- Email: [email protected]
Experience with primary neoplasms of the trachea and carina
J
THoRAc CARDIOVASC SURG
88:511-518, 1984
Experience with primary neoplasms of the trachea and carina From 1963 to 1983, 44 patients presented with a primary tracheal neoplasm that was amenable to
surgical treatment Forty-two of the 44 tumors were ma6gnant Thirty-three patients were managed by resection and primary anastomosis. The following reseetiees were done: trachea only, 12; trachea plus carina, 13; trachea plus cricoid cartilage, four; and trachea plus larynx, four. There were two operative death'l in these 33 patients. Prosthetic recOititruction with heavy-duty Marlex mesh was done in six patients. Three of the six died of erosion of the innominate artery during the postoperative period. In three patients with nonresectable tumors, a silicone-coated Montgomery T-tube provided transient but worthwhile palliation. In two patients with nonobstructive adenoid cystic carcinoma involving the subgIottis, irradiation was chosen as the initial treatment, since resection would necessitate laryngectomy. Resection, including laryngectomy, may be required in the future. The foUowing points are emphasized: (1) A majority of operable neoplasms can be resected through a cervical coUar incision and median sternotomy. Median sternotomy is the optimal operative exposure in most neoplasms necessitating resection of the carina. (2) Partial resection of the cricoid with sparing of the recurrent laryngeal nerves and larynx is possible in some patients with primary malignant tumors involvingthe proximal trachea and subglottic region. (3) In patients with adenoid cystic carcinoma, resection may afford exceUent, long-term palliation even when the resection is incomplete. Pulmonary metastases are common in patients with adenoid cystic tumors. However, they usually progress slowly, may remain asymptomatic for many years, and are not necessarily a contraindication to resection of the primary tumor even when they are synchronous. Our experience suggests that adjunctive radiotherapy is beneficial in patients with adenoid cystic carcinoma.
F. G. Pearson, M.D., T. R. J. Todd, M.D. (by invitation), and J. D. Cooper, M.D., Toronto, Ontario, Canada
Aside from anecdotal case reports, only a few centers throughout the world have reported a significant experience with the surgical management of primary tumors of the trachea." These publications identify that the majority of primary tracheal tumors are malignant and are frequently diagnosed at a stage when curative resection is no longer possible. Unfortunately, tumors of the trachea are not readily apparent in the plain chest film, and in many cases slowly progressive symptoms of upper airway obstruction are misdiagnosed as asthma or chronic bronchitis for long periods. The present report reviews our experience with the From the Division of Thoracic Surgery. University of Toronto and Toronto General Hospital, Toronto, Ontario, Canada. Read at the Sixty-fourth Annual Meeting of The American Association for Thoracic Surgery, New York, N. Y, May 7-9, 1984. Address for reprints: Dr. F. G. Pearson, Bell Wing 1-636, Toronto General Hospital, Toronto, Ontario, Canada M5G IL7.
surgical management of primary tracheal tumors at Toronto General Hospital between 1963 and 1983. The review does not include patients with nonreseetable tumors who received such palliative treatment as radiotherapy, chemotherapy, palliative endoscopic resections, or more recently endoscopic laser therapy. Patients Pathology. In this group of 44 patients, there were 28 with adenoid cystic carcinoma, nine with squamous cell carcinoma, four with leiomyosarcoma, and one each with malignant melanoma, neurofibroma, and chrondroma. Adenoid cystic carcinoma and squamous cell carcinoma are by far the commonest primary neoplasms encountered and account for 37 of our 44 cases. In a recent publication, Grillo" reported on 91 patients with primary tracheal tumors seen between 1962 and 1981 and managed by surgical resection of the trachea
5 11
The Journal of Thoracic and Cardiovascular Surgery
512 Pearson, Todd. Cooper
Table I. Management of 44 patients No. of
Therapy Segmental resection Primary anastomosis Prosthetic reconstruction Insertion of T-tube stent Radiotherapy-resection pending
patients 39 33
6 3
2
including larynx and carina. His series was composed of 39 adenoid cystic carcinomas, 25 squamous cell carcinomas, and 26 miscellaneous tumors, the majority of which were malignant. Management (Table I). The tumor was resected in 39 of the 44 patients. In 33 patients the defect was reconstructed by primary anastomosis, and in six patients requiring extensive resections a porous prosthesis of heavy-duty Marlex mesh was used. Eighteen patients required resection of a segment of trachea only, 13 patients required the addition of a carinal resection, four patients required the addition of subglottic, partial cricoid resection, and in four patients requiring resection of both the trachea and larynx the airway was restored by a permanent cervical tracheostomy. Three patients had technically unresectable tumors. Their symptoms were palliated by a silicone T-tube stent that was introduced through the anterior wall of the cervical trachea by way of a cervical incision. In two of these three patients, a T-Y tube was used because the obstructed airway extended to or beyond the main carina. All three patients had squamous cell carcinoma. The technique used for the introduction of the T-Y tube has been described previously.' Two patients had nonobstructing adenoid cystic carcinoma with involvement of the subglottis at a level necessitating laryngectomy for complete and potentially curative resection. In both of these patients the lesion was technically operable, but we chose to use radiotherapy as the initial treatment in order to preserve the voice for as long as possible. The status of the tumor is being monitored by regular interval bronchoscopy, and both patients may require a tracheal resection with laryngectomy and terminal tracheostomy at some future date. Anesthetic technique. Thirty of the 39 resected cases were managed by previously described techniques" that require the use of sterile anesthetic connections introduced directly through the operative field for intubation of the distal end(s) of the divided airway. In most cases the sterile transoperative circuit was maintained with cuffed, armored endotracheal tubes. We 9 have previous-
ly reported specific techniques for the anesthetic management of patients requiring carinal resection. During the past 2 years, we have used the newer techniques of jet ventilation in six patients and have identified some significant advantages with this method compared with the standard techniques of intermittent intubation. Operative exposure. A cervical collar incision was used in 10 patients, all of whom had lesions restricted to the cervical trachea. Four of the 10 patients required a partial cricoid resection, an additional four required a laryngectomy, and two patients with small lesions were managed by resection of a short circumferential segment of cervical trachea. Although a generous collar incision will provide access to a considerable length of the mediastinal trachea, this limited incision will not provide the type of exposure required for block resection of a malignant tumor when the tumor extends into the upper mediastinum. A right posterolateral thoracotomy was used in 11 patients. Six of the 11 patients had a carinal resection, and the remaining five patients had resections of relatively short segments of distal mediastinal trachea. The combination of a cervical collar incision and full median sternotomy provides access to the airway from thyroid cartilage to carina. With experience, we have selected this operative exposure for an increasing number of patients seen during the past decade. Median sternotomy was used in 18 of the 39 resections. Eleven of these 18 resections were limited to the trachea, and the remaining seven included the carina. Thus, median sternotomy was used in seven of our 13 carinal resections and is currently our incision of choice for the majority of such cases. Exposure of the carina and main bronchi through a median sternotomy requires a transpericardial approach. The anterior pericardium is divided vertically to permit circumferential mobilization of the ascending aortic arch, which is then retracted laterally and to the left. The superior vena cava is displaced laterally and to the right, and the right main pulmonary artery is exposed and displaced inferiorly. The posterior pericardium, which is displayed by these latter maneuvers, is then divided vertically and the entire mediastinal trachea and carina are clearly exposed and accessible. This technique of operative exposure has several distinct advantages over a right posterolateral thoracotomy in selected cases: 1. Any pulmonary resection is possible through a median sternotomy. Five of the patients having a combined tracheal and carinal resection required excision of the right main bronchus and right upper lobe.
Volume 88 Number 4 October. 1984
Two additional patients having a tracheal and carinal resection required a concomitant left pneumonectomy. 2. A cervical collar incision combined with a median sternotomy provides easy access for a superior laryngeal or suprahyoid release procedure when required. This incision also provides adequate exposure for an intrapericardial mobilization of the right pulmonary hilum, which may also be necessary to minimize tension on a tracheal or tracheobronchial anastomosis. Tension-reducing techniques. Mobilization and release procedures to reduce tension on the anastomosis were used in any case in which undue tension was anticipated. This decision was usually made during the operation after attempts to approximate the ends of the divided airway suggested that undue tension would result. The decision to use right hilar mobilization or a superior release procedure was not based on the length of airway resected, since the degree of tension on an anastomosis after circumferential resection of any given length of trachea varies considerably from person to person. On occasion, a satisfactory tension-free anastomosis was achieved without mobilization after resection of more than 5 em of the adult tracheal length. In this series of 39 resections, a superior laryngeal release'? was used in eight patients. During the past 10 years, we have favored the suprahyoid release procedure described by Montgomery" and have used it in three patients. Intrapericardial mobilization of the right pulmonary hilum was used to reduce anastomotic tension in 12 patients. In all patients, neck flexion was maintained for approximately 1 week after operation by the use of a heavy suture (No.5 gauge) running between the skin and subcutaneous tissue over the point of the chin to the skin overlying the upper sternum." Results Mortality and morbidity. There were no intraoperative deaths. Five postoperative deaths occurred in the 39 patients undergoing resection. There were only two postoperative deaths in the 33 patients in whom a primary reconstruction was done. In one of these two patients, death resulted after 40 days from an anastomotic dehiscence and subsequent secondary pneumonia. This patient with adenoid cystic carcinoma had the most extensive resection with primary reconstruction of any patient in the series: All but the upper two tracheal rings were resected along with the carina, and the right main bronchus was anastomosed to the cervical trachea. The left main bronchus could not be elevated for any method of anastomosis, and the left lung was removed. The anastomosis was obviously under
Primary neoplasms of trachea and carina 5 1 3
considerable tension and ultimately failed despite reinforcement with a vascularized pedicle of greater omentum. In retrospect, either a lesser and incomplete resection, leaving residual disease at the primary anastomosis, or prosthetic replacement should have been used. The second postoperative death in this group occurred at 4 days as a result of pulmonary embolism from a thrombus (due to operative trauma) in the innominate vein, which had not been divided. As in the previous case, this patient with an adenoid cystic carcinoma had required carinal and extensive tracheal resection that precluded reanastomosis of the left main bronchus, and the left lung was removed at the time of reconstruction. The postoperative mortality was forbidding in the six patients in whom a Marlex prosthesis was used. Three of the six patients died within 2 weeks of operation because of erosion of the innominate artery at the site where the artery overlies this relatively rigid prosthesis. Death was due to an abrupt and massive hemorrhage into the reconstructed airway. There were few serious, nonlethal complications. Most patients had transient difficulty in raising secretions during the early postoperative period, and although transient atelectasis and pneumonitis, occurred there was only one death from prieumonia secondary to the anastomotic dehiscence previously described. Aspiration, particularly with swallowed liquids, was observed in five of the eight patients in whom a superior laryngeal release was added and in one of the three patients having a suprahyoid release procedure. This complication resolved in every. patient within a period ranging from I week to a maximum of 3 months after operation. The most severe and prolonged problems with aspiration occurred in the elderly group. One recurrent laryngeal nerve. was deliberately resected in four patients because of the location and proximity of the primary tumor. In no instance, however, was there a permanent inadvertent injury to the recurrent laryngeal nerves. Only one of the 34 surviving patients has a functionally significant anastomotic stricture. In this patient, an adenoid cystic carcinoma was resected by removal of the carina and by a sleeve resection of the right upper lobe. The stricture occurred at the anastomosis between the distal trachea and bronchus intermedius and was satisfactorily controlled by interval endoscopic dilatation. Survival. Adenoid cystic carcinoma. Twenty-six of the 28 patients with adenoid cystic tumors were managed by resection. All five postoperative deaths in the series
The Journal of
5 14
Pearson, Todd. Cooper
occurred in this group. Twelve of the 21 patients who survived operation had a complete and potentially curative resection. The resection was incomplete, with proved residual tumor at the time of operation, in the remaining nine patients. All of the 21 survivors with adenoid cystic carcinoma received adjuvant radiotherapy before or after operation. For those patients in whom preoperative radiotherapy was electively used, the dosage was restricted to between 3,000 and 3,500 rads administered in 15 treatments during a 3 week period. Any patient who had an incomplete resection and who had received preoperative irradiation was referred for additional postoperative radiotherapy. During the past 10 years we have chosen to initiate this combined treatment approach with resection. Frozen section and formal histologic assessment by the pathologist are more accurate when the tissues have not been irradiated. There is no radiation damage that might prejudice healing, and a higher dosage (4,000 rads in twenty treatments during 4 weeks) can be administered safely within 1 to 2 months of operation. Of the 12 patients who survived operation and had a complete and potentially curative resection, nine are alive and clinically free of local recurrence at follow-up periods of from 1 to 20 years, with a mean survival of 8.3 years. Three of these 12 patients have died of unrelated disease between 6 and 18 years after operation, with a mean survival of 12 years. Two patients in this group have asymptomatic pulmonary metastases, and both are living and clinically well at 13 and 16 years. Nine of the 21 patients surviving operation had incomplete resections. Seven of these nine patients have died of recurrent tumor at intervals from 2 to 9 years after resection. Five of these seven deaths were due to a local recurrence of the tumor, two were due to cerebral metastases, and one was due to diffuse pulmonary metastases. One of the nine patients with an incomplete resection is alive and clinically free of recurrence. IRRADIATION ONLy-oPERATION PENDING. The two previously described patients with nonobstructive adenoid cystic carcinoma extending to the subglottic larynx were deliberately treated initially with radical local irradiation with the intention of preserving a normal voice for as long as possible. Both patients have had an excellent response to radiotherapy and are alive and clinically free of recurrent disease at 36 and 40 months, respectively. Their follow-up evaluation includes bronchoscopy at 6 month intervals. If and when evidence of locally recurrent tumor develops, each will require a resection with laryngectomy and a permanent suprasternal end tracheostomy.
Thoracic and Cardiovascular Surgery
SURVIVAL IN PATIENTS WITH METASTASES. Lymphatic spread was not observed in any patient with adenoid cystic carcinoma. Hematogenous metastases have occurred in 10 of the 28 patients: pulmonary in nine, brain in two, and bone in one. The two patients with metastatic brain tumor died of their metastases at 1 and 6.5 years, respectively. The natural history of pulmonary metastases is of important practical interest. In the nine patients with pulmonary involvement, the metastases were synchronous with the primary tumor in two instances and metachronous in seven. Metastases were multiple and bilateral in all cases. Seven of the nine patients with pulmonary metastases have died, but in only two of the nine were the pulmonary metastases responsible for the deaths. These two patients died of their pulmonary spread 68 and 108 months after operation. The causes of death in the other five patients were concomitant brain metastases in two instances (both had asymptomatic pulmonary spread), local recurrence in two instances (again both with asymptomatic pulmonary spread), and a postoperative complication in one instance. Two of the nine patients with pulmonary metastases are alive 13 and 16 years after operation. Neither patient is seriously disabled. One is asymptomatic and the other has a mild intermittent cough and poorly defmed chest discomfort. Squamous cell carcinoma. Six of the nine patients with squamous cell tumors were treated by resection. There were no operative deaths. Two of the six undergoing resection have died at 6 months (pancreatic metastases), and 46 months (local recurrence). Four patients are living and clinically free of recurrence 6, 16,21, and 56 months after operation. Three of the nine patients had unresectable squamous cell tumors and were managed by the introduction of a silicone T-tube stent (one T-tube and two T-Y-tubes). All three patients had received maximal doses of radiotherapy prior to this treatment. A satisfactory transient palliation was achieved in each patient, but all three died within 3 to 6 months of the introduction of the internal stent. Miscellaneous tumors. Three of the four patients with leiomyosarcoma are alive and clinically free of tumor 16, 63, and 125 months after operation. One patient died of a rapid and massive local recurrence 22 months after resection. The patient with a primary malignant melanoma of the trachea died of widespread hematogenous metastases after 14 months. The patient with a tracheal chondroma died of an unrelated squamous cell carcino-
Volume 88 Number 4 October, 1984
ma of the esophagus 18 months after tracheal resection, and the patient with a neurofibroma is alive and clinically disease free at 53 months. Subglottic resections. In four patients, the operation included a partial resection of the cricoid cartilage by a previously reported technique that preserves the recurrent laryngeal nerves when feasible." There was one case each of adenoid cystic carcinoma, squamous cell carcinoma, leiomyosarcoma, and neurofibroma. In only one of the four patients was it necessary to sacrifice one recurrent laryngeal nerve because of tumor involvement. The alternative to a subglottic resection in the three patients with malignant tumors would have been a total laryngectomy. They have been maintained on a regular follow-up program, and all four patients are alive without clinical evidence of recurrence: adenoid cystic carcinoma, 11 years; squamous cell carcinoma, 11 months; leiomyosarcoma, 10 months, neurofibroma, 53 months. Prosthetic reconstruction. Our initial experience with the experimental and subsequent clinical use of heavyduty Marlex mesh as a tracheal prosthesis has been reported previously.v" Three of these six patients died postoperatively of innominate artery erosion. A safe and satisfactory airway was restored in the three surviving patients. One of these three survivors died of locally recurrent tumor at 25 months. In one patient the prosthesis provided an excellent airway during the first 4.5 years after operation. Thereafter, a progressive accumulation of granulation and scar developed at the proximal anastomosis, and the prosthesis was removed 5.5 years after the initial resection. This second operation done in 1968 when it had become possible to reconstruct extensive tracheal defects by primary anastomosis by the currently available mobilization and release techniques. This patient remains alive and clinically free of recurrent tumor 21 years after the initial operation. In the third survivor, a local recurrence developed following an initial resection of the mediastinal trachea with reconstruction by primary anastomosis. At a second operation, it was necessary to resect the entire trachea including the larynx and the cervical esophagus. The esophageal defect was reconstructed by elevation of the stomach through the posterior mediastinum and pharyngogastrostomy. The innominate artery was resected and the mediastinal trachea was replaced with a cylinder of heavy-duty Marlex mesh, which was anastomosed distally at the level of the carina and secured proximally as a cutaneous stoma in the suprasternal notch. The patient died of brain metastases 68 months after the second operation.
Primary neoplasms of trachea and carina
5 15
Discussion Our experience with the surgical management of primary tumors of the trachea is similar to that reported by others. Most of the primary neoplasms are malignant, and adenoid cystic carcinoma and squamous cell carcinoma are by far the most common histologic types. Using the currently available techniques of mobilization and "release," the surgeon can safely resect long circumferential segments of the trachea and achieve primary reconstruction without the need of a prosthesis in the majority of operable patients. As experience with carinal resection has gradually increased, the innovations in techniques of reconstruction have advanced the capability to manage tumors at this level. A cervical collar incision with median sternotomy is now our incision of choice in most patients requiring a carinal resection. Worthwhile survival can be obtained in patients with adenoid cystic carcinoma, squamous cell carcinoma, and sarcomas of the trachea when a complete and potentially curative resection is possible. In the majority of patients with adenoid cystic carcinoma, even an incomplete resection may provide long periods of symptomfree palliation. Most adenoid cystic carcinomas are radiosensitive, and adjuvant radiotherapy is recommended in all resectable tumors of this cell type. Pulmonary metastases are relatively common in patients with adenoid cystic carcinoma and frequently progress very slowly. A majority of our patients with such metastases have remained asymptomatic for many years, and we do not consider synchronous pulmonary metastases to be inoperable if the primary tumor is resectable. Safe and reliable methods of prosthetic replacement remain elusive. Our limited experience with a semirigid porous prosthesis of heavy-duty Marlex mesh is clearly unsatisfactory. Innominate artery erosion from the prosthesis was a fatal complication in half of our patients. With such a limited number of cases, however, the three good results obtained with a Marlex prosthesis are anecdotal. Neville, Hamouda, and Anderson" are the only group that has reported a large clinical experience with a silicone rubber prosthesis, and no other authors have yet reported a significant number of similar favorable results.
REFERENCES Houston HE, Payne WS, Harrison EG Jr: Primary cancers of the trachea. Arch Surg 99:132, 1969
The Journal of Thoracic and Cardiovascular Surgery
516 Pearson. Todd. Cooper
2 Eschapasse H: Les tumeurs tracheales primitives traitement chirurgical. Rev Fr Mal Respir 2:245, 1974 3 Pearson FG, Thompson DW, Weissberg D, Simpson WJK, Kergin FG: Carcinoma of the trachea. Ann Thorac Surg 18:16, 1974 4 Grillo HC: tracheal tumors. Surgical management. Ann Thorac Surg 26: 112, 1978 5 Perelman M, Koroleva N: Surgery ofthe trachea. World J Surg 4:583, 1980 6 Grillo HC: Tracheal surgery. Scand J Thorac Cardiovasc Surg 17:67, 1983 7 Westaby S, Jackson J, Pearson FG: A bifurcated silicone rubber stent for relief of tracheobronchial obstruction. J THoRAc CARDIOVASC SURG 83:414, 1982 8 Grillo HC: Surgery of the trachea, Current Problems in Surgery, Chicago, 1970, Year Book Medical Publishers, Inc. 9 Theman TE, Kerr JH, Nelems JM, Pearson FG: Carinal resection. A report of two cases and a description of the anesthetic technique. J THORAC CARDIOVASC SURG71:314, 1976 10 Dedo HH, Fishman NH: Laryngeal release and sleeve resection for tracheal stenosis. Ann Otol Rhinol Laryngol 78:285, 1968 II Montgomery WW: Suprahyoid release for tracheal stenosis. Arch Otolaryngol 99:255, 1974 12 Pearson FG, Andrews MJ: Detection and management of tracheal stenosis following cuffed tube tracheostomy. Ann Thorac Surg 12:359, 1971 13 Pearson FG, Cooper JD, Nelems JM, Van Nostrand A WP: Primary tracheal anastomosis after resection of the cricoid cartilage with preservation of recurrent laryngeal nerves. J THoRAc CARDIOVASC SURG 70:806, 1975 14 Pearson FG, Henderson RD, Gross AE, Ginsberg RJ, Stone RM: The reconstruction of circumferential tracheal defects with a porous prosthesis. An experimental and clinical study using heavy Marlex mesh. J THoRAc CARDIOVASC SURG 55:605, 1968 15 Neville WE, Hamouda F, Anderson J: Replacement of the intrathoracic trachea and both stem bronchi with a molded Silastic prosthesis. J THORAC CARDIOVASC SURG 63:569, 1972
Discussion DR. HERMES C. GRILLO Boston. Mass.
We are grateful to Dr. Pearson and his colleagues for reporting this excellent series, since there have been few series of tracheal tumors reported from single institutions (or even collected series) treated by current techniques. When we separate the tumors into types, namely, adenoid cystic, squamous, and the others, and further consider the very long follow-up necessary for adenoid cystic carcinomas, it is difficult to determine the results precisely. However, the fact that the three major series reported (from Boston, Moscow, and Toronto) plus a fourth combined series (from Toulouse) show similar results gives reassurance that progress is being made.
At Massachusetts General Hospital, I have seen 122 primary tracheal tumors in about the same period of time. However, the distribution has been a little different in that 46 were squamous and 45 were adenoid cystic. The other 31 included six carcinoids and multiple other varieties. Sixty-two primary reconstructions were done. I strongly concur that the use of radiation for adenoid cystic carcinoma is critical. In some early cases in which irradiation was not used because margins were not diseased, recurrence appeared as long as 18 years later. With combined surgical resection and postoperative irr~di ation, survival has been quite good both for adenoid cystic carcinoma and, surprisingly, for squamous cell carcinoma. As Dr. Pearson pointed out, the presence or absence of microscopic tumor at resection margins or in lymphatics did not seem to be the critical factor. Obviously, more data are yet required. DR. LESTER BRY ANT Johnson City. Tenn.
Our own past experience with primary squamous carcinoma of the trachea has been sufficiently discouraging that, when presented with a 60-year-old man with a squamous carcinoma last year, we decided to try a new approach. The carcinoma was located on the right lateral and posterior walls of the trachea just above the carina. The chest radiograph showed enlargement of the left hilum consistent with lymph node metastases, and this was confirmed by computed tomographic scan. The patient's initial treatment consisted of external beam cobalt teletherapy to the primary tumor and lymph node drainage areas in the mediastinum and paraclavicular regions (total dose 3,000 rads). This was followed by bilateral staged thoracotomies during which we implanted a total of 97 seeds of radioactive iodine (iodine 125) in the wall ofthe trachea and in the adjacent mediastinal tissues. The seeds will provide 15,000 rads of gamma therapy to the target volume over a period of I year, but the contiguous organs will receive less than 2,000 rads at a distance of 2 to 5 em from the implants. It is now 16 months since the patient received external beam therapy and I year after his second thoracotomy for interstitial radioactive iodine therapy. He has no evidence of tumor recurrence, but he does have a significant radiation tracheitis which has produced an intractable cough. Biopsy specimens of the tracheal mucosa have demonstrated the typical dysplastic changes of radiation effect. It is difficult for anyone surgeon to gain much experience with primary squamous tumors of the trachea because of the rarity of this neoplasm. However, we are encouraged with the early results in this patient, and we are hopeful that other surgeons might try this approach in order to determine if the technique has merit. DR. W. CLARK HARGROVE Philadelphia. Pa.
We at the Presbyterian Hospital in Philadelphia agree that amenable lesions should be treated by surgical extirpation.
Volume 88 Number 4 October. 1984
However, our experience with this problem has been in the palliative management by laser therapy of a group of patients with obstructive lesions of the tracheobronchial tree. Over the past 18 months we have performed some 200 laser resections in 89 patients. Of these 89, 25 had primary squamous neoplasms of the trachea and carina. Surgical resection was contraindicated in all patients because of evidence of metastatic disease or poor general medical condition. All patients had respiratory distress on admission to the hospital. In most cases, it was moderate to severe. Of the 25 patients with tracheal obstructions, 20 (80%) had good results. We defined good results as relief of airway obstruction and ability to ambulate without supplemental oxygen. There were no hospital deaths. Each of the patients who did not respond to treatment died within 1 month of discharge from the hospital. The longest survival period has been 10 months after the patient's admission to the hospital with an endotracheal tube in place. Dr. Pearson, in reviewing your experience with squamous carcinomas of the trachea, do you think surgical resection should remain the primary form of therapy, or might the laser offer an alternative form of treatment in the majority of patients? Might the laser also have an adjunctive role in the management of adenoid cystic lesions? DR. EDWARD W. HUMPHREY Minneapolis. Minn.
I rise to ask if Dr. Pearson has observed a phenomenon that I have seen in patients undergoing an extensive resection of the carina including the lower trachea and both main bronchi with a consequent wide lymph node dissection in that area. In two patients treated in this manner, refractory ventricular arrhythmias developed and lasted for about 3 weeks after the operation. The problem then resolved and never recurred. These patients had no prior history of cardiac abnormalities. Since this phenomenon does not usually occur in transplanted hearts, I am uncertain whether these have been isolated instances or whether they might have resulted from the interruption of all the lymphatic drainage of an innervated heart. DR. WILLIAM E. NEVILLE Newark. N. J.
I certainly agree that a resection and pulmonary anastomosis is the ideal method of dealing with malignant tracheal tumors. However, is some patients a prosthesis must be used to reconstruct the airway. One big problem that has been alluded to is that no one surgeon with the exception of Dr. Grillo and the.group at Massachusetts General, the Mayo Clinic, Memorial, and now Toronto have a vast experience in the management of tracheal cancers, since they comprise only about 1% of all malignant tumors. I am going to confine my remarks to the prosthesis. Dr. Pearson did have good results in using a prosthesis, but he became disenchanted when three patients died of innominate artery hemorrhage and one patient developed a stenosis. At present, we are all aware of the possibility of arterial erosion whether a suture line or a prosthesis is behind the innominate
Primary neoplasms of trachea and carina 5 1 7
artery, and there are methods to prevent this. Also, with a nonporous tube, stenosis cannot occur. Theoretically, all of Dr. Pearson's complications are preventable at the present time. My second point is that excellent palliation can be achieved in nonreseetable lesions or recurrent cancer causing airway obstruction by using an intraluminal stent of the silicone rubber prosthesis. In one man with complete airway obstruction, we made a collar incision and were able to place two prostheses through and beyond the obstruction by "piggy-backing" them. A technique that is efficacious is to look down from above with a flexible scope and position the lower tube so that it does not obstruct the carina. Three years postoperatively, this man is doing well. In another patient with a right pneumonectomy and recurrent cancer, we used a median sternotomy. The trachea was opened, a straight prosthesis was inserted through the obstructed area, and a satisfactory airway was established. In the past 15 years, we have used the straight prosthesis for malignancy in eight patients; four are alive and well after 1 to 6 years. We have used the prosthesis as an intraluminal stent in three patients with cancer, and one is alive at 3 years. The bifurcated prosthesis has been used in 14 patients. There have been complications, but five of the patients are alive and well after 1 to 11 years. DR. QUENTIN R. STILES Los Angeles. Calif
Dr. Pearson, I am very interested in your use of jet ventilation. Did you use that because you found the ventilation was actually better or simply because it was less cumbersome than the two tubes? DR. PEARSON (Closing) We have used jet ventilation recently, Dr. Stiles, because it is simpler and less cumbersome than the standard techniques. Jet ventilation is delivered through a small-bore catheter which can be left in place in the distal airway during reconstruction and anastomosis. There is an added advantage in that a respiratory frequency of 120 cycles per minute causes minimal movement in the lungs and tracheobronchial tree. The "moving field" created by normal spontaneous respiration is abolished. Dr. Grillo has the largest reported experience in the world with primary neoplasms of the trachea. He has recently reported on a group of more than 100 patients with primary tracheal tumors. Squamous cell carcinoma was slightly more common than adenoid cystic carcinoma in his series. The adenoid cystic tumors, however, were more frequently operable than the squamous cell cancers. Dr. Bryant's report on the use of radioactive iodine and radioactive implant techniques is interesting-and novel in the management of tracheal tumors. I believe a median sternotomy might provide better access for this surgical procedure than bilateral thoracotomies. It is also important to emphasize that most adenoid cystic cancers will respond to conventional techniques of irradiation, and unresectable tumors of this kind
5 1 8 Pearson, Todd, Cooper
should still, 1 think, be treated initially with such standard radiation techniques. Dr. Hargrove has reported on an extensive experience with the use of a laser for tracheobronchial lesions. Twenty-five of lis patients had primary squamous tumors of the trachea, or trachea and carina. Our experience with the laser in such cases is very limited and has only recently been initiated. We would reserve such treatment for patients with inoperable tumors, for the symptomatic relief of upper airways obstruction. I do not anticipate, however, that the laser will completely replace the older techniques of endoscopic removal with biopsy forceps. In many cases, endoscopic resection with biopsy forceps will be as effective as laser resection (though not quite as neat), and it is much more rapidly achieved. I would still consider laser therapy for primary squamous cell carcinoma of the trachea as a palliative modality. In patients with a technically resectable squamous cell carcinoma, I believe there is no question but that operative removal is the best current therapy. Dr. Humphrey, we did not record any incidence of refractory ventricular arrhythmias in our patients who had a carinal
The Journal of Thoracic and Cardiovascular Surgery
resection. I have no explanation for your personal observations. Dr. Neville has reported the world's largest experience with prosthetic replacement and has achieved some excellent longterm results. I believe that some of his patients could have been managed by resection and primary anastomosis, and I still think this is the preferable mode of reconstruction when possible. All but one of our six prosthetic reconstructions (using heavy-duty Marlex mesh) were done during the 19608, at a time when extensive resection with primary reconstruction was at an early stage of development. One of our six patients was operated upon in 1973, and because of our unfavorable experience with innominate artery erosion, the innominate artery was deliberately resected in this patient at the time of the original operation. He survived for 6 years with this prosthesis before dying of cerebral metastases. If primary tracheal tumors were as common as coronary artery disease, I am certain we would have developed, by now, a very effective prosthesis for tracheal replacement. Unfortunately, it still remains an elusive goal.
THoRAc CARDIOVASC SURG
88:511-518, 1984
Experience with primary neoplasms of the trachea and carina From 1963 to 1983, 44 patients presented with a primary tracheal neoplasm that was amenable to
surgical treatment Forty-two of the 44 tumors were ma6gnant Thirty-three patients were managed by resection and primary anastomosis. The following reseetiees were done: trachea only, 12; trachea plus carina, 13; trachea plus cricoid cartilage, four; and trachea plus larynx, four. There were two operative death'l in these 33 patients. Prosthetic recOititruction with heavy-duty Marlex mesh was done in six patients. Three of the six died of erosion of the innominate artery during the postoperative period. In three patients with nonresectable tumors, a silicone-coated Montgomery T-tube provided transient but worthwhile palliation. In two patients with nonobstructive adenoid cystic carcinoma involving the subgIottis, irradiation was chosen as the initial treatment, since resection would necessitate laryngectomy. Resection, including laryngectomy, may be required in the future. The foUowing points are emphasized: (1) A majority of operable neoplasms can be resected through a cervical coUar incision and median sternotomy. Median sternotomy is the optimal operative exposure in most neoplasms necessitating resection of the carina. (2) Partial resection of the cricoid with sparing of the recurrent laryngeal nerves and larynx is possible in some patients with primary malignant tumors involvingthe proximal trachea and subglottic region. (3) In patients with adenoid cystic carcinoma, resection may afford exceUent, long-term palliation even when the resection is incomplete. Pulmonary metastases are common in patients with adenoid cystic tumors. However, they usually progress slowly, may remain asymptomatic for many years, and are not necessarily a contraindication to resection of the primary tumor even when they are synchronous. Our experience suggests that adjunctive radiotherapy is beneficial in patients with adenoid cystic carcinoma.
F. G. Pearson, M.D., T. R. J. Todd, M.D. (by invitation), and J. D. Cooper, M.D., Toronto, Ontario, Canada
Aside from anecdotal case reports, only a few centers throughout the world have reported a significant experience with the surgical management of primary tumors of the trachea." These publications identify that the majority of primary tracheal tumors are malignant and are frequently diagnosed at a stage when curative resection is no longer possible. Unfortunately, tumors of the trachea are not readily apparent in the plain chest film, and in many cases slowly progressive symptoms of upper airway obstruction are misdiagnosed as asthma or chronic bronchitis for long periods. The present report reviews our experience with the From the Division of Thoracic Surgery. University of Toronto and Toronto General Hospital, Toronto, Ontario, Canada. Read at the Sixty-fourth Annual Meeting of The American Association for Thoracic Surgery, New York, N. Y, May 7-9, 1984. Address for reprints: Dr. F. G. Pearson, Bell Wing 1-636, Toronto General Hospital, Toronto, Ontario, Canada M5G IL7.
surgical management of primary tracheal tumors at Toronto General Hospital between 1963 and 1983. The review does not include patients with nonreseetable tumors who received such palliative treatment as radiotherapy, chemotherapy, palliative endoscopic resections, or more recently endoscopic laser therapy. Patients Pathology. In this group of 44 patients, there were 28 with adenoid cystic carcinoma, nine with squamous cell carcinoma, four with leiomyosarcoma, and one each with malignant melanoma, neurofibroma, and chrondroma. Adenoid cystic carcinoma and squamous cell carcinoma are by far the commonest primary neoplasms encountered and account for 37 of our 44 cases. In a recent publication, Grillo" reported on 91 patients with primary tracheal tumors seen between 1962 and 1981 and managed by surgical resection of the trachea
5 11
The Journal of Thoracic and Cardiovascular Surgery
512 Pearson, Todd. Cooper
Table I. Management of 44 patients No. of
Therapy Segmental resection Primary anastomosis Prosthetic reconstruction Insertion of T-tube stent Radiotherapy-resection pending
patients 39 33
6 3
2
including larynx and carina. His series was composed of 39 adenoid cystic carcinomas, 25 squamous cell carcinomas, and 26 miscellaneous tumors, the majority of which were malignant. Management (Table I). The tumor was resected in 39 of the 44 patients. In 33 patients the defect was reconstructed by primary anastomosis, and in six patients requiring extensive resections a porous prosthesis of heavy-duty Marlex mesh was used. Eighteen patients required resection of a segment of trachea only, 13 patients required the addition of a carinal resection, four patients required the addition of subglottic, partial cricoid resection, and in four patients requiring resection of both the trachea and larynx the airway was restored by a permanent cervical tracheostomy. Three patients had technically unresectable tumors. Their symptoms were palliated by a silicone T-tube stent that was introduced through the anterior wall of the cervical trachea by way of a cervical incision. In two of these three patients, a T-Y tube was used because the obstructed airway extended to or beyond the main carina. All three patients had squamous cell carcinoma. The technique used for the introduction of the T-Y tube has been described previously.' Two patients had nonobstructing adenoid cystic carcinoma with involvement of the subglottis at a level necessitating laryngectomy for complete and potentially curative resection. In both of these patients the lesion was technically operable, but we chose to use radiotherapy as the initial treatment in order to preserve the voice for as long as possible. The status of the tumor is being monitored by regular interval bronchoscopy, and both patients may require a tracheal resection with laryngectomy and terminal tracheostomy at some future date. Anesthetic technique. Thirty of the 39 resected cases were managed by previously described techniques" that require the use of sterile anesthetic connections introduced directly through the operative field for intubation of the distal end(s) of the divided airway. In most cases the sterile transoperative circuit was maintained with cuffed, armored endotracheal tubes. We 9 have previous-
ly reported specific techniques for the anesthetic management of patients requiring carinal resection. During the past 2 years, we have used the newer techniques of jet ventilation in six patients and have identified some significant advantages with this method compared with the standard techniques of intermittent intubation. Operative exposure. A cervical collar incision was used in 10 patients, all of whom had lesions restricted to the cervical trachea. Four of the 10 patients required a partial cricoid resection, an additional four required a laryngectomy, and two patients with small lesions were managed by resection of a short circumferential segment of cervical trachea. Although a generous collar incision will provide access to a considerable length of the mediastinal trachea, this limited incision will not provide the type of exposure required for block resection of a malignant tumor when the tumor extends into the upper mediastinum. A right posterolateral thoracotomy was used in 11 patients. Six of the 11 patients had a carinal resection, and the remaining five patients had resections of relatively short segments of distal mediastinal trachea. The combination of a cervical collar incision and full median sternotomy provides access to the airway from thyroid cartilage to carina. With experience, we have selected this operative exposure for an increasing number of patients seen during the past decade. Median sternotomy was used in 18 of the 39 resections. Eleven of these 18 resections were limited to the trachea, and the remaining seven included the carina. Thus, median sternotomy was used in seven of our 13 carinal resections and is currently our incision of choice for the majority of such cases. Exposure of the carina and main bronchi through a median sternotomy requires a transpericardial approach. The anterior pericardium is divided vertically to permit circumferential mobilization of the ascending aortic arch, which is then retracted laterally and to the left. The superior vena cava is displaced laterally and to the right, and the right main pulmonary artery is exposed and displaced inferiorly. The posterior pericardium, which is displayed by these latter maneuvers, is then divided vertically and the entire mediastinal trachea and carina are clearly exposed and accessible. This technique of operative exposure has several distinct advantages over a right posterolateral thoracotomy in selected cases: 1. Any pulmonary resection is possible through a median sternotomy. Five of the patients having a combined tracheal and carinal resection required excision of the right main bronchus and right upper lobe.
Volume 88 Number 4 October. 1984
Two additional patients having a tracheal and carinal resection required a concomitant left pneumonectomy. 2. A cervical collar incision combined with a median sternotomy provides easy access for a superior laryngeal or suprahyoid release procedure when required. This incision also provides adequate exposure for an intrapericardial mobilization of the right pulmonary hilum, which may also be necessary to minimize tension on a tracheal or tracheobronchial anastomosis. Tension-reducing techniques. Mobilization and release procedures to reduce tension on the anastomosis were used in any case in which undue tension was anticipated. This decision was usually made during the operation after attempts to approximate the ends of the divided airway suggested that undue tension would result. The decision to use right hilar mobilization or a superior release procedure was not based on the length of airway resected, since the degree of tension on an anastomosis after circumferential resection of any given length of trachea varies considerably from person to person. On occasion, a satisfactory tension-free anastomosis was achieved without mobilization after resection of more than 5 em of the adult tracheal length. In this series of 39 resections, a superior laryngeal release'? was used in eight patients. During the past 10 years, we have favored the suprahyoid release procedure described by Montgomery" and have used it in three patients. Intrapericardial mobilization of the right pulmonary hilum was used to reduce anastomotic tension in 12 patients. In all patients, neck flexion was maintained for approximately 1 week after operation by the use of a heavy suture (No.5 gauge) running between the skin and subcutaneous tissue over the point of the chin to the skin overlying the upper sternum." Results Mortality and morbidity. There were no intraoperative deaths. Five postoperative deaths occurred in the 39 patients undergoing resection. There were only two postoperative deaths in the 33 patients in whom a primary reconstruction was done. In one of these two patients, death resulted after 40 days from an anastomotic dehiscence and subsequent secondary pneumonia. This patient with adenoid cystic carcinoma had the most extensive resection with primary reconstruction of any patient in the series: All but the upper two tracheal rings were resected along with the carina, and the right main bronchus was anastomosed to the cervical trachea. The left main bronchus could not be elevated for any method of anastomosis, and the left lung was removed. The anastomosis was obviously under
Primary neoplasms of trachea and carina 5 1 3
considerable tension and ultimately failed despite reinforcement with a vascularized pedicle of greater omentum. In retrospect, either a lesser and incomplete resection, leaving residual disease at the primary anastomosis, or prosthetic replacement should have been used. The second postoperative death in this group occurred at 4 days as a result of pulmonary embolism from a thrombus (due to operative trauma) in the innominate vein, which had not been divided. As in the previous case, this patient with an adenoid cystic carcinoma had required carinal and extensive tracheal resection that precluded reanastomosis of the left main bronchus, and the left lung was removed at the time of reconstruction. The postoperative mortality was forbidding in the six patients in whom a Marlex prosthesis was used. Three of the six patients died within 2 weeks of operation because of erosion of the innominate artery at the site where the artery overlies this relatively rigid prosthesis. Death was due to an abrupt and massive hemorrhage into the reconstructed airway. There were few serious, nonlethal complications. Most patients had transient difficulty in raising secretions during the early postoperative period, and although transient atelectasis and pneumonitis, occurred there was only one death from prieumonia secondary to the anastomotic dehiscence previously described. Aspiration, particularly with swallowed liquids, was observed in five of the eight patients in whom a superior laryngeal release was added and in one of the three patients having a suprahyoid release procedure. This complication resolved in every. patient within a period ranging from I week to a maximum of 3 months after operation. The most severe and prolonged problems with aspiration occurred in the elderly group. One recurrent laryngeal nerve. was deliberately resected in four patients because of the location and proximity of the primary tumor. In no instance, however, was there a permanent inadvertent injury to the recurrent laryngeal nerves. Only one of the 34 surviving patients has a functionally significant anastomotic stricture. In this patient, an adenoid cystic carcinoma was resected by removal of the carina and by a sleeve resection of the right upper lobe. The stricture occurred at the anastomosis between the distal trachea and bronchus intermedius and was satisfactorily controlled by interval endoscopic dilatation. Survival. Adenoid cystic carcinoma. Twenty-six of the 28 patients with adenoid cystic tumors were managed by resection. All five postoperative deaths in the series
The Journal of
5 14
Pearson, Todd. Cooper
occurred in this group. Twelve of the 21 patients who survived operation had a complete and potentially curative resection. The resection was incomplete, with proved residual tumor at the time of operation, in the remaining nine patients. All of the 21 survivors with adenoid cystic carcinoma received adjuvant radiotherapy before or after operation. For those patients in whom preoperative radiotherapy was electively used, the dosage was restricted to between 3,000 and 3,500 rads administered in 15 treatments during a 3 week period. Any patient who had an incomplete resection and who had received preoperative irradiation was referred for additional postoperative radiotherapy. During the past 10 years we have chosen to initiate this combined treatment approach with resection. Frozen section and formal histologic assessment by the pathologist are more accurate when the tissues have not been irradiated. There is no radiation damage that might prejudice healing, and a higher dosage (4,000 rads in twenty treatments during 4 weeks) can be administered safely within 1 to 2 months of operation. Of the 12 patients who survived operation and had a complete and potentially curative resection, nine are alive and clinically free of local recurrence at follow-up periods of from 1 to 20 years, with a mean survival of 8.3 years. Three of these 12 patients have died of unrelated disease between 6 and 18 years after operation, with a mean survival of 12 years. Two patients in this group have asymptomatic pulmonary metastases, and both are living and clinically well at 13 and 16 years. Nine of the 21 patients surviving operation had incomplete resections. Seven of these nine patients have died of recurrent tumor at intervals from 2 to 9 years after resection. Five of these seven deaths were due to a local recurrence of the tumor, two were due to cerebral metastases, and one was due to diffuse pulmonary metastases. One of the nine patients with an incomplete resection is alive and clinically free of recurrence. IRRADIATION ONLy-oPERATION PENDING. The two previously described patients with nonobstructive adenoid cystic carcinoma extending to the subglottic larynx were deliberately treated initially with radical local irradiation with the intention of preserving a normal voice for as long as possible. Both patients have had an excellent response to radiotherapy and are alive and clinically free of recurrent disease at 36 and 40 months, respectively. Their follow-up evaluation includes bronchoscopy at 6 month intervals. If and when evidence of locally recurrent tumor develops, each will require a resection with laryngectomy and a permanent suprasternal end tracheostomy.
Thoracic and Cardiovascular Surgery
SURVIVAL IN PATIENTS WITH METASTASES. Lymphatic spread was not observed in any patient with adenoid cystic carcinoma. Hematogenous metastases have occurred in 10 of the 28 patients: pulmonary in nine, brain in two, and bone in one. The two patients with metastatic brain tumor died of their metastases at 1 and 6.5 years, respectively. The natural history of pulmonary metastases is of important practical interest. In the nine patients with pulmonary involvement, the metastases were synchronous with the primary tumor in two instances and metachronous in seven. Metastases were multiple and bilateral in all cases. Seven of the nine patients with pulmonary metastases have died, but in only two of the nine were the pulmonary metastases responsible for the deaths. These two patients died of their pulmonary spread 68 and 108 months after operation. The causes of death in the other five patients were concomitant brain metastases in two instances (both had asymptomatic pulmonary spread), local recurrence in two instances (again both with asymptomatic pulmonary spread), and a postoperative complication in one instance. Two of the nine patients with pulmonary metastases are alive 13 and 16 years after operation. Neither patient is seriously disabled. One is asymptomatic and the other has a mild intermittent cough and poorly defmed chest discomfort. Squamous cell carcinoma. Six of the nine patients with squamous cell tumors were treated by resection. There were no operative deaths. Two of the six undergoing resection have died at 6 months (pancreatic metastases), and 46 months (local recurrence). Four patients are living and clinically free of recurrence 6, 16,21, and 56 months after operation. Three of the nine patients had unresectable squamous cell tumors and were managed by the introduction of a silicone T-tube stent (one T-tube and two T-Y-tubes). All three patients had received maximal doses of radiotherapy prior to this treatment. A satisfactory transient palliation was achieved in each patient, but all three died within 3 to 6 months of the introduction of the internal stent. Miscellaneous tumors. Three of the four patients with leiomyosarcoma are alive and clinically free of tumor 16, 63, and 125 months after operation. One patient died of a rapid and massive local recurrence 22 months after resection. The patient with a primary malignant melanoma of the trachea died of widespread hematogenous metastases after 14 months. The patient with a tracheal chondroma died of an unrelated squamous cell carcino-
Volume 88 Number 4 October, 1984
ma of the esophagus 18 months after tracheal resection, and the patient with a neurofibroma is alive and clinically disease free at 53 months. Subglottic resections. In four patients, the operation included a partial resection of the cricoid cartilage by a previously reported technique that preserves the recurrent laryngeal nerves when feasible." There was one case each of adenoid cystic carcinoma, squamous cell carcinoma, leiomyosarcoma, and neurofibroma. In only one of the four patients was it necessary to sacrifice one recurrent laryngeal nerve because of tumor involvement. The alternative to a subglottic resection in the three patients with malignant tumors would have been a total laryngectomy. They have been maintained on a regular follow-up program, and all four patients are alive without clinical evidence of recurrence: adenoid cystic carcinoma, 11 years; squamous cell carcinoma, 11 months; leiomyosarcoma, 10 months, neurofibroma, 53 months. Prosthetic reconstruction. Our initial experience with the experimental and subsequent clinical use of heavyduty Marlex mesh as a tracheal prosthesis has been reported previously.v" Three of these six patients died postoperatively of innominate artery erosion. A safe and satisfactory airway was restored in the three surviving patients. One of these three survivors died of locally recurrent tumor at 25 months. In one patient the prosthesis provided an excellent airway during the first 4.5 years after operation. Thereafter, a progressive accumulation of granulation and scar developed at the proximal anastomosis, and the prosthesis was removed 5.5 years after the initial resection. This second operation done in 1968 when it had become possible to reconstruct extensive tracheal defects by primary anastomosis by the currently available mobilization and release techniques. This patient remains alive and clinically free of recurrent tumor 21 years after the initial operation. In the third survivor, a local recurrence developed following an initial resection of the mediastinal trachea with reconstruction by primary anastomosis. At a second operation, it was necessary to resect the entire trachea including the larynx and the cervical esophagus. The esophageal defect was reconstructed by elevation of the stomach through the posterior mediastinum and pharyngogastrostomy. The innominate artery was resected and the mediastinal trachea was replaced with a cylinder of heavy-duty Marlex mesh, which was anastomosed distally at the level of the carina and secured proximally as a cutaneous stoma in the suprasternal notch. The patient died of brain metastases 68 months after the second operation.
Primary neoplasms of trachea and carina
5 15
Discussion Our experience with the surgical management of primary tumors of the trachea is similar to that reported by others. Most of the primary neoplasms are malignant, and adenoid cystic carcinoma and squamous cell carcinoma are by far the most common histologic types. Using the currently available techniques of mobilization and "release," the surgeon can safely resect long circumferential segments of the trachea and achieve primary reconstruction without the need of a prosthesis in the majority of operable patients. As experience with carinal resection has gradually increased, the innovations in techniques of reconstruction have advanced the capability to manage tumors at this level. A cervical collar incision with median sternotomy is now our incision of choice in most patients requiring a carinal resection. Worthwhile survival can be obtained in patients with adenoid cystic carcinoma, squamous cell carcinoma, and sarcomas of the trachea when a complete and potentially curative resection is possible. In the majority of patients with adenoid cystic carcinoma, even an incomplete resection may provide long periods of symptomfree palliation. Most adenoid cystic carcinomas are radiosensitive, and adjuvant radiotherapy is recommended in all resectable tumors of this cell type. Pulmonary metastases are relatively common in patients with adenoid cystic carcinoma and frequently progress very slowly. A majority of our patients with such metastases have remained asymptomatic for many years, and we do not consider synchronous pulmonary metastases to be inoperable if the primary tumor is resectable. Safe and reliable methods of prosthetic replacement remain elusive. Our limited experience with a semirigid porous prosthesis of heavy-duty Marlex mesh is clearly unsatisfactory. Innominate artery erosion from the prosthesis was a fatal complication in half of our patients. With such a limited number of cases, however, the three good results obtained with a Marlex prosthesis are anecdotal. Neville, Hamouda, and Anderson" are the only group that has reported a large clinical experience with a silicone rubber prosthesis, and no other authors have yet reported a significant number of similar favorable results.
REFERENCES Houston HE, Payne WS, Harrison EG Jr: Primary cancers of the trachea. Arch Surg 99:132, 1969
The Journal of Thoracic and Cardiovascular Surgery
516 Pearson. Todd. Cooper
2 Eschapasse H: Les tumeurs tracheales primitives traitement chirurgical. Rev Fr Mal Respir 2:245, 1974 3 Pearson FG, Thompson DW, Weissberg D, Simpson WJK, Kergin FG: Carcinoma of the trachea. Ann Thorac Surg 18:16, 1974 4 Grillo HC: tracheal tumors. Surgical management. Ann Thorac Surg 26: 112, 1978 5 Perelman M, Koroleva N: Surgery ofthe trachea. World J Surg 4:583, 1980 6 Grillo HC: Tracheal surgery. Scand J Thorac Cardiovasc Surg 17:67, 1983 7 Westaby S, Jackson J, Pearson FG: A bifurcated silicone rubber stent for relief of tracheobronchial obstruction. J THoRAc CARDIOVASC SURG 83:414, 1982 8 Grillo HC: Surgery of the trachea, Current Problems in Surgery, Chicago, 1970, Year Book Medical Publishers, Inc. 9 Theman TE, Kerr JH, Nelems JM, Pearson FG: Carinal resection. A report of two cases and a description of the anesthetic technique. J THORAC CARDIOVASC SURG71:314, 1976 10 Dedo HH, Fishman NH: Laryngeal release and sleeve resection for tracheal stenosis. Ann Otol Rhinol Laryngol 78:285, 1968 II Montgomery WW: Suprahyoid release for tracheal stenosis. Arch Otolaryngol 99:255, 1974 12 Pearson FG, Andrews MJ: Detection and management of tracheal stenosis following cuffed tube tracheostomy. Ann Thorac Surg 12:359, 1971 13 Pearson FG, Cooper JD, Nelems JM, Van Nostrand A WP: Primary tracheal anastomosis after resection of the cricoid cartilage with preservation of recurrent laryngeal nerves. J THoRAc CARDIOVASC SURG 70:806, 1975 14 Pearson FG, Henderson RD, Gross AE, Ginsberg RJ, Stone RM: The reconstruction of circumferential tracheal defects with a porous prosthesis. An experimental and clinical study using heavy Marlex mesh. J THoRAc CARDIOVASC SURG 55:605, 1968 15 Neville WE, Hamouda F, Anderson J: Replacement of the intrathoracic trachea and both stem bronchi with a molded Silastic prosthesis. J THORAC CARDIOVASC SURG 63:569, 1972
Discussion DR. HERMES C. GRILLO Boston. Mass.
We are grateful to Dr. Pearson and his colleagues for reporting this excellent series, since there have been few series of tracheal tumors reported from single institutions (or even collected series) treated by current techniques. When we separate the tumors into types, namely, adenoid cystic, squamous, and the others, and further consider the very long follow-up necessary for adenoid cystic carcinomas, it is difficult to determine the results precisely. However, the fact that the three major series reported (from Boston, Moscow, and Toronto) plus a fourth combined series (from Toulouse) show similar results gives reassurance that progress is being made.
At Massachusetts General Hospital, I have seen 122 primary tracheal tumors in about the same period of time. However, the distribution has been a little different in that 46 were squamous and 45 were adenoid cystic. The other 31 included six carcinoids and multiple other varieties. Sixty-two primary reconstructions were done. I strongly concur that the use of radiation for adenoid cystic carcinoma is critical. In some early cases in which irradiation was not used because margins were not diseased, recurrence appeared as long as 18 years later. With combined surgical resection and postoperative irr~di ation, survival has been quite good both for adenoid cystic carcinoma and, surprisingly, for squamous cell carcinoma. As Dr. Pearson pointed out, the presence or absence of microscopic tumor at resection margins or in lymphatics did not seem to be the critical factor. Obviously, more data are yet required. DR. LESTER BRY ANT Johnson City. Tenn.
Our own past experience with primary squamous carcinoma of the trachea has been sufficiently discouraging that, when presented with a 60-year-old man with a squamous carcinoma last year, we decided to try a new approach. The carcinoma was located on the right lateral and posterior walls of the trachea just above the carina. The chest radiograph showed enlargement of the left hilum consistent with lymph node metastases, and this was confirmed by computed tomographic scan. The patient's initial treatment consisted of external beam cobalt teletherapy to the primary tumor and lymph node drainage areas in the mediastinum and paraclavicular regions (total dose 3,000 rads). This was followed by bilateral staged thoracotomies during which we implanted a total of 97 seeds of radioactive iodine (iodine 125) in the wall ofthe trachea and in the adjacent mediastinal tissues. The seeds will provide 15,000 rads of gamma therapy to the target volume over a period of I year, but the contiguous organs will receive less than 2,000 rads at a distance of 2 to 5 em from the implants. It is now 16 months since the patient received external beam therapy and I year after his second thoracotomy for interstitial radioactive iodine therapy. He has no evidence of tumor recurrence, but he does have a significant radiation tracheitis which has produced an intractable cough. Biopsy specimens of the tracheal mucosa have demonstrated the typical dysplastic changes of radiation effect. It is difficult for anyone surgeon to gain much experience with primary squamous tumors of the trachea because of the rarity of this neoplasm. However, we are encouraged with the early results in this patient, and we are hopeful that other surgeons might try this approach in order to determine if the technique has merit. DR. W. CLARK HARGROVE Philadelphia. Pa.
We at the Presbyterian Hospital in Philadelphia agree that amenable lesions should be treated by surgical extirpation.
Volume 88 Number 4 October. 1984
However, our experience with this problem has been in the palliative management by laser therapy of a group of patients with obstructive lesions of the tracheobronchial tree. Over the past 18 months we have performed some 200 laser resections in 89 patients. Of these 89, 25 had primary squamous neoplasms of the trachea and carina. Surgical resection was contraindicated in all patients because of evidence of metastatic disease or poor general medical condition. All patients had respiratory distress on admission to the hospital. In most cases, it was moderate to severe. Of the 25 patients with tracheal obstructions, 20 (80%) had good results. We defined good results as relief of airway obstruction and ability to ambulate without supplemental oxygen. There were no hospital deaths. Each of the patients who did not respond to treatment died within 1 month of discharge from the hospital. The longest survival period has been 10 months after the patient's admission to the hospital with an endotracheal tube in place. Dr. Pearson, in reviewing your experience with squamous carcinomas of the trachea, do you think surgical resection should remain the primary form of therapy, or might the laser offer an alternative form of treatment in the majority of patients? Might the laser also have an adjunctive role in the management of adenoid cystic lesions? DR. EDWARD W. HUMPHREY Minneapolis. Minn.
I rise to ask if Dr. Pearson has observed a phenomenon that I have seen in patients undergoing an extensive resection of the carina including the lower trachea and both main bronchi with a consequent wide lymph node dissection in that area. In two patients treated in this manner, refractory ventricular arrhythmias developed and lasted for about 3 weeks after the operation. The problem then resolved and never recurred. These patients had no prior history of cardiac abnormalities. Since this phenomenon does not usually occur in transplanted hearts, I am uncertain whether these have been isolated instances or whether they might have resulted from the interruption of all the lymphatic drainage of an innervated heart. DR. WILLIAM E. NEVILLE Newark. N. J.
I certainly agree that a resection and pulmonary anastomosis is the ideal method of dealing with malignant tracheal tumors. However, is some patients a prosthesis must be used to reconstruct the airway. One big problem that has been alluded to is that no one surgeon with the exception of Dr. Grillo and the.group at Massachusetts General, the Mayo Clinic, Memorial, and now Toronto have a vast experience in the management of tracheal cancers, since they comprise only about 1% of all malignant tumors. I am going to confine my remarks to the prosthesis. Dr. Pearson did have good results in using a prosthesis, but he became disenchanted when three patients died of innominate artery hemorrhage and one patient developed a stenosis. At present, we are all aware of the possibility of arterial erosion whether a suture line or a prosthesis is behind the innominate
Primary neoplasms of trachea and carina 5 1 7
artery, and there are methods to prevent this. Also, with a nonporous tube, stenosis cannot occur. Theoretically, all of Dr. Pearson's complications are preventable at the present time. My second point is that excellent palliation can be achieved in nonreseetable lesions or recurrent cancer causing airway obstruction by using an intraluminal stent of the silicone rubber prosthesis. In one man with complete airway obstruction, we made a collar incision and were able to place two prostheses through and beyond the obstruction by "piggy-backing" them. A technique that is efficacious is to look down from above with a flexible scope and position the lower tube so that it does not obstruct the carina. Three years postoperatively, this man is doing well. In another patient with a right pneumonectomy and recurrent cancer, we used a median sternotomy. The trachea was opened, a straight prosthesis was inserted through the obstructed area, and a satisfactory airway was established. In the past 15 years, we have used the straight prosthesis for malignancy in eight patients; four are alive and well after 1 to 6 years. We have used the prosthesis as an intraluminal stent in three patients with cancer, and one is alive at 3 years. The bifurcated prosthesis has been used in 14 patients. There have been complications, but five of the patients are alive and well after 1 to 11 years. DR. QUENTIN R. STILES Los Angeles. Calif
Dr. Pearson, I am very interested in your use of jet ventilation. Did you use that because you found the ventilation was actually better or simply because it was less cumbersome than the two tubes? DR. PEARSON (Closing) We have used jet ventilation recently, Dr. Stiles, because it is simpler and less cumbersome than the standard techniques. Jet ventilation is delivered through a small-bore catheter which can be left in place in the distal airway during reconstruction and anastomosis. There is an added advantage in that a respiratory frequency of 120 cycles per minute causes minimal movement in the lungs and tracheobronchial tree. The "moving field" created by normal spontaneous respiration is abolished. Dr. Grillo has the largest reported experience in the world with primary neoplasms of the trachea. He has recently reported on a group of more than 100 patients with primary tracheal tumors. Squamous cell carcinoma was slightly more common than adenoid cystic carcinoma in his series. The adenoid cystic tumors, however, were more frequently operable than the squamous cell cancers. Dr. Bryant's report on the use of radioactive iodine and radioactive implant techniques is interesting-and novel in the management of tracheal tumors. I believe a median sternotomy might provide better access for this surgical procedure than bilateral thoracotomies. It is also important to emphasize that most adenoid cystic cancers will respond to conventional techniques of irradiation, and unresectable tumors of this kind
5 1 8 Pearson, Todd, Cooper
should still, 1 think, be treated initially with such standard radiation techniques. Dr. Hargrove has reported on an extensive experience with the use of a laser for tracheobronchial lesions. Twenty-five of lis patients had primary squamous tumors of the trachea, or trachea and carina. Our experience with the laser in such cases is very limited and has only recently been initiated. We would reserve such treatment for patients with inoperable tumors, for the symptomatic relief of upper airways obstruction. I do not anticipate, however, that the laser will completely replace the older techniques of endoscopic removal with biopsy forceps. In many cases, endoscopic resection with biopsy forceps will be as effective as laser resection (though not quite as neat), and it is much more rapidly achieved. I would still consider laser therapy for primary squamous cell carcinoma of the trachea as a palliative modality. In patients with a technically resectable squamous cell carcinoma, I believe there is no question but that operative removal is the best current therapy. Dr. Humphrey, we did not record any incidence of refractory ventricular arrhythmias in our patients who had a carinal
The Journal of Thoracic and Cardiovascular Surgery
resection. I have no explanation for your personal observations. Dr. Neville has reported the world's largest experience with prosthetic replacement and has achieved some excellent longterm results. I believe that some of his patients could have been managed by resection and primary anastomosis, and I still think this is the preferable mode of reconstruction when possible. All but one of our six prosthetic reconstructions (using heavy-duty Marlex mesh) were done during the 19608, at a time when extensive resection with primary reconstruction was at an early stage of development. One of our six patients was operated upon in 1973, and because of our unfavorable experience with innominate artery erosion, the innominate artery was deliberately resected in this patient at the time of the original operation. He survived for 6 years with this prosthesis before dying of cerebral metastases. If primary tracheal tumors were as common as coronary artery disease, I am certain we would have developed, by now, a very effective prosthesis for tracheal replacement. Unfortunately, it still remains an elusive goal.