Document XzZLxkX0yQ7dV9n4EvMjOdN4B

MEDIASTINOSCOPY FOR ASSESSING MEDIASTINAL SPREAD IN CLINICAL STAGING OF CARCINOMA OF T H E LUNG EDWARMD. GOLDBERMCD,,* ARVINS. GLICKSMAMND,,+ FAZLURR. KHAN,M D , ~AND JAMES J. NICKSONM,D* A clinical staging for carcinoma of the lung utilizing conventional diagnostic procedures and mediastinoscopy for careful assessment of the upper mediastinum has been developed. In a n analysis of 144 patients, one of 83 right-sided tumors had contralateral spread compared to 7 of 61 for the left side. Bilateral spread was equal. Forty-three of 52 patients with squamous cell carcinoma had negative upper mediastinal nodes, 6 had ipsilateral, and 3 contralateral node involvement. Forty of 64 poorly differentiated tumors had involved nodes, and 28 had either contralateral or bilateral spread. Twelve of the 16 oat-cell cancers had involved nodes; 8 were bilateral. Eight of the 12 adenocarcinomas had involved nodes; 3 were bilateral. Seventy-three percent of the well-differentiated tumors fell into Stages 1 and 2; 83 percent of the anaplastic tumors in Stages 3 or 4. TH E MANAGEMENT OF CARCINOMA OF T H E lung is a taxing and unrewarding problem to the surgeon, radiotherapist, chemotherapist, and others involved in patient care. Even with the most exacting diagnostic techniques, sufficient information may not be available in suspected cases to provide a basis for rational treatment. In addition to the difficulty in providing a histologic diagnosis, the usual diagnostic procedures commonly fail to indicate the extent of spread or of operability. Often exploratory thoracotomy is considered necessary to obtain tissue for diagnosis and to determine the extent of disease.6.10,16,25-27 Thoracotomy as an exploratory procedure is not an innocuous procedure. Mortality sta- Presented at the twenty-second annual scientific session of the James Ewing Soaety, New York, N.Y., April 24-26, 1969. *From the Oncology Council and Department of Surgery, Michael Reese Hospital and Medical Center; Department of Surgery, Cook County Hospital, Chicago, Ill. , t Department of Radiotherapy, Mount Sinai Hospital, New York, N.Y. Z Oncology Council and Department of Radiation Therapy, Michael Reese Hospital and Medical Center, Chicago. Supported, in part, by USPHS Grant No. 59892 and N.I.H. Grant No.FR 5476. Address for reprints: Edward M. Goldberg. MD, Michael Reese Hospital. 2929 So. Ellis Ave., Chicago, Ill. 60616. Received for publication August 3, 1969. tistics may be as high as 10 per~ent.26~C~o7mplications following this procedure are another serious deterrent.26*26 Also, thoracotomy allows adequate evaluation only of the ipsilateral mediastinal nodes. Exploratory thoracotomies cannot be justified if less hazardous techniques are available to make a histologic diagnosis and to more accurately determine operability. For rational planning of therapy in patients with lung cancer, it is essential to know the extent of local growth and the involvement of lymph nodes, in general, and, in particular, those in the upper mediastinum. In 1959, Carlens examined mediastinal structures through a lighted instrument which he named the mediastinoscope.4 This approach had its origin in the techniques introduced by Harkens in 1954 and Radner in 1955. Harkens introduced the cervicomediastinal exploration. A lighted laryngoscope was utilized to expose a limited area on one side of the mediastinum to obtain material for tissue diagnosis.l3 Radner approached the mediastinum through a suprasternal incision. He opened the pretracheal fascia and with blunt dissection entered the upper mediastinum. He was thus able to remove accessible 'mediastinal lymph n0des.~4Since its introduction in 1959, mediastinoscopy has been used in many parts of the world and the 347 348 CANCERFebruary 1970 VOl. 25 results of well over 6,000 cases have been reported.5. 15. 17. 23 Mediastinoscopy can be carried out under general endotracheal or local anesthesia.12 Through a small transverse suprasternal notch incision, the strap muscles are separated along the mid-line to expose the anterior wall of the trachea. After opening the pretracheal fascia, blunt dissection is carried down along the trachea posterior to the innominate artery and the aortic arch (Fig. 1). T h e tissue is loose, permitting easy exposure of the entire trachea to its bifurcation (Fig. 2). T h e instrument is then introduced to visualize the trachea and paratracheal areas, the carina, right and left main bronchi, the right upper lobe bronchus, the innominate artery, the aortic arch, the right pulmonary artery, the azygos vein, the left recurrent nerve, and the esophagus from the neck to the level of the carina (Figs. 3, 4). Tumor masses and lymph nodes are readily enucleated in entirety or a specimen may be taken for biopsy. Bleeding is minimal and easily controlled by cautery, silver clip or gelfoam. Silver clips may also be used to identify the site of biopsy. T h e incision is closed primarily without drainage. Treatment of carcinoma of the lung is a major undertaking. Correct assessment of the indications for treatment demand an understanding of the prognostic factors which can be determined preoperatively. Berg, Gibbon, Noel, and Wernig have shown that patients with cancers confined to the lung parenchyma have a better prognosis than do patients who have tumors which extended outside the lung18 11.22935 Although a number of surgeons such as Abbey-Smith, Chamberlain, and PriceThomas have advocated extended resection for some of the more locally advanced cases, very few patients survive S years.6030~32 I n Berg's series, none were alive 5 years after resection, and no patient with extensive tumor and with mediastinal lymph node involvement was alive 12 months after pulmonary resection. Long-term survival in bronchogenic carcinoma has been correlated with the extent of lymph node involvement. When no lymph nodes were present, or only segmental or interlobar nodes were found, the prognosis was appreciably better than when the upper mecliastinal nodes were involved.', 22 Furthermore, the possibility of contralateral spread of cancer to nodes in the mediastinum has FIG. 1. Represents a view through the mediastinoscope at the level of the innominate artery with left paratracheal lymph nodes. No. 2 -MEDIASTINOSCOAPNYD STAGING Goldberg et al. 349 FIG. 2. Represents a view through the mediastinoscope at the level of the tracheal bifurcation with subcarinal lymph nodes and the right pulmonary artery. been recognized for some time. The first reported observation of the crossing of lymphatics from the left main bronchus to the right paratracheal nodes was made by Mascagni, in 1787.'8 Many morbid anatomists since then have demonstrated lymphatic crossover in the mediastinum.8.9. 2O. 28, 33.34 This was best classified by Rouviere, in 1932, who demon- FIG. 3. Represents the related upper mediastinal FIG. 4. Represents the related upper mediastha1 anatomy from an anterior view. anatomy from a right lateral view. 350 CANCERFebiw a y 1970 Vol. 25 strated that bifurcation nodes drain entirely ing bronchus wall at least 1 cm distal to into the right paratracheal chain and, thus, the carina; may have associated density of all drainage pathways from left lower lobe an entire lung. nodes cross over to the right paratracheal T3Tumors involving the bronchus wall nodes.28 within 1 cm of the carina or any tumors Cordier, in 1958, carried this a step fur- involving the parietal pleura. Multiple ther. He reported that crossover drainage oc- primary neoplasms in one lung. curs both from left to right and right to left. T4All tumors involving by contiguous growth Crossover occurred not only at the level of structures beyond the parietal pleura. the carinal nodes, but between the right and (No includes no lymph nodes proximal to the left paratracheal chain of nodes as well.8 right second division of the trachea.) From the above considerations, it seems No No nodes involved. clear that a clinical staging program, to be N1 Ipsilateral nodes only involved. of value, must identify with a high proba- N2 Contralateral or bilateral nodes involved. bility those patients who will benefit from Mo No evidence of extrathoracic disease. the established' modes of treatment and M1Evidence of extrathoracic disease, e.g., those for whom this would be a futile or supraclavicular nodes, liver, bone, etc. even hurtful exercise. The tumor is classified in the highest possible T group. The position and extent of growth PROPOSECDLASSIFICATION of the tumor rather than the size determine the T classification. This is in accord with the As stated by the International Union conclusions reached by Bignall and Moon, Against Cancer: "The immediate purpose of Noel, Berg, Cliffton, and Nickson and others.1, staging is to facilitate the accurate, concise 2871 2 ~ 2 C2 linically, i t is not always possible to description of the apparent extent of the distinguish between lesions which involve the disease in a way that can readily be com- parietal pleura but do not go beyond. This is municated to others or reproduced by them. so even with the most exacting x-ray planog- What is needed is simply an agreement on raphy. However, medial lesions within reach the recording of such precise information of the mediastinoscope may be so delineated. on the extent of the disease as to make pos- I n the past, Noel and Berg demonstrated sible the combination or re-combination of that segmental and interlobar nodes did not cases according to any agreed plan."l4 appreciably influence the prognosis one way Ideally, a staging technique should describe or the other.l.22 However, the involvement of the primary site, the presence of cancerous mediastinal nodes has been recognized by nodes and their location, and the presence Saltzer, Cliffton, and others as being im- or absence of metastasis. This makes an as- portant indicators of the prognosis for individ- sessment of treatment plan possible and per- ual cases.'. 7922.29 This involvement may in- mits comparisons of treatments by different clude any one of the 3 potential chains de- modalities and in different institutions. The scribed by Rouviere-the anterior mediastinal classification of carcinoma of the lung pro- or prevascular, the posterior mediastinal or posed here is based upon the T, N, and` M esophageal, or the paratracheal chains. I n the system. It utilizes the conventional diagnos- classification proposed here, nodes proximal to tic procedures plus mediastinoscopy. the right second division of the trachea are not considered mediastinal. The class No is CLASSIFICATOIOF NLUNGCANCER assigned to those cases in which no upper mediastinal nodes are recognized. T h e class T = Tumor N = Lymph nodes involved N1 indicates ipsilateral nodal involvement, N2 contralateral or bilateral involvement. M = Distant metastasis (Tumor is classified in the highest possible T group) T1 Discrete tumor of any size in the lung periphery not in contact with chest wall, Stage I STAGESOF LUNGCANCER TN M mediastinum, or hilum. May have asso- ciated density limited to a lobe or less. T2 Tumors involving hilum; tumors involv- I1 No. 2 A I11 MEDIASTINOSCAONPDYSTAGING * Gold berg et al. 35 1 CARCINOMA OF LUNG SITE PIA S T I N O S C O P Y IUMBER OF B IV Any T , N, Combination with M1 Generally, it is customary to group together the combinations of T, N, and M in such a way that a reproducible assessment of extent of disease can be made in sufficient numbers of patients to be useful for prognostic purposes. T h e suggested stages based upon the proposed classifications is as fol- lows: Stage 1 includes the T1,T2 tumors in which no upper mediastinal nodes had been recognized. Stage 2 are those TI, T2tumors with ipsilateral upper mediastinal nodes. Stage 3A includes T1,T2 tumors with contralateral nodes, and T3 tumors with no nodes or with ipsilateral upper mediastinal nodes. Stage 3B are the T3tumors with contralateral or bilateral nodes and all of the T4tumors. Stage 4 is used for any combina- tion of T and N plus extrathoracic disease. It is clearly understood that as experience accumulates with these groupings, the need for regroupings may become necessary. No consideration is given to cell type in the above system of classification. RESULTS These data are derived from an analysis of 144 patients with a diagnosis of primary carcinoma of the lung established by exploratory thoracotomy, bronchoscopy, or mediastinoscopy. They are available from a series of 200 patients who had mediastinoscopy and a clinical diagnosis of carcinoma of the lung. I n this series there were 83 tumors located in the right lung and 61 located in the left lung. Although only one of 83 had solely contralateral spread from the right to the left as compared to 7 out of 61 for the left to the right, bilateral spread was approximately the same right to left as left to right. T h e percentage of negative mediastinal nodes was the same on both sides (Table 1). Four .RIGHT LUNG RUL RLL COMBINED LEFT LUNG LUL LLL COMBINED 12 8 2 2 TABL1E. broad categories of histology are identified (with which practically every pathologist will find something to disagree): 1. the welldifferentiated squamous cell carcinoma; 2. the more poorly differentiated; the undifferentiated, the anaplastic large cell carcinoma; 3. the very undifferentiated small cell (oat cell) carcinoma, and 4. the adenocarcinoma.3~22 Mediastinal lymph node involvement was determined in this group of 144 patients (Table 2). It was found that 43 of 52 patients with squamous cell carcinomas had negative upper mediastinal nodes. Six were found to have positive ipsilateral nodes, and CARCINOMA OF L U N G PATHOLOGY NUMBER MEDIASTINOSCOPY I WELL DIFFERENTIATED LARGE CELL POORLY DIFFERENTIATED 1 16anapiasiic SMALL CELL UNDIFFERENTIATED ~ 001 cell 44 ~ ADENOCARCINOWA 11'21 41 TABL2E. 352 CANCERFebruary 1970 Vol. 25 3 had contralateral positive nodes. On the other hand, only 20 of 64 patients with poorly differentiated tumors had negative upper med'iastinal nodes and 28 had evidence of contralateral or bilateral spread. Eight of 16 oat-cell carcinomas showed bilateral lymph node involvement. Four of the 12 patients with adenocarcinomas had negative mediastinal nodes; 5 had ipsilateral lymph node involvement, and 3 had bilateral lymph node spread. Of the 52 patients with well-differentiated squamous cell carcinoma, almost three fourths (73%) fell into Stages 1 o r 2. However, of the patients with large cell poorly differentiated, undifferentiated, or anaplastic carcinoma, over three fourths of these patients (83y0)fell into Stages 3 and 4.T h e small number of patients with small cell (oat cell) carcinoma had a staging distribution similar to those with anaplastic carcinoma. Seven of the 12 patients with adenocarcinoma fell into the lower stages (Table 3). It seems apparent that the degrees of differentiation representated by the 4 histo- logic groups does suggest that the biological character is accounted for by the proposed staging system. Although sufficient data is not available to us at this time for a valid analysis of survival, the data to date suggests that Stages CARCINOMA OF LUNG WELL DIFFERENTIATED squomous cell LARGE CELL POORLY DIFFERENTIATED __ onoplastic ~~ ~~ SMALL CELL UNDIFFERENTIATED oat cell ADENOCARCINQYA 1 12 1 I1 TABL3E. 1 and 2 have a better survival than Stage 3, and Stage 3 is better than Stage 4.I t is hoped that by utilizing this staging procedure in a prospective study, more valid prognostication can be reported at a future date. I n 200 consecutive patients with carcinoma of the lung, there were 3 bleeding episodes. Two patients lost approximately 75 cc of blood, and one patient lost approximately 250 cc of blood. Bleeding in all cases was controlled with pressure or hemostatic agents. Four patients developed a bradycardia which was readily alleviated by atropine, and 2 patients developed' temporary left recurrent nerve paresis. Three patients developed a pneumothorax or mediastinal emphysema. Treatment was required in only one case, and this was accomplished by drainage through the mediastinal incision to a water seal. Thus, the morbidity rate in this series was 6%. There were no mortalities in this series. T h e clinical staging proposed here requires careful assessment of the upper mediastinum. It does not take into account spread into the lower mediastinum or below the diaphragm. In reality this staging technique with respect to "N" is based on histologic proof of tumor in the mediastinal lymph nodes. Of great interest was the correlation of mediastinal spread, contralaterality, and bilaterality which occurred with increasing frequency as the degree of differentiation was lost in the tumor. T h e usefulness of the staging procedure lies in its bringing together cases with similar extent of disease. Value judgments as to the efficacy of treatment from one institution to another can be made in a more rational manner. Staging as proposed, we believe, will contribute significantly in assessing operability. It is equally important in those cases for whom radiotherapy and chemotherapy may be required either alone or as adjunct to the surgical therapy. As more cases are done, survival statistics for each of the T, N, and M categories will be accumulated so that valid conclusions as to the usefulness of each of the classification categories will emerge. No. 2 -MEDIASTINOSCOAPNYD STAGING Goldberg et al. 353 REFERENCES 1. Bergh. N. P., and Schersten, T.: Bronchogenic 17. Maassen, W.: Ergebnisse und bedeutung der carcinoma: a follow upsstudy of a surgically treated mediastnoskopie. Berlin, Springer-Verlag, 1967. series with special reference to the prognostic sig- 18. Mascagni, P.: Vasorum lymphaticorum corporis nificance of lymph node metastases. Acta Chir. Scand. humani historia et ichnographia. Senis 1787. (Supp1.)347; 1965. 19. McCort, J. J., and Robbins, L. L.: Roentgen 2. Bignall, J. R., and Moon, A. J.: Survival after diagnosis of intrathoracic lymph node metastases in lung resection for bronchial carcinoma. Thorax 10: carcinoma of the lung. Radiology 57:339-360, 1951. 183-190,1955. 20. Most, A.: Die lymphgefasse der lunge. Biblio- 3. Bloom, H. J. G.: The influence of tumor grade theca Medica, Abteil. C. Heft 1621, 1908. on radiotherapy results. Brit. J. Radio2. 38:227-240, 1965. 21. Nickson, J. J., Cliffton, E. E., and Selby, H.: Carcinoma of the lung. Amer. J. Roentgen. 77:826-835, 4. Carlens, E.: Mediastinoscopy: a method for in- 1957. spection and tissue biopsy in the superior mediastinum. Dis. Chest 36:343-352, 1959. 22. Nohl, H. C.: T h e spread of carcinoma of the bronchus. London, Lloyd-Luke Ltd., 1962. 5. Carlens, E., and Jepson, 0.: Mediastinoscopy. OtoZaaryng. Clin. NA 1:171-193, 1968. 23. Palva, T., and Viikari, S.: Mediastinoscopy. J. Thorac. Cardiov. Surg. 42206-212, 1961. 6. Chamberlain, J. M., McNeill, T. M.. Parnassa, P., and Edsall, J. R.: Bronchogenic carcinoma. J . Thorac. Cardiov. Surg. 38:727-745,1959. 24. Radner, S.: Suprasternal node biopsy in lymph spreading intrathoracic disease. Acta Med. Scand. 152: 413415,1955. 7. Cliffton, E. E.: T h e criteria for operability and resectability in lung cancer. JAMA 195:1031-1032, 1966. 25. Reynders, H.: The value of mediastinoscopic study in ascertaining the inoperability of pulmonary carcinoma. J. Int. Colt. Surg. 39:597-610, 1963. 8. Cardier, G., Papamiltiades, M., and Cedard, C.: Les lymphatiques des bronches et des segments pulmonaires. Bronches 8:8-52, 1958. 9. Ghon, A.: Der primare Lungenherd bei der T u berkulose der Kinder. Berlin, Urban und Schwarzenberg, 1912. 26. Reynders, H.: Mediastinoscopy in bronchogenic cancer. Dis. Chest 45:605-612, 1964. 27. Rodinov, V. V.: Immediate and late results of exploratory thoracotomy for lung carcinoma. V O ~ . Onkol. 8:12-18, 1962. 28. Rouviere, H.: Anatomie des lymphatiques de 10. Gibbon, J. H.,Allbritten, F. F., Templeton, J. Y., 1'Homme. Paris, Masson et Cie, 1932. and Nealon, T. F.: Cancer of the lung. An analysis of 532 consecutive cases. Ann. Surg. 138:489-501, 1953. 11. -, Templeton, J. Y., and Nealon, T. F.: 29. Salzer, G., Wenzl, M., Jenny, R. H., and Stangl, A.: Das Bronchuscarcinom Wien. Berlin. Springer-Verlag, 1952. Factors which influence the long term survival of patients with cancer of the lung. Ann. Surg. 145:637-643, 1957. 30. Smith, R. A.: T h e results of raising the resec- tability rate in operations for lung carcinoma. J. Thorac. Cardiov. Surg. 48:418-429, 1964. 12. Goldberg, E. M., and Radner, D. B.: Mediastinoscopy. Chicago Med. Sch. Quart. 25:60-62, 1965. 31. Steinert, R.: Untersuchungen des Lymphsystems der Lunge, zugleich ein Beitrag zur Frage der Topo- 13. Harkens, D. E., Black, H., Claus, R., and Far- graphie der bmnchialen Lymphknoten. Beitr. Klin. rand, R. E.: A simple cervico mediastinal exploration Tuberk. 68:497-510, 1928. for tissue diagnosis of intrathoracic diseases. New Eng. 32. Thomas, C. P.: Lobectomy with sleeve resection. J . Med., 251:1041-1044, 1954. Thorax: 159-11,1960. 14. Harmer, M. H.:Classification and Staging Clinical Evaluation in Breast Cancer. J. L. Hayward, and R. D. Bulbrook, eds. London and New York, Academic Press, 1966; pp. 109-123. 15. Jepson, 0.: Mediastinoscopy. Copenhagen, Munksgaard, 1966. 33. Warren, M. F., and Drinker, C. K.: Flow of lymph from lungs of dog. Amer. J. Physiol. 136:207221,1942. 34. Weinberg, J. A.: Identification of regional lymph nodes in the treatment of bronchogenic carcinoma. J . Thorac. Surg. 22517-526,1951. 16. Kirklin, J. W., McDonald, J. R., Glagett, 0.T., 35. Wurnig, P.: Zur Methode der Beurteilung KuraMoersch, A. J.. and Gage, P. P. Bronchogenic car- tiver Erfolge der Carcinomchirurgie an Hand des cinoma. Surg. Gynec. Obstet. 100:429-438, 1955. Bronchuscarcinoms. Thoraxchirurgie 2:281-289, 1954.