J
THoRAc CARDIOVASC SURG
84:489-496, 1982
Survival in patients undergoing tracheal sleeve pneumonectomy for bronchogenic carcinoma Thirty-four patients underwent tracheal sleeve pneumonectomy during the years 1964 through 1981. In 30 patients. simultaneous resection of the right lung and carina was performed with an anastomosis established between the left main bronchus and trachea. In four patients a previous pneumonectomy had been performed (two right and tlI'o left) and. because of recurrence in the stump. resection of the stump and carina was carried out with either right or left bronchotracheal anastomosis. Preoperative irradiation was given in 28 of the patients in the group, and tumor sterilization occurred in six of those so treated. Four of 30 patients are long-term survivors: Two are still alive and free of cancer at 7 and more than 11 years postoperatively. One patient in the group offour with stump recurrence is alive more than 11 years postoperatively. All long-term survivors had epidermoid carcinoma and all received preoperative irradiation. The most frequent complications were fistula formation and pneumonia, resulting in 10 deaths in the perioperative period for a mortality of 29%. The 5 year survival rate of the entire group was 15%, and the survival rate by life-table analysis in the 30 patients with lung and carinal resection was 13.3% at 5 and 8.8% at 10 years.
Robert J. Jensik, M.D., L. Penfield Faber, M.D., C. Frederick Kittle, M.D., Chicago, Ill., Robert W. Miley, M.D. (by invitation), Fort Worth, Texas, W. Craig Thatcher, M.D. (by invitation), and Nabil EI-Baz, M.D. (by invitation), Chicago, Ill.
o
uri initial series of patients undergoing tracheal sleeve pneumonectomy was reported in 1972. The subsequent 10 year period has provided follow-up time for this group as well as the opportunity to perform an additional 17 procedures, for a current total of 34 patients operated upon during the years from 1964 through 1981. This procedure represents an extended type of resection, a term suggested by Chamberlain and associates.! It is to be considered in a patient in whom the tumor is centrally located at the hilus of the lung with extension to involve the orifice of the main bronchus or the lateral aspect of the lower trachea. In 30 patients the right lung was resected along with the lower trachea and carina, and the left main bron-
From the Department of Cardiovascular Thoracic Surgery, RushPresbyterian-St. Luke's Medical Center, Chicago, Ill., and the Department of Surgery, Abraham Lincoln Medical School of the University of Illinois, Chicago, Ill. 60612. Read at the Sixty-second Annual Meeting of The American Association for Thoracic Surgery, Phoenix, Ariz., May 3-5, 1982. Address for reprints; Robert J. Jensik , M.D., 1725 W. Harrison St., Chicago, Ill. 60612. 0022-5223/82/100489+08$00.80/0
© 1982 The
C. V. Mosby Co.
chus was anastomosed to the trachea. Four patients had undergone a previous pneumonectomy (two right and two left) and, because of recurrence of tumor in the bronchial stump, the stump and carina were resected and either the right or left main bronchus was anastomosed to the trachea. Preoperative radiation therapy was administered in 28 of the patients in the study. It was considered to be a major factor in reducing the extent and volume of the tumor to facilitate the procedure.
Patients and methods The sex and age range are shown in Table I. There were 26 male and eight female patients with ages ranging from 45 to 69 years. The various histologic types of cancer are shown in Table II. Also listed is the number of cases of each cell type in which the tumor was sterilized by preoperative radiotherapy, as the resected specimen in six patients did not reveal identifiable cancer cells. The procedure done most frequently in this group is shown in Fig. 1, A. At thoracotomy the azygos vein is ligated posteriorly and anteriorly at its entrance to the vena cava and excised to give complete exposure to the 489
The Journal of Thoracic and Cardiovascular Surgery
490 Jensik et al.
Table II. Histology
Table I. Sex and age Age (yr) Sex
40-49
Male Female
8 2
I
50-59
10
4
I
60-69
Totals
8 2
26 8
tracheobronchial angle. After ligation and division of the pulmonary artery and pulmonary veins, usually within the pericardial sac, the lower trachea and left main bronchus are mobilized and freed from surrounding tissues. Both structures are divided, and frozen tissue studies of the margins are made to be certain they are free of tumor. After the lung is removed, anastomosis is then performed between the trachea and left main bronchus. Anesthesia is maintained with the endotracheal tube advanced into the left main bronchus. Anesthesia has been modified in our last four patients by means of a small catheter with high-frequency ventilation." This method permits greater accuracy in placing and tying sutures and eliminates intermittent withdrawal and reinsertion of the endotracheal tube into the left main bronchus. Twenty-five of the 30 patients in this group received radiation therapy administered either by cobalt or linear accelerator sources. The dose varied from 3,200 to 5,000 rads, but most individuals received 4,000 rads administered over a 4 week period. During the past 10 years all patients underwent preoperative radionuclide scanning prior to the procedure to rule out distant metastases. Fig. 2, A shows a radiograph typical of the patients in this group. The tumor is located at the hilus of the lung, and in the tomograms the tumor mass and narrowing of the distal trachea are clearly demonstrated (Fig. 3, A). The patient, a 59-year-old man, had a cough and hemoptysis. Bronchoscopic examination revealed the narrowing of the distal trachea above the carina with proximal extension of the tumor from the upper lobe to the orifice of the stem bronchus. Biopsy specimens were interpreted as epidermoid carcinoma. Preoperative radiation was administered over a 4 week period with the patient receiving 4,000 rads. Although this was the total dose, the radiograph prior to operation demonstrated pronounced fibrosis along the mediastinum (Fig. 2, B). However, in the tomogram, there was significant reduction in the size of the tumor mass and a more normal tracheal lumen (Fig. 3, B). At thoracotomy, residual cancer was identified at the tracheobronchial angle and prevented division of the main bronchus without cutting through tumor. Tracheal
Radiation sterilization
Cell type
No.
Epidermoid Adenocarcinoma Mixed Small cell
23 2 3 6
4
Totals
34
6
2
sleeve pneumonectomy was performed after intrapericardial division ofthe right pulmonary artery and veins. In the tomogram 2 months postoperatively, the patent tracheobronchial airway can be seen without evidence of stenosis (Fig. 3, C). Two patients underwent stump and carinal resection for recurrent tumor at intervals of 5 months and 2 years, 7 months, respectively, following left pneumonectomy (Fig. 1, B). Both operations were performed through a right thoracotomy with partial cardiopulmonary bypass. The second patient was treated with preoperative irradiation and received 6,000 rads. Two patients had stump and carinal resection for recurrent cancer at intervals of 18 months and 3 years, 2 months, respectively, after right pneumonectomy (Fig. 1, C). Fig. 4 shows a radiograph typical of the patients in this group, in this case, taken 11 years, 5 months postoperatively. The preoperative study was essentially the same. The patient had undergone right pneumonectomy at another institution in March of 1969, and tumor cells were identified at the bronchial margin of the resected specimen. Not until 10 months later was a positive biopsy specimen obtained from the stump on bronchoscopy. The patient was treated with linear accelerator sources over the next 2 months and received a total dose of 5,000 rads. Stump and carinal resection was performed 6 months later in September of 1970, with the intent of completely controlling the lung cancer. The other patient with right stump recurrence had stump and carinal resection 3 years, 2 months following right pneumonectomy. The involvement extended superiorly approximately 2 em above the stump. Two courses of radiation were administered, 4,000 rads initially and, because of a poor response as demonstrated by repeat bronchoscopy, an additional 2,200 rads. Six months later resection was done under extremely difficult circumstances because of the reduction in size of the right pleural cavity. A Neville tracheal prosthesis! was used for replacement since the extent of the tracheal resection was estimated to be from 4 to 4.5 em.
Volume84 Number4 OCtober, 1982
Tracheal sleeve pneumonectomy
49 I
Fig. 1. Types of resections. A, Right tracheal sleeve pneumonectomy. B, Left bronchial stump and carinal resection. C, Right bronchial stump and carinal resection.
Results There were four long-term survivors in the group of 30 patients undergoing right tracheal sleeve pneumonectomy. The first patient in the series lived 9 years, 10 months, and death was due to metastatic carcinoma to the liver, although it was never determined whether another primary cancer had developed as a source for the hepatic spread. The other long-term survivor who died remained free of carcinoma and died 12 years, I month postoperatively of cardiorespiratory failure. He was of unusual interest because at thoracotomy, done elsewhere, the lesion was considered inoperable owing to extensive involvement at the hilus of the lung, and the chest was closed. Following completion of cobalt irradiation, a large cavity developed at the site of the tumor and the patient had considerable bloody, purulent expectoration. Tracheal sleeve pneumonectomy was done to palliate symptoms in June of 1969. The patient died in July of 1981 at 73 years of age. The two additional long-term survivors are free of recurrent cancer 12 years, 4 months and 7 years, respectively, since their operations (Fig. 2). Three patients lived between 2 and 3'/2 years, and death was attributed to metastatic disease to the skeletal system or the brain. One patient with brain metastasis underwent two craniotomy procedures in the remaining year of his life. The three patients who lived between I and 2 years also died of metastasis of the lung cancer to bone, liver, or kidney. There was only one instance of
recurrence of tumor at the anastomotic site in the entire series. The cell type in all of the long-term survivors was epidermoid carcinoma, and all survivors had been treated with preoperative irradiation. Specimens removed from two patients revealed no evidence of viable tumor. There was one long-term survivor in the group of four patients with stump recurrence (Fig. 4). This patient had received preoperative irradiation with resulting tumor sterilization. Two of the three remaining patients who had undergone carina and stump resection died in the perioperative period, one of pneumonia and the other of bronchopulmonary fistula. Both patients had recurrence in the left bronchial stump. The other patient with right stump recurrence died shortly after readmission to the hospital of bronchopneumonia and hemorrhage resulting from erosion of the right atrium by the tracheal prosthesis, which had become separated from both the tracheal and bronchial margins. The survival time of all 34 patients is shown in Fig. 5. The major complications are listed in Table III and were the cause of perioperative death in 10 patients. The most frequent complication was fistula formation, which occurred in seven patients. In six the communication was with the pleural cavity through the anastomosis and in the other, with the pulmonary ar-
492 Jensik et al.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 2. Case I. Carcinoma at hilus . right lung. A. Prior to radiation . B. Postradiation fibrosis. C. Film taken 6 years, II months postoperatively.
tery. The last patient and four with pleural communication died in the perioperative period . It was possible to save one patient by reoperation and closure of the fistula 2 weeks postoperatively with a pedicle graft of intercostal muscle. However, this patient died of pulmonary embolism 2 months after release from the hospital. Fistula formation in the remaining patient was associated with the prosthesis. All but one of the patients who had a fistula had received preoperative irradiation. The one exception was a patient whose anastomosis was difficult and could never be made airtight. The prolonged ventila-
tory support probably contributed to the increase in size of the fistula, and the patient died of pleuropulmonary infection and cardiac failure. Pneumonia developed without evidence of bronchopleural fistula in the four patients listed as having this complication. The infection developed shortly after the operation and progressed rapidly. All of the patients died despite vigorous ventilatory support and antibiotic therapy . In the one patient listed as having a gastrointestinal hemorrhage, the problem developed probably as a result of a stress ulcer. He also had a fistula but died of
Volume 84 Number 4
Tracheal sleeve pneumonectomy
49 3
OCtober, 1982
Fig. 3. Case I. Tomograms. A. Prior to radiation; tracheobronchial narrowing. B. After radiation: normal airway. C. After operation: tapering tracheobronchial airway. massive bleeding. Hepatitis occurred in one patient, but he was discharged with the problem apparently under control. He died within 6 months, however, probably of an exacerbation of hepatitis. The death resulting from a perforated ulcer was the result of an error in diagnosis . An extremely high amylase determination prompted a course of management for acute pancreatitis. Possibly this death might have been averted. Perioperative deaths are listed in Table IV. The eight deaths in the 30 patients undergoing tracheal sleeve pneumonectomy are at least double the number for standard pneumonectomy. The higher number of deaths among the group having bronchial stump and carinal resections is a probable reflection of the greater technical difficulties associated with the procedure because of the interval between operations. Discussion Procedures of this type are examples of extended resections. One of the earliest reports of a tracheal sleeve pneumonectomy, with a survival time of 6 months, was mentioned by Gibborr' in his discussion of a report by Chamberlain and associates-: "Bronchogenic Carcinoma-An Aggressive Surgical Attitude. " A similar example was described by Bjork and Rodriguez" in their experimental study of tracheal and carinal reconstruction. They gave no survival time but did mention subsequent stricture formation.
2-18-82 Fig. 4. Case 2. Right stump and carinal resection II years, 5 months postoperatively. Isolated examples were described by Thompson," Correll (personal communication), Nelson," and Ferrato," and Ferrato reported a 6 year survival period . One of the two patients mentioned by Mathey' ? survived 4V2 years .
The Journal of Thoracic and Cardiovascular Surgery
494 Jensik et al.
YEARS SURVIVAL TO JAN. -1982 CASES
3
2
4
5
6
7
8
'64
9
10
12 13
II
9 yr. 10mo
ALIVE
DEAD
0
65
0
3
'66
0
3
4
'67
0
4
'68
0
12 yr I mo
'69 II yr. 3mo. - - II yr. 4 mo.
0
~
6
'70
5
'71
3
'72~
~
D
'73 0
'74
2
'75
DEATHS SURVIVORS
0
5
0
3
0
0
7 yr 0 0
TRACHEAL -SLEEVE PNEUMONECTOMY /964 TO ////82
0
0
0
0
0
'80
0
3
'81
0
0
TOTAL
4
0
'76
3
2
34
TOTALS
3
31
Fig. 5. Bar graph: Survival (34 patients).
Table III. Complications Complication Fistula Bronchopleural Bronchoarterial Pneumonia Gastrointestinal bleeding Hepatitis Perforated ulcer Total
No. 7 6 I 4 I I I
14
Grillo I \, 12 has the largest experience with tracheal and carinal resections, although only four cases of pneumonectomy are listed in his recent report. 13 There were no perioperative deaths, one patient died at 2 years, and the others are alive from 5 months to 1 year, 9 months postoperatively. He has expressed concern about the curability of pulmonary lesions extending to the trachea and approaches these situations conservatively, stressing the importance of mediastinoscopy to
assess nodal involvement and a careful search to exclude distant metastases. Mediastinoscopy has been used in two of our last three patients, but the procedures were done prior to the patient's admission to Rush-Presbyterian-St. Luke's Medical Center. Mediastinoscopy is not done for routine assessment since it is accepted that nodal involvement will be present adjacent to the tumor and, if present at a higher level, all nodes as well as the tumor are in the field of irradiation. Nodal disease was present in the six patients not receiving preoperative irradiation, but the nodes were removed with the specimen. Those patients who received radiation therapy seldom had diseased nodes present. Perelman and Koroliva'" described 6 cases but did not mention survival times. Deslauriers and colleagues" reported a series of 16 patients in 1979, with one of eight survivors alive at 7 years, 4 months and four alive at 2 years. According to a personal communication, their present total is 26 patients with a survival rate of 27% at 3 years. Nine patients are alive and free of cancer, one at 9 years, two
Volume 84
Tracheal sleeve pneumonectomy
Number 4 October. 1982
495
1.0 0.9 0.8
>to-
0.7
~
m < m 0
a: e, ...J
< s>a:
0.6 0.5 0.4
::J
en
If.
0.3
I
0.2
SYR-n.3':r,
I
0.1
IOYR-8.8%
0.0 0
20
40
60
80
100
120
140
160
180
200
MONTHS
Fig. 6. Life-table analysis. Survival probability after tracheal sleeve pneumonectomy, 1964 to January, 1982 (N = 30).
at 5 years, and several at 3 years. The perioperative mortality has decreased from 50% in the first 10 patients to 12% in the last 16. Somewhat disappointing is the greater increase in mortality in the second group of 17 patients in our study, the major causative factor being related to the formation of fistula. In an effort to reduce this complication, we have changed suture material to the polyglycolic acid types. Although one fistula did develop after this change was made, the size did not increase and repair was possible. It is difficult to incriminate the preoperative radiation as a causative factor even though it was given to three of the four patients in the second group who died postoperatively as a result of this complication. Of the first 17 patients, 15 were treated preoperatively with radiation and only one patient had a fistula. However, the radiation dose of more than 6,000 rads in this individual may have been excessive. The benefits resulting from reduction in volume of tumor and the improvement in the appearance of the lesion on bronchoscopic examination would still justify its use in our opinion. In addition to the survival results of all patients shown in the bar graph (Fig. 5), the probability of survival by life-table analysis of the 30 patients with
Table IV. Perioperative deaths
Deaths
Patients
~
- - _..... Tracheal sleeve pneumonectomy Resection of bronchial stump and carina
30 4
Totals
34
No.
%
8 2
26 50
10
29
tracheal sleeve pneumonectomy was calculated to be 13.3% at 5 years and 8.8% at 10 years (Fig. 6). Excluding the eight patients who died in the perioperative period increases the figures to 18% and 10%, respectively. The results achieved were better in the first 17 patients operated upon in a 7 year period from 1964 to 1971. Although a similar number of patients constitute the second group of patients, the time frame from 1971 through 1981 is an 11 year period and accordingly may suggest that fewer patients will be seen with such advanced disease. This may be a result of greater awareness of the problem of lung cancer and more vigorous surveillance in an effort to detect cases early. Despite an anticipated decline, however, patients will still appear having advanced disease with exten-
The Journal of
496 Jensik et al.
sion to the trachea. These individuals will pose a challenge to the surgeon, who will need an aggressive surgical philosophy since complications are frequent, perioperative mortality is high, and the number of long-term survivors is fewer than with standard resections. Since alternative measures of treatment offer so little and since we have achieved a 15% 5 year survival rate in the 34 patients treated by this extended effort, we believe tracheal sleeve pneumonectomy is justified. We wish to express our appreciation to Ms. Meryl Gale, Research Data Analyst, Department of Preventive Medicine, Section of Biostatistics and Epidemiology, for her valuable assistance in statistical analysis.
2
3
4
5 6
7
8 9 10
II
12 13 14 15
REFERENCES lensik Rl, Faber LP, Milloy Fl, Goldin MD: Tracheal sleeve pneumonectomy for advanced carcinoma of the lung. Surg Gynecol Obstet 134: 231, 1972 Chamberlain 1M, McNeil TM, Parnassa P, Edsall MB: Bronchogenic carcinoma. An aggressive surgical attitude. 1 THoRAc CARDIOVASC SURG 38:727, 1959 EI-Baz N. Faro RS, Ivankovich AD, lensik Rl, Faber LP, Kittle CF: One-lung high-frequency ventilation for tracheoplasty and bronchoplasty. A new technique. Ann Thorac Surg (in press) Neville WE, Bolanowski P1P, Soltanzadeh H: Prosthetic reconstruction of the trachea and carina. 1 THoRAc CARDIOVASC SURG 72:525, 1976 Gibbon lH: Discussion of Chamberlain et aF Bjork YO, Rodriguez LE: Reconstruction of the trachea and bifurcation. 1 THoRAc CARDIOVASC SURG 35:596, 1958 Thompson DT: Tracheal resection with left lung anastomosis following right pneumonectomy. Thorax 21:560, 1966 Nelson KG: Discussion of Deslauriers et al':' Ferrato Pl: Discussion of Deslauriers et al':' Mathey 1, Binet lP, Galey 11, Evrard C, Lemoine G, Denis B: Tracheal and tracheobronchial resections. Technique and results in 20 cases. 1 THoRAc CARDIOVASC SURG 51: I, 1966 Grillo HC, Bendixen HH, Gephart T: Resection of the carina and lower trachea. Ann Surg 158:889, 1963 Grillo HC: Tracheal tumors. Surgical management. Ann Thorac Surg 26: 112, 1978 Grillo HC: 1. Maxwell Chamberlain Memorial Paper. Carinal reconstruction. Ann Thorac Surg (in press) Perelman M, Koroleva N: Surgery of the trachea. World 1 Surg 4:583, 1980 Deslauriers 1, Beaulieu M, Benazera A, McClish A: Sleeve pneumonectomy for bronchogenic carcinoma. Ann Thorac Surg 28:465, 1979
Thoracic and Cardiovascular Surgery
Discussion DR. lEAN DESLAURIERS Ste-Foy, Quebec. Canada
Our experience with this procedure correlates with the one presented today. From 1968 to 1982, we have done 26 right sleeve pneumonectomies for bronchogenic carcinoma. The overall operative mortality is 27% (7/26); in the past 5 years, however, it has decreased to 12.5% (2/16). We think that the improvement is related to better preoperative staging and more aggressive postoperative respiratory care. Mediastinoscopy should be included in the preoperative staging; in our opinion, neoplastic involvement of superior mediastinal modes is a contraindication to this operation. In our series, the absolute I year survival rate (including operative deaths) is 46% (12/26), the absolute 3 year survival rate 22% (4/18), and the absolute 5 year survival rate 23% (3/13). My questions for the authors relate to the importance of mediastinoscopy and the possible relationship between preoperative radiation and the incidence of postoperative fistulas. DR. 1 ENS I K (Closing) The results reported by Dr. Deslauriers are similar to ours but differ in one respect: (two deaths in 17 patients) the perioperative mortality was lower in the first half of our study and substantially higher in the second half, opposite to the results of Dr. Deslaurier. It almost appears as if their group is improving and we are getting poorer results. This may be due to our enthusiasm over initial results, prompting intervention in more difficult cases. About half of our patients were operated upon before 1972, when staging criteria had not yet been established. In our second group of patients, only two had preoperative mediastinoscopy for staging prior to their admission to Rush-Presbyterian- St. Luke's Medical Center. Since we resort to preoperative irradiation so frequently, I believe that if nodes with metastases are present in the mediastinum, the radiotherapy will control these areas of spread. When involved lymph nodes were encountered in those patients who did not receive prior radiation, it was possible to remove them at the time of the procedure. These patients were Stage III candidates. It would appear that the irradiation was responsible for the greater incidence of fistula formation in the second group of patients; yet, in the first group, only one of a similar number of patients had this complication. It is difficult for me to explain the difference. In an effort to reduce the number of fistulas developing, we have changed suture material from silk and synthetic varieties to polyglycolic acid sutures. It does appear that we are seeing fewer patients, as the bar graph demonstrates, with advanced cancer involving the tracheal wall or carina. This may be a result of greater awareness of the problem of lung cancer and more vigorous efforts of surveillance for early detection.