M, Levy M. Unilateral increased transradiancy of the lung caused by bronchial carcinoid tumor. Thorax 1979;34:73944. 13. Grant JL, Naylor RW, Crandell WB. Bronchial adenoma resection with relief of hypoxic pulmonary vasoconstriction. Chest 1980;77:44&9. 14. DeLima R. Bronchial adenoma. Chest 1980;77814. 15. Yamazaki K, Kubo Y, Kawabata M. A case of fibrous histiocytoma in the left main bronchus. Jpn J Thorac Surg 1988;41: 34G2. 16. Caligiari PA, Banner AS, Jensik RJ. Tuberculous mainstem bronchial stenosis treated with sleeve resection. Arch Intern Med 1984;1244:1302-3. 17. Keszler P. Sleeve resection and other bronchoplasties in the surgery of bronchogenic tumors. lnt Surg 1986;71:229-32. 18. Lowe JE, Bridgman AH, Sabiston DC. The role of broncho-
1280
NEWTON ET AL MAIN BRONCHIAL RESECTION
plastic procedures in the surgical management of benign and malignant pulmonary lesions. J Thorac Cardiovasc Surg 1982;83:227-34. 19. Gebauer PW. Plastic reconstruction of tuberculous bronchostenosis with dermal grafts. J Thorac Surg 1950;19:604-28. 20. Thomas CP. Conservative resection of the bronchial tree. J R Coll Surg Edinb 1956;3:16&86. 21. Bjork VO, Rodriguez LE. Reconstruction of the trachea and its bifurcation. J Thorac Surg 1958;35:596-603.
Ann Thorac Surg 1991;52:1272-80
22. Nicks R. Restoration and reconstruction of the trachea and main bronchi. J Thorac Surg 1956;32:22645. 23. Mathey J, Galey JJ, Lemoire G. Tracheal and tracheobronchial resections. J Thorac Cardiovasc Surg 1966;51:1-13. 24. Keshishian JM, Blades B, Beattie EJ. Tracheal reconstruction. J Thorac Surg 1956;32:707-27. 25. Paulson DL, Urschel HC, McNamara JJ, Shaw RR. Bronchoplastic procedures for bronchogenic carcinoma. J Thorac Cardiovasc Surg 1970;59:38-48.
DISCUSSION DR HAROLD C. URSCHEL, JR (Dallas, TX): I congratulate Dr Newton and his colleagues on a landmark paper. I want to make two technical points. I talked to Hermes Grillo about a recent patient of mine. This 26-year-old patient had a totally occluded left main bronchus and underwent laser removal of a mucoepidermoid carcinoma of the left main bronchus. A stent was placed to support the bronchus, allowing time for drainage of the infected lung before resection. This is a valuable technique for total occlusions if an attempt is to be made to save the whole lung. The second point is that he described the inferior pericardial release, which we use. But in this case we performed a total circular pericardial release, which we use for lung transplants. The whole lung comes up easily without any tension at all on the tracheal suture line. I think the principles that you have illustrated technically are very valuable to all of us. DR NEWTON: Thank you very much.
DR ROBERT J. GINSBERG (New York, NY): I want to emphasize one technical point that Newton and associates mentioned briefly, and that is the type of tube used. Those who use a right-sided double-lumen tube will find that they cannot always mobilize the carina the way Newton and associates have demonstrated so nicely. It is very important to use something like the Roger Wilson adaptation, or a bronchial blocker [I], to allow the carina to be pulled down through the aortic arch. A doublelumen tube will fix the carina on the other side of the arch, making the anastomosis extremely difficult. I think this is a very important point in performing left main bronchial anastomoses. DR NEWTON: I agree with that point. We found the flexibility of the Wilson tube to be very important.
Reference 1. Ginsberg RJ. New technique for one-lung anesthesia using endobronchial blocker. J Thorac Cardiovasc Surg 1981;82: 542-6.
