Prospective evaluation of mediastinoscopy for assessment of carcinoma of the lung

J

THORAC CARDIOVASC SURG

91:53-56, 1986

Prospective evaluation of mediastinoscopy for assessment of carcinoma of the lung Between 1979 and 1984, mediastinoscopy was performed on 1,000 of the 1,500 patientsadmitted to the Thoracic Surgical Service of the Toronto General Hospital with the diagnosis of carcinomaof the lung. In 144 cases, concomitant anterior mediastinoscopy was also performed. Abnormal mediastinal nodeswere found in 296 (29.6% ~ The overall compHcation rate was 2.3 %, with no deaths, Mediastinoscopy revealed diseased nodes in 24% of patients with squamous ceU carcinoma, 29% with adenocarcinoma, 54% with smaU ceO lDIdifferentlated carcinoma, 31% with large cell lDIdifferentlated carcinoma, and 12% with bronchoalveolar carcinoma. Abnomma.l mediastinal nodes werefound with equalfrequency in right- and left-sided tumors and occurred in 31% of tumors in the main bronch~ 25% of upper lobe tumors, and 17% of lower lobe tumors. Of the 704 patients having negative results of mediastinoscopy, 590 were subjected to thoracotomy. Ninety-three percent underwent resection (85% curative, 7% palliative) and 7% bad unresectable tumors. Of the resections, 20 % werepneumonectomies. At thoracotomy, 52 of the 590 patients with negative mediastinoscopic results were found to have abnormal mediastinal nodes. Sixty-two of the 296 patients with positive results of mediastinoscopy were selected for thoracotomy. Fighty-six percent had resectable lesio~ (67% curative, 18% palliative) and 14% unresectable. The pneumonectomy rate in this groupwas 35 %. These current data supportour previous opinion that routine mediastinoscopy can be done with negligible morbidity and provides essential information for the classification and management of cancer of the lung.

William P. Luke, M.D.* (by invitation), F. Griffith Pearson, M.D., Thomas R. J. Todd, M.D. (by invitation), G. Alec Patterson, M.D. (by invitation), and Joel D. Cooper, M.D., Toronto, Ontario, Canada

Although surgical resection remains the mainstay of therapy for carcinoma of the lung, the overall 5 year survival rate for all cases of lung cancer has improved only imperceptibly over the past five decades and is currently 13%.I This poor prognosis is largely related to the fact that only approximately 20% of patients with lung cancer have localized disease when first seen.' Because of this low incidence of localized disease, strict staging is essential to provide therapeutic guidelines and to permit

standardization and comparison of therapeutic modalities. Since 1963, the Thoracic Surgical Service at the Toronto General Hospital has used mediastinoscopy routinely as a method for staging presumably operable carcinoma of the lung. The present study is a prospective analysis of 1,000 consecutive mediastinoscopies performed over the past 5 years to evaluate the role of this procedure in the current management of carcinoma of the lung.

Patients and methods From the Division of Thoracic Surgery, Toronto General Hospital, Toronto, Ontario, Canada. Read at the Sixty-fifth Annual Meeting of The American Association for Thoracic Surgery, New Orleans, La., April 29-May 1, 1985.

Address for reprints: Joel D. Cooper, M.D. 10-226 Eaton Building, Toronto General Hospital, 200 Elizabeth St., Toronto, Ontario, Canada M5G I L7. 'Current address: William P. Luke, M.D., Cape Cod Hospital, Hyannis, Mass. 0260 I.

Between 1979 and 1984, mediastinoscopy was performed in 1,000 of the 1,500 patients admitted to the Thoracic Surgical Service of the Toronto General Hospital. Results were prospectively entered into a previously established data base. In all patients, staging was done before and after operation according to the TNM classification of the American Joint Committee for Cancer Staging and End Results Reporting.' For patients having a thoracotomy, hilar and mediastinal 53

54

The Journal of Thoracic and Cardiovascular Surgery

Luke et af.

Table

Table I. Pathology by cell type eel/type Squamous cell Adenocarcinoma Small cell anaplastic Large cell undifferentiated Bronchoalveolar Mixed Other

Table

Negative mediastinoscopy

Positive mediastinoscopy

338

112 84 67

191

SS 46

22

23

4 3 4

14 37

n Reason lor inoperability

No.

cy;

Extrathoracic spread Cell type Pulmonary status Age Cardiac status Patient refused operation Coexisting disease Positive pleural cytology

32 28

2S

28

]6 ]4 10

]4 12

8

7 4 1

9

5

m. Resectability rate after mediastinoscopy Thoracotomies

Resections

Negative mediastinoscopies (N = 704) Positive mediastinoscopies (N = 296)

590 62

549

52

931 85.5

Totals

6S2

601

92.2

Table IV. Type of resection

Nonresectable Curative Palliative Not recorded

Negative mediastinoscopy

Positive mediastinoscopy

IN = 590)

IN=(j2)

41 (6.9'k) 504 (85A'Yr) 37 (6.2'Yr) 8 (1.3%)

9 (146%) 42 (67.3'Yr) 11 (18.1%)

o

lymph nodes were sampled and mapped according to the American Thoracic Society's "Clinical Staging of Primary Cancer of the Lung.?' There were 778 men and 222 women. The mean age was in the seventh decade of life. It is the policy of our service to do cervical mediastinoscopy in virtually all patients with presumably operable lung cancer. When the tumor is located in the upper lobe of the left lung, an anterior mediastinoscopy is usually added if results of the cervical mediastinoscopy are negative. In this group of 1,000 patients undergoing cervical mediastinoscopy, 144 had concomitant anterior mediastinoscopy.

Results Of the 1,000 mediastinoscopies, 704 (70.4%) revealed no evidence of tumor within the mediastinum. In 296 patients (29.6%), lymph nodes were found to contain tumor. These percentages are consistent with other large series and have not changed over the past two decades." The chest x-ray films were reviewed for all patients with

Resection rate (

positive mediastinoscopic findings to determine the status of the mediastinum. In 72% no abnormality of the mediastinum was noted. In the group of 590 patients with negative mediastinoscopic findings who underwent thoracotomy, 52 patients were found to have abnormal mediastinal nodes that were not recognized at the initial mediastinoscopy. This represents a false-negative rate of 8.9%. In most cases, however, the abnormal nodes found at thoracotomy were in locations that are inaccessible at mediasti noscopy.' The cell type of the 1,000 tumors is listed in Table I. Diseased mediastinal nodes were found in 54% of patients with small cell anaplastic carcinoma, 31% with large cell undifferentiated carcinoma, 29% with adenocarcinoma, 24% with squamous cell carcinoma, and 12% with bronchoalveolar carcinoma. Abnormal mediastinal nodes were associated with equal frequency with right- and left-sided tumors. They occurred in conjunction with 34% of tumors of the main bronchus, 26% of upper lobe tumors, and 17% of lower lobe tumors.

Operability A total of 652 patients underwent thoracotomy. This represents a resection rate of 41 % of all patients seen with carcinoma of the lung and 65% of all patients who underwent mediastinoscopy. In the group of 704 patients having negative resultsof mediastinoscopy, 590 (84%) underwent thoracotomy and 114 (16%) were considered inoperable. The reason for inoperability in this group of patients are summa,

Volume 91 Number 1

Mediastinoscopy 55

January, 1986

Table V. Extent of resection Negative mediastinoscopy Curative

Wedge

Lobectomy Wedge plus lobectomy Bilobectomy Sleeve lobectomy Pneumonectomy Other

12 265 24 32 39 100 39

I

rized in Table II. The resection rate in these patients was 93%; 86%of these resections wereconsidered "curative" and 7% palliative (Table III). A positive mediastinoscopic result is considered a relative, but not absolute contraindication to thoracotomy by our service," Sixty-two of 296 patients with positive resultswere selected for thoracotomy. In general, these were patients with non-small cell carcinomas with ipsilateral, intranodal disease only. This highly selected subgroup constitutes approximately 20% of all patients with positive results of mediastinoscopy. In this selected group of 62 patients, 53 (85%) had technically resectable lesions. Forty-two (67%) of the resections were considered "curative" and 11 (18%) palliative. The type and extent of resections are summarized in Tables IV and V. Pneumonectomy was required in 20% oftheresections in the group with negative mediastinoscopic results compared with 35% in the group with positive results.

ComplicatioQS of mediastinoscopy Inthisseries of 1,000mediastinoscopies there were no deaths attributable to the procedure. There were 23 complications for an overall morbidity rate of 2.3%. Three major complications necessitated thoracotomy, two for hemorrhage and one for tracheal injury. The remaining complications were minor and included infection, pneumothorax, and postoperative arrhythmias. Although recurrent nerve injury has been observed and reported by us in a previous series,' it did not occur in this series. The complications are listed in Table VI. All patients with resectable N2 disease were referred for postoperative radiation therapy and those with N2 disease whowas not operated upon were treated with a variety of chemotherapy and radiotherapy protocols.

Discussion Prognosis in patients with lung cancer depends upon the stage of the disease at the time of treatment as well

Positive mediastinoscopy

Palliative

Curative

4 12 1 4 4 11 2

0 15 I 1 6 16 2

I

Palliative

1 5 0 0 I 2 2

Table VI. Complications in 1,000 consecutive mediastinoscopies* Complication Major Hemorrhage Tracheal injury Minor Pneumothorax Wound infection Recurrent nerve palsy Other Total

No. 2 I

6

5

o 9 23 (2.3%)

'There were no deaths as a result of mediastinoscopy.

as the cell type. It is nowwidely recognized that patients with NO disease have an excellent prognosis if a curative resection is performed. T1 NO tumor will be cured in about 80%of patients.Patients with T2 NOtumors have a somewhat less favorable prognosis. Once tumor spreads to lymph nodes, the prognosis is significantly worse. In patients having thoracotomy for NI disease, the 5 year survival rate is 30% or better. In contrast, the 5 year survival rate of patients undergoing thoracotomy with N2 disease is considerably worse. In several seriesof patientswith N2 disease whounderwent thoracotomy without preoperative selection by mediastinoscopy, overall survival was no more than 10% if operative mortality and unresectable cases are included in the analysis. We9 have previously demonstrated that patients having negative mediastinoscopic fmdings but N2 disease at thoracotomy have a cumulative 5 year survival rate of 24%. However, in a highly selected group of patients undergoing thoracotomy after having positive fmdings on mediastinoscopy, the cumulative 5 year survival rate was only 9%. One can presume, therefore, that if thoracotomy were undertaken in all patients having positive mediastinoscopic fmdings, the 5 year survival rate would be exceedingly low. Thus, the use of mediastinoscopy contributes to accurate preoper-

56

The Journal of Thoracic and Cardiovascular Surgery

Luke et al.

ative staging as well as subclassification of Stage III tumors. This information is useful in selecting patients for operation, and it is important when comparing results of therapy between different institutions and for planning appropriate protocols for adjuvant therapy. Controversy continues about the use of invasive versus noninvasive staging." Mediastinal tomography, computed tomographic scanning, nuclear magnetic imaging, and transbronchoscopic needle aspiration all have their advocates.":" In our opinion, however, none of these methods has yet been sufficiently accurate to replace mediastinoscopy. We are currently conducting a prospective, blind study comparing sensitivity and specificity of computed tomographic scanning, magnetic resonance imaging, and mediastinoscopy in 100 consecutive patients with carcinoma of the lung. We hope this will further clarify the respective values of each method. Our current study confirms that mediastinoscopy is safe and has a specificity of 100% and a sensitivity of 91 %. It remains the most accurate method for pretreatment staging and classification of mediastinal lymph node involvement. We are indebted to Mr. Mel Cohen, Mrs. Wendy McGonigal, and Ms. Ann Botelho for assistance with data retrieval and preparation of this manuscript. REFERENCES I Silverberg E: Cancer statistics 1985. CA 35: 19-35, 1985 2 Anderson RW: Carcinoma of the lung. Surg Clin North Am 60:793-814, 1980 3 Task Force on Carcinoma of the Lung: Staging of lung cancer 1979, Chicago, 1979, The American Joint Committee for Cancer Staging and End Results Reporting 4 American Thoracic Society Medical Section of the American Lung Association: Clinical Staging of Primary Lung Cancer. Am Rev Respir Dis 127:654-659, 1983 5 Pearson FG: An evaluation of mediastinoscopy in the management of presumably operable bronchial carcinoma. J THORAC CARDIOVASC SURG 55:617-625, 1968 6 Deneffe G, Daenen W, Stalpaert G: Experience with loo consecutive anterior mediastinotomies in lung cancer. Acta Chir Belg 6:509-513, 1982 7 Ashraf MH, Milson PL, Walesby RK: Selection by

8

9

10

II

12 13

14

15

mediastinoscopy and long-term survival in bronchogenic carcinoma. Ann Thorac Surg 30:208-214, 1980 Coughlin M, Deslauriers J, Beaulieu M, Fournier B, Piroux M, Rouleau J, Tardif A: Role of mediastinoscopy in pretreatment staging of patients with primary lung cancer. Ann Thorac Surg (in press) Pearson FG, Delarue NC, lives R, Todd TRJ, Cooper JD: Significance of positive superior mediastinal nodes identified at mediastinoscopy in patients with resectable cancer of the lung. J THORAC CARDIOVASC SURG 83: I-II, 1982 Cooper JD, Ginsberg RJ: Mediastinoscopy in the preoperative evaluation in patients with bronchogenic carcinoma, in Current Controversies in Thoracic Surgery, Philadelphia, W. B. Saunders Company (in press) Ginsberg RJ, Nelerns JM, Pearson FB: Mediastinal tomography as an aid in surgical staging in lung cancer. Ann R Coli Phys Surg Can 7:45, 1974 Konn A, Konn FA, Garvey J: Oblique hilar tomography in mediastinoscopy. Chest 86:424-429, 1984 Breyer RH, Karstaedt N, Mills SA: Computed tomography for evaluation of mediastinal lymph nodes in lung cancer. Correlation with surgical staging. Ann Thorae Surg 38:15-20, 1984 Rea HH, Shevland JE, House AJS: Accuracy of computed tomographic scanning in assessment of the mediastinum in bronchial carcinoma. J THORAC CARDIOVASC SURG 81:825-829, 1981 Wang KP, Bower R, Huponik EF, Siegelman S: Flexible transbronchial needle aspiration for staging of bronchogenic carcinoma. Chest 84:571-576, 1983

Discussion DR. COOPER (Closing) I want to emphasize that we think mediastinoscopy not only stages tumor but classifies the N2 disease. It is universally accepted among surgeons that a diseased supraclavicular node is clearly a sign of inoperability; yet, among some, the presence of an abnormal node just a few centimeters more proximal in the same lymphatic chain, which can be discovered only by mediastinoscopy, is considered not worth knowing about. We disagree with that. I think as surgeons we become frustrated over the current lack of effective alternative therapy for patients with lung cancer. However, that frustration should not I be the rationale for inappropriate selection of patients for operations.