VOLUME 19 NUMBER 2 * FEBRUARY 1975

Transcription

1 THE ANNALS OF THORACIC SURGERY Journal of The Society of Thoracic Surgeons and the Southern Thoracic Surgical Association VOLUME 9 NUMBER 2 * FEBRUARY 975 Medias tinoscopy Its Application in Central Versus Peripheral Thoracic Lesions William Stanford, Col, Stephen Steele, Capt, Raymond G. Armstrong, Col, and Gordon L. Larsen, Col, all MC, USAF ABSTRACT In an attempt to ascertain the value of mediastinoscopy in peripheral lung lesions, records of 57 patients undergoing cervicomediastinal exploration (CME) at Wilford Hall USAF Medical Center were reviewed. Among patients with benign lesions, CME was positive in 90.6% of those who had central lesions and 58.3% of those with peripheral lesions. It was positive in all 7 patients who had peripheral lesions with associated mediastinal nodes on roentgenogram and negative in all 5 who had peripheral lesions without nodes. In the patients with malignant lesions, CME was positive in 72.9% of those who had central lesions and 58.% of those with peripheral lesions. It was positive in 24 of 27 patients who had peripheral lesions with associated mediastinal nodes and negative in 5 of 6 patients with peripheral lesions without nodes. Although we recognize this to be a selected series, CME does appear to be valuable in patients with central lesions and peripheral lesions with mediastinal nodal involvement on roentgenogram. It does not appear to be as useful in those with peripheral lesions who do not have central nodal involvement. M ediastinoscopy is a well-recognized and effective method for diagnosing benign and malignant central thoracic lesions [l, 2. Its value in peripheral lung lesions is not as clear. In an attempt to ascertain the value of routine mediastinoscopy in patients with peripheral lung lesions, all patients undergoing mediastinoscopy at Wilford Hall USAF Medical Center, Lackland AFB, Tex., were studied. From the Thoracic Surgery Service, Department of Surgery, Wilford Hall USAF Medical Center (AFSC), Lackland AFB, Tex. Accepted for publication Sept. 24, 974. Address reprint requests to Dr. Stanford, Thoracic Surgery Service, Wilford Hall USAF Medical Center (AFSC), Lackland AFB, Tex

2 STANFORD ET AL. C in i ca Mat eria Between April, 967, and May, 972, 44 patients with mediastinal or pulmonary lesions or both were seen at Wilford Hall USAF Medical Center. Of the 44, 60 underwent mediastinoscopy and cervicomediastinal exploration (CME). Records were available for 57 of the 60 (98.yo), and data gathered from these records form the basis of this report. Of the 57 patients, 30 were men and 27 were women; ages ranged from 6 through 74 years. In this series a positive tissue diagnosis by either thoracotomy or CME was made in every patient but ; 5% of the lesions were classified as malignant (80 of 56) and 48% as benign (76 of 56). A diagnosis by CME was possible in 7 of the 57 (74.5y0), including a positive diagnosis in 85.5% of patients with benign lesions and 65y0 of those with malignant lesions. The group was further subdivided according to central versus peripheral lesions. Any lesion in the hilum, mediastinum, or area adjacent to it was classified as a central lesion, while any lesion separated from the hilum or mediastinum by intervening lung tissue was classified as a peripheral lesion. Utilizing these criteria we found 56 of 56 lesions (35.8y0) to be peripheral and 0 of the 56 to be central (64.7%). BENIGN DISEASE Twelve of 76 patients with benign disease had peripheral lesions (5.7y0). Of these, CME was diagnostic in 7 (58.3y0). Each of these 7 showed associated mediastinal nodes on roentgenogram. CME was not diagnostic in any of the 5 patients with no mediastinal involvement on roentgenogram. Of the 0 patients with central lesions, 64 (84.270) had benign disease; CME was diagnostic in 58 (90.6y0). MALIGNANT DISEASE Among those with malignant disease, 43 of 80 patients had peripheral lesions (53.7y0) and CME was diagnostic in 25 (58.y0). This group includes 24 of the 27 patients with peripheral lesions and mediastinal involvement and of the 6 patients with a peripheral lesion and no mediastinal involvement. Three of the 27 with peripheral lesions and mediastinal widening had a negative CME, whereas 5 of the 6 with peripheral lesions without mediastinal widening had a negative CME. In the 37 patients with malignant central lesions, mediastinoscopy was diagnostic in 27 (72.9y0). There was patient with a peripheral lesion and a negative CME in whom a diagnosis was never made. The number of benign versus malignant lesions tabulated by patient age is seen in Table. The percentage of patients with positive nodes on 22 THE ANNALS OF THORACIC SURGERY

3 Mediast in oscopy TABLE. TYPE OF LESION TABULATED WITH PATIENT AGE Type of Lesion Age Benign Malignant Below Over 60a Diagnosis unknown, CME categorized by cell type is seen in Table 2. The breakdown of central versus peripheral lesions with respect to diagnosis is seen in Table 3. Comment This series represents a selected group of patients typical of those seen in a large referral practice. Selection played a role, in that those patients suspected of having carcinoma were more often subjected to mediastinoscopy than those in whom the lesion was thought to be benign. Patients with central lesions and those with large peripheral lesions underwent mediastinoscopy more often than those with small lesions located well out in the periphery. Our policy is not to explore peripheral lesions with laminated calcification, since we believe these are most likely to be benign. However, we do explore peripheral lesions that contain a speck of calcification, especially in patients who smoke and in those over 40 in whom no previous films are available for comparison. In the latter group we found 2 patients with a malignancy which happened to encompass an old granuloma. Finally, TABLE 2. NUMBER OF POSITIVE NODES BY CELL TYPE Disease No. of Patients Sarcoidosis Squamous cell carcinoma 20 3 Adenocarcinoma 7 9 Undifferentiated carcinoma 0 9 Oat cell carcinoma 9 6 Small cell carcinoma 8 3 Hodgkin s disease 6 6 Caseating granuloma 6 5 Miscellaneous malignant lesions 0 6 Miscellaneous benign lesions 5 6 Unknown 0 Positive Nodes (CME) No. 9, I VOL. 9, NO. 2, FEBRUARY,

4 STANFORD ET AL. TABLE 3. LOCATION OF LESION WITH RESPECT TO DIAGNOSIS IN 57 PATIENTS Disease Peripheral, Peripheral, with Nodes without Nodes Central on X-ray on X-ray Benign Sarcoidosis 52(5) 3(3) Caseating granuloma (TB) 2(2) 3(3) (0) Reactive hyperplasia 40)) (0) Abscess (Corynebacterium) (0) Coccidioidom ycosis 2(2) V) Pseudolymphoma (0) Hamartoma () Histoplasmosis 2() Thymic cyst () Atypical tuberculosis (0) Total 64(58) 7(7) 5(0) = 76(65) Malignant Squamous cell *(4) 8(8) 4() Adenocarcinoma 4(2) 9(7) 4(0) Undifferentiated 4(4) 5(5) (0) Oat cell 6(4) 2(2) (0) Small cell 4(2) 2() 2(0) Hodgkin s disease 6(6) Alveolar cell l(0) Malignant thymoma () Large cell () () (0) Metastatic 2(2) 204 Clear cell () Total 37(27) Unknown ( ) =CME positive. 27(24) 6() = 80(52) (0)= l(0) 57( 7) we see a large number of patients with sarcoidosis because of the large recruit population at Lackland AFB. With these factors in mind, some comparisons can be made. BENIGN DISEASE As can be seen from Table 3, CME was diagnostic in 5 of 52 patients with central involvement from sarcoidosis as well as in all 3 patients with peripheral lesions who had associated central nodal involvement on roentgenogram. In caseating granuloma (tuberculosis), CME was useful in the 2 patients with central and 3 with peripheral lesions with nodes. It was not helpful in the individual who had a single peripheral lesion without nodes. The same was true for patients with coccidioidomycosis, histoplasmosis, thymic cyst, pseudolymphoma, corynebacterial abscess, hamartoma, and atypical tuberculosis. Among the patients with benign lesions, CME was positive in all 7 who had peripheral lesions with central nodes; it was negative in the 5 who had peripheral lesions without nodes. 24 THE ANNALS OF THORACIC SURGERY

5 Mediastinoscopy MALIGNANT DISEASE In patients with squamous cell carcinoma, CME was positive in 4 of the 8 with central lesions and in all 8 who had peripheral lesions with nodes. It was negative in 3 of 4 patients who had peripheral lesions without nodes. In patients with adenocarcinoma, CME was positive in 2 of 4 who had central lesions and in 7 of 9 who had peripheral lesions with nodes. It was negative in all 4 who had peripheral lesions without nodes. In the patients with anaplastic disease, comprising undifferentiated, large cell, small cell, and oat cell carcinomas, CME was positive in of 5 patients with central lesions and 9 of 0 who had peripheral lesions with nodes. It was negative in 5 of those who had peripheral lesions without nodes. The patients with Hodgkin s disease, malignant thymoma, and clear cell carcinomas all had central lesions, and CME was diagnostic in all. Two patients with central metastatic lesions had positive results on CME, and 2 who had peripheral metastatic lesions without nodes showed negative results. The single patient with peripheral alveolar cell carcinoma without central nodes also had a negative test. Therefore, in the category of malignant lesions, CME was negative in 5 of 6 patients with peripheral lesions without central nodal involvement and was positive in 24 of 27 patients with peripheral lesions with central nodal involvement. INFLUENCE OF SELECTION In an attempt to ascertain the influence of selection-i.e., whether we were missing patients because we did not do a CME in every case-we looked at the 28 patients who underwent primary thoracotomy without preliminary CME over the same time period. In this group there were 47 malignant lesions. Of the 47 patients, 03 underwent curative resection (no mediastinal involvement), leaving only 44 with inoperable disease or mediastinal involvement, or both. Four of these had primary mediastinal tumors that should have yielded a positive diagnosis on CME. The tumors in the remaining 40 consisted of both peripheral and central inoperable lesions. Conceivably some of these could have been missed on CME, but it is unlikely that any large patient population would have been eliminated by our selection process. To see how this selection affected the yield rate of positive biopsies, this series was compared to one by Trinkle and associates [3] which included 300 consecutive CMEs on patients with thoracic lesions. In their series two-thirds of the lesions were malignant and one-third benign, whereas in ours they were about equal. Of the 03 patients with benign disease reported by Trinkle, a positive diagnosis was gained in 26y0 versus our 84.4y0. In individual groupings they achieved 93.7y0 positive results in patients with sarcoidosis versus our 98.7y0, and 42.8y0 in those with granulomatous disease versus our 75y0 (combining typical and atypical tuberculosis, coccidioidomycosis, and histoplasmosis). VOL. 9, NO. 2, FEBRUARY,

6 STANFORD ET AL. TABLE 4. SUMMARY OF DATA GATHERED FROM 57 PATIENTS UNDERGOING CME Description of Lesion Central lesions Benign Malignant Peripheral lesions Benign with normal mediastinum Malignant with normal mediastinurn Benign with enlarged nodes Malignant with enlarged nodes No. of Patients CME Diagnostic % In patients with malignant disease Trinkle and colleagues reported an overall yield of 32y0 positive results as compared with our 65%. In cases of squamous cell carcinoma they reported yo positive versus our 65y0. For adenocarcinoma they found 50y0 positive versus our 57.9y0; and in patients with anaplastic lesions they found 60% versus our 63Y0. Our selection process would thus appear to improve the overall percentage yield for CME. However, in certain categories (sarcoidosis, adenocarcinoma, and undifferentiated carcinoma) the differences were less marked. From these data, summarized in Table 4, it appears that CME is of value in central lesions and in peripheral lesions associated with central nodal involvement on roentgenogram. It does not appear to be of value in peripheral lesions with no mediastinal enlargement. References. Ashbaugh, D. G. Mediastinoscopy. Arch Surg 00:568, Pearson, F. G., Nelems, J. M., Henderson, R. D., and Delarue, N. C. The role of mediastinoscopy in the selection of treatment for bronchial carcinoma with involvement of superior mediastinal lymph nodes. J Thorac Cardiovusc Surg 64382, Trinkle, J. K., Bryant, L. R., Hiller, A. J., and Playforth, R. H. Mediastinoscopy-experience with 300 consecutive cases. J Thorac Cardiovasc Surg 60:297, THE ANNALS OF THORACIC SURGERY