Ann Thorac Surg 1995;59:173-177
© 1995 The Society of Thoracic Surgeons
Can Esophagectomy Cure Cancer of the Thoracic Esophagus Involving the Major Airways?
Toshiki Matsubara, MD,
Mamoru Ueda, MD,
Toshifusa Nakajima, MD,
Ken Nakagawa, MD,
Takashi Yamashita, MD,
Noboru Horikoshi, MD,
Akio Yanagisawa, MD
Departments of Surgery, Radiology, Medical Oncology, and Pathology, Cancer Institute Hospital, Tokyo, Japan
Accepted for publication August 18, 1994.
 |
Abstract
|
|---|
To evaluate the effects of aggressive operation for esophageal cancer invading the trachea and main bronchi, we investigated retrospectively 62 patients with proven tracheobronchial involvement who underwent thoracotomy for esophagectomy between 1973 and 1993. We operated unless the tumor was assessed to be definitely unresectable. Esophagectomy was possible in 55 patients, and the resectability rate was 95% after preoperative computed tomography and bronchoscopy became routine. After esophagectomy, no residual cancer lesion was recognizable macroscopically in 53% of patients. The hospital mortality rate in esophagectomy cases was 7% in the past 8 years. The outcome in patients who underwent curative resection was significantly favorable (p < 0.0001), and the 2-year survival was 51%. The patients with nonresectable cancer all died within 6 months compared with a 23% 1-year survival rate for palliative esophagectomy cases (p < 0.006). Among patients with tracheobronchial involvement assessed as resectable on computed tomography and bronchoscopy, a considerable proportion benefited from aggressive therapy with esophagectomy. The possibility of complete cure was high, especially when the cancer responded well to preoperative therapy and no lymph nodes were involved.
|
Introduction
|
|---|
Involvement of the trachea or main bronchi (the major airways) is evidence of markedly advanced disease in cancer of the thoracic esophagus. Many surgeons consider esophagectomy to be contraindicated in all such cases [1, 2]. However, we have tried to perform esophagectomy combined with adjuvant therapy when the disease was assessed to be technically resectable. The purpose of this study was to evaluate the effect of aggressive therapy for cancer of the thoracic esophagus with resectable tracheobronchial involvement.
|
Material and Methods
|
|---|
Patients
During the period from 1973 to 1993, thoracotomy with the intention of esophagectomy was undertaken in 536 patients with cancer of the thoracic esophagus. Esophagectomy was performed unless distant metastasis or technically unresectable involvement of adjacent organs was found at operation. Of 527 patients in whom esophagectomy was possible, 55 (10%) had histologically confirmed invasion to the trachea or main bronchi. Of the 9 patients who could not undergo esophagectomy, 8 had unresectable involvement of the mediastinal organs including the major airways. The total of 63 cases with tracheobronchial involvement were analyzed in this study. All had squamous cell carcinoma, except for one with adenosquamous carcinoma.
It was difficult to obtain a definite pathologic evidence of tracheobronchial involvement, unless the involved portion was extensively removed. In this study, the following patients undergoing esophagectomy were included, as cancer invasion was considered definite: (1) patients in whom macroscopically recognizable invading lesion was left at operation, and the operative margin of the resected specimen was positive on microscopic examination; and (2) those in whom the tracheobronchial tree was involved macroscopically and the specimen was excised just along the margin of the invading lesion. No residual cancer nests were found on macroscopic inspection, but the operative margin of the specimen was histologically positive.
In patients with effective preoperative therapy, granulomas caused by necrosis of cancer cells were considered as evidence of cancer invasion before treatment. When cancer cells or granulomas were found within 0.2 mm of the operative margin, the margin was defined as positive. Three patients with negative operative margins were excluded from the present analysis, although tracheobronchial invasion was strongly suspected on the basis of macroscopic findings at esophagectomy.
Preoperative Evaluation
We performed a thoracotomy unless the lesion was assessed as definitely unresectable on routine preoperative examinations or on inspection at the preceding laparotomy. Since 1981 we have routinely done preoperative computed tomography (CT) and flexible bronchoscopy for advanced cancer adjacent to the tracheobronchial tree. The preoperative assessment of tracheobronchial invasion was categorized into three grades: negative, probable, and positive (Fig 1
).
View larger version (128K):
[in this window]
[in a new window]
|
Fig 1. . Patient with positive findings on bronchoscopy. Invasion was recognized immediately beneath the mucosa of the membranous portion of the trachea. This patient underwent irradiation therapy followed by successful esophagectomy. No residual cancer cells were found in the specimen, and the patient was alive 8 years after operation.
|
|
Operation
In the present analysis, we defined esophagectomy as either curative (C) or palliative (P) according to the absence or presence of macroscopically recognizable residual invading cancer, regardless of microscopic findings. The numbers of C and P group patients were 29 and 26, respectively. The demographic factors of each group are shown in Table 1.
View this table:
[in this window]
[in a new window]
|
Table 1. . Demographic Factors in Patients With Invasion of the Trachea or Main Bronchi Who Underwent Thoracotomya
|
|
Esophagectomy was done through a right thoracotomy in 52 patients, and through a left thoracotomy in 3 patients. Mediastinal and upper perigastric lymph nodes were dissected systematically in all patients, excluding 11 in the P group. The cervical nodes were dissected in the most recent 17 patients, 4 in group P and 13 in group C. To avoid pulmonary complications, the bronchial arteries and the pulmonary branches of the vagal nerves were preserved when possible in procedures performed since 1987.
Whenever technically possible, we removed the involved portion of the tracheobronchial tree along with the esophagus. To prevent postoperative complication, we tried to resect a minimal amount of the major airway and avoid segmental resection of the lower trachea or main bronchus. But in 1 patient in group C, in whom the tumor had completely penetrated the mid-trachea for a length of 35 mm, we resected the proximal 65 mm of the trachea along with the laryngopharynx. In 8 group C patients, the wall of the trachea or left main bronchus was excised partially for almost the entire thickness (thick wall resection), and the defect in the wall was repaired with a pedunculated muscle flap: the latissimus dorsi muscle, 6; and the intercostal muscle, 2. The resected wall was 10 to 42 mm (20 mm on average) in length, and 7 to 25 mm (12 mm on average) in width. In 7 of these patients, who underwent operation recently, the mucosa of the trachea or main bronchus was preserved as cancer had not invaded the mucosa macroscopically (Fig 2). In the remaining 20 group C patients with shallow invasion, we excised only the external wall from the trachea or bronchus along the macroscopic margin of the invading cancer lesion, preserving a greater part of the cartilage and muscle layer of the trachea and main bronchi (thin wall resection); therefore, microscopically they all had cancer cells or granuloma caused by cancer necrosis in the margin of the specimen. In the P group, none underwent thick wall or segmental resection of the large airway.
View larger version (110K):
[in this window]
[in a new window]
|
Fig 2. . Patient undergoing thick wall resection of the trachea 42 x 15 mm in area. (A) The arrows indicate the resection margin. Except for the central portion, the mucosal layer was preserved. The defect was covered with the latissimus dorsi muscle. (B) Bronchoscopy 6 months after operation. Neither stenosis nor cancer recurrence was found.
|
|
Combined Therapy for Patients Undergoing Esophagectomy
The therapeutic strategy was selected according to risk factors and extent of disease. Preoperative adjuvant therapy was given to patients with good status when assessed as having tracheobronchial involvement; radiotherapy was given to 9 patients, chemotherapy to 2 patients, and both to 23 patients. The standard regimen varied with the decade: from 1973 to 1980, radiation therapy, a total of 40 to 50 Gy over 4 to 5 weeks, concomitantly with bleomycin, 5 mg intramuscularly at each radiation fraction; from 1981 to 1989, 50 Gy over a period of 5 weeks without chemotherapy; and since 1990, 40 Gy over 4 weeks concurrently with cisplatin 75 mg/m2 on day 1 followed by a continuous infusion of 5-fluorouracil 350 mg/m2 daily for 28 days. In 7 of the group C patients, no viable cancer cells remained in the surgical margin of the specimen. In 5 of these, no residual cancer cells were found in the resected specimen, but in 1 patient the tracheal wall was exposed directly in the ulcer floor. Postoperative adjuvant therapy was given to 18 patients: radiotherapy in 10, chemotherapy in 5, and both in 3 patients.
Statistics
Differences in frequency were evaluated by the 
2 test or Fisher's exact test. The survival curves were calculated by the Kaplan-Meier method. The difference in survival was evaluated by the generalized Wilcoxon method. All patients censored within 5 years were living at the follow-up time from March to July 1994.
|
Results
|
|---|
After preoperative CT and bronchoscopy became routine, 39 (95%) of 41 patients who underwent thoracotomy underwent esophagectomy. Among patients undergoing thoracotomy, the sensitivity of preoperative evaluation of tracheobronchial invasion on CT and bronchoscopy was not high (Table 2). Computed tomography was more sensitive than bronchoscopy. Bronchoscopic results were positive only in 10 of 20 CT-positive patients, whereas CT results were positive in 10 of 11 patients positive on bronchoscopy. In no patients except for one who underwent wide tracheal resection, cancer cells were bronchoscopically confirmed with biopsy.
View this table:
[in this window]
[in a new window]
|
Table 2. . Preoperative Assessment on Computed Tomography and Bronchoscopy in Patients With Tracheobronchial Involvement Operated on Between 1981 and 1993a
|
|
In patients who underwent esophagectomy, a total of 10 patients died during hospitalization; 4 due to cancer relapse, 3 due to mediastinitis, 2 due to pneumonia, and 1 as a result of hepatic failure. The overall hospital mortality rate in the past 8 years was 7% (2 of 28) compared with 30% (8 of 27) in the preceding period (p < 0.04). Stenosis of the major airways attributable to causes other than cancer recurrence was not observed in any case (Fig 2B
). Of 8 patients with unresectable disease, 4 died during hospitalization.
The 1- and 2-year survival after esophagectomy were 44% and 25%, respectively. The difference in survival between the C and P groups was significant (Fig 3; p < 0.0001). In the P group, 1-year survival was 23%, and no one survived more than 2 years. In the C group, the 1-, 2-, and 5-year survivals were 64%, 51%, and 19%, respectively. Among the group C patients who underwent thin wall resection, the patients without any residual viable cancer cells at the surgical margin after preoperative therapy had significantly better outcomes (Fig 4; p < 0.024). Patients who underwent thick wall resection or wide tracheal resection had more favorable results than group P patients (p < 0.03). The patient who underwent wide tracheal resection was alive at 2 years and 5 months after operation. The patients who could not undergo esophagectomy all died within 181 days (average, 82 days), which was significantly worse than the group P patients (Fig 3; p < 0.006). In 7 of 10 patients who survived more than 2 years, preoperative adjuvant therapy had been given.
View larger version (18K):
[in this window]
[in a new window]
|
Fig 4. . Cumulative survival curves in patients undergoing curative esophagectomy. (a, b) Thin wall resection with (b) or without (a) viable cancer cells remaining at the resection margin; (c) thick wall resection or wide tracheal resection.
|
|
Lymph node involvement was found in 18 patients (62%) in group C, the total number of positive nodes being 0 to 18 (average, 3.4). In group C patients without node involvement, 7 of 10 survived 2 years, compared with 2 of 9 of those with node involvement (p < 0.06). In particular, of 7 patients without any viable cancer cells at the resection margin, 5 had no involved nodes. Of these 5 patients, none died of cancer recurrence, and 4 survived more than 3 years after operation.
In group C, cancer recurrence was clinically recognized in 15 cases (Table 3). Local recurrence was found in 7 patients, 6 in the left main bronchus and 1 in the trachea. In all patients with local recurrence, marked cancer invasion to the extrabronchial tissues was also recognized on CT or plain roentgenography.
|
Comment
|
|---|
In the International Union Against Cancer TNM classification of esophageal cancer, involvement of the respiratory tract is categorized as T4 [3], and many surgeons have regarded that esophagectomy does not benefit such patients [2]. However, the outcomes in our series of patients with tracheobronchial involvement were not totally unfavorable when technically resectable. Although few showed long survival, the overall 1-year survival held some hope (44%). Recent advances in preoperative evaluation, adjuvant therapy, surgical technique, and postoperative care have decreased significantly the mortality after esophagectomy, although we have done more aggressive lymph node dissection in curative operation. The hospital mortality rate in recently treated cases with tracheobronchial involvement was acceptable (7%). These results suggested that our aggressive approach benefited a certain portion of these patients with such advanced cancer who would have died in a short time without aggressive therapies [2, 4, 5]. In particular, once an esophagobronchial fistula has developed, the prognosis is extremely poor [6].
Because the outcomes in unresectable cancer were always poor, it is essential to assess resectability accurately before treatment. The role of noninvasive diagnostic tests is still controversial [1, 7, 8]. Because the operative risk is now lower and tragic outcomes are probable without aggressive therapy, it is more important not to exclude patients with resectable cancer from operation than to exclude unresectable patients. For this reason, we performed thoracotomy unless the disease was evaluated to be definitely unresectable either technically or because of distant metastasis. Nevertheless, our overall resectability rate in those undergoing thoracotomy was 95% after we incorporated CT and bronchoscopy into the routine preoperative workup. Although the sensitivity of CT and bronchoscopy for assessing the presence of infiltration of the tracheobronchial tree was not as high, they still played a significant role in identifying technically unresectable patients, thereby avoiding unnecessary and ineffective thoracotomy.
Sleeve resection of the lower trachea or main bronchi is now commonly done without significant risk to obtain a sufficient free resection margin in lung cancer [9, 10]. Several authors also reported esophageal cancer cases in which this operation was done [11–13]. However, this extensive operation seems to be indicated in only certain limited cases of esophageal cancer for the following reasons. First, the risk of this operation concomitant with esophagectomy is considered to be high. This procedure further decreases the tracheobronchial blood circulation that has already been impaired after esophagectomy, because the tracheal and bronchial arteries run through and around the upper esophagus [14] and the principal regional lymph nodes of the esophagus [15]. Second, patients with wide involvement of the major airway usually also had more extensive involvement of the surrounding connective tissue, nerves, and neighboring organs, because the esophageal cancer invades the tracheobronchial wall from the outside compared with the invasion along the bronchial wall in bronchogenic cancer. In most of our patients, the extent of extrabronchial infiltration revealed by CT was markedly greater than suggested by the intrabronchial findings on bronchoscopy. According to the review by Kawahara and colleagues [16] on 18 Japanese patients with esophageal cancer undergoing sleeve resection of the tracheobronchial tree, 8 died within 1 month and none were alive at 1 year except for one who died at 18 months.
Because it was difficult to obtain a sufficient free operative margin and because hematogenic recurrence was as common as local recurrence, the adjuvant therapy was expected to yield significant results. Now routinely we provide preoperative radiochemotherapy to good risk patients when tracheobronchial invasion is strongly suspected. When residual cancer cells are not found microscopically in either the operative margin or lymph nodes after preoperative adjuvant therapy, complete cure is highly probable. It might be argued that 3 patients with no viable cancer cells found in the resected specimen might have had the same favorable results without esophagectomy. However, eradication of cancer cells cannot be confirmed except by resection. In addition, in these patients deep ulcer or stenosis had been still present at operation.
It is our contention that aggressive therapy involving esophagectomy is preferable when the tracheobronchial involvement is assessed as technically resectable. Preoperative assessment on CT and bronchoscopy played a significant role in selecting resectable patients. A considerable proportion of such patients benefit from aggressive treatment and possible cure may still be offered. In particular, when the cancer responds well to preoperative therapy and no lymph nodes are involved, complete cure is probable.
|
Acknowledgments
|
|---|
We are grateful to Professor J. Patrick Barron of the International Medical Communications Center of Tokyo Medical College for his review of the manuscript.
|
Footnotes
|
|---|
Address reprint requests to Dr Matsubara, Department of Surgery, Cancer Institute Hospital, 1-37-1 Kami-Ikebukuro, Toshima-Ku, Tokyo 170, Japan.
|
References
|
|---|
- Rice TW, Kirby TJ. The assessment of patients undergoing esophagectomy. Semin Thorac Cardiovasc Surg 1992;4: 263–9.[Medline]
- Müller JM, Erasmi H, Stelzner M, Zieren U, Pichlmaier H. Surgical therapy of oesophageal carcinoma. Br J Surg 1990;77:845–57.[Medline]
- International Union Against Cancer. TNM classification of malignant tumors. 4th ed. New York: Springer-Verlag, 1987: 40--2.
- Baulieux J, Barth X, Boulez J, Peix JL, Adeleine P, Maillet P. The advantage of palliative resection in squamous cell carcinoma of the esophagus. Int Surg 1985;70:197–204.[Medline]
- Bluett MK, Sawyers JL, Healy D. Esophageal carcinoma-improved quality of survival with resection. Am Surg 1987;53:126–32.[Medline]
- Gschossmann JM, Bonner JA, Foote RL, Shaw EG, Martenson JA, Su J. Malignant tracheoesophageal fistula in patients with esophageal cancer. Cancer 1993;72:1513–21.[Medline]
- Choi TK, Siu KF, Lam KH, Wong J. Bronchoscopy and carcinoma of the esophagus. II-carcinoma of the esophagus with tracheo-bronceal involvement. Am J Surg 1984;147:760–2.[Medline]
- Lehr L, Rupp N, Siewert JR. Assessment of resectability of esophageal cancer by computed tomography and magnetic resonance imaging. Surgery 1988;103:344–50.[Medline]
- Barclay RS, McSwan N, Welsh TM. Tracheal reconstruction without the use of grafts. Thorax 1957;12:177–80.
- Grillo HC, Bendixen HH, Gephart T. Resection of the carina and lower trachea. Ann Surg 1963;158:889–93.
- Thompson DT. Lower tracheal and carinal resection associated with subtotal oesophagectomy for carcinoma of oesophagus involving trachea. Thorax 1973;28:257–60.[Abstract/Free Full Text]
- Tomita M, Ayabe H, Kawahara K, et al. Surgical consideration of oesophagectomy combined with tracheobronchoplastic procedure for treatment of esophageal cancer. Acta Med Nagasaki 1986;31:143–51.
- Ikeda T, Sakai T, Sakai S, Kaseda S, Obitsu R, Iwatsuka M. Resection of the carina and oesophagus for malignant tumours of the oesophagus or tracheo-bronchial tree. Thorax 1984;39:201–5.[Abstract/Free Full Text]
- Salassa JR, Pearson BW, Payne WS. Gross and microscopic blood supply of the trachea. Ann Thoracic Surg 1977;24: 100–7.[Abstract]
- Matsubara T, Ueda M, Yanagida O, et al. How extensive should lymph node dissection be for cancer of the thoracic esophagus? J Thorac Cardiovasc Surg 1994;107:1073–8.[Abstract/Free Full Text]
- Kawahara H, Shiraishi T, Yoshida Y, et al. Surgical treatment for esophageal cancer invading the adjacent organs. Shujutsu (Operation) 1992;46:865–978.
Top
Abstract
Introduction
