Eur J Cardio-thorac Surg (1992) 6:498-502
0 SpringeTVerlag i992
Squamous cell carcinoma of the lung: prognostic factors affecting the long-term outcome after surgical resection H. Huwer, I. Volkmer, A. Wiegand, G. Kalweit, and P. Feindt Department of Thoracic and Cardiovascular Surgery, Surgery Division, Saarland University Medical School, Homburg, FRG
Abstract. From 1976 to 1989,638 patients with squamous cell carcinoma were operated on. Reliable information concerning survival could be obtained in 540 of these cases via follow-up examinations. All tumors were retrospectively classified according to the TNM system of the fourth edition of UICC classification. The average survival time after potentially curative operation was 7.32 years in patients with stage I tumors, the 5-year survival rate was 55.3%. The corresponding values were 2.42 years and 29.1% respectively in patients with stage II tumors, compared to 1.22 years and 12.0% respectively in those with stage III a tumors. The differences between tumor stages are statistically significant. Analyzing the influence of the T and N factors on the prognosis following potentially curative surgery, the classification in the various tumor stages was largely verified. Stage III a is inhomogeneous in terms of prognosis. T and N factors showed prognostic value in this group. In the group with T3 tumors there was a significant difference in the mean survival times, independently of the lymph node status, when the patients were divided into those with T3 tumors close to the main carina (16 months) and those who had infiltrated adjacent structures (7.5 months). When differentiating the Nl location, a significant difference was found between the group of patients with T2 tumors and true lymph node metastases and the group with T2 and direct tumor infiltration of the lymph node (mean survival time 27.5 vs 42.3 months). In cases with N2 disease, prognostic differences were shown in as much as the mean survival time in patients with superior mediastinal N2 was significantly shorter than in those with only tracheobronchial lymph node metastases (12.3 vs 21.6 months). The age of the patients and type of resection treatment within the same tumor stage group had no influence on the prognosis. Tendency to keratinize, carcinomatous lymphangiosis, and tumorous invasion of vessels also had no prognostic value in any given stage. Stage III a patients with poorly differentiated tumors had a worse prognosis than those with moderately differentiated ones. Following incomplete resection or exploratory thoracotomy, the short- und long-term prognoses were equally bad (mean survival time 9 and 7 months, respectively, and 5-year survival rate 0%). Mean survival time could be prolonged by external irradiation only when tumorous infiltration in the distal cut end of the bronchial stump was microscopically ascertained (13.0 months vs 7.3 months without irradiation treatment). However, there was no 5-year survivor in either group. [Eur J Cardio-thorac Surg (1992) 6:498-5021 Key words: Squamous
cell carcinoma
of the bronchus
- Staging
Surgical resection is the therapy of choice in patients with squamous cell carcinoma of the bronchus. The prognosis
- Prognostic
factors
of the TNM system and additional investigated in the present study.
parameters
will be
of surgically treated patients appears to depend primarily on the extent of the tumor and the lymph node metastases. However, the weighting of these factors is still in a state of flux, especially since a uniform worldwide scheme for TNM classification and staging of bronchial carcinomas, enabling comprehensive comparisons, was only published in 1987 in the fourth edition of the UICC classification. The prognostic significance of the parameters Received for publication: Accepted
for publication:
December 9, 1991 March 26, 1992
Patients and methods From 1976 to 1989, 638 patients with squamous cell carcinoma of the bronchus were operated on. A follow-up investigation was possible in 540 of those patients. For.direct comparison, retrospective classification of the tumors in accordance with the TNM scheme of the fourth edition of the UICC classification was carried out in all patients; there was no reevaluation of the histologic material. Incomplete resection is defined as one documented by histological proof of residual tumor at one of the margins of the resected
specimen. All other resections were considered complete and poten-
tially curative (191. The mean age of the patients. of whom 516 were men and 24 women, was 61 years at the time of operation. Tumor locations were equally distributed between sides (left lung 262, right lung 278). Table 1 shows the distribution of the kinds of curative resection according to tumor stage. After potentially curative operation. in the group of stage I tumors there were 31 keratinizing squamous cell carcinomas (16.2%), 45 partly keratinizing (23.6%), and 85 non-keratinizing (44.5%). In the stage II group there were 19 keratinizing (14.8%). 34 partly keratinizing (26.6%). and 49 non-keratinizing (38.3%) squamous cell carcinomas and in the stage III a group there were 30 keratinizing ( 19.1%), 45 partly keratinizing (28.7%), and 54 nonkeratinizing (34.4%) squamous cell carcinomas. There was no mention of keratinization tendency in 30 cases (15.7%) in stage I, 26 (20.3%) in stage II. and 28 (17.8%) in stage III a. Carcinomatous lymphangiosis was present in 21 stage I patients (I 1.O%), 35 (27.3%) stage II, and 42 (26.8%) stage III a. Invasion of blood vessels was recorded 7 times (3.7%) in stage I, 9 times (7.0%) in stage II. and 11 times (7.0%) in stage III a. Table 2 shows the degrees of differentiation of potentially curatively resected tumors. Lymph node dissection in every patient was carried out by removing preparatively visualized mediastinal lymph node groups. Complete skeletization of the mediastinal structures was not performed. From 1976 to 1985 each patient was examined preoperatively by mediastinoscopy. In patients with mediastinal lymph node involvement, postoperative irradiation was carried out routinely. After incomplete resection. postoperative irradiation was performed occasionally. Tumorous infiltration in the distal cut end of the bronchial stump was found microscopically following otherwise potentially curative intervention in 28 patients. Seventeen of these underwent postoperative irradiation. The survival rates were determined by means of actuarial analysis and compared for statistical significance (P < 0.05) by the K sample test
Results Out of the 540 patients with complete follow-up, 476 (i.e., 88.2%) underwent potentially curative resection. Fortyeight patients had an incomplete resection (8.9%) and 16 an exploratory thoracotomy (2.9%). The mean survival times for these two groups were 9 and 7 months, respectively, and were not significantly different. Incomplete resection because of infiltration of the distal cut end of the bronchial stump was identified microscopically in 28 patients. Seventeen of these patients underwent irradiation therapy postoperatively, while the remaining 11 had no further treatment. The mean survival time in postoperatively irradiated patients was 13 months, significantly higher than that in the 11 non-treated patients (7.3 months). No patient in either group survived 5 years. After potentially curative operation the 30-day mortality for all types of resection and all tumor stages was 7.2%: 8.4% after pneumonectomy and 5.8% after lobectomy and bilobectomy. Comparing the survival times in the various age groups without regard to tumor stage, there was no significant difference between the groups. After potentially curative resection treatment, tendency to keratinize. carcinomatous lymphangiosis, and tumorous infiltration into vessels did not affect the survival times within each tumor stage group.
Table 1. Distribution of the kinds of curative resection among the various tumor stages Stage I
Lobectomy Bilobectomy Pneumonectomy
Stage III
12
%
,1
%
n
%
114 18 59
59.7 9.4 30.9
48 8 72
37.5 6.3 56.2
33 5 119
21.0 3.2 75.8
Table 2. Degrees of differentiation tumors
Well differentiated Moderately differentiated Poorly differentiated No information
Stage II
of potentially curatively resected
Stage I
Stage 11
Stage III
I1
%I
n
(’ 0
n
9c
38 88
19.9 46.1
3 58
2.3 45.3
5 76
3.2 48.4
56 9
29.3 4.7
38 29
29.7 22.7
51 25
32.5 15.9
Table 3. Mean survival times and 5-year actuarial survivals according to tumor stage Stage
n
Mean survival time (years)
5-year actuarial survival (% )
I
191 128 157
7.32 2.42 1.22
59.9 34.2 16.1
II IIIa
Patients with moderately differentiated squamous cell carcinomas lived significantly longer than patients with poorly differentiated tumors after complete resection (mean survival time 17.7 months vs 10.1 months) only in the tumor stage III a group (P < 0.05). Considering the tumors in terms of their localization in the right or left side, there was no significant difference with respect to survival time within the same tumor stage group. The type of resection treatment (Table 1) had no effect on the survival times within each tumor stage group. In regard to prognosis there were clear differences between the various tumor stage groups. The differences in mean survival times between patients with stage I, stage II, and stage III a tumors are significant (P < 0.01; Table 3). In analyzing the influence of the size of the tumor (T) and lymph node involvement (N) on the prognosis following potentially curative surgery. one variable (T or N) was kept constant. Among patients with consistently absent lymph node involvement (NO) the mean survival times of those with Tl and T2 tumors differed significantly from those with T3 tumors (Table 4). Spreading of the tumor to the chest wall, diaphragm, mediastinal pleura, and pericardium, or a growth in the main bronchus less than 2 cm distal to the carina but without involvement of the carina. or associated atelectasis or
500
Table 4. Mean survival times and 5-year survival rates in patients with constant N and variable T classification
Table 5. Mean survival times and 5-year survival rates in patients with constant T and variable N classification
Tumor classification
n
Mean survival time (months)
S-year actuarial survival (%)
Tumor classification
TIN0 T2NO T3NO
64 127 37
122.7 84.4 11.1
67.6 55.2 23.1
TIN0 TIN1 TIN2
TlNl T2Nl T3Nl
28 100 25
28.0 28.8 13.2
38.9 32.6 21.6
TIN2 T2N2 T3N2
13 59 23
21.0 16.5 9.9
23.1 6.8 0
obstructive pneumonitis of the entire lung resulted in a marked deterioration of the prognosis (P < 0.01). Nl disease resulted in a significant difference between the mean survival times of patients with Tl or T2 tumors and those with T3 (P < 0.05; Table 4). In the presence of mediastinal lymph node metastases (N2 disease), the prognosis of patients with Tl tumors is significantly better than that of patients with T3 tumors (P < 0.05). Classifying by tumor size (Table 5), a significant dependence of the prognosis on the lymph node stage can be demonstrated for T2 tumors; the mean survival times show significant differences between T2N0, T2N1, and T2N2 tumors (P < 0.01). By contrast, in Tl disease there were only significant differences between TIN0 and TIN1 (P < 0.05) and again between TIN0 and TIN2 (P < 0.01). The mean survival times of patients with TIN1 disease did not differ significantly from those of patients with TlN2 disease. In the patients with T3 tumors, the mean survival times were approximately the same irrespective of lymph node involvement. It was noticeable, however, that none of the patients with mediastinal lymph node metastases survived 5 years (Table 5). Nl disease is defined as metastases in the ipsilateral peribronchial lymph nodes and/or in the ipsilateral hilus lymph nodes, including direct spread of the primary tumor. Table 6 subdivides the T2Nl classification into true lymph node metastases and direct lymph node infiltration by the tumor. The differences in the mean survival times between the two groups are significant (P < 0.05). Comparing the mean survival times of patients with N2 metastases, there was a significant difference between the N2 localizations “tracheobronchial” (n = 13) and “superior mediastinal” (n = 39) (21.6 vs 12.3 months, respectively; P < 0.05). Subdividing the N2 cases into Tl, T2, and T3 showed no further significant difference. In 157 patients with stage III a tumors who had undergone potentially curative resection (Table 7), we found that the long-term prognosis tended to be worse when the mediastinal lymph nodes were involved. The 5-year actuarial survival of patients with T3NO and T3Nl disease differs from patients with T2N2 and T3N2 disease. In stage III a, however, tumor size and extent was also important (Table 7): patients with TlN2 tumors had a
n
Mean survival time (months)
5-year actuarial survival (%)
64 28 13
122.7 28.0 21.0
67.6 38.9 23.1
T2NO T2Nl T2N2
127 100 59
84.4 28.8 16.5
55.2 32.6 6.8
T3NO T3Nl T3N2
37 25 23
11.1 13.2 9.9
23.1 21.6 0
Table 6. Mean survival times and 5-year actuarial survival rates in patients with true lymph node metastases and those with direct lymph node infiltration in T2Nl disease Tumor classification
n
Mean survival time (months)
5-year actuarial survival (%)
T2Nl (infiltration) T2Nl (metastases)
29 71
43.3 27.5
47.3 27.6
Table 7. Five-year actuarial survival in patients curatively resected stage IIIa tumors
with potentially
Tumor classification
n
5-year actuarial survival (%)
T3NO T3Nl TIN2 T2N2 T3N2
37 25 13 59 23
23.1 27.7 23.1 6.8 0
Table 8. Mean survival times in patients with potentially curatively resected T3 tumors Tumor in the main bronchus less than 2 cm distal to the carina
T3NO T3NO/Nl T3NO-2
Tumor directly invading chest wall, diaphragm, mediastinal pleura pericardium
n
Mean survival time (months)
n
Mean survival time (months)
17 27 36
22 27 16
20 35 49
8.5 7.0 7.5
higher 5-year actuarial survival than patients with T2N2 and T3N2, and they also had a significantly longer mean survival time than patients with T3N2 tumors (P < 0.05). In total, 85 T3 tumors were potentially curatively resected. Table 8 shows these patients divided into two groups: those in whom tumor was in the main bronchus less than 2 cm distal to the carina and those in whom tumor infiltrated the chest wall, diaphragm, mediastinal pleura, or pericardium. The mean survival times in pa-
Table 9. Tumor classifications of patients 10 years or more after surgery Tumor classification
n
TlNO T2NO T2Nl TIN2 T3NO T3Nl
5 14 8
who have survived for
1 2 1 31
tients with T3NO-2 disease in whom the tumor had grown closer than 2 cm to the main carina without invading it were significantly longer than those of patients with T3NO-2 tumors in whom the tumor had infiltrated the adjacent structures (Table 8). The true i.e. not calculated, IO-year survival rate in our patients after potentially curative surgery is 19.1% (31 out of 162 patients who were operated on between 1976 and 1980). Table 9 shows these long-term survivors according to TNM classification. Discussion Surgical resection is the therapy of choice in squamous cell carcinoma of the bronchus. The study of Enn Noti [I 61 describes the natural course of this disease. The mean survival time of patients with squamous cell carcinoma was 7.1 months and the treatments given, excluding surgical therapy. could be regarded as only palliative and as having practically no influence on survival. We found comparable poor short- and long-term results in our patient population after exploratory thoracotomy and incomplete resection (mean survival time 7 and 9 months respectively, 5-year survival rate 0%). Only when tumorous invasion in the distal cut end of the bronchial stump was identified by postoperative microscopy could the mean survival time be significantly increased by means of external irradiation; however, 5 years’ survival could not be achieved. Naruke et al. [15] also showed improved prognosis for those patients in whom microscopically identified tumorous invasion of the bronchial stump was irradiated, whereas Law et al. [5] and McCaughan et al. [lo] saw no positive influence of external irradiation in such cases. The prognosis for patients with squamous cell carcinoma following potentially curative resection treatment seems above all to depend on the lymph node status, i.e., the N factor in the TNM system [3, 4, 8, 12, 171. In our study this was shown by the fact that, especially in the T2 tumor group, the step from NO to Nl and from Nl and N2 significantly worsened the prognosis in each instance, just as the prognosis for patients with TlNl and TIN2 diseases was significantly worse than that for patients with TlNO. By contrast, there was no significant difference between the mean survival times for patients with TlNl and Tl N2. The numbers of patients with these two conditions were relatively small (28 and 13, respectively),
however, so this could be a reason for the lack of demonstrated significant difference. Martini et al. (93 showed that the prognosis does not depend upon the specific level or site of lymph node metastasis within the Nl region, nor on whether gross or microscopic involvement is present. We found significantly longer survival in cases of T2Nl tumor when there is no true metastatic spread to the lymph nodes, but only infiltration. In the case of N2 disease, patients with the “N2 tracheobronchial” variant had a significantly longer mean survial time than those with the “N2 superior mediastinal” variant in our study. After subdividing these N2 patients according to Tl, T2, and T3 disease, we saw no further significant differences, among other things probably because the groups had become too small. Naruke et al. [14], on the other hand, found a significantly poorer prognosis for patients with subcarinal lymph node metastases within the N2 disease group. The North American Lung Cancer Study Group [6] has chosen to define one criterion, among others, of an incomplete resection as the presence of metastatic tumor in the highest resected ipsilateral superior mediastinal node, even if no gross or microscopic residual disease is documented as remaining in the ipsilateral hemithorax. Pearson et al. [18] reported a significantly poorer longterm prognosis for patients in whom mediastinal lymph node involvement could be verified by mediastinoscopy. In a retrospective study by the Lung Cancer Study Group of 1988 [7] concerning N2 disease in non-smallcell bronchial carcinoma, paired comparison of the survival rates of patients with variously located mediastinal lymph node involvement demonstrated the highest death rate in the patients with metastases in several mediastinal lymph node locations. The prognostic influence of the T tactor in our study was found in a significant difference between the mean survival time of the Tl-2N0 group and that of the T3NO group, as well as between that of the Tl-2Nl group and that of the T3Nl group. The T factor also had prognostic value in stage III a tumors: patients with TlN2 tumors have better short- and long-term prognoses than patients with T3N2 tumors. Similar results have been reported by other authors [I 1, 15, 211. In our study there was a significant difference in mean survival times within the T3 group, independent of lymph node status. when the tumors were divided into those in close proximity to the main carina and those invading the adjacent structures (mean survival times 16 vs. 7.5 months respectively). Nakahashi et al. [13#] found the 4-year survival of five patients with T3 non-small-cell lung cancer with invasion of the main bronchus to be 80%. The 5-year survival for patients with involvement of adjacent structures was 17%. The staging of the fourth edition (1987) of the UICC TNM classification was confirmed in our study with regard to its prognostic information: the mean survival times of patients with stage I, II, and III a tumors differed significantly. The true IO-year survival rate was 19%. The patients (31 out of 162) had tumors in the following stages: I: 19 patients: II: 8 patients; III a: 4 patients.
502
The patients with tumors in stage III a are an inhomogeneous group in terms of their long-term prognosis. This has been described by McCaughan et al. [lo], Trastek et al. [20], and Naruke et al. [15]. In our study, the patients with T3N0, T3N1, and TlN2 tumors had 5-year survival rates of 21.6-23.1%, while the 5-year survival rates for patients with T3N2 and T2N2 were 0% and 6.8%, respectively. Of the patients who survived 10 years, two had T3NO tumors, one a T3N1, and another a TlN2 tumor. The nature of resection treatment (lobectomy or pneumonectomy) after potentially curative operation of squamous cell carcinoma had no effect on the prognosis within a given tumor stage in our patient population. By contrast, Deneffe et al. [2] found the 5-year survival rate tended to drop in patients if a pneumonectomy was performed. The age of the patient at the time of operation also had no significant influence on the mean survival times in our patient population. Besides tumor size and lymph node metastases, we investigated other possible factors which might affect the prognosis after potentially curative resection of a squamous cell carcinoma of the bronchus. A significant difference in the prognosis between tumors with moderate differentiation and those with poor differentiation was found only in stage IIIa tumors. Trastek et al. [20] made a similar observation in patients with T3 tumors. Other factors, e.g., infiltration of tumor cells into vessels, tendency to keratinize, or carcinomatous lymphangiosis, did not affect the prognosis in our investigation, whereas other authors have attributed significance to such factors
VI. References 1. Chung CK, Zaino R, Stryker JA, O’Neil M Jr, DeMuth WE Jr (1982) Carcinoma of the lung: evaluation of histological grade and factors influencing prognosis. Ann Thorac Surg 33: 599604 2. Deneffe G, Lacquet LM, Verbeken E, Vermaut G (1988) Surgical treatment of bronchogenic carcinoma: a retrospective study of 720 thoracotomies. Ann Thorac Surg 45: 380-383 3. Greschuchna D. Maassen W (1982) Stadieneinteilung und Ergebnisse der operativen Behandlung des Bronchialkarzinoms. Prax Pneumol 36: 281-284 4. Immerman SC, Vanecko RM, Fry WA. Head LR, Shields TW (1981) Site of recurrence in patients with stages I and II carcinoma of the lung resected for cure. Ann Thorac Surg 32: 23 21 5. Law MR, Henk JM, Lennox SC, Hodson ME (1982) Value of radiotherapy for tumour on the bronchial stump after resection of bronchial carcinoma. Thorax 37: 496-499
6.
7.
8.
9.
10.
11.
12. 13.
14.
15.
16. 17.
18.
19. 20.
21.
Lung Cancer Study Group (1988) The benefit of adjuvant treatment for resected locally advanced non-small cell lung cancer. J Clin Oncol 6: 9- 17 Lung Cancer Study Group (1988) Should subcarinal lymph nodes be routinely examined in patients with non small cell lung cancer? J Thorac Cardiovasc Surg 95: 883-887 Martini N, Flehinger BJ, Zaman MB, Beattie EJ Jr (1980) Prospective study of 445 lung carcinomas with mediastinal lymph-node metastases. J Thorac Cardiovasc Surg 80: 390-399 Martini N, Flehinger BJ, Nagasaki F, Hart B (1983) Prognostic significance of Nl disease in carcinoma of the lung. J Thorac Cardiovasc Surg 86: 646-653 McCaughan B, Martini N, Bains MS, McCormack P (1985) Chest wall invasion in carcinoma of the lung. J Thorac Cardiovast Surg 89: 836-841 Merkle NM, Biilzebruck H, Probst G, Vogt-Moykopf I(l988) Das Bronchialkarzinom mit mediastinalen Lymphknotenmetastasen - eine Operationsindikation? Prax Klin Pneumol 42: 360-362 Mountain CF (1977) Assessment of the role of surgery for control of lung cancer. Ann Thorac Surg 24: 365-373 Nakahashi H, Yasumoto K, Ishida T, Nagashima A, Nishino T, Oka T, Sugimachi K (1988) Results of surgical treatment of patients with T3 non-small cell lung cancer. Ann Thorac Surg 46:178-181 Naruke T, Suemasu K, Ishikawa S (1978) Lymph-node mapping and curability at various levels of metastasis in resected lung cancer. J Thorac Cardiovasc Surg 76: 832-839 Naruke T, Goya T, Tsuchiya R, Suemasu K (1988) The importance of surgery to non-small cell carcinoma of lung with mediastinal lymph-node metastasis. Ann Thorac Surg 46: 603 -610 Noti E (1984) The natural five-year course in bronchial carcinoma. Cancer 53: 221 I-2216 Pairolero PC, Williams DE, Bergstralh EJ, Piehler JM, Bernatz PE, Payne WS (1984) Postsurgical stage I bronchogenic carcinoma: morbid implications of recurrent disease. Ann Thorac Surg 38: 331-336 Pearson FG, Delarue NC, Ilves R, Todd TR, Cooper JD (1982) Significance of positive superior mediastinal nodes identified at mediastinoscopy in patients with resectable cancer of the lung. J Thorac Cardiovasc Surg 83: 1- 11 Shields TW (1989) The “incomplete” resection. Ann Thorac Surg 47: 487-488 Trastek VF, Pairolero PC, Piehler JM, Weiland LH, O’Brien PC, Payne WS, Bernatz PE (1984) En bloc (non-chest wall) resection for bronchogenic carcinoma with parietal fixation. J Thorac Cardiovasc Surg 87: 352-358 Watanabe Y, Shimizu J, Oda M, Hayashi Y, Watanabe S, Iwa T (1991) Results of surgical treatment in patients with stage IIIa non-small-cell lung cancer. Thorac Cardiovasc Surg 39: 44-49
Dr H. Huwer and Priv.-Doz. Dr. I. Volkmer Klinik fur Thorax-, Herz- und Gefasschirurgie Universitat des Saarlandes W-6650 Homburg/Saar Federal Republic of Germany
0 SpringeTVerlag i992
Squamous cell carcinoma of the lung: prognostic factors affecting the long-term outcome after surgical resection H. Huwer, I. Volkmer, A. Wiegand, G. Kalweit, and P. Feindt Department of Thoracic and Cardiovascular Surgery, Surgery Division, Saarland University Medical School, Homburg, FRG
Abstract. From 1976 to 1989,638 patients with squamous cell carcinoma were operated on. Reliable information concerning survival could be obtained in 540 of these cases via follow-up examinations. All tumors were retrospectively classified according to the TNM system of the fourth edition of UICC classification. The average survival time after potentially curative operation was 7.32 years in patients with stage I tumors, the 5-year survival rate was 55.3%. The corresponding values were 2.42 years and 29.1% respectively in patients with stage II tumors, compared to 1.22 years and 12.0% respectively in those with stage III a tumors. The differences between tumor stages are statistically significant. Analyzing the influence of the T and N factors on the prognosis following potentially curative surgery, the classification in the various tumor stages was largely verified. Stage III a is inhomogeneous in terms of prognosis. T and N factors showed prognostic value in this group. In the group with T3 tumors there was a significant difference in the mean survival times, independently of the lymph node status, when the patients were divided into those with T3 tumors close to the main carina (16 months) and those who had infiltrated adjacent structures (7.5 months). When differentiating the Nl location, a significant difference was found between the group of patients with T2 tumors and true lymph node metastases and the group with T2 and direct tumor infiltration of the lymph node (mean survival time 27.5 vs 42.3 months). In cases with N2 disease, prognostic differences were shown in as much as the mean survival time in patients with superior mediastinal N2 was significantly shorter than in those with only tracheobronchial lymph node metastases (12.3 vs 21.6 months). The age of the patients and type of resection treatment within the same tumor stage group had no influence on the prognosis. Tendency to keratinize, carcinomatous lymphangiosis, and tumorous invasion of vessels also had no prognostic value in any given stage. Stage III a patients with poorly differentiated tumors had a worse prognosis than those with moderately differentiated ones. Following incomplete resection or exploratory thoracotomy, the short- und long-term prognoses were equally bad (mean survival time 9 and 7 months, respectively, and 5-year survival rate 0%). Mean survival time could be prolonged by external irradiation only when tumorous infiltration in the distal cut end of the bronchial stump was microscopically ascertained (13.0 months vs 7.3 months without irradiation treatment). However, there was no 5-year survivor in either group. [Eur J Cardio-thorac Surg (1992) 6:498-5021 Key words: Squamous
cell carcinoma
of the bronchus
- Staging
Surgical resection is the therapy of choice in patients with squamous cell carcinoma of the bronchus. The prognosis
- Prognostic
factors
of the TNM system and additional investigated in the present study.
parameters
will be
of surgically treated patients appears to depend primarily on the extent of the tumor and the lymph node metastases. However, the weighting of these factors is still in a state of flux, especially since a uniform worldwide scheme for TNM classification and staging of bronchial carcinomas, enabling comprehensive comparisons, was only published in 1987 in the fourth edition of the UICC classification. The prognostic significance of the parameters Received for publication: Accepted
for publication:
December 9, 1991 March 26, 1992
Patients and methods From 1976 to 1989, 638 patients with squamous cell carcinoma of the bronchus were operated on. A follow-up investigation was possible in 540 of those patients. For.direct comparison, retrospective classification of the tumors in accordance with the TNM scheme of the fourth edition of the UICC classification was carried out in all patients; there was no reevaluation of the histologic material. Incomplete resection is defined as one documented by histological proof of residual tumor at one of the margins of the resected
specimen. All other resections were considered complete and poten-
tially curative (191. The mean age of the patients. of whom 516 were men and 24 women, was 61 years at the time of operation. Tumor locations were equally distributed between sides (left lung 262, right lung 278). Table 1 shows the distribution of the kinds of curative resection according to tumor stage. After potentially curative operation. in the group of stage I tumors there were 31 keratinizing squamous cell carcinomas (16.2%), 45 partly keratinizing (23.6%), and 85 non-keratinizing (44.5%). In the stage II group there were 19 keratinizing (14.8%). 34 partly keratinizing (26.6%). and 49 non-keratinizing (38.3%) squamous cell carcinomas and in the stage III a group there were 30 keratinizing ( 19.1%), 45 partly keratinizing (28.7%), and 54 nonkeratinizing (34.4%) squamous cell carcinomas. There was no mention of keratinization tendency in 30 cases (15.7%) in stage I, 26 (20.3%) in stage II. and 28 (17.8%) in stage III a. Carcinomatous lymphangiosis was present in 21 stage I patients (I 1.O%), 35 (27.3%) stage II, and 42 (26.8%) stage III a. Invasion of blood vessels was recorded 7 times (3.7%) in stage I, 9 times (7.0%) in stage II. and 11 times (7.0%) in stage III a. Table 2 shows the degrees of differentiation of potentially curatively resected tumors. Lymph node dissection in every patient was carried out by removing preparatively visualized mediastinal lymph node groups. Complete skeletization of the mediastinal structures was not performed. From 1976 to 1985 each patient was examined preoperatively by mediastinoscopy. In patients with mediastinal lymph node involvement, postoperative irradiation was carried out routinely. After incomplete resection. postoperative irradiation was performed occasionally. Tumorous infiltration in the distal cut end of the bronchial stump was found microscopically following otherwise potentially curative intervention in 28 patients. Seventeen of these underwent postoperative irradiation. The survival rates were determined by means of actuarial analysis and compared for statistical significance (P < 0.05) by the K sample test
Results Out of the 540 patients with complete follow-up, 476 (i.e., 88.2%) underwent potentially curative resection. Fortyeight patients had an incomplete resection (8.9%) and 16 an exploratory thoracotomy (2.9%). The mean survival times for these two groups were 9 and 7 months, respectively, and were not significantly different. Incomplete resection because of infiltration of the distal cut end of the bronchial stump was identified microscopically in 28 patients. Seventeen of these patients underwent irradiation therapy postoperatively, while the remaining 11 had no further treatment. The mean survival time in postoperatively irradiated patients was 13 months, significantly higher than that in the 11 non-treated patients (7.3 months). No patient in either group survived 5 years. After potentially curative operation the 30-day mortality for all types of resection and all tumor stages was 7.2%: 8.4% after pneumonectomy and 5.8% after lobectomy and bilobectomy. Comparing the survival times in the various age groups without regard to tumor stage, there was no significant difference between the groups. After potentially curative resection treatment, tendency to keratinize. carcinomatous lymphangiosis, and tumorous infiltration into vessels did not affect the survival times within each tumor stage group.
Table 1. Distribution of the kinds of curative resection among the various tumor stages Stage I
Lobectomy Bilobectomy Pneumonectomy
Stage III
12
%
,1
%
n
%
114 18 59
59.7 9.4 30.9
48 8 72
37.5 6.3 56.2
33 5 119
21.0 3.2 75.8
Table 2. Degrees of differentiation tumors
Well differentiated Moderately differentiated Poorly differentiated No information
Stage II
of potentially curatively resected
Stage I
Stage 11
Stage III
I1
%I
n
(’ 0
n
9c
38 88
19.9 46.1
3 58
2.3 45.3
5 76
3.2 48.4
56 9
29.3 4.7
38 29
29.7 22.7
51 25
32.5 15.9
Table 3. Mean survival times and 5-year actuarial survivals according to tumor stage Stage
n
Mean survival time (years)
5-year actuarial survival (% )
I
191 128 157
7.32 2.42 1.22
59.9 34.2 16.1
II IIIa
Patients with moderately differentiated squamous cell carcinomas lived significantly longer than patients with poorly differentiated tumors after complete resection (mean survival time 17.7 months vs 10.1 months) only in the tumor stage III a group (P < 0.05). Considering the tumors in terms of their localization in the right or left side, there was no significant difference with respect to survival time within the same tumor stage group. The type of resection treatment (Table 1) had no effect on the survival times within each tumor stage group. In regard to prognosis there were clear differences between the various tumor stage groups. The differences in mean survival times between patients with stage I, stage II, and stage III a tumors are significant (P < 0.01; Table 3). In analyzing the influence of the size of the tumor (T) and lymph node involvement (N) on the prognosis following potentially curative surgery. one variable (T or N) was kept constant. Among patients with consistently absent lymph node involvement (NO) the mean survival times of those with Tl and T2 tumors differed significantly from those with T3 tumors (Table 4). Spreading of the tumor to the chest wall, diaphragm, mediastinal pleura, and pericardium, or a growth in the main bronchus less than 2 cm distal to the carina but without involvement of the carina. or associated atelectasis or
500
Table 4. Mean survival times and 5-year survival rates in patients with constant N and variable T classification
Table 5. Mean survival times and 5-year survival rates in patients with constant T and variable N classification
Tumor classification
n
Mean survival time (months)
S-year actuarial survival (%)
Tumor classification
TIN0 T2NO T3NO
64 127 37
122.7 84.4 11.1
67.6 55.2 23.1
TIN0 TIN1 TIN2
TlNl T2Nl T3Nl
28 100 25
28.0 28.8 13.2
38.9 32.6 21.6
TIN2 T2N2 T3N2
13 59 23
21.0 16.5 9.9
23.1 6.8 0
obstructive pneumonitis of the entire lung resulted in a marked deterioration of the prognosis (P < 0.01). Nl disease resulted in a significant difference between the mean survival times of patients with Tl or T2 tumors and those with T3 (P < 0.05; Table 4). In the presence of mediastinal lymph node metastases (N2 disease), the prognosis of patients with Tl tumors is significantly better than that of patients with T3 tumors (P < 0.05). Classifying by tumor size (Table 5), a significant dependence of the prognosis on the lymph node stage can be demonstrated for T2 tumors; the mean survival times show significant differences between T2N0, T2N1, and T2N2 tumors (P < 0.01). By contrast, in Tl disease there were only significant differences between TIN0 and TIN1 (P < 0.05) and again between TIN0 and TIN2 (P < 0.01). The mean survival times of patients with TIN1 disease did not differ significantly from those of patients with TlN2 disease. In the patients with T3 tumors, the mean survival times were approximately the same irrespective of lymph node involvement. It was noticeable, however, that none of the patients with mediastinal lymph node metastases survived 5 years (Table 5). Nl disease is defined as metastases in the ipsilateral peribronchial lymph nodes and/or in the ipsilateral hilus lymph nodes, including direct spread of the primary tumor. Table 6 subdivides the T2Nl classification into true lymph node metastases and direct lymph node infiltration by the tumor. The differences in the mean survival times between the two groups are significant (P < 0.05). Comparing the mean survival times of patients with N2 metastases, there was a significant difference between the N2 localizations “tracheobronchial” (n = 13) and “superior mediastinal” (n = 39) (21.6 vs 12.3 months, respectively; P < 0.05). Subdividing the N2 cases into Tl, T2, and T3 showed no further significant difference. In 157 patients with stage III a tumors who had undergone potentially curative resection (Table 7), we found that the long-term prognosis tended to be worse when the mediastinal lymph nodes were involved. The 5-year actuarial survival of patients with T3NO and T3Nl disease differs from patients with T2N2 and T3N2 disease. In stage III a, however, tumor size and extent was also important (Table 7): patients with TlN2 tumors had a
n
Mean survival time (months)
5-year actuarial survival (%)
64 28 13
122.7 28.0 21.0
67.6 38.9 23.1
T2NO T2Nl T2N2
127 100 59
84.4 28.8 16.5
55.2 32.6 6.8
T3NO T3Nl T3N2
37 25 23
11.1 13.2 9.9
23.1 21.6 0
Table 6. Mean survival times and 5-year actuarial survival rates in patients with true lymph node metastases and those with direct lymph node infiltration in T2Nl disease Tumor classification
n
Mean survival time (months)
5-year actuarial survival (%)
T2Nl (infiltration) T2Nl (metastases)
29 71
43.3 27.5
47.3 27.6
Table 7. Five-year actuarial survival in patients curatively resected stage IIIa tumors
with potentially
Tumor classification
n
5-year actuarial survival (%)
T3NO T3Nl TIN2 T2N2 T3N2
37 25 13 59 23
23.1 27.7 23.1 6.8 0
Table 8. Mean survival times in patients with potentially curatively resected T3 tumors Tumor in the main bronchus less than 2 cm distal to the carina
T3NO T3NO/Nl T3NO-2
Tumor directly invading chest wall, diaphragm, mediastinal pleura pericardium
n
Mean survival time (months)
n
Mean survival time (months)
17 27 36
22 27 16
20 35 49
8.5 7.0 7.5
higher 5-year actuarial survival than patients with T2N2 and T3N2, and they also had a significantly longer mean survival time than patients with T3N2 tumors (P < 0.05). In total, 85 T3 tumors were potentially curatively resected. Table 8 shows these patients divided into two groups: those in whom tumor was in the main bronchus less than 2 cm distal to the carina and those in whom tumor infiltrated the chest wall, diaphragm, mediastinal pleura, or pericardium. The mean survival times in pa-
Table 9. Tumor classifications of patients 10 years or more after surgery Tumor classification
n
TlNO T2NO T2Nl TIN2 T3NO T3Nl
5 14 8
who have survived for
1 2 1 31
tients with T3NO-2 disease in whom the tumor had grown closer than 2 cm to the main carina without invading it were significantly longer than those of patients with T3NO-2 tumors in whom the tumor had infiltrated the adjacent structures (Table 8). The true i.e. not calculated, IO-year survival rate in our patients after potentially curative surgery is 19.1% (31 out of 162 patients who were operated on between 1976 and 1980). Table 9 shows these long-term survivors according to TNM classification. Discussion Surgical resection is the therapy of choice in squamous cell carcinoma of the bronchus. The study of Enn Noti [I 61 describes the natural course of this disease. The mean survival time of patients with squamous cell carcinoma was 7.1 months and the treatments given, excluding surgical therapy. could be regarded as only palliative and as having practically no influence on survival. We found comparable poor short- and long-term results in our patient population after exploratory thoracotomy and incomplete resection (mean survival time 7 and 9 months respectively, 5-year survival rate 0%). Only when tumorous invasion in the distal cut end of the bronchial stump was identified by postoperative microscopy could the mean survival time be significantly increased by means of external irradiation; however, 5 years’ survival could not be achieved. Naruke et al. [15] also showed improved prognosis for those patients in whom microscopically identified tumorous invasion of the bronchial stump was irradiated, whereas Law et al. [5] and McCaughan et al. [lo] saw no positive influence of external irradiation in such cases. The prognosis for patients with squamous cell carcinoma following potentially curative resection treatment seems above all to depend on the lymph node status, i.e., the N factor in the TNM system [3, 4, 8, 12, 171. In our study this was shown by the fact that, especially in the T2 tumor group, the step from NO to Nl and from Nl and N2 significantly worsened the prognosis in each instance, just as the prognosis for patients with TlNl and TIN2 diseases was significantly worse than that for patients with TlNO. By contrast, there was no significant difference between the mean survival times for patients with TlNl and Tl N2. The numbers of patients with these two conditions were relatively small (28 and 13, respectively),
however, so this could be a reason for the lack of demonstrated significant difference. Martini et al. (93 showed that the prognosis does not depend upon the specific level or site of lymph node metastasis within the Nl region, nor on whether gross or microscopic involvement is present. We found significantly longer survival in cases of T2Nl tumor when there is no true metastatic spread to the lymph nodes, but only infiltration. In the case of N2 disease, patients with the “N2 tracheobronchial” variant had a significantly longer mean survial time than those with the “N2 superior mediastinal” variant in our study. After subdividing these N2 patients according to Tl, T2, and T3 disease, we saw no further significant differences, among other things probably because the groups had become too small. Naruke et al. [14], on the other hand, found a significantly poorer prognosis for patients with subcarinal lymph node metastases within the N2 disease group. The North American Lung Cancer Study Group [6] has chosen to define one criterion, among others, of an incomplete resection as the presence of metastatic tumor in the highest resected ipsilateral superior mediastinal node, even if no gross or microscopic residual disease is documented as remaining in the ipsilateral hemithorax. Pearson et al. [18] reported a significantly poorer longterm prognosis for patients in whom mediastinal lymph node involvement could be verified by mediastinoscopy. In a retrospective study by the Lung Cancer Study Group of 1988 [7] concerning N2 disease in non-smallcell bronchial carcinoma, paired comparison of the survival rates of patients with variously located mediastinal lymph node involvement demonstrated the highest death rate in the patients with metastases in several mediastinal lymph node locations. The prognostic influence of the T tactor in our study was found in a significant difference between the mean survival time of the Tl-2N0 group and that of the T3NO group, as well as between that of the Tl-2Nl group and that of the T3Nl group. The T factor also had prognostic value in stage III a tumors: patients with TlN2 tumors have better short- and long-term prognoses than patients with T3N2 tumors. Similar results have been reported by other authors [I 1, 15, 211. In our study there was a significant difference in mean survival times within the T3 group, independent of lymph node status. when the tumors were divided into those in close proximity to the main carina and those invading the adjacent structures (mean survival times 16 vs. 7.5 months respectively). Nakahashi et al. [13#] found the 4-year survival of five patients with T3 non-small-cell lung cancer with invasion of the main bronchus to be 80%. The 5-year survival for patients with involvement of adjacent structures was 17%. The staging of the fourth edition (1987) of the UICC TNM classification was confirmed in our study with regard to its prognostic information: the mean survival times of patients with stage I, II, and III a tumors differed significantly. The true IO-year survival rate was 19%. The patients (31 out of 162) had tumors in the following stages: I: 19 patients: II: 8 patients; III a: 4 patients.
502
The patients with tumors in stage III a are an inhomogeneous group in terms of their long-term prognosis. This has been described by McCaughan et al. [lo], Trastek et al. [20], and Naruke et al. [15]. In our study, the patients with T3N0, T3N1, and TlN2 tumors had 5-year survival rates of 21.6-23.1%, while the 5-year survival rates for patients with T3N2 and T2N2 were 0% and 6.8%, respectively. Of the patients who survived 10 years, two had T3NO tumors, one a T3N1, and another a TlN2 tumor. The nature of resection treatment (lobectomy or pneumonectomy) after potentially curative operation of squamous cell carcinoma had no effect on the prognosis within a given tumor stage in our patient population. By contrast, Deneffe et al. [2] found the 5-year survival rate tended to drop in patients if a pneumonectomy was performed. The age of the patient at the time of operation also had no significant influence on the mean survival times in our patient population. Besides tumor size and lymph node metastases, we investigated other possible factors which might affect the prognosis after potentially curative resection of a squamous cell carcinoma of the bronchus. A significant difference in the prognosis between tumors with moderate differentiation and those with poor differentiation was found only in stage IIIa tumors. Trastek et al. [20] made a similar observation in patients with T3 tumors. Other factors, e.g., infiltration of tumor cells into vessels, tendency to keratinize, or carcinomatous lymphangiosis, did not affect the prognosis in our investigation, whereas other authors have attributed significance to such factors
VI. References 1. Chung CK, Zaino R, Stryker JA, O’Neil M Jr, DeMuth WE Jr (1982) Carcinoma of the lung: evaluation of histological grade and factors influencing prognosis. Ann Thorac Surg 33: 599604 2. Deneffe G, Lacquet LM, Verbeken E, Vermaut G (1988) Surgical treatment of bronchogenic carcinoma: a retrospective study of 720 thoracotomies. Ann Thorac Surg 45: 380-383 3. Greschuchna D. Maassen W (1982) Stadieneinteilung und Ergebnisse der operativen Behandlung des Bronchialkarzinoms. Prax Pneumol 36: 281-284 4. Immerman SC, Vanecko RM, Fry WA. Head LR, Shields TW (1981) Site of recurrence in patients with stages I and II carcinoma of the lung resected for cure. Ann Thorac Surg 32: 23 21 5. Law MR, Henk JM, Lennox SC, Hodson ME (1982) Value of radiotherapy for tumour on the bronchial stump after resection of bronchial carcinoma. Thorax 37: 496-499
6.
7.
8.
9.
10.
11.
12. 13.
14.
15.
16. 17.
18.
19. 20.
21.
Lung Cancer Study Group (1988) The benefit of adjuvant treatment for resected locally advanced non-small cell lung cancer. J Clin Oncol 6: 9- 17 Lung Cancer Study Group (1988) Should subcarinal lymph nodes be routinely examined in patients with non small cell lung cancer? J Thorac Cardiovasc Surg 95: 883-887 Martini N, Flehinger BJ, Zaman MB, Beattie EJ Jr (1980) Prospective study of 445 lung carcinomas with mediastinal lymph-node metastases. J Thorac Cardiovasc Surg 80: 390-399 Martini N, Flehinger BJ, Nagasaki F, Hart B (1983) Prognostic significance of Nl disease in carcinoma of the lung. J Thorac Cardiovasc Surg 86: 646-653 McCaughan B, Martini N, Bains MS, McCormack P (1985) Chest wall invasion in carcinoma of the lung. J Thorac Cardiovast Surg 89: 836-841 Merkle NM, Biilzebruck H, Probst G, Vogt-Moykopf I(l988) Das Bronchialkarzinom mit mediastinalen Lymphknotenmetastasen - eine Operationsindikation? Prax Klin Pneumol 42: 360-362 Mountain CF (1977) Assessment of the role of surgery for control of lung cancer. Ann Thorac Surg 24: 365-373 Nakahashi H, Yasumoto K, Ishida T, Nagashima A, Nishino T, Oka T, Sugimachi K (1988) Results of surgical treatment of patients with T3 non-small cell lung cancer. Ann Thorac Surg 46:178-181 Naruke T, Suemasu K, Ishikawa S (1978) Lymph-node mapping and curability at various levels of metastasis in resected lung cancer. J Thorac Cardiovasc Surg 76: 832-839 Naruke T, Goya T, Tsuchiya R, Suemasu K (1988) The importance of surgery to non-small cell carcinoma of lung with mediastinal lymph-node metastasis. Ann Thorac Surg 46: 603 -610 Noti E (1984) The natural five-year course in bronchial carcinoma. Cancer 53: 221 I-2216 Pairolero PC, Williams DE, Bergstralh EJ, Piehler JM, Bernatz PE, Payne WS (1984) Postsurgical stage I bronchogenic carcinoma: morbid implications of recurrent disease. Ann Thorac Surg 38: 331-336 Pearson FG, Delarue NC, Ilves R, Todd TR, Cooper JD (1982) Significance of positive superior mediastinal nodes identified at mediastinoscopy in patients with resectable cancer of the lung. J Thorac Cardiovasc Surg 83: 1- 11 Shields TW (1989) The “incomplete” resection. Ann Thorac Surg 47: 487-488 Trastek VF, Pairolero PC, Piehler JM, Weiland LH, O’Brien PC, Payne WS, Bernatz PE (1984) En bloc (non-chest wall) resection for bronchogenic carcinoma with parietal fixation. J Thorac Cardiovasc Surg 87: 352-358 Watanabe Y, Shimizu J, Oda M, Hayashi Y, Watanabe S, Iwa T (1991) Results of surgical treatment in patients with stage IIIa non-small-cell lung cancer. Thorac Cardiovasc Surg 39: 44-49
Dr H. Huwer and Priv.-Doz. Dr. I. Volkmer Klinik fur Thorax-, Herz- und Gefasschirurgie Universitat des Saarlandes W-6650 Homburg/Saar Federal Republic of Germany
