- Email: [email protected]
Survival with residual tumor on the bronchial margin after resection for bronchogenic carcinoma
Survival with residual tumor on the bronchial margin after resection for bronchogenic carcinoma Sixty-four (/4.7 percent) of 434 consecutive patients having pulmonary resection for bronchogenic carcinoma were found to have microscopic residual tumor on the cut margins of the resected specimens. These subjects were further subdivided histologically into those with direct extension of the tumor (34 patients). lymphatic permeation (/4 patients). clumps of cancer cells in parabronchial tissues (six patients). and the presence of carcinoma in situ change (IO patients). Bronchopleural fistulas developed in eight (/2.5 percent) of 64 patients. The operative mortality rate was /5.6 percent. with four of the deaths occurring as the result of bronchopleural fistulas. Thirty-two patients (50 percent) survived / year. 2/ (32.8 percent) survived 3 years, and /5 (23.4 percent) livedfor 5 years or more. The patients with tumor in the submucosal and peribronchial lymphatics had the worst prognosis. 78.6 percent having died within / year and the remainder within 3 years. All 5-year survivors were men with squamous cell carcinoma and had relatively small tumors (mean diameter 2.9 cm.). No direct relationship between the length of the resected bronchial stump and survival could be established; a short stump did not preclude long survival. The possible factors involved in the relatively high 5 year survival rate in this group of patients and the therapeutic implications of these factors are discussed.
A. S. Soorae, M.B., F.R.C.S.(Ed.), F.R.C.S.(Eng.), and H. M. Stevenson, M.B., F.R.C.S.(Ed.), Belfast, Northern Ireland
T
he proximal spread of bronchogenic carcinoma was studied by early workers like Griess and associates' and Cotton," but they made no attempt to correlate this spread with survival in their patients. More recently, favorable reports of long-term survival of patients with microscopic residual carcinoma at the bronchial margin have appeared,": 4 and it is becoming apparent that these patients have a better survival rate than those having an incomplete resection, in whom gross tumor is left behind elsewhere in the hemithorax." This study was designed to determine the incidence and pattern of residual carcinoma at the bronchial margin after pulmonary resection and to analyze the factors that may have contributed to the 5 year survival. From the Thoracic Surgical Unit, Royal Victoria Hospital, Grosvenor Road, Belfast BTl 2 6BA, Northern Ireland. Received for publication Oct. 4. 1978. Accepted for publication Feb. 19, 1979. Address for reprints: A. S. Soorae, M.B. F.R.C.S.(Ed.) F.R.C.S. (Eng.), Consultant Cardiothoracic Surgeon, Fazakerley Hospital, Longmoor Lane. Liverpool L9 7AL, England.
Patients and methods A total of 556 consecutive patients underwent thoracotomy for bronchogenic carcinoma between January, 1968, and December, 1972, and the results were analyzed retrospectively in June, 1978. This period for study was chosen to obtain a minimum follow-up of 5 years. The carcinoma was found to be unresectable in 122 patients (100 men and 22 women), for a resectability rate of 78.1 percent. Of the 434 patients who had pulmonary resection, 374 were men and 60 women, for a male-to-female ratio of 6.2: I. This group of patients will be referred to as the main series. Sixty-four patients (57 men and seven women) were found to have residual microscopic tumor at the line of section. These patients form the basis of this study, and they will be referred to as the residual tumor group. Ages ranged from 30 to 77 years in the main series, with a mean over-all age of 60'12 years in the main series and 59 years in the residual tumor group (Table I). The numbers and types of resection are shown in Table II. The distribution of patients in each decade and the numbers and types of
0022-5223/79(080175+06$00.60(0 © 1979 The C. V. Mosby Co.
175
The Journal of
1 76 Soorae and Stevenson
Thoracic and Cardiovascular Surgery
Table I. Age distribution by decades Residual tumor group
Main series Age by decades (yr.)
No. of patients
30-39 40-49 50-59 60-69 70-79
10 64 136 185 39
Totals
434
Mean age
I Percent 2.3 14.8 31.3 42.6 9.0 100
No. of patients
I
Percent
2 13 12 33 4
3.1 20.3 18.7 51.6 6.2 100
64
60.5 yr.
59 yr.
Table II. Analysis of numbers and types of resection Main series (n = 434) Operation
Pneumonectomy Lobectomy Sleeve resection Wedge resection Segmental resection
I
No. of patients
210 195 17 8 4
Residual tumor group (n = 64)
Percent
No. of patients
48.4 45.0 3.9 1.8 0.9
28 27 8 0 1
I
Percent
45.3 42.2 12.5 0 1.5
Table III. Analysis of histology Main series (n = 434)
Cell type
Squamous cell Anaplastic Adenocarcinoma Oat cell
No. of patients
252 84 53 45
I
Residual tumor group (n = 64)
Percent
No. of patients
58.0 19.4 12.2 10.4
41 9 8 6
I Percent 64 14 12.5 9.5
resection were similar in the two groups of patients, with the majority of patients being in the fifth and sixth decades of life. The clinical histories, surgical records, and the follow-up data were obtained from the patients' case notes. The size of the carcinoma in the resected specimen was measured and the distance from the line of section to the nearest visible margin of the tumor was recorded. An inadequate bronchial stump was defined as one in which the bronchial margin was less than 1.5 ern. from the main tumor mass. The histologic report from each surgical specimen was reviewed. The sectioned margin of the bronchus was carefully studied for the presence of residual carcinoma in the wall and
the parabronchial tissues, in addition to any mucosal changes. The presence or absence of lymph node metastases was noted. Operative death was defined as death occurring within 30 days of the operation. For determining survival, all those patients who died were considered to have died as the result of carcinoma. After discharge from the hospital the patients were examined at 6 weeks, 3 months, 6 months, 1 year, 1V2 years, and subsequently, if the patients remained free from recurrence or metastases, at yearly intervals. Postoperative radiotherapy and chemotherapy were not used routinely in any of the patients for residual tumor on the bronchial margin. There was complete follow-up in all 64 cases. Results The 64 patients with residual tumor were divided into four subgroups depending upon the distance of the tumor from the bronchial line of section: Group A, less than 0.5 cm., Group B, 0.5 to 1.5 cm., Group C, 1.6 to 2.5 cm., and Group D, more than 2.5 cm. On gross examination the bronchial margin was apparently tumor free in all of these cases. The mean diameter of the carcinoma in the main series was 4.2 cm. and that in the residual tumor group, 4.8 ern. The 5 year survivors had smaller carcinomas with a mean diameter of 2.9 cm. Histologic data are analyzed in Table III. Squamous and adenocarcinomas included all degrees of differentiation. Hilar lymph node metastases were present in 212 patients (48.8 percent) in the main series and 46 patients (71. 9 percent) in the residual tumor group. Four distinct patterns of residual microscopic carcinoma at the bronchial line of section were observed: 1. Direct extension (34 patients). This pattern was seen as a direct extension of the main tumor mass proximally in the bronchial wall, mainly in patients with an inadequate bronchial stump (Table IV). 2. Lymphatic permeation (14 patients). This pattern was characterized by submucosal (six patients) and peribronchial (eight patients) permeation. 3. Parabronchial tissues (six patients). In these patients there were clumps of cancer cells separate from the main tumor mass. 4. Carcinoma in situ (10 patients). In these cases there was an intact basement membrane, cellular disorganization, and a great variation in size, shape, and chromatin content of the nucleus with frequent hyperchromatism. There was an increase in atypical mitoses. In six cases there was a variable amount of squamous metaplasia and carcinoma in situ change immediately
Volume 78 Number 2 August. 1979
177
Bronchogenic carcinoma
Table IV. Analysis of distance of tumor from the bronchial margin in the 64 patients Lymphatic permeation Distance of tumor from the bronchial margin A. Less than 0.5 em. B. 0.5-1.5 em. 1.6-2.5 em. D. More than 2.5 em. Totals
e.
Parabronchial tissues
Direct extension
Submucosal
18 10
o
I
I
2
3 1 3 8
2 2 1 6
5
2
I 34
2
Peribronchial
6
Carcinoma in situ change
Totals 22
2 4 3
21 13
1 10
8 64
Table V. Analysis of cell type and hilar node metastases in the 64 patients Lymphatic permeation
A. Cell type B. Positive hilar nodes
Direct extension
Submucosal
Percentage with positive nodes
Peribronchial
Parabronchial tissues
Carcinoma in situ change
Total
7 7
3 2
10 4
41 28
68.3 44.5
A. Squamous cell B. Positive nodes
20 14
A. Anaplastic B. Positive nodes
8 3
0 0
o o
0 0
9 4
A. Adenocarcinoma B. Positive nodes
2 2
5 5
o o
0 0
8 8
100
A. Oat cell B. Positive nodes
4 4
0 0
0 0
6 6
100
34
6 6
10 4
46
Total No. of patients Positive nodes
23
around the tumor, which continued up to the bronchial margin. In the remaining four, there were satellite foci of carcinoma in situ, some of which involved the bronchial line of section with apparently no continuity with the main tumor mass. Histologically, all tumors in this subgroup were of squamous cell type. In addition to the 64 patients, 14 patients had squamous metaplasia at the line of section and five (35.7 percent) of them survived for more than 5 years. These patients have been excluded from any further analysis. The length of the resected bronchial stump in the four subdivisions of the residual tumor group are analyzed in Table IV. Twenty-eight of 34 patients with direct extension had an inadequate bronchial stump. Histologic characteristics and hilar lymph node metastases are analyzed in Table V. All patients who had residual tumor in the lymphatics or who had adenocarcinoma or oat cell carcinoma had hilar lymph node metastases. Bronchopleural fistulas developed in eight patients (12.5 percent), four of whom died in the hospital postoperatively. In five of these patients the line of section was less than 0.5 em. from the tumor and there was
8 8
6 5
64 71.9
direct extension of the carcinoma. The remaining three patients were evenly divided in the other three subgroups. Only one patient with a bronchopleural fistula and a window thoracostomy is alive at 9 3.4 years, the others having died at 6 months, 10 months, and 18 months. Ten patients (15.6 percent) died in the hospital postoperati vely. Four of these deaths occurred as a result of bronchopleural fistula, two respiratory failure, one myocardial infarction, one pulmonary embolus, one hemorrhage, and one chest infection. Of the 54 patients who survived the initial operation, 15 (23.4 percent) lived for more than 5 years (Table VI). If postoperative deaths are excluded from analysis, then 27.8 percent of these patients survived 5 years or more. All gross macroscopic tumor was removed at operation in all the survivors except one (Table VI, Patient 6), who was given a course of radiotherapy postoperatively for residual chest wall tumor. All 5 year survivors were men with a mean age of 62 years and a duration of presenting symptoms ranging from no symptoms to I year. The mean diameter of the tumor was 2.9 em. and all were of squamous cell type.
I 78
The Journal of Thoracic and Cardiovascular Surgery
Soorae and Stevenson
Table VI. Details of 5 year survivors Case No.
Age (yr.}, sex
Duration of symptoms
Size of carcinoma (em.)
Operation
Distance from bronchial margin (em.)
56, M
3 mo.
Pneumonectomy
5
2
48, M
I mo.
Lobectomy
2
3
58, M
3 mo.
Lobectomy
4
4
76, M
2 mo.
Pneumonectomy
1.5
5
69, M
8 mo.
Sleeve resection
2
1.5
6
62, M
Asympt.
Lobectomy
3.5
2
7
62, M
6 mo.
Sleeve resection
3
<0.5
8
51, M
4 mo.
Lobectomy
3
3
9
64, M
4 mo.
Pneumonectomy
3.5
10
73, M
IV:! mo.
Sleeve resection
2
1.5
II
62, M
3 mo.
2
<0.5
12
55, M
Asympt.
Segmental resection Lobectomy
2.5
<0.5
13
65, M
Asympt.
Lobectomy
2
<0.5
14
61, M
I yr.
Sleeve resection
3
2
15
68, M
I mo.
Lobectomy
4
<0.5
2
<0.5
<0,5
Histology Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca
Hilar node metastases Yes
No Yes
No Yes Yes Yes
No No No No Yes
No No Yes
Legend: Ca, Carcinoma. A & W, Alive and well.
Table VII. Analysis of survival * Survival Pattern of residual tumor
No. of patients
Direct extension Lymphatic permeation (submucosal and peribronchial) Parabronchial Carcinoma in situ Totals
34
7
14
1
6
Postop. deaths
2
10
o
64
10
1 yr.
13 yr.
18 3
12
o
I
5 yr. and over
7
o
2
I
1
9
8
32
21
7 IS
survivors had direct extension of the tumor, and in all of them the resected bronchial stump was less than 0.5 em. long. Seven 5 year survivors had carcinoma in situ on the resected bronchial stump, which varied from I to 3 em. in length; clumps of tumor cells in the parabronchial tissues were seen in only one patient. There were no long-term survivors among patients with tumor in the submucosal and peribronchial lymphatics, 78.6 percent of these patients having died within I year and the rest having died within 3 years (Table VII). Overall, 32 patients (50 percent) survived I year and 21 patients (32.8 percent) survived 3 years.
'The over-all 5 year survival rate was 23.5 percent.
Discussion
Seven patients (46.6 percent) had hilar lymph node metastases. Seven patients were treated by lobectomy, four had sleeve resection, three pneumonectomy, and one segmental resection. Six patients were alive and clinically tumor free at the conclusion of the series, the longest survival period being 9 3,4 years. Seven 5-year
It would be reasonable to suppose that patients with residual tumor at the line of section would have a poor prognosis. However, long-term survival after incomplete resection of carcinoma of the lung has often been documented. In addition to the fifteen 5 year survivors in this series, Jeffery" reported six 5 year survivors of
Volume 78
Bronchogenic carcinoma
Number 2
August, 1979
Partern of residual tumor Ca in situ Ca in situ Dir. extension Ca in situ Ca in situ Ca in situ Dir. extension Ca in situ Dir. extension Parabronchial Dir. extension Dir. extension Dir. extension Ca in situ Dir. extension
Survival A&W, 9¥.. yr. A& W, 9'h yr. A& W, 9'12 yr. A&W, 8'1.1 yr. Died, 7 34 yr. Died, 7'/2 yr. Died, 6'/2 yr. Died, 6Y2 yr. Died, 6'/2 yr. Died, 6 yr. Died, 6 yr. A& W, 5¥.. yr. A&W, 5 34 yr. Died, 5 34 yr.
Comment Bronchopleural fistula
Postop. radiotherapy for residual chest wall tumor
Radical pneumonectomy with wedge of lower trachea
Stump recurrence at 3'1.1 yr.treated with radiotherapy
Died, 5'4 yr.
18 patients with residual tumor on the bronchial stump, and Shields" reported nine such survivors of 54 patients with microscopic residual tumor on the bronchial margin. Shields,4 upon analyzing 221 incomplete resections, concluded that the patients with microscopic residual tumor at a grossly normal cut margin do better than patients with other types of incomplete resections. In contrast, Abbey Smith" reported cure in three patients with gross residual tumor on the left atrial wall. Cotton," on studying the proximal spread of bronchial carcinoma in 100 consecutive cases of resection, confirmed the earlier work of Griess and associates' and suggested that at least 1.9 em. of apparently normal bronchus proximal to the macroscopic tumor should be resected if the risk of residual tumor at the line of section is to be eliminated. This is essentially true for cases of direct extension in the bronchial wall, but not so in other types of residual tumor: for example, tumor in the lymphatic channels, clumps of cancer cells in the
179
parabronchial tissues, and carcinoma in situ change, which may be contiguous with the main tumor or may be in separate foci. Residual tumor in all of these cases can be far removed from the main tumor mass (Table I V), as there is no way of identifying these cases at the time of resection. The inability to resect an adequate length of bronchial stump should not be considered as a contraindication to resection or as an indication for converting a lobectomy to a pneumonectomy in a patient who might not tolerate this procedure. A short bronchial stump in the resected lung does not preclude long-term survival. In seven of the 15 survivors the line of section was less than 0.5 ern. from the tumor, and all of these patients had direct extension of the carcinoma. No direct relationship between the length of the resected stump and survival could be established. In the series reported by Cotton," there was obvious invasion of submucosal lymphatics in six patients, and all of them had extensive intrapulmonary and mediastinallymph node invasion. On this basis, he postulated a very unfavorable prognosis for this group of patients. His theory has been amply confirmed by this study, wherein of 14 such patients, all with hilar node metastases, II (78.6 percent) died within I year and the remaining three died in the next 18 months. Lymphatic permeation by tumor at the line of section, therefore, is one of the worst prognostic factors. The immediate postoperative mortality rate of 15.6 percent is similar to that reported in other series of incomplete resections. Abbey Smith" reported a mortality rate of 20 percent and Shields," in a comparable series of 67 patients with involvement of the bronchial margin, reported a 19.4 percent mortality rate. Bronchopleural fistulas developed postoperatively in eight patients (12.5 percent), four of whom died in the hospital within 30 days. This incidence, again, is similar to the 10.4 percent incidence reported by Shields." In five of these patients with bronchopleural fistulas, the resected bronchial stump was less than 0.5 em. long and had direct extension of tumor. The high mortality and morbidity in this series can be attributed directly to the high incidence of bronchopleural fistula and hence to the presence of residual tumor. Resecting a longer bronchial stump may reduce the immediate morbidity and mortality, but the length of the bronchial stump has no direct influence on survival. The over-all 5 year survival rate of 23.4 percent (27.8 percent if hospital deaths are excluded) is not much different from the over-all survival rate after resection in general, which is reported to vary between 22 and 41 percent. 7-9 The common factors in the 5 year survivors were that they were all men with relatively small squa-
The Journal of Thoracic and Cardiovascular Surgery
I 80 Soorae and Stevenson
mous cell tumors of varying degrees of differentiation. The mean tumor diameter of 2.9 cm. (range 1.5 to 5 cm.) in the survivors is much smaller than the mean diameter of 4.8 ern. in the residual tumor group as a whole. A relationship between the size of the tumor and survival is still controversial, although Soorae and Abbey Smith'? recently showed that larger tumors have a worse prognosis after resection. There are conflicting reports about the importance of cell type upon the survival rates in bronchial carcinoma,II-13 but there is uniform agreement concerning the over-all poor results with tumors of the oat cell type.:" Nonetheless, it is interesting that all 5 year survivors had squamous cell tumors. Shields" also made a similar observation in his series of incomplete resections, in which all nine 5 year survivors had squamous cell tumors. Therefore, it can be safely concluded that, even in the presence of residual tumor, squamous cell type exerts a beneficial influence on survival. The relatively high incidence (46.7 percent) of hilar node metastases in the survivors is difficult to explain. This incidence is similar to that in the main series (48.8 percent) but lower than that in the residual tumor group as a whole (71.9 percent). Spontaneous regression of bronchogeric carcinoma without treatment is extremely rare. Local mechanical trauma to residual neoplastic cells, interference with their nutrition, endocrine factors, pyrexia, and infection all have been proposed by various workers to explain long survival in the presence of tumor after resection. A more likely explanation for the survival with residual tumor in this series is probably a positive immunologic response by the patient when the main tumor mass has been excised. The most useful role of irradiation in carcinoma of the bronchus is in palliation of symptoms caused by the primary tumor or metastases and in radical treatment of oat cell carcinoma.P Routine postoperative radiotherapy has not been beneficial"; the role of irradiation, chemotherapy, or immunotherapy in patients with residual microscopic foci remains undetermined. 4 For this reason, routine adjunct therapy was not employed in the management of patients with residual tumor on the bronchial margin. Radiotherapy was given only to patients in whom gross tumor was left behind following resection (Table VI, Patient 6) or to those who had subsequent recurrence of the tumor at the line of section (Table VI, Patient 14). However, there is increasing evidence that nonspecific active immunization can improve survival in cases of bronchial carcinoma with minimal residual tumor load. 16. 17 All of the patients in
our study are in this category; therefore, adjunct immunotherapy may play an important role in the management of these cases. REFERENCES Griess DF, McDonald JR, Clagett OT: The proximal extension of carcinoma of the lung in the bronchial wall. J THoRAc SURG 14:362-368, 1945 2 Cotton RE: The bronchial spread of lung cancer. Br J Dis Chest 53:142-150, 1959 3 Jeffery RM: Survival in bronchial carcinoma. Tumour remaining in the bronchial stump following resection. Ann R Coli Surg Engl 51:55-59, 1972 4 Shields TW: The fate of patients after incomplete resection of bronchial carcinoma. Surg Gynecol Obstet
139:569-572, 1974 5 Abbey Smith R: Cure of lung cancer from incomplete surgical resection. Br Med J 2:563 -565, 1971 6 Abbey Smith R: The results of raising the resectability rate in operations for lung carcinoma. Thorax 12:79-86,
1957 7 Bignall JR, Martin M, Smithers DW: Survival in 6,086 cases of bronchial carcinoma. Lancet 1: 1067-1070, 1967 8 Brock L: Long survival after operation for cancer of the lung. Br J Surg 62:1-5, 1975 9 Stanford W, Spivey CG, Larsen GL, Alexander JA, Besich WJ: Results of treatment of primary carcinoma of the lung. Analysis of 3,000 cases. J THORAC CARDlOVASC SURG 72:441-449, 1976 10 Soorae AS, Abbey Smith R: Tumour size as a prognostic factor after resection of lung carcinoma. Thorax 32: 19-
25, 1977 II Collins NP: Bronchogenic carcinoma. Importance of the cell type. Arch Surg 7:925-932, 1958 12 Belcher JR, Anderson R: Surgical treatment of carcinoma of the bronchus. Br Med J 1:948-954, 1965 13 Kirsh MM, Prior M, Gago 0, Moores WY, Kahn DR, Pellegrini R V, Sloan H: The effect of histological cell type on the prognosis of patients with bronchogenic carcinoma. Ann Thorac Surg 13:303-310, 1972 14 Kato Y, Ferguson TB, Bennett DE, Burford TH: Oat cell carcinoma of the lung. A review of 138 cases. Cancer
23:517-524, 1969 15 Spittle MF: Radiotherapy in the treatment of carcinoma of the bronchus. Proceedings of a Symposium on Carcinoma of the Bronchus held at the University of Nottingham, March 24,1977, KC Kalman, ed., published by M. C. S. Consultants, 1977, p 21-31 16 Price Evans DA: Immunology of bronchial carcinoma. Thorax 31:493-506, 1976 17 Jansen HM, The TH, deGast GC, Esselink MT, van der Wal AM, Orie NGM: Adjuvant immunotherapy with BCG in squamous-cell bronchial carcinoma. Immunereactivity in relation to immunostimulation (preliminary results in a controlled trial). Thorax 33:429-438, 1978
A. S. Soorae, M.B., F.R.C.S.(Ed.), F.R.C.S.(Eng.), and H. M. Stevenson, M.B., F.R.C.S.(Ed.), Belfast, Northern Ireland
T
he proximal spread of bronchogenic carcinoma was studied by early workers like Griess and associates' and Cotton," but they made no attempt to correlate this spread with survival in their patients. More recently, favorable reports of long-term survival of patients with microscopic residual carcinoma at the bronchial margin have appeared,": 4 and it is becoming apparent that these patients have a better survival rate than those having an incomplete resection, in whom gross tumor is left behind elsewhere in the hemithorax." This study was designed to determine the incidence and pattern of residual carcinoma at the bronchial margin after pulmonary resection and to analyze the factors that may have contributed to the 5 year survival. From the Thoracic Surgical Unit, Royal Victoria Hospital, Grosvenor Road, Belfast BTl 2 6BA, Northern Ireland. Received for publication Oct. 4. 1978. Accepted for publication Feb. 19, 1979. Address for reprints: A. S. Soorae, M.B. F.R.C.S.(Ed.) F.R.C.S. (Eng.), Consultant Cardiothoracic Surgeon, Fazakerley Hospital, Longmoor Lane. Liverpool L9 7AL, England.
Patients and methods A total of 556 consecutive patients underwent thoracotomy for bronchogenic carcinoma between January, 1968, and December, 1972, and the results were analyzed retrospectively in June, 1978. This period for study was chosen to obtain a minimum follow-up of 5 years. The carcinoma was found to be unresectable in 122 patients (100 men and 22 women), for a resectability rate of 78.1 percent. Of the 434 patients who had pulmonary resection, 374 were men and 60 women, for a male-to-female ratio of 6.2: I. This group of patients will be referred to as the main series. Sixty-four patients (57 men and seven women) were found to have residual microscopic tumor at the line of section. These patients form the basis of this study, and they will be referred to as the residual tumor group. Ages ranged from 30 to 77 years in the main series, with a mean over-all age of 60'12 years in the main series and 59 years in the residual tumor group (Table I). The numbers and types of resection are shown in Table II. The distribution of patients in each decade and the numbers and types of
0022-5223/79(080175+06$00.60(0 © 1979 The C. V. Mosby Co.
175
The Journal of
1 76 Soorae and Stevenson
Thoracic and Cardiovascular Surgery
Table I. Age distribution by decades Residual tumor group
Main series Age by decades (yr.)
No. of patients
30-39 40-49 50-59 60-69 70-79
10 64 136 185 39
Totals
434
Mean age
I Percent 2.3 14.8 31.3 42.6 9.0 100
No. of patients
I
Percent
2 13 12 33 4
3.1 20.3 18.7 51.6 6.2 100
64
60.5 yr.
59 yr.
Table II. Analysis of numbers and types of resection Main series (n = 434) Operation
Pneumonectomy Lobectomy Sleeve resection Wedge resection Segmental resection
I
No. of patients
210 195 17 8 4
Residual tumor group (n = 64)
Percent
No. of patients
48.4 45.0 3.9 1.8 0.9
28 27 8 0 1
I
Percent
45.3 42.2 12.5 0 1.5
Table III. Analysis of histology Main series (n = 434)
Cell type
Squamous cell Anaplastic Adenocarcinoma Oat cell
No. of patients
252 84 53 45
I
Residual tumor group (n = 64)
Percent
No. of patients
58.0 19.4 12.2 10.4
41 9 8 6
I Percent 64 14 12.5 9.5
resection were similar in the two groups of patients, with the majority of patients being in the fifth and sixth decades of life. The clinical histories, surgical records, and the follow-up data were obtained from the patients' case notes. The size of the carcinoma in the resected specimen was measured and the distance from the line of section to the nearest visible margin of the tumor was recorded. An inadequate bronchial stump was defined as one in which the bronchial margin was less than 1.5 ern. from the main tumor mass. The histologic report from each surgical specimen was reviewed. The sectioned margin of the bronchus was carefully studied for the presence of residual carcinoma in the wall and
the parabronchial tissues, in addition to any mucosal changes. The presence or absence of lymph node metastases was noted. Operative death was defined as death occurring within 30 days of the operation. For determining survival, all those patients who died were considered to have died as the result of carcinoma. After discharge from the hospital the patients were examined at 6 weeks, 3 months, 6 months, 1 year, 1V2 years, and subsequently, if the patients remained free from recurrence or metastases, at yearly intervals. Postoperative radiotherapy and chemotherapy were not used routinely in any of the patients for residual tumor on the bronchial margin. There was complete follow-up in all 64 cases. Results The 64 patients with residual tumor were divided into four subgroups depending upon the distance of the tumor from the bronchial line of section: Group A, less than 0.5 cm., Group B, 0.5 to 1.5 cm., Group C, 1.6 to 2.5 cm., and Group D, more than 2.5 cm. On gross examination the bronchial margin was apparently tumor free in all of these cases. The mean diameter of the carcinoma in the main series was 4.2 cm. and that in the residual tumor group, 4.8 ern. The 5 year survivors had smaller carcinomas with a mean diameter of 2.9 cm. Histologic data are analyzed in Table III. Squamous and adenocarcinomas included all degrees of differentiation. Hilar lymph node metastases were present in 212 patients (48.8 percent) in the main series and 46 patients (71. 9 percent) in the residual tumor group. Four distinct patterns of residual microscopic carcinoma at the bronchial line of section were observed: 1. Direct extension (34 patients). This pattern was seen as a direct extension of the main tumor mass proximally in the bronchial wall, mainly in patients with an inadequate bronchial stump (Table IV). 2. Lymphatic permeation (14 patients). This pattern was characterized by submucosal (six patients) and peribronchial (eight patients) permeation. 3. Parabronchial tissues (six patients). In these patients there were clumps of cancer cells separate from the main tumor mass. 4. Carcinoma in situ (10 patients). In these cases there was an intact basement membrane, cellular disorganization, and a great variation in size, shape, and chromatin content of the nucleus with frequent hyperchromatism. There was an increase in atypical mitoses. In six cases there was a variable amount of squamous metaplasia and carcinoma in situ change immediately
Volume 78 Number 2 August. 1979
177
Bronchogenic carcinoma
Table IV. Analysis of distance of tumor from the bronchial margin in the 64 patients Lymphatic permeation Distance of tumor from the bronchial margin A. Less than 0.5 em. B. 0.5-1.5 em. 1.6-2.5 em. D. More than 2.5 em. Totals
e.
Parabronchial tissues
Direct extension
Submucosal
18 10
o
I
I
2
3 1 3 8
2 2 1 6
5
2
I 34
2
Peribronchial
6
Carcinoma in situ change
Totals 22
2 4 3
21 13
1 10
8 64
Table V. Analysis of cell type and hilar node metastases in the 64 patients Lymphatic permeation
A. Cell type B. Positive hilar nodes
Direct extension
Submucosal
Percentage with positive nodes
Peribronchial
Parabronchial tissues
Carcinoma in situ change
Total
7 7
3 2
10 4
41 28
68.3 44.5
A. Squamous cell B. Positive nodes
20 14
A. Anaplastic B. Positive nodes
8 3
0 0
o o
0 0
9 4
A. Adenocarcinoma B. Positive nodes
2 2
5 5
o o
0 0
8 8
100
A. Oat cell B. Positive nodes
4 4
0 0
0 0
6 6
100
34
6 6
10 4
46
Total No. of patients Positive nodes
23
around the tumor, which continued up to the bronchial margin. In the remaining four, there were satellite foci of carcinoma in situ, some of which involved the bronchial line of section with apparently no continuity with the main tumor mass. Histologically, all tumors in this subgroup were of squamous cell type. In addition to the 64 patients, 14 patients had squamous metaplasia at the line of section and five (35.7 percent) of them survived for more than 5 years. These patients have been excluded from any further analysis. The length of the resected bronchial stump in the four subdivisions of the residual tumor group are analyzed in Table IV. Twenty-eight of 34 patients with direct extension had an inadequate bronchial stump. Histologic characteristics and hilar lymph node metastases are analyzed in Table V. All patients who had residual tumor in the lymphatics or who had adenocarcinoma or oat cell carcinoma had hilar lymph node metastases. Bronchopleural fistulas developed in eight patients (12.5 percent), four of whom died in the hospital postoperatively. In five of these patients the line of section was less than 0.5 em. from the tumor and there was
8 8
6 5
64 71.9
direct extension of the carcinoma. The remaining three patients were evenly divided in the other three subgroups. Only one patient with a bronchopleural fistula and a window thoracostomy is alive at 9 3.4 years, the others having died at 6 months, 10 months, and 18 months. Ten patients (15.6 percent) died in the hospital postoperati vely. Four of these deaths occurred as a result of bronchopleural fistula, two respiratory failure, one myocardial infarction, one pulmonary embolus, one hemorrhage, and one chest infection. Of the 54 patients who survived the initial operation, 15 (23.4 percent) lived for more than 5 years (Table VI). If postoperative deaths are excluded from analysis, then 27.8 percent of these patients survived 5 years or more. All gross macroscopic tumor was removed at operation in all the survivors except one (Table VI, Patient 6), who was given a course of radiotherapy postoperatively for residual chest wall tumor. All 5 year survivors were men with a mean age of 62 years and a duration of presenting symptoms ranging from no symptoms to I year. The mean diameter of the tumor was 2.9 em. and all were of squamous cell type.
I 78
The Journal of Thoracic and Cardiovascular Surgery
Soorae and Stevenson
Table VI. Details of 5 year survivors Case No.
Age (yr.}, sex
Duration of symptoms
Size of carcinoma (em.)
Operation
Distance from bronchial margin (em.)
56, M
3 mo.
Pneumonectomy
5
2
48, M
I mo.
Lobectomy
2
3
58, M
3 mo.
Lobectomy
4
4
76, M
2 mo.
Pneumonectomy
1.5
5
69, M
8 mo.
Sleeve resection
2
1.5
6
62, M
Asympt.
Lobectomy
3.5
2
7
62, M
6 mo.
Sleeve resection
3
<0.5
8
51, M
4 mo.
Lobectomy
3
3
9
64, M
4 mo.
Pneumonectomy
3.5
10
73, M
IV:! mo.
Sleeve resection
2
1.5
II
62, M
3 mo.
2
<0.5
12
55, M
Asympt.
Segmental resection Lobectomy
2.5
<0.5
13
65, M
Asympt.
Lobectomy
2
<0.5
14
61, M
I yr.
Sleeve resection
3
2
15
68, M
I mo.
Lobectomy
4
<0.5
2
<0.5
<0,5
Histology Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca Squamous cell Ca
Hilar node metastases Yes
No Yes
No Yes Yes Yes
No No No No Yes
No No Yes
Legend: Ca, Carcinoma. A & W, Alive and well.
Table VII. Analysis of survival * Survival Pattern of residual tumor
No. of patients
Direct extension Lymphatic permeation (submucosal and peribronchial) Parabronchial Carcinoma in situ Totals
34
7
14
1
6
Postop. deaths
2
10
o
64
10
1 yr.
13 yr.
18 3
12
o
I
5 yr. and over
7
o
2
I
1
9
8
32
21
7 IS
survivors had direct extension of the tumor, and in all of them the resected bronchial stump was less than 0.5 em. long. Seven 5 year survivors had carcinoma in situ on the resected bronchial stump, which varied from I to 3 em. in length; clumps of tumor cells in the parabronchial tissues were seen in only one patient. There were no long-term survivors among patients with tumor in the submucosal and peribronchial lymphatics, 78.6 percent of these patients having died within I year and the rest having died within 3 years (Table VII). Overall, 32 patients (50 percent) survived I year and 21 patients (32.8 percent) survived 3 years.
'The over-all 5 year survival rate was 23.5 percent.
Discussion
Seven patients (46.6 percent) had hilar lymph node metastases. Seven patients were treated by lobectomy, four had sleeve resection, three pneumonectomy, and one segmental resection. Six patients were alive and clinically tumor free at the conclusion of the series, the longest survival period being 9 3,4 years. Seven 5-year
It would be reasonable to suppose that patients with residual tumor at the line of section would have a poor prognosis. However, long-term survival after incomplete resection of carcinoma of the lung has often been documented. In addition to the fifteen 5 year survivors in this series, Jeffery" reported six 5 year survivors of
Volume 78
Bronchogenic carcinoma
Number 2
August, 1979
Partern of residual tumor Ca in situ Ca in situ Dir. extension Ca in situ Ca in situ Ca in situ Dir. extension Ca in situ Dir. extension Parabronchial Dir. extension Dir. extension Dir. extension Ca in situ Dir. extension
Survival A&W, 9¥.. yr. A& W, 9'h yr. A& W, 9'12 yr. A&W, 8'1.1 yr. Died, 7 34 yr. Died, 7'/2 yr. Died, 6'/2 yr. Died, 6Y2 yr. Died, 6'/2 yr. Died, 6 yr. Died, 6 yr. A& W, 5¥.. yr. A&W, 5 34 yr. Died, 5 34 yr.
Comment Bronchopleural fistula
Postop. radiotherapy for residual chest wall tumor
Radical pneumonectomy with wedge of lower trachea
Stump recurrence at 3'1.1 yr.treated with radiotherapy
Died, 5'4 yr.
18 patients with residual tumor on the bronchial stump, and Shields" reported nine such survivors of 54 patients with microscopic residual tumor on the bronchial margin. Shields,4 upon analyzing 221 incomplete resections, concluded that the patients with microscopic residual tumor at a grossly normal cut margin do better than patients with other types of incomplete resections. In contrast, Abbey Smith" reported cure in three patients with gross residual tumor on the left atrial wall. Cotton," on studying the proximal spread of bronchial carcinoma in 100 consecutive cases of resection, confirmed the earlier work of Griess and associates' and suggested that at least 1.9 em. of apparently normal bronchus proximal to the macroscopic tumor should be resected if the risk of residual tumor at the line of section is to be eliminated. This is essentially true for cases of direct extension in the bronchial wall, but not so in other types of residual tumor: for example, tumor in the lymphatic channels, clumps of cancer cells in the
179
parabronchial tissues, and carcinoma in situ change, which may be contiguous with the main tumor or may be in separate foci. Residual tumor in all of these cases can be far removed from the main tumor mass (Table I V), as there is no way of identifying these cases at the time of resection. The inability to resect an adequate length of bronchial stump should not be considered as a contraindication to resection or as an indication for converting a lobectomy to a pneumonectomy in a patient who might not tolerate this procedure. A short bronchial stump in the resected lung does not preclude long-term survival. In seven of the 15 survivors the line of section was less than 0.5 ern. from the tumor, and all of these patients had direct extension of the carcinoma. No direct relationship between the length of the resected stump and survival could be established. In the series reported by Cotton," there was obvious invasion of submucosal lymphatics in six patients, and all of them had extensive intrapulmonary and mediastinallymph node invasion. On this basis, he postulated a very unfavorable prognosis for this group of patients. His theory has been amply confirmed by this study, wherein of 14 such patients, all with hilar node metastases, II (78.6 percent) died within I year and the remaining three died in the next 18 months. Lymphatic permeation by tumor at the line of section, therefore, is one of the worst prognostic factors. The immediate postoperative mortality rate of 15.6 percent is similar to that reported in other series of incomplete resections. Abbey Smith" reported a mortality rate of 20 percent and Shields," in a comparable series of 67 patients with involvement of the bronchial margin, reported a 19.4 percent mortality rate. Bronchopleural fistulas developed postoperatively in eight patients (12.5 percent), four of whom died in the hospital within 30 days. This incidence, again, is similar to the 10.4 percent incidence reported by Shields." In five of these patients with bronchopleural fistulas, the resected bronchial stump was less than 0.5 em. long and had direct extension of tumor. The high mortality and morbidity in this series can be attributed directly to the high incidence of bronchopleural fistula and hence to the presence of residual tumor. Resecting a longer bronchial stump may reduce the immediate morbidity and mortality, but the length of the bronchial stump has no direct influence on survival. The over-all 5 year survival rate of 23.4 percent (27.8 percent if hospital deaths are excluded) is not much different from the over-all survival rate after resection in general, which is reported to vary between 22 and 41 percent. 7-9 The common factors in the 5 year survivors were that they were all men with relatively small squa-
The Journal of Thoracic and Cardiovascular Surgery
I 80 Soorae and Stevenson
mous cell tumors of varying degrees of differentiation. The mean tumor diameter of 2.9 cm. (range 1.5 to 5 cm.) in the survivors is much smaller than the mean diameter of 4.8 ern. in the residual tumor group as a whole. A relationship between the size of the tumor and survival is still controversial, although Soorae and Abbey Smith'? recently showed that larger tumors have a worse prognosis after resection. There are conflicting reports about the importance of cell type upon the survival rates in bronchial carcinoma,II-13 but there is uniform agreement concerning the over-all poor results with tumors of the oat cell type.:" Nonetheless, it is interesting that all 5 year survivors had squamous cell tumors. Shields" also made a similar observation in his series of incomplete resections, in which all nine 5 year survivors had squamous cell tumors. Therefore, it can be safely concluded that, even in the presence of residual tumor, squamous cell type exerts a beneficial influence on survival. The relatively high incidence (46.7 percent) of hilar node metastases in the survivors is difficult to explain. This incidence is similar to that in the main series (48.8 percent) but lower than that in the residual tumor group as a whole (71.9 percent). Spontaneous regression of bronchogeric carcinoma without treatment is extremely rare. Local mechanical trauma to residual neoplastic cells, interference with their nutrition, endocrine factors, pyrexia, and infection all have been proposed by various workers to explain long survival in the presence of tumor after resection. A more likely explanation for the survival with residual tumor in this series is probably a positive immunologic response by the patient when the main tumor mass has been excised. The most useful role of irradiation in carcinoma of the bronchus is in palliation of symptoms caused by the primary tumor or metastases and in radical treatment of oat cell carcinoma.P Routine postoperative radiotherapy has not been beneficial"; the role of irradiation, chemotherapy, or immunotherapy in patients with residual microscopic foci remains undetermined. 4 For this reason, routine adjunct therapy was not employed in the management of patients with residual tumor on the bronchial margin. Radiotherapy was given only to patients in whom gross tumor was left behind following resection (Table VI, Patient 6) or to those who had subsequent recurrence of the tumor at the line of section (Table VI, Patient 14). However, there is increasing evidence that nonspecific active immunization can improve survival in cases of bronchial carcinoma with minimal residual tumor load. 16. 17 All of the patients in
our study are in this category; therefore, adjunct immunotherapy may play an important role in the management of these cases. REFERENCES Griess DF, McDonald JR, Clagett OT: The proximal extension of carcinoma of the lung in the bronchial wall. J THoRAc SURG 14:362-368, 1945 2 Cotton RE: The bronchial spread of lung cancer. Br J Dis Chest 53:142-150, 1959 3 Jeffery RM: Survival in bronchial carcinoma. Tumour remaining in the bronchial stump following resection. Ann R Coli Surg Engl 51:55-59, 1972 4 Shields TW: The fate of patients after incomplete resection of bronchial carcinoma. Surg Gynecol Obstet
139:569-572, 1974 5 Abbey Smith R: Cure of lung cancer from incomplete surgical resection. Br Med J 2:563 -565, 1971 6 Abbey Smith R: The results of raising the resectability rate in operations for lung carcinoma. Thorax 12:79-86,
1957 7 Bignall JR, Martin M, Smithers DW: Survival in 6,086 cases of bronchial carcinoma. Lancet 1: 1067-1070, 1967 8 Brock L: Long survival after operation for cancer of the lung. Br J Surg 62:1-5, 1975 9 Stanford W, Spivey CG, Larsen GL, Alexander JA, Besich WJ: Results of treatment of primary carcinoma of the lung. Analysis of 3,000 cases. J THORAC CARDlOVASC SURG 72:441-449, 1976 10 Soorae AS, Abbey Smith R: Tumour size as a prognostic factor after resection of lung carcinoma. Thorax 32: 19-
25, 1977 II Collins NP: Bronchogenic carcinoma. Importance of the cell type. Arch Surg 7:925-932, 1958 12 Belcher JR, Anderson R: Surgical treatment of carcinoma of the bronchus. Br Med J 1:948-954, 1965 13 Kirsh MM, Prior M, Gago 0, Moores WY, Kahn DR, Pellegrini R V, Sloan H: The effect of histological cell type on the prognosis of patients with bronchogenic carcinoma. Ann Thorac Surg 13:303-310, 1972 14 Kato Y, Ferguson TB, Bennett DE, Burford TH: Oat cell carcinoma of the lung. A review of 138 cases. Cancer
23:517-524, 1969 15 Spittle MF: Radiotherapy in the treatment of carcinoma of the bronchus. Proceedings of a Symposium on Carcinoma of the Bronchus held at the University of Nottingham, March 24,1977, KC Kalman, ed., published by M. C. S. Consultants, 1977, p 21-31 16 Price Evans DA: Immunology of bronchial carcinoma. Thorax 31:493-506, 1976 17 Jansen HM, The TH, deGast GC, Esselink MT, van der Wal AM, Orie NGM: Adjuvant immunotherapy with BCG in squamous-cell bronchial carcinoma. Immunereactivity in relation to immunostimulation (preliminary results in a controlled trial). Thorax 33:429-438, 1978