J THoRAc CARDIOVASC SURG 1987;94:69-74
Superior sulcus lung tumors Results of combined treatment (irradiation and radical resection) Twenty-one patients underwent combined therapy (irradiation and radical resection) for a Pancoast tumor at the Massachusetts General Hospital between 1976 and 1985. All patients underwent en bloc removal of the apical chest waD and underlying lung. In addition four patients required subclavian artery resection, and in five patients a portion of the vertebral body was resected. There were three operative deaths. Median survival was 24 months and actuarial survival rate was 55 % at 3 years and 27 % at 5 years. Long-term palliation of pain was achieved in 72 % of the patients. Involvement of the subclavian artery, vertebral body, or rib did not preclude long-term survival. Computed tomographic scanning in these patients is often indeterminate regarding invasion of chest waD structures but is more helpful than plain films alone. When compared to recent series in which irradiation alone was used, the combined approach appears to produce better results.
Cameron D. Wright, M.D., Ashby C. Moncure, M.D., Jo-Anne O. Shepard, M.D., Earle W. Wilkins, Jr., M.D., Douglas J. Mathisen, M.D., and Hermes C. Grillo, M.D., Boston, Mass.
results in the treatment of superior sulcus tumors of the lung (Pancoast tumor) with a combined approach (irradiation and radical resection) have been reported by Paulson,' Shaw,' Miller,' Attar,' Stanford,' and their associates. Irradiation alone has been reported to achieve nearly equivalent results.v? Because treatment remains somewhat controversial, we have reviewed our results with the combined approach with an emphasis on local factors that influence survival, the use of the CT scan in preoperative staging, and comparison with the results of patients treated with irradiation alone. Patients and methods The records of 21 patients who underwent surgical therapy for a Pancoast tumor during the period 1976 to 1985 at the Massachusetts General Hospital were reviewed. The diagnosis of a Pancoast tumor was made on the basis of a small apical lung mass involving the apex of the chest (pleura and/or rib) usually associated From the General Thoracic Surgical Unit and the Departments of Surgery and Radiology (l.S.), Massachusetts General Hospital and Harvard Medical School, Boston, Mass. Received for publication June 12, 1986. Accepted for publication July 25, 1986. Address for reprints: Hermes C. Grillo, M.D., Massachusetts General Hospital, Boston, Mass. 02114.
with pain in the shoulder and/or arm. Eight patients had computed tomographic (CT) scans performed. CT was not available in patients seen earlier in the series. CT scans were performed on an EMI 7070 or Techicare 2060 scanner at 10 mm intervals without contrast. The CT scans were reviewed by a thoracic radiologist (l.S.) and correlated with the operative findings. There were 11 women and 10 men with a mean age of 53 years (range 32 to 69). All but two patients had pain about the shoulder, 11 patients had pain radiating down the arm, and five had Homer's syndrome. One patient had bilateral Pancoast tumors treated 6 years apart. The tumor was on the right side in 14 and on the left in eight. Twenty patients had a preoperative needle biopsy, and the results indicated carcinoma in 19. All patients underwent preoperative bronchoscopy and mediastinoscopy with normal results. All but one had preoperative irradiation with a mean dose of 3,800 rads (range 3,000 to 4,500). The operation consisted of en bloc removal of the apical chest wall with the underlying lung as reported by Shaw, Paulson, and Kee.' Eighteen lobectomies and four segmentectomies or wedge excisions were performed. In four patients ribs 1 and 2 were removed, in seven patients ribs 1 through 3 were removed, in eight patients ribs 1 through 4 were removed, and in three patients ribs 69
The Journal of
Wright et al.
Thoracic and Cardiovascular Surgery
(From 30 Days After Surgery)
Fig. 1. Probability of survival on the basis of 21 patients with Pancoast's tumor after combined therapy.
1 through 5 were removed. Chest wall reconstruction was not performed. Thirteen patients required removal of the lower trunk of the brachial plexus. In four patients the subclavian artery was resected (in three of them the artery was repaired, end to end in two instances and with an interposition graft in one instance) and in five patients a portion of the vertebral body was resected because of local invasion. In five patients tumor had to be left behind within the brachial plexus or the subclavian artery. Eight tumors were squamous cell carcinoma, two were adenosquamous, seven were adenocarcinoma, two were large cell, and three were undifferentiated. All lymph nodes were normal both at mediastinoscopy and at pathologic examination. Five patients had marked radionecrosis of the tumor on pathologic review. Ten patients had invasion of the rib confirmed pathologically, and the remaining 12 had documented pleural invasion. Eleven patients received postoperative irradiation, approximately 2,200 R, because of diseased margins (five patients) or close margins (six patients). Survival was calculated by the Kaplan-Meier" actuarial method using data on operative survivors only.
Results There were three postoperative deaths, of which one was probably preventable. This patient was a 59year-old man whose death was caused by a bronchopleural fistula and empyema developing in the early postoperative period. An unusual fulminant pneumonia and an
empyema (without a bronchopleural fistula) leading to sepsis were the other two causes of death.There were two complications. One patient had a deep venous thrombosis. The other patient had a tumor invading both the subclavian artery and vein, which were resected en bloc but not reconstructed. Gangrene of the arm developed and necessitated amputation. Five patients had persistent local pain postoperatively that was difficult to eradicate. Four of these patients had tumor involving the brachial plexus, which necessitated resection. Follow-up was complete on all but one patient. Median survival was 24 months and actuarial survival rate was 63% at 1 year, 55% at 3 years, and 27% at 5 years (Fig. 1). Four of the five patients with radionecrosis of the tumor survived more than 2 years. Three patients of 10 with rib invasion were long-term survivors. Of five patients who had vertebral body resection for documented invasion, one is alive at 12 months and one at 36 months. Of the four patients with subclavian artery resection, one patient is alive at 13 months and one at 29 months. Cell type did not appear to influence survival. Palliation of pain was achieved in 72% of the patients and lasted for the duration of their survival. Three patients were studied by CT before and after radiotherapy. In each of these patients CT demonstrated a decrease in the size of the mass (Figs. 2 and 3). Of the eight patients who had CT scans, two had pathologically confirmed chest wall invasion, in both instances correctly diagnosed by CT as evidenced by rib destruction. The_ remainder of the scans were considered indeterminate for chest wall invasion (the mass abutted the chest wall without a discrete chest wall mass or rib destruction), and the tumors were found to be noninvasive at operation. None of the patients who underwent operation had evidence of vertebral body invasion on CT. CT results in three of the patients were considered indeterminate for invasion of the vertebrae because the mass abutted the spine, but one of these tumors was invasive at operation (Fig. 4). Four patients who had documented invasion of the subclavian artery underwent CT scanning, and only one of these scans was correctly interpreted as suggesting vascular invasion (Fig. 5). Two CT scans disclosed enlarged right paratracheal nodes, 1 to 1.5 cm in size, which were considered possibly malignant but were subsequently proved benign. Six patients had normal-sized nodes «1 em) on CT and all these were benign as well.
Discussion Our results are similar to those reported by Paulson,' Shaw,' Miller,' Attar," Stanford,' and their colleagues,
Volume 94 Number 1 July 1987
Superior sulcus lung tumors
Fig. 2. CT scan of right apical mass invading a rib posteriorly (arrow) obtained before irradiation. (From Dedrich CG. The Solitary Pulmonary Nodule and Staging of Lung Cancer. In: McLoud TC, ed. Clinics in Chest Medicine. Vol 5. Ph!ladelphia: WB Saunders, 1984:345-64. Reproduced with permission of WB Saunders.)
Fig. 3. CT scan from patient in Fig. 2 after irradiation, demonstrating a decrease in the size of the mass.
asseen in Table I. We strongly believe that mediastinoscopy should be included in the preoperative staging, as both Paulson and Miller have reported no long-term survivors with nodal disease. Of interest are our results with subclavian artery and vertebral body involvement, which suggest that there is some benefit to local resection. Our good results in those patients with radionecrosis of the tumor confirm the findings of Paulson. On the basis of the results of these five surgical series, it is difficult to explain the poor results by Kirsh and associates? in 1973, who reported no 5 year survivors with combined therapy. Of additional interest are the recent results reported
with radical irradiation alone in operable disease. Komalei and colleagues" in 1981 reported a 23% overall 5 year survival rate and a 45% 5 year survival rate in 17 patients in whom they could obtain local control (Table I). Symptomatic relief was obtained in 86% of the patients for a median duration of 12 months. Van Houtte and co-workers?in 1984 reported an 18% overall 5 year survival rate and a 40% 5 year survival rate in 13 patients who had no rib erosion and noted that there were no survivors if the rib was involved (Table I). Symptomatic relief was obtained in two thirds of the patients for a median duration of 11 months. Both groups of authors concluded that high-dose (>5,000
7 2 Wright et al.
The Journal 01 Thoracic and Cardiovascular Surge!)
Fig. 4. Right apical mass abutting vertebral body and chest wall without evidence of bone erosion on CT scan. Vertebral body invasion was demonstrated at operation but rib was not invaded.
Fig. 5. CT scan demonstrating encasement of the inominate artery with obliteration of mediastinal fat planes (arrow).
cGy), wide-margin irradiation was required and neither reported any major complication. The brain was noted to be the most common site of distant relapse. Both groups had local failure rates of about 50%, and there were no long-term survivors among patients with local failure. In Komaki's series," of the 19 patients with local failure, 10 had local failure alone without extrathoracic metastases. In Van Houtte's series, 14 patients had local failure and 10 had local failure alone. These results indicate a significant shortcoming in the ability of irradiation to sterilize the local tumor area without the addition of surgical intervention. Furthermore, combined therapy provided longer palliation of pain than
irradiation alone. All seven of our patients who died with disease postoperatively had distant metastases and none had local failure only. We interpret these results to mean that the best attempt at curing the patient involves both irradiation and radical resection, because neither surgical therapy'? nor irradiation alone produces satisfactory local control rates, which are prerequisites for long-term survival. Whereas others have advised that local arterial involvement precludes resection and long-term survival, two of four patients in our series survived more than 1 year after local arterial resection and reconstruction. Despite a presumed abundant collateral blood supply
Superior sulcus lung tumors
Number 1 July 1987
Fig. 6. CT scan illustrating left apical mass invading mediastinal fat (arrow), vertebral body (arrowheads), and a posterior rib (open arrow). From Shepard JO. Computed Tomography of the Mediastinum. In: McLoud TC, ed. Clinics in Chest Medicine. Vol 5. Philadelphia: WB Saunders, 1984:291-305. Reproduced with permission of WB Saunders.)
about the subclavian artery, we would advise immediate reconstruction, as the procedure is relatively simple and the result was unacceptable in the one patient in whom the artery was not reconstructed. Needle biopsy was helpful in our series in providing a diagnosis of carcinoma in 19 of 20 patients. No pneumothoraces occurred, probably in part because of adherence of the tumor to the chest wall. Although there were no cases of small cell carcinoma in our series, others have reported an incidence of 6%6 to 10%9 among patients with a Pancoast tumor. Among patients with small cell carcinoma, Pancoast's syndrome is reported to occur with an incidence from 3%11 to 4.5%.12 Although the majority of patients with Pancoast's syndrome have a pulmonary malignancy, some have metastatic nonpulmonary malignancies," nonpulmonary malignancies, 14 orbenign causes of the syndrome." Thus in a minority of patients treatment would be markedly altered by a needle biopsy result. It is also reassuring to have a positive biopsy before irradiation, especially in the rare case of a patient whose condition becomes inoperable during irradiation. CT scanning was helpful in assessing the extent of chest wall and vertebral body involvement and suggesting the absence of mediastinal nodal disease. For example, patients with gross invasion of the vertebral
Table I. Results of combined therapy and irradiation alone in patients with Pancoast tumors Authors
Combined irradiation and operation Attar et al. (1979) 19 Miller et al. (1979) 25 Paulson (1985) 78 Stanford et al. (1979) 16 Wright (1987) 21 Irradiation alone Komaki et al. (1981) 36 Van Houtte et al. (1984) 31
Survival at 5 yr 23% at 3 yr 40% 35% (44% if NO)* 50% 27% 23% 18%
'NO denotes no lymph nodal metastases.
body (Fig. 6) are not operative candidates, whereas those with tumor abutment are. CT scanning was not particularly predictive of subclavian artery invasion, although with coronal reconstruction and intravenous contrast it may be more definitive. 16. 17 Abutment on CT scan does not necessarily indicate nonresectability and in fact often predicts successful resection. Because of the multiple overlying structures at the apex of the chest, CT scanning gives the surgeon the best estimate of the extent of the lesion":" and probably allows for more accurate prediction of ability to do a total resection. The
Wright et of.
use of magnetic resonance imaging, which provides direct coronal and sagittal sections, will probably be of additional benefit and we are currently investigating this imaging method. It may be of particular value in showing arterial invasion. Our results support the combined approach by irradiation and surgical therapy to Pancoast tumors, including those locally involving the vertebral body or subclavian artery. Needle biopsy is a helpful adjunct. CT scanning, while helpful and better than plain films, is often indeterminate regarding invasion of chest wall structures. Irradiation alone does not produce equivalent results. 1.
REFERENCES Paulson DL. The "superior sulcus" lesion. In: Delarue N, Eschapasse H, eds. International trends in general thoracic surgery, vol. I. Lung cancer, Philadelphia: WB Saunders, 1985: 121-31. Shaw RR, Paulson DI, Kee JL. Treatment of the superior sulcus tumor by irradiation followed by resection. Ann Surg 1961;154:29-40. Miller JI, Mansour KA, Hatcher CR. Carcinoma of the superior pulmonary sulcus. Ann Thorac Surg 1979;28:447. Attar S, Miller JE, Satterfield J, et al. Pancoast's tumor: irradiation or surgery? Ann Thorac Surg 1979;28:57886. Stanford W, Barnes RP, Tucker AR. Influence of staging in superior sulcus (Pancoast) tumors of the lung. Ann Thorac Surg 1979;29:406-9. Komaki R, Roh J, Cox JD, daConceicao AL. Superior sulcus tumors: results of irradiation of 36 patients. Cancer 1981;48:1563-8.
The Journal of Thoracic and Cardiovascular Surgery
7. Van Houtte P, MacLennon I, Poulter C, Rubbin P. External radiation in the management of superior sulcus tumor. Cancer 1984;54:223-7. 8. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:45781. 9. Kirsh MM, Dickerman R, Fayos J, et al. The value of chest wall resection in the treatment of superior sulcus tumors of the lung. Ann Thorac Surg 1973;15;339-46. 10. Paulson DL. Carcinoma in the lung. Curr Probl Surg I 967:(Nov):44. 11. Byrd RB, Carr DT, Miller WE, Payne WS, Woolner LB. Radiographic abnormalities in carcinoma of the lung as related to histological cell type. Thorax 1969;24:573-5. 12. Johnson DH, Hainsworth JD, Greco FA. Pancoast's syndrome and small cell lung cancer. Chest 1982;82:6026. 13. Omenn GS. Pancoast syndrome due to metastatic carcinoma from the uterine cervix. Chest 1971;60:268-70. 14. Wilson KS, Cunningham TA, Alexander S. Myeloma presenting as Pancoast's syndrome. Br Med J 1979; 1:20. 15. Stathatos C, Kontaxis AN, Zafiracopoulos P. Pancoast's syndrome due to hydatid cysts of the thoracic outlet. J THORAC CARDIOVASC SURG 1969;58:764-8. 16. Webb RW, Jeffery RB, Godwin JD. Thoracic computed tomography in superior sulcus tumors. J Comput Assist Tomogr 1981;5:361-5. 17. Hamlin DJ, Burgenen A. CT, including sagittal and coronal reconstruction, in the evaluation of Pancoast tumors. CT 1982;6:43-50. 18. O'Connell RS, McLoud TC, Wilkins EW. Superior sulcus tumor: radiographic diagnosis and workup. AJR 1983;140:25-30.