World Journal of Surgery

, Volume 32, Issue 8, pp 1657–1662

Pulmonary Resection for Metastatic Head and Neck Cancer

Authors

  • Fengshi Chen
    • Department of Thoracic SurgeryKyoto University
  • Makoto Sonobe
    • Department of Thoracic SurgeryKyoto University
  • Kiyoshi Sato
    • Department of Thoracic SurgeryKyoto University
  • Takuji Fujinaga
    • Department of Thoracic SurgeryKyoto University
  • Tsuyoshi Shoji
    • Department of Thoracic SurgeryKyoto University
  • Hiroaki Sakai
    • Department of Thoracic SurgeryKyoto University
  • Ryo Miyahara
    • Department of Thoracic SurgeryKyoto University
  • Toru Bando
    • Department of Thoracic SurgeryKyoto University
  • Kenichi Okubo
    • Department of Thoracic SurgeryKyoto University
  • Toshiki Hirata
    • Department of Thoracic SurgeryKyoto University
    • Department of Thoracic SurgeryKyoto University
Article

DOI: 10.1007/s00268-008-9631-8

Cite this article as:
Chen, F., Sonobe, M., Sato, K. et al. World J Surg (2008) 32: 1657. doi:10.1007/s00268-008-9631-8

Abstract

Background

Pulmonary metastasectomy has become the standard therapy for various metastatic malignancies to the lungs; however, little data have been available about lung metastasectomy for head and neck cancers. To confirm a role for resection of pulmonary metastases for such tumors, we reviewed our institutional experience.

Methods

Between 1991 and 2007, 20 patients with pulmonary metastases from head and neck cancers underwent complete pulmonary resection. All patients had obtained or had obtainable locoregional control of their primary head and neck cancers. Various perioperative variables were investigated retrospectively to analyze the prognostic factors for overall survival and disease-free survival after metastasectomy.

Results

Of the 20 patients, 10 (50%) had squamous cell carcinoma, 7 (35%) had adenoid cystic carcinoma, and 3 had miscellaneous carcinomas. The median disease-free interval from the time of treatment of the head and neck primary cancers to the development of pulmonary metastases was 27 months. Overall survival rate after metastasectomy was 59.4% at 5 years and 47.5% at 10 years, respectively. Disease-free survival rate was 25.0% at 5 years after pulmonary resection. A disease-free interval equal to or longer than 12 months was a significantly favorable prognostic factor for both overall survival and disease-free survival (p = 0.02 and 0.01, respectively). Patients with squamous cell carcinoma and male sex showed a worse overall survival (p = 0.04 and 0.03, respectively).

Conclusion

The current practice of pulmonary metastasectomy for head and neck cancers in our institution was well justified. A disease-free interval equal to or longer than 12 months, nonsquamous cell carcinoma, and female sex might be relevant to a better prognosis.

Introduction

Head and neck cancers usually spread first to the regional lymph nodes, but more rarely may metastasize to distant sites. The lungs, followed by bone and liver, are the most common sites of distant metastases [1, 2]. The incidence of pulmonary metastases from head and neck cancers is reported to be 10–30% [13]. Although pulmonary metastasectomy has become the standard therapy for various metastatic malignancies to the lungs [47], there have been a limited number of studies examining the role of pulmonary metastasectomy for patients with head and neck cancers [810]. Herein we report our experience with pulmonary metastasectomy in patients with head and neck cancers and assess its role in their treatment.

Patients and methods

A retrospective analysis of patients with lung metastases from head and neck cancers, who were referred for pulmonary resection to Kyoto University Hospital between April 1991 and March 2007, was performed. In this period, complete resection of pulmonary metastases from head and neck cancers was performed on 20 patients. Patients with skin or thyroid carcinoma were excluded in this study. Chest X-rays or CT scans were considered positive if they showed new, rounded pulmonary lesions without calcifications in patients previously known to have head and neck cancer. All specimens obtained from metastasectomies were reviewed by several pathologists. The pulmonary lesions were considered metastatic only if the pathologists could exclude the possibility of a primary lung malignancy and define the tumor as metastatic based upon histologic similarity to the primary tumor, especially for squamous cell carcinomas. Tumors originating from the bronchial mucosa were diagnosed as lung primaries. In every patient we collected data on primary treatment and on the interval between primary treatment and the diagnosis of pulmonary metastases (disease-free interval = DFI). Further data were collected on the characteristics of metastases, on the method of treatment, and on details concerning the metastasectomy. The endpoint was overall survival (OS) and disease-free survival (DFS) after the first pulmonary metastasectomy.

All patients who underwent resection of their pulmonary metastases met the following criteria: (1) pulmonary lesions were deemed resectable not only by radiologic examinations but also by the patients’ general conditions, (2) metastatic disease was limited to the lungs, and (3) locoregional control of their head and neck primary cancer was obtained or obtainable. Twenty-one patients met these criteria and metastasectomy was performed. Complete clinical and pathologic data were reviewed. Last follow-up categories were defined as no evidence of disease (NED), alive with disease (AWD), dead of other causes (DOC), and dead of disease (DOD). To analyze the prognostic factors for OS and DFS, several parameters were investigated retrospectively, such as, age, sex, DFI, characteristics of primary tumors (histology and location), recurrence of primary tumors before detection of the lung metastasis, characteristics of pulmonary metastases (number, size, and location), and number of metastasectomies for pulmonary recurrence.

Statistical analysis was performed using the StatView (version 4.5) software package (Abacus Concepts, Berkeley, CA). The postoperative survival rate was analyzed by the Kaplan-Meier method, and the differences in survival rate were assessed by the log-rank test. Differences were considered significant when p < 0.05.

Results

Of 20 patients, 9 patients were male and 11 were female. The median age was 63 years (range = 38–80 years). Ten patients (50%) had squamous cell carcinomas, 7 patients (35%) had adenoid cystic carcinomas, and three patients had miscellaneous tumors, which included a clear cell carcinoma, a pleomorphic carcinoma, and a mucoepidermoid carcinoma. The sites of the primary cancers were the oral cavity in 11 cases, salivary glands in 4 cases, pharynx in 3 cases, larynx in 1 case, and paranasal sinus in 1 case (Table 1). In all cases, the primary cancers had been completely resected. DFI varied from 2 months to 27 years, with a median of 27 months, except for one case with simultaneous resection of the pulmonary metastasis and the primary tumor. In 8 patients (40%), locoregional recurrence was observed and resected completely before detection of the lung metastasis. In 5 patients (25%), lung metastases were found to be bilateral. Nine patients (45%) had a solitary pulmonary metastasis. For the 11 patients with multiple metastases, which were confirmed by pathology, the median number of lesions was 2.7 (range = 2–6). There was an interesting discrepancy in terms of the number of pulmonary nodules between preoperative evaluation by CT and the intraoperative assessment. Preoperative evaluation by CT presumed one nodule in 9 patients, two nodules in six, three in four, and four in one. All patients with one nodule in the preoperative evaluation showed the same number of nodules intraoperatively, while six of 11 patients with more than two nodules showed a different number of nodules at the time of surgery. Of these six patients, three with two nodules identified preoperatively were found to have between three and six nodules intraoperatively; two patients with three nodules preoperatively were found to have five and six nodules, respectively; and one patient with four nodules identified preoperatively had only three identified intraoperatively.
Table 1

Histology and location of head and neck primaries

Tumor type

Location

Number of patients

Total

Squamous cell carcinoma

  

10

Oral cavity

6

 

Pharynx

3

 

Larynx

1

 

Adenoid cystic carcinoma

  

7

Salivary gland

3

 

Oral cavity

3

 

Paranasal sinus

1

 

Miscellaneous

  

3

Mucoepidermoid carcinoma

Oral cavity

1

 

Pleomorphic carcinoma

Oral cavity

1

 

Clear cell carcinoma

Salivary gland

1

 

Fifteen patients (75%) with unilateral tumors underwent a unilateral thoracotomy for resection of their metastases, while the remaining five patients had a bilateral thoracotomy. Three of five patients underwent simultaneous bilateral thoracotomy, while two of five patients had a staged bilateral thoracotomy for bilateral lesions. An attempt was made to conserve as much lung tissue as possible. This was reflected by the fact that a wedge resection was the most common procedure performed. In fact, lobectomy was performed in only four of 20 cases because of the size of the tumor. The median maximum diameter of the resected tumors was 10 mm (range = 2–70 mm). No patients died directly of surgery and all the patients were discharged home without any lesions after surgery. Postoperative chemotherapy was performed for the patients with squamous cell carcinomas, while chemotherapy was not considered for the patients with adenoid cystic carcinomas.

The median time at follow-up examination was 29 months (range = 4–190 months). Six patients (30%) remained free of disease after their first thoracic operation and 14 patients (70%) developed recurrences, the majority of which were located in the lungs (Table 2). Five patients (25%) could undergo resection for pulmonary recurrences. In detail, five of seven patients (71%) with adenoid cystic carcinoma had recurrence after lung metastasectomy and three patients could undergo pulmonary resection two to four times including the first metastasectomy. By contrast, seven of ten patients with squamous cell carcinoma had recurrence, but only one patient could undergo pulmonary resection once more. In addition, one patient with clear cell carcinoma also underwent additional pulmonary resection for recurrent tumors. The median time to recurrence after pulmonary resection was 21 months (range = 1–83 months).
Table 2

Sites of recurrence after initial thoracotomy (n = 14 patients)

Location

No. of patients with site involved

Squamous cell carcinoma (n = 7)

Adenoid cystic carcinoma (n = 5)

Miscellaneous cancers (n = 2)

Locoregional

3

1

1

1

Distant

    

Lung

9

5

4

0

Bone

2

2

0

0

Brain

2

1

1

0

Liver

2

1

0

1

Skin

1

1

0

0

The OS rate after metastasectomy was 59.4% at 5 years and 47.5% at 10 years. The DFS rate was 25.0% at 5 years after pulmonary resection (Fig. 1). Figure 2a shows that OS for patients with squamous cell carcinoma was significantly worse than that with nonsquamous cell carcinoma (p = 0.04). OS for patients with various histologies is shown in Fig. 2b. There was no difference in OS for patients with squamous cell carcinoma and for patients with adenoid cystic carcinoma (p = 0.11). Eight (40%) patients are currently alive without evidence of disease (NED) and 3 (15%) are alive with disease (AWD). Nine patients (45%) died of disease (DOD) and none died of other causes (DOC). On survival analysis, patients with DFI equal to or longer than 12 months had a better OS and DFS (p = 0.02 and 0.01, respectively, Fig. 3). A better OS was also observed for patients with female sex (p = 0.03, Table 3). In addition, patients with squamous cell carcinoma showed a worse OS than those with nonsquamous cell carcinoma (p = 0.04, Fig. 2b).
https://static-content.springer.com/image/art%3A10.1007%2Fs00268-008-9631-8/MediaObjects/268_2008_9631_Fig1_HTML.gif
Fig. 1

Overall and disease-free survival of patients following pulmonary resection (n = 20)

https://static-content.springer.com/image/art%3A10.1007%2Fs00268-008-9631-8/MediaObjects/268_2008_9631_Fig2_HTML.gif
Fig. 2

(a) Overall survival for patients with squamous cell carcinoma and nonsquamous cell carcinoma. (b) Overall survival according to histology

https://static-content.springer.com/image/art%3A10.1007%2Fs00268-008-9631-8/MediaObjects/268_2008_9631_Fig3_HTML.gif
Fig. 3

(a) Overall survival according to disease-free interval. (b) Disease-free survival according to disease-free interval. DFI, disease-free interval

Table 3

Analyses for overall survival and disease-free survival

Variables

No. of patients

p value

Overall survival

Disease-free survival

Age (years)

    <60

8

0.87

0.36

    ≥60

12

  

Gender

    Male

9

0.03

0.24

    Female

11

  

Disease-free interval (years)

    <1

7

0.02

0.01

    ≥1

13

  

Histology of primary

    Squamous cell carcinoma

10

0.03

0.87

    Nonsquamous cell carcinoma

10

  

Location of primary

    Oral cavity

10

0.78

0.46

    Other than oral cavity

10

  

Recurrence of primary ahead of lung metastasis

    Yes

8

0.56

0.39

    No

12

  

No. of pulmonary metastases

    Solitary

9

0.39

0.56

    Multiple

11

  

Largest size of metastases (cm)

    <1

6

0.12

0.55

    ≥1

14

  

Location of metastases

    Unilateral

15

0.15

0.80

    Bilateral

5

  

No. of metastasectomy for pulmonary recurrence

    1

15

0.26

0.12

    ≥2

5

  

Discussion

There have been many studies on pulmonary metastasectomy for various primary cancers such as colorectal cancer, breast cancer, and osteosarcoma [4, 5]. Pulmonary metastasectomy has become the standard therapy for various metastatic malignancies to the lungs. However, there have been only a small number of studies examining the role of pulmonary metastasectomy for patients with head and neck tumors [3, 815]. According to the study on 138 cases of pulmonary metastases from head and neck cancers by Wedman et al. [8], the prognosis of the patients who had metastasectomy was proven to be significantly better than that for those who did not undergo metastasectomy. Herein, we reported our experience with pulmonary metastasectomy in patients with head and neck cancers and evaluated its role in their treatment, focusing on OS and DFS.

The 5-year OS rate for patients who underwent a pulmonary metastasectomy for head and neck cancers was up to around 50% in several reports [810]. According to the OS rate, our attitude toward pulmonary metastasectomy for head and neck cancers was thought to be acceptable [3, 815]. To date, several parameters such as age, DFI, histology, nodal status, and laterality have been reported as prognostic factors.

In this study, a better OS was observed for patients with DFI equal to or longer than 12 months. There have been several reports investigating the relationship between survival and DFI for pulmonary metastasectomy for head and neck cancers [811]. Wedman et al. [8] reported that DFI longer than 12 months would be a positive prognostic factor for OS, while Liu et al. [9] stated that DFI longer than 24 months might be a positive prognostic factor. Furthermore, Locati et al. [12] commented that DFI equal to or longer than 36 months would be a positive prognostic factor for the OS of patients who had pulmonary metastasectomy for adenoid cystic carcinoma in the salivary glands. Thus, the intervals were different among the studies, but there is a significant relationship between DFI and OS. We also found that a better DFS was observed for patients with DFI equal to or longer than 12 months. To the best of our knowledge, there were no reports describing a significant relationship between DFI and DFS.

In addition, patients with squamous cell carcinoma showed a worse OS than those with nonsquamous cell carcinoma (p = 0.04). Wedman et al. [8] also found the presence of squamous cell carcinoma to be a significant prognostic factor for a poor survival rate after the diagnosis of pulmonary metastases. Furthermore, our study also showed that male sex was an adverse prognostic factor, but this might be related to the fact that seven of ten patients with squamous cell carcinoma were male and that three of ten patients with nonsquamous cell carcinoma were male. However, more cases are needed to analyze the true relationship between them.

Finley et al. [3] identified locoregional control of the head and neck primary tumor as an important prognostic factor. However, in our study locoregional recurrence of the primary disease before lung metastasectomy did not affect either OS or DFS (Table 3). This is because, despite recurrences, all the recurrent lesions were resected completely and were well controlled after metastasectomy as well as before pulmonary resection. However, it is of interest that seven of eight patients (89%) with locoregional recurrence of the primary disease before lung metastasectomy showed recurrent disease after pulmonary metastasectomy. In detail, five of ten patients with squamous cell carcinoma had locoregional recurrence of the primary disease before lung metastasectomy, while one of seven patients with adenoid cystic carcinoma had locoregional recurrence of the primary disease before lung metastasectomy. Seven of ten patients with squamous cell carcinoma had recurrence after pulmonary metastasectomy, while five of seven patients with adenoid cystic carcinoma had recurrence. Patients with squamous cell carcinoma had recurrence not only in the lungs but also in other organs, but the most frequent recurrence sites were the lungs in patients with squamous cell carcinoma and patients with adenoid cystic carcinoma (Table 2).

There seem to be different biological characteristics between squamous cell carcinomas and adenoid cystic carcinomas, even if there were no significant differences between them in several aspects including OS or DFS in our study, probably because of the small number of cases. Patients with squamous cell carcinoma are reported to develop recurrence most likely within 2 years after pulmonary metastasectomy [9]. As shown in Figure 2b, our patients with squamous cell carcinoma who had survived more than 2 years also could expect to achieve long-term survival. Adenoid cystic carcinomas are rare tumors, characterized by a slow, protracted course with multiple recurrences [9, 14]. As we have shown before, in our study almost all adenoid cystic carcinomas recurred in the lungs. Moreover, we had a patient with a recurrent tumor after four resections from the lungs who survived well with pulmonary metastasis more than 16 years after the first lung resection; however, it is still unclear whether pulmonary resection will prolong the survival of such patients. According to several reports, it is thought that resection did not cure patients with pulmonary metastases of adenoid cystic carcinoma [9, 12]. Because Locati et al. [12] regarded more than 3 years of DFI a positive prognostic factor, we plan to follow the patients with this in mind.

In conclusion, the current practice of pulmonary metastasectomy for head and neck cancers in our institution was well justified. A disease-free interval equal to or longer than 12 months, nonsquamous cell carcinoma, and female sex were relevant to a better prognosis; however, a greater number of cases is necessary to evaluate a prognostic factor properly and to determine the selection criteria for resection. In addition, we need to keep in mind that there seems to be different biological characteristics between squamous cell carcinomas and adenoid cystic carcinomas.

Copyright information

© Société Internationale de Chirurgie 2008