- Email: [email protected]
Surgical treatment of lung metastases of head and neck tumors
Surgical Riad N. Younes,
Treatment of Lung Metastases Head and Neck Tumors MD, PhD, Jefferson L. Gross, Luis Paulo Kowalski,
BACKGROUND: Head and neck tumors often spread to the lungs, with a variety of presentations. The ideal treatment for those patients is still controversial. Resection of lung metastases was shown to significantly influence overall survival of patients. ELECTIVE: To evaluate results of surgical resection of lung nodules in patients with head and neck primary tumors. METHODS: A retrospective analysis was made of 53 patients with head and neck tumors and lung nodules (no other metastases detected in other organs) admitted to our department. They were separated into two groups: OPER (thoracotomy, n = 26), and NOTOPER (no thoracotomies, n = 27). Overall survival was compared (Kaplan-Meier, log-rank) between groups. RESULTS: Overall median SUtViVal of all patients was 10 months, of OPER 29 months, and of NOTOPER 6 months (P
H
ead and neck tumors are locally aggressive neoplasms, with high potential of invasion and regional lymph node metastases. Surgical resection with or without adjuvant radiation therapy is the treatment of choice, with a good cure rate in many cases. When distant metastases develop in patients with head and neck car-
L From the Departments of Thoracic and Head and Neck Surgery, Hospital AC Camargo, University of Sao Paula School of Medicine, Universidade Paulista (UNIP), Ssio Paulo, Brazil. Requests for reprints should be addressed to Riad N. Younes, MD, PhD, Department of Thoracic Surgery, Hospital AC Camargo, R. Prof Antonio Prudente, 211, Slo Paula, SP Brazil 01509-010. Presented at the 43rd Annual Meeting of the Society of Head and Neck Surgeons, Cancun, Mexico, April 10-12, 1997.
71: 0 1997 by Excerpta All rights reserved.
Medica,
Inc.
of
MD, Jordao F. Silva, MD, Jose A.P. Fernandez, MD, PhD, Sao Pa&, Brazil
MD,
cinemas, the likelihood of long-term survival becomes very low. The incidence of metastatic spread following the treatment of the primary head and neck tumors depends basically on the initial stage of the disease, as well as on the completeness of local treatment.’ Most of the patients have metastases primarily identified in the lungs (8O%).2,3 Although hematogenous spread of the tumor is considered a systemic disease, most of the patients have metastases confined to the pulmonary tissue. Most chemotherapeutic regimens currently in use for stage IV head and neck tumors are considered to be palliative. The overall response rate is low, with a less than 5% 5-year survival rate in patients exclusively submitted to chemotherapy. Patients with pulmonary metastases of head and neck tumors present with a poor prognosis, with a less than 10% survival rate following diagnosis of the metastatic spread.4 A number of studies previously showed that a surgical approach to lung metastases could be an option for patients with nodules preferentially located in the lungs, and with no effective systemic treatment available.5-9 Overall 5-year survival rates of patients with head and neck tumors with completely resected pulmonary metastases range between 31% and 43%. Although most studies are retrospective, with few cases described over several decades (with the changing diagnostic sensitivity, operative technique and mortality, and adjuvant therapy), the encouraging results present the best hope for long-term control and survival of metastatic head and neck tumors. The role of surgical treatment of lung metastases remains unclear. The present study evaluates the influence on management and outcome of resection of pulmonary nodules in patients with a diagnosis of head and neck tumors. This is a retrospective consecutive study of all patients treated in the last 6 years at our institution.
PATIENTS
AND METHODS
All patients with locoregionally controlled head and neck tumors presenting with nodules confined to the lungs were included in a protocol designed to evaluate the results of the treatment of this patient population. The study started in January 1991, and analyzed patients accrued until December 1996. All patients underwent a complete preoperative workup and were completely staged by the Department of Head and Neck Surgery before they were considered for thoracotomy. Patients were included in the protocol if they had (1) locoregional control of the primary head and neck tumor, (2) nodules diagnosed only in the lungs, with no extrapulmonary spread of the disease, (3) technically resectable nodules, and (4) clinical condition adequate for the procedure. The number of metastases and the laterality of the nodules were not considered as contraindications for the operation. Wedge resection of the nod0002-9610/97/$17.00 PII SOOO2-961 0(97)00164-5
499
TABLE
I Characteristics
of Patients
Submitted
Characteristic
to Operation
Number
Site of primary tumor Lip/oral cavity Oropharynx Tongue Larynx Other Operated on Gender (M/F) Age (years) Complete resection Incomplete resection Disease-free interval
Follow-up (months) Alive/dead Not operated on Follow-up (months) Age (years) Disease-free interval
of Patients
10 15 8 13 2
0
2412 Range 38-69, median n = 19 n=7 Range O-93 <12, 27% 12-24,31% >24, 42% Range 2.8-50, median 1 O/l 6 (38% alive)
(months)
Gender (M/F) Alive/dead
10
15
20
Figure 2. Survival mors to the lungs:
of patients with metastatic completeness of resection.
P PW IFi 60 40 20 0 10
15
20
25
30
35
45
so
Time (months)
p < 0.0001 Figure 1. Survival mors to the lungs:
40
of patients with me&static type of treatment.
head
and neck;u-
ules was the operative procedure more frequently indicated. Solitary pulmonary nodules that could not be defined by the pathologist as metastases were considered to be second primary lung tumors, and treated accordingly by lobectomy and mediastinal lymph node dissection. Tumors originating in the bronchial mucosa were diagnosed as lung primaries. We did not perform a systematic mediastinal lymph node dissection on patients with multiple nodules. Enlarged mediastinal lymph nodes were sampled and sent for pathological analysis. Bilateral metastases were submitted to bilateral staged thoracotomy, separated by 2 to 3 weeks. Recurrent metastases confined to the lungs were considered for multiple thoracotomies, as long as they met the inclusion criteria. The patients were followed up for complications (fever >38”C, atelectasis, respiratory, renal, cardiac, neurologic or 500
THE AMERICAN
JOURNAL
OF
SURGERY@
VOLUME
10
45
50
head
and neck
tu-
-DFI < 12 months -DFI>12mmths
40 20 0 0
5
10
15
20
174
25
30
35
40 The
Figure 3. Survival of patients mors to the lungs: disease-free the lung metastases.
5
35
15
100
0
30
p < 0.0001
53
p = 0.14 (Log Rank)
$
25
Ttme (mmtha)
Range O-l 7, median 2 Range 33-80, median 50 Range 3-74 <12,52% 12-24,37% >24, 11% 2512 7/20 (26% alive)
(months)
5
p = 0.006
45
50
(months)
(Breslow)
with metastatic head and neck tuinterval between the primary and
hepatic complications, wound infection), recurrence (diagnosed by symptoms, physical examination, radiology, confirmed by the pathologist), and survival. Survival was determined for all patients from the date of the thoracotomy. If the patient presented with bilateral disease, survival was calculated from the date of the second thoracotomy. Follow-up of the patients was done every 3 months with chest roentgenography. Computed tomography (CT) was ordered whenever there was a new symptom or radiological finding suggestive of recurrent disease. Actuarial survival curves were calculated using the method of Kaplan and Meier.” The survival curves were analyzed for significance using the log-rank and Breslow tests. The Cox proportional hazards model was used to determine the significance of prognostic factors. Differences were considered significant for P t0.05.
RESULTS Fifty-three patients with head and neck carcinomas and nodules confined to the lungs were included in this study. They were separated into two groups: patients submitted to thoracotomy and resection (OPER, n = 26), and patients not submitted to thoracotomy (NOTOPER, n = 27). Table I shows the characteristics of the patients included in this report. Overall operative complication rate was 17%, with no 30-day operative mortality. All nodules resected were analyzed by the Pathology Department: 22% of all lesions resected were confirmed to be benign disease (granulomas, NOVEMBER
1997
t SURGICAL TABLE II Results
Author Finley’ McCormack’ Maze? Mountain9 Rendina’ Present study
of Resection of Lung Thoracotomy Years
Metastases of Head and Neck Tumors: and Complete Resection of Metastases
of Accrual
1970-1989
Number
of Patients
11 Complete 27 44 48 11 26
1960-l 977 1957-1986 1947-1980 1976-1981 1991-1996
fibrosis, pneumonitis, bonchiolitis). Two patients had second primary tumors, diagnosed at stage I for lung cancer. Overall median survival of all patients was 10 months. Patients submitted to operation had a median survival time of 19.6 months, significantly greater than NOTOPER, which was 6 months (I’24 months (log-rank test P = 0.27). However, the Breslow test of the survival curves showed significantly greater survival for patients with metastases appearing more than a year after the primary tumor (P = 0.006): Those results suggest a beneficial impact of DFI on survival for the first months of follow-up (Figure 3). Multivariate analysis showed that operation and complete resection were the only independent variables with significant impact on overall survival.
COMMENTS Surgical resection of pulmonary metastases is not a new procedure, being reported sporadically since the first decades of this century. Recently, this approach has been widely indicated for a variety of primary tumor histologies, mainly in cancers with no effective alternative treatment. The dismal prognosis associated with lung spread of solid tumors (usually less than 1% 5year survival rate,4’” and THE
TREATMENT
AMERICAN
OF
LUNG
METASTASESNOUNES
Survival
ET AL 1
Following
B-Year Survival Rate (Median Months) 29% (28) 43% (30) 47% (31) 43% 41%
7
ii
the paucity of chemotherapeutic regimens with reasonable impact on survival, encouraged thoracic surgeons to resect pulmonary nodules with more aggressive indications. Autopsy data showing a 30% to 60% incidence of metastases confined only to the lungs supported this approach, and paved the way for the rationale of surgical resection of hematogenous disseminated tumors. Most series in the literature describe the results of surgery for pulmonary metastases, regardless of the histology of the primary tumors. Although the results shown by those reports are encouraging, better selection of patients for individual primary tumors is paramount for proper management. Several studies showed the results of resection of lung metastases in patients with head and neck tumors. Most of these studies were retrospective, with patients treated over a wide span of time (usually decades). The available diagnosis techniques (CT scans, magnetic resonance imaging), postoperative follow-up, and treatment results (chemotherapy, radiotherapy, intensive care) changed significantly during the years, affecting the homogeneity of the series (Table II). Overall 5year results varied from 29% to 43% following resection of metastatic pulmonary disease. The present study is a prospective protocol, evaluated in a retrospective analysis, including consecutive patients with the same pathological diagnosis techniques (including immunohistochemistry), radiological staging methods, surgical and adjuvant treatment of the primary disease, and systemic therapy for disseminated disease. All patients were included in the study during the 6-year period, which insured a more homogeneous population. The results of surgical resection were compared with patients exclusively submitted to systemic therapy during the same period of the study. Those patients did not accept the indication of thoracotomy, or they were not clinically fit for the operation (respiratory limitation due to chronic obstructive pulmonary disease). Chemotherapy was based on regimens containing cisplatin. Overall survival rate of all patients presenting with head and neck tumors and isolated lung metastases reached a median of 10 months in our patient population, comparable with previously published data. Median survival of patients not submitted to surgical operation of the thoracic lesions was 6 months; no patient survived beyond 17 months following the diagnosis of lung metastases in that group. Only 7 patients received chemotherapy as the exclusive treatment for their metastases, with 7 months median survival rate. Patients with pulmonary metastases and no treatment (refused treatment or were clinically unfit) had a median JOURNAL
OF
SURGERY@
VOLUME
174
NOVEMBER
1997
501
SURGICAL
++
survival of 4 months. Patients submitted to complete resection of their lung metastases had a median survival of 23 months, comparable with the results of 7 patients reported by Finley et al6 (median survival following complete resection was 30 months). Patients (n = 7) with incomplete resection of lung metastases presented with poor prognosis, although median survival of this subgroup reached 15 months. Most patients were shown during the operation to have multiple small lesions spread throughout the lungs, and not detectable at CT scan. Resection of lung metastases is gaining widespread acceptance as an attractive option for treating solid tumors, including head and neck primaries. Despite complete resection of all gross disease, the majority of patients in all series still die of neoplastic dissemination. Several studies tried to identify prognostic factors that correlate directly with outcome, in order to select patients more likely to benefit from the operation or to include subsets of patients in adjuvant or neoadjuvant therapy that might influence prognosis. The series that reported on head and neck tumors with lung metastases presented small numbers of patients, over a wide range of accrual period (Table II). Few specific prognostic factors could be clearly defined and analyzed in those series. Locoregional control of the primary tumor is paramount for better prognosis, and the disease-free interval was shown to impact on survival following resection of lung metastases. In our study, all patients had complete control of locoregional disease before resection of the lung nodules, so we could not assess this factor in our series.
CONCLUSIONS Resection of metastases confined to the lungs from head and neck tumors has been shown to be effective, mainly in patients in whom locoregional primary tumor was completely controlled and all identifiable pulmonary nodules completely resected. The number of nodules seen at imag ing studies in the preoperative evaluation does not correlate with the number of malignant nodules confirmed by pathology. Second lung primary tumors could be completely treated in patients with head and neck tumors and not rep-
502
THE AMERICAN
JOURNAL
OF
SURGERY”
VOLUME
174
resent actual metastatic dissemination of the previously treated neoplasm. Although the disease-free interval between the treatment of the primary tumor and the appearance of lung metastases was shown to marginally affect survival on univariate analysis, this factor should not be considered as an excluding criterion for the operative approach to an individual patient. Continuing evaluation of the results of resection of lung metastases should provide clearer identification and selection of candidates for surgical resection, as well as for associated systemic treatment, neoadjuvant or adjuvant.
REFERENCES 1. Merino OR, Lindbera RD, Fletcher GH. An analysis of distant metastases from squamous cell carcinoma of the upper respiratory and digestive tracts. Cancer. 1977;40:145-151. 2. Kotwall C, Sako K, Razack MS, et al. Metastatic patterns in squamous cell cancer of the head and neck. Am J Surg. 1987;154:439-442. 3. Papac RJ. Distant metastases from head and neck cancer. Cancer. 1984;53:342-345. 4. Dennington ML, Carter DR, Meyers AD. Distant metastases in head and neck epidermoid carcinoma. Laryngoscope. 1980;90:196201. 5. Mazer TM, Robbins KT, McMurtrey MJ, Byers RM. Resection of pulmonary metastases from squamous carcinoma of the head and neck. Am J Surg. 1988;156:238-242. 6. Finley RK, Verazin GT, Driscoll DL, et al. Results of surgical resection of pulmonary metastases of squamous cell carcinoma of the head and neck. Am .I Surg. 1992;164:594-598. 7. Rendina EA, Vincentiis M, Primerano G, et al. Pulmonary resection for metastatic laryngeal carcinoma. J Thorax Cardiovac Surg. 1986;92:114-117. 8. McCormack, Martini N. The changing role of surgery for pulmonary metastases. Ann Thorns Surg. 1979;28:139-145. 9. Mountain CF, McMurtrey MJ, Hermes KE. Surgery for pulmonary metastases: a 20-year experience. Ann Thorac Surg. 1984; 38~323-330. 10. Kaplan EL, Meier P. Non-parametric estimations from incomplete observations. J Am Stat Assoc. 1958;53:457-481. 11. Lefor AJ, Bredenburg CE, Kellman RM, Ayst JC. Multiple malignancies of the lung and head and neck: second primary tumor or metastasis? Arch Surg. 1986;121:265-270.
NOVEMBER
1997
Treatment of Lung Metastases Head and Neck Tumors MD, PhD, Jefferson L. Gross, Luis Paulo Kowalski,
BACKGROUND: Head and neck tumors often spread to the lungs, with a variety of presentations. The ideal treatment for those patients is still controversial. Resection of lung metastases was shown to significantly influence overall survival of patients. ELECTIVE: To evaluate results of surgical resection of lung nodules in patients with head and neck primary tumors. METHODS: A retrospective analysis was made of 53 patients with head and neck tumors and lung nodules (no other metastases detected in other organs) admitted to our department. They were separated into two groups: OPER (thoracotomy, n = 26), and NOTOPER (no thoracotomies, n = 27). Overall survival was compared (Kaplan-Meier, log-rank) between groups. RESULTS: Overall median SUtViVal of all patients was 10 months, of OPER 29 months, and of NOTOPER 6 months (P
H
ead and neck tumors are locally aggressive neoplasms, with high potential of invasion and regional lymph node metastases. Surgical resection with or without adjuvant radiation therapy is the treatment of choice, with a good cure rate in many cases. When distant metastases develop in patients with head and neck car-
L From the Departments of Thoracic and Head and Neck Surgery, Hospital AC Camargo, University of Sao Paula School of Medicine, Universidade Paulista (UNIP), Ssio Paulo, Brazil. Requests for reprints should be addressed to Riad N. Younes, MD, PhD, Department of Thoracic Surgery, Hospital AC Camargo, R. Prof Antonio Prudente, 211, Slo Paula, SP Brazil 01509-010. Presented at the 43rd Annual Meeting of the Society of Head and Neck Surgeons, Cancun, Mexico, April 10-12, 1997.
71: 0 1997 by Excerpta All rights reserved.
Medica,
Inc.
of
MD, Jordao F. Silva, MD, Jose A.P. Fernandez, MD, PhD, Sao Pa&, Brazil
MD,
cinemas, the likelihood of long-term survival becomes very low. The incidence of metastatic spread following the treatment of the primary head and neck tumors depends basically on the initial stage of the disease, as well as on the completeness of local treatment.’ Most of the patients have metastases primarily identified in the lungs (8O%).2,3 Although hematogenous spread of the tumor is considered a systemic disease, most of the patients have metastases confined to the pulmonary tissue. Most chemotherapeutic regimens currently in use for stage IV head and neck tumors are considered to be palliative. The overall response rate is low, with a less than 5% 5-year survival rate in patients exclusively submitted to chemotherapy. Patients with pulmonary metastases of head and neck tumors present with a poor prognosis, with a less than 10% survival rate following diagnosis of the metastatic spread.4 A number of studies previously showed that a surgical approach to lung metastases could be an option for patients with nodules preferentially located in the lungs, and with no effective systemic treatment available.5-9 Overall 5-year survival rates of patients with head and neck tumors with completely resected pulmonary metastases range between 31% and 43%. Although most studies are retrospective, with few cases described over several decades (with the changing diagnostic sensitivity, operative technique and mortality, and adjuvant therapy), the encouraging results present the best hope for long-term control and survival of metastatic head and neck tumors. The role of surgical treatment of lung metastases remains unclear. The present study evaluates the influence on management and outcome of resection of pulmonary nodules in patients with a diagnosis of head and neck tumors. This is a retrospective consecutive study of all patients treated in the last 6 years at our institution.
PATIENTS
AND METHODS
All patients with locoregionally controlled head and neck tumors presenting with nodules confined to the lungs were included in a protocol designed to evaluate the results of the treatment of this patient population. The study started in January 1991, and analyzed patients accrued until December 1996. All patients underwent a complete preoperative workup and were completely staged by the Department of Head and Neck Surgery before they were considered for thoracotomy. Patients were included in the protocol if they had (1) locoregional control of the primary head and neck tumor, (2) nodules diagnosed only in the lungs, with no extrapulmonary spread of the disease, (3) technically resectable nodules, and (4) clinical condition adequate for the procedure. The number of metastases and the laterality of the nodules were not considered as contraindications for the operation. Wedge resection of the nod0002-9610/97/$17.00 PII SOOO2-961 0(97)00164-5
499
TABLE
I Characteristics
of Patients
Submitted
Characteristic
to Operation
Number
Site of primary tumor Lip/oral cavity Oropharynx Tongue Larynx Other Operated on Gender (M/F) Age (years) Complete resection Incomplete resection Disease-free interval
Follow-up (months) Alive/dead Not operated on Follow-up (months) Age (years) Disease-free interval
of Patients
10 15 8 13 2
0
2412 Range 38-69, median n = 19 n=7 Range O-93 <12, 27% 12-24,31% >24, 42% Range 2.8-50, median 1 O/l 6 (38% alive)
(months)
Gender (M/F) Alive/dead
10
15
20
Figure 2. Survival mors to the lungs:
of patients with metastatic completeness of resection.
P PW IFi 60 40 20 0 10
15
20
25
30
35
45
so
Time (months)
p < 0.0001 Figure 1. Survival mors to the lungs:
40
of patients with me&static type of treatment.
head
and neck;u-
ules was the operative procedure more frequently indicated. Solitary pulmonary nodules that could not be defined by the pathologist as metastases were considered to be second primary lung tumors, and treated accordingly by lobectomy and mediastinal lymph node dissection. Tumors originating in the bronchial mucosa were diagnosed as lung primaries. We did not perform a systematic mediastinal lymph node dissection on patients with multiple nodules. Enlarged mediastinal lymph nodes were sampled and sent for pathological analysis. Bilateral metastases were submitted to bilateral staged thoracotomy, separated by 2 to 3 weeks. Recurrent metastases confined to the lungs were considered for multiple thoracotomies, as long as they met the inclusion criteria. The patients were followed up for complications (fever >38”C, atelectasis, respiratory, renal, cardiac, neurologic or 500
THE AMERICAN
JOURNAL
OF
SURGERY@
VOLUME
10
45
50
head
and neck
tu-
-DFI < 12 months -DFI>12mmths
40 20 0 0
5
10
15
20
174
25
30
35
40 The
Figure 3. Survival of patients mors to the lungs: disease-free the lung metastases.
5
35
15
100
0
30
p < 0.0001
53
p = 0.14 (Log Rank)
$
25
Ttme (mmtha)
Range O-l 7, median 2 Range 33-80, median 50 Range 3-74 <12,52% 12-24,37% >24, 11% 2512 7/20 (26% alive)
(months)
5
p = 0.006
45
50
(months)
(Breslow)
with metastatic head and neck tuinterval between the primary and
hepatic complications, wound infection), recurrence (diagnosed by symptoms, physical examination, radiology, confirmed by the pathologist), and survival. Survival was determined for all patients from the date of the thoracotomy. If the patient presented with bilateral disease, survival was calculated from the date of the second thoracotomy. Follow-up of the patients was done every 3 months with chest roentgenography. Computed tomography (CT) was ordered whenever there was a new symptom or radiological finding suggestive of recurrent disease. Actuarial survival curves were calculated using the method of Kaplan and Meier.” The survival curves were analyzed for significance using the log-rank and Breslow tests. The Cox proportional hazards model was used to determine the significance of prognostic factors. Differences were considered significant for P t0.05.
RESULTS Fifty-three patients with head and neck carcinomas and nodules confined to the lungs were included in this study. They were separated into two groups: patients submitted to thoracotomy and resection (OPER, n = 26), and patients not submitted to thoracotomy (NOTOPER, n = 27). Table I shows the characteristics of the patients included in this report. Overall operative complication rate was 17%, with no 30-day operative mortality. All nodules resected were analyzed by the Pathology Department: 22% of all lesions resected were confirmed to be benign disease (granulomas, NOVEMBER
1997
t SURGICAL TABLE II Results
Author Finley’ McCormack’ Maze? Mountain9 Rendina’ Present study
of Resection of Lung Thoracotomy Years
Metastases of Head and Neck Tumors: and Complete Resection of Metastases
of Accrual
1970-1989
Number
of Patients
11 Complete 27 44 48 11 26
1960-l 977 1957-1986 1947-1980 1976-1981 1991-1996
fibrosis, pneumonitis, bonchiolitis). Two patients had second primary tumors, diagnosed at stage I for lung cancer. Overall median survival of all patients was 10 months. Patients submitted to operation had a median survival time of 19.6 months, significantly greater than NOTOPER, which was 6 months (I’
COMMENTS Surgical resection of pulmonary metastases is not a new procedure, being reported sporadically since the first decades of this century. Recently, this approach has been widely indicated for a variety of primary tumor histologies, mainly in cancers with no effective alternative treatment. The dismal prognosis associated with lung spread of solid tumors (usually less than 1% 5year survival rate,4’” and THE
TREATMENT
AMERICAN
OF
LUNG
METASTASESNOUNES
Survival
ET AL 1
Following
B-Year Survival Rate (Median Months) 29% (28) 43% (30) 47% (31) 43% 41%
7
ii
the paucity of chemotherapeutic regimens with reasonable impact on survival, encouraged thoracic surgeons to resect pulmonary nodules with more aggressive indications. Autopsy data showing a 30% to 60% incidence of metastases confined only to the lungs supported this approach, and paved the way for the rationale of surgical resection of hematogenous disseminated tumors. Most series in the literature describe the results of surgery for pulmonary metastases, regardless of the histology of the primary tumors. Although the results shown by those reports are encouraging, better selection of patients for individual primary tumors is paramount for proper management. Several studies showed the results of resection of lung metastases in patients with head and neck tumors. Most of these studies were retrospective, with patients treated over a wide span of time (usually decades). The available diagnosis techniques (CT scans, magnetic resonance imaging), postoperative follow-up, and treatment results (chemotherapy, radiotherapy, intensive care) changed significantly during the years, affecting the homogeneity of the series (Table II). Overall 5year results varied from 29% to 43% following resection of metastatic pulmonary disease. The present study is a prospective protocol, evaluated in a retrospective analysis, including consecutive patients with the same pathological diagnosis techniques (including immunohistochemistry), radiological staging methods, surgical and adjuvant treatment of the primary disease, and systemic therapy for disseminated disease. All patients were included in the study during the 6-year period, which insured a more homogeneous population. The results of surgical resection were compared with patients exclusively submitted to systemic therapy during the same period of the study. Those patients did not accept the indication of thoracotomy, or they were not clinically fit for the operation (respiratory limitation due to chronic obstructive pulmonary disease). Chemotherapy was based on regimens containing cisplatin. Overall survival rate of all patients presenting with head and neck tumors and isolated lung metastases reached a median of 10 months in our patient population, comparable with previously published data. Median survival of patients not submitted to surgical operation of the thoracic lesions was 6 months; no patient survived beyond 17 months following the diagnosis of lung metastases in that group. Only 7 patients received chemotherapy as the exclusive treatment for their metastases, with 7 months median survival rate. Patients with pulmonary metastases and no treatment (refused treatment or were clinically unfit) had a median JOURNAL
OF
SURGERY@
VOLUME
174
NOVEMBER
1997
501
SURGICAL
++
survival of 4 months. Patients submitted to complete resection of their lung metastases had a median survival of 23 months, comparable with the results of 7 patients reported by Finley et al6 (median survival following complete resection was 30 months). Patients (n = 7) with incomplete resection of lung metastases presented with poor prognosis, although median survival of this subgroup reached 15 months. Most patients were shown during the operation to have multiple small lesions spread throughout the lungs, and not detectable at CT scan. Resection of lung metastases is gaining widespread acceptance as an attractive option for treating solid tumors, including head and neck primaries. Despite complete resection of all gross disease, the majority of patients in all series still die of neoplastic dissemination. Several studies tried to identify prognostic factors that correlate directly with outcome, in order to select patients more likely to benefit from the operation or to include subsets of patients in adjuvant or neoadjuvant therapy that might influence prognosis. The series that reported on head and neck tumors with lung metastases presented small numbers of patients, over a wide range of accrual period (Table II). Few specific prognostic factors could be clearly defined and analyzed in those series. Locoregional control of the primary tumor is paramount for better prognosis, and the disease-free interval was shown to impact on survival following resection of lung metastases. In our study, all patients had complete control of locoregional disease before resection of the lung nodules, so we could not assess this factor in our series.
CONCLUSIONS Resection of metastases confined to the lungs from head and neck tumors has been shown to be effective, mainly in patients in whom locoregional primary tumor was completely controlled and all identifiable pulmonary nodules completely resected. The number of nodules seen at imag ing studies in the preoperative evaluation does not correlate with the number of malignant nodules confirmed by pathology. Second lung primary tumors could be completely treated in patients with head and neck tumors and not rep-
502
THE AMERICAN
JOURNAL
OF
SURGERY”
VOLUME
174
resent actual metastatic dissemination of the previously treated neoplasm. Although the disease-free interval between the treatment of the primary tumor and the appearance of lung metastases was shown to marginally affect survival on univariate analysis, this factor should not be considered as an excluding criterion for the operative approach to an individual patient. Continuing evaluation of the results of resection of lung metastases should provide clearer identification and selection of candidates for surgical resection, as well as for associated systemic treatment, neoadjuvant or adjuvant.
REFERENCES 1. Merino OR, Lindbera RD, Fletcher GH. An analysis of distant metastases from squamous cell carcinoma of the upper respiratory and digestive tracts. Cancer. 1977;40:145-151. 2. Kotwall C, Sako K, Razack MS, et al. Metastatic patterns in squamous cell cancer of the head and neck. Am J Surg. 1987;154:439-442. 3. Papac RJ. Distant metastases from head and neck cancer. Cancer. 1984;53:342-345. 4. Dennington ML, Carter DR, Meyers AD. Distant metastases in head and neck epidermoid carcinoma. Laryngoscope. 1980;90:196201. 5. Mazer TM, Robbins KT, McMurtrey MJ, Byers RM. Resection of pulmonary metastases from squamous carcinoma of the head and neck. Am J Surg. 1988;156:238-242. 6. Finley RK, Verazin GT, Driscoll DL, et al. Results of surgical resection of pulmonary metastases of squamous cell carcinoma of the head and neck. Am .I Surg. 1992;164:594-598. 7. Rendina EA, Vincentiis M, Primerano G, et al. Pulmonary resection for metastatic laryngeal carcinoma. J Thorax Cardiovac Surg. 1986;92:114-117. 8. McCormack, Martini N. The changing role of surgery for pulmonary metastases. Ann Thorns Surg. 1979;28:139-145. 9. Mountain CF, McMurtrey MJ, Hermes KE. Surgery for pulmonary metastases: a 20-year experience. Ann Thorac Surg. 1984; 38~323-330. 10. Kaplan EL, Meier P. Non-parametric estimations from incomplete observations. J Am Stat Assoc. 1958;53:457-481. 11. Lefor AJ, Bredenburg CE, Kellman RM, Ayst JC. Multiple malignancies of the lung and head and neck: second primary tumor or metastasis? Arch Surg. 1986;121:265-270.
NOVEMBER
1997