Diagnosis and management of synchronous lung cancers

J

THoRAc CARDIOVASC SURG

89:378-385, 1985

Diagnosis and management of synchronous lung cancers The fmdings in 28 patients with synchronous lung cancers are reviewed. Mediastinoscopy and systemic staging were performed to exclude the possibilities that one pulmonary lesion was metastatic from the other or that both represented systemic metastases from another tumor. Nineteen patients underwent resection of both tumors. Median survival was 25 months for four patients with defmite Stage I synchronouscancers (no nodalinvolvement; different ceU types, bronchoscopicaUy separate endobronchial lesions or arising from separate foci of carcinoma in situ) and was 27 months for seven patients with possible synchronousStage I cancers (no nodal involvement; similar ceU types; located in separate lobes). Median survival was 11 months for 16 patients having Stage II or ill lung cancer accompanied by a second synchronous lung cancer. In the absence of hilar or mediastinal nodal involvement and systemic metastases, synchronoustumors should be considered separate primaries when located in different lobes, even if they have similar histologic features. Prognosis of synchronous cancers is related to the presence or absence of nodal metastases. Pneumonectomy is the operation of choice for synchronous unilateral tumors. With bilateral tumors, sequentialresectionstarting with the most advanced lesionis appropriate. Preservation of lung tissue witbout compromising the cancer operation is critical.

Mark K. Ferguson, M.D.,* Tom R. DeMeester, M.D.,* Jean DesLauriers, M.D.,** Alex G. Little, M.D.,* Michael Piraux, M.D.,** and Harvey Golomb, M.D.,* Chicago, Ill., and Ste-Foy, Quebec, Canada

h e incidence of lung cancer has increased in all segments of the Western population over the past 25 years. Coincident with this increase has been an increase in reports of multiple lung cancers.I The problem of evaluating two pulmonary masses presenting simultaneously is complex. The lesions may both be metastatic disease from a tumor arising elsewhere in the body, one lesion may be a metastasis from the other, or both lesions may be benign or malignant processes in their own right. The incidence of synchronous lung cancers is sufficiently rare that all of these possibilities must be considered carefully. Over the past 10 years more than 2,100 patients have From the Departments of Surgery and Medicine, The University of Chicago Pritzker School of Medicine, Chicago, Ill., and the Department of Surgery, Le Centre de Pneumonologie de Laval, Ste-Foy, Quebec, Canada. Read at the Tenth Annual Meeting of The Western Thoracic Surgical Association, Maui, Hawaii. June 20-23. 1984. Address for reprints: Mark K. Ferguson, M.D., Department of Surgery, Box 255, The University of Chicago, 5841 S. Maryland Ave., Chicago, Ill. 60637. *The University of Chicago Pritzker School of Medicine. **Le Centre de Pneumonologie de Laval.

378

been treated for bronchogeniccarcinoma at the University of Chicago Hospitals and Le Centre de Pneumonologie de Laval. All patients were reviewed who were found to have two separate malignant lung nodules presenting simultaneously. This paper summarizes our techniques used in evaluation of these patients, our criteria for differentiating synchronous lung cancers from the other possibilities, and the appropriate management of these tumors. Patients Twenty-eight patients with possible synchronous lung cancers were identified. There were 25 men and three women with a mean age of 62 years (range 39 to 82 years). Two separate lesions were identified by chest radiograph in 20 patients. In the other eight, chest x-ray films revealed only one pulmonary mass, and the second lesion was diagnosed by bronchoscopy in three and during thoracotomy in five. The locations of the two separate tumors are illustrated in Fig. 1. There was a predilection for occurrence in the upper lobes, especially when the masses presented bilaterally. Preoperative evaluation included bronchoscopy in 25 patients, which provided a histologic diagnosis of both

Volume 89

Synchronous lung cancers

Number 3

March. 1985

tumors in three patients and of one tumor in nine patients. Needle biopsy was used to obtain the histologic diagnosis of both tumors in four patients, all with bilateral masses, and of one tumor in 13 patients, eight of whom had bilateral masses. Preoperative histologic diagnosis of both masses was thus obtained in 12 patients, all of whom had bilateral tumors. Six patients underwent a gallium scan, which showed both tumors in three patients, one mass in two patients, and no masses in the final patient. Computed axial tomography (CT) was available only during the last few years included in this review. CT was performed in three patients and showed the two separate nodules in two patients, whereas the third scan was read as normal. Bone scans were performed in 26 patients and revealed no abnormalities. Brain scans were performed in 13 patients and also disclosed no abnormalities. Results of liver/spleen scans were normal in 12 patients. Mediastinoscopy was normal in 24 patients and was not performed in the remaining four.

379

BILATERAL

~d3 ~63 9

5

2

1 UNILATERAL

Treatment Twenty-four patients underwent operation other than mediastinoscopy and bronchoscopy, with no operative deaths. Nineteen had resection of both pulmonary masses. Of those patients with unilateral tumors, eight had pneumonectomy, one had right upper lobectomy with wedge resection of the right lower lobe, one had combined right middle and lower lobectomy, and the final patient underwent wedge resections of the left upper and left lower lobes. Of the 13 patients with bilateral tumors undergoing operation, seven received staged resection of both tumors with intervals of 1 to 6 months between operations. Two patients had left pneumonectomy followed by segmental resection of the right upper lobe, two patients underwent lobectomy on one side followed by a lobectomy on the contralateral side, two patients had lobectomy followed by segmental resection on the opposite side, and the final patient received wedge resection of a nodule in each lung. A sternotomy was performed with wedge resections of two nodules in one patient. In a second patient in whom sternotomy was performed, one tumor was wedge resected and the other was unresectable. Two patients underwent lobectomy, which disclosed abnormal mediastinal nodes, and did not undergo further surgical treatment. A lingulectomy was performed in one patient, but postoperative pulmonary function would not permit the pneumonectomy required to resect the contralateral tumor. The final patient underwent exploratory thoracotomy and but had unresectable disease.

4

1

1

5

Fig. 1. Distribution of synchronous lung cancers with a list of the number of patients in each category.

Four patients did not have an operation. One died during coronary artery bypass grafting. Two patients had insufficient pulmonary function to permit pulmonary resection, and the fourth patient had poorly differentiated lymphoma in addition to two lung nodules and was not believed to be an operative candidate. Combination chemotherapy and radiation therapy were administered to five patients, four patients had chemotherapy alone, and four patients had radiation therapy alone. Histology The histologic findings of the tumors in these patients are listed in Table I. The combination of squamous cell carcinoma in both tumors was by far the most common, followed by the pairing of squamous cell carcinoma and large cell carcinoma in six. Adenocarcinoma in both tumors was found in four patients. The presence of more than one cell type within a single mass was seen in only two patients, both of whom had squamous cell carcinoma as the dominant element with features of adenocarcinoma also present. In neither of these patients was adenocarcinoma the cell type of the second tumor.

380

The Journal of Thoracic and Cardiovascular Surgery

Ferguson et al.

Table I. Histologic type of synchronous lung cancers First tumor

Second tumor

Unilateral

Bilateral

Total

Squamous cell Adenocarcinoma Squamous cell Squamous cell Adenocarcinoma Squamous cell

Squamous cell Adenocarcinoma Large cell Oat cell Large cell Adenocarcinoma

5 2

10

15

2 5

4 6 I I I

I I I I

Table II. Classification of patients with synchronous lung cancers* Stage Ill: T3 and/or N2

Stage I TI NO or T2 NO Definite Possible Total

4 6 10

I 4 5

6 5 II

'Staging based on extent of most advanced tumor. See text for further explanation.

In most cases, pathological examination did not indicate whether a specific tumor arose from a focus of carcinoma in situ, which would render that tumor a primary carcinoma. In no cases, however, was there an indication that any of the synchronous lesions in the pulmonary parenchyma arose from lymphatic or vascular embolic spread. Results of treatment Patients were classified according to the TNM staging system. They were further subdivided into groups having definite synchronous lung cancers or possible synchronous lung cancers (Table II). In two patients, the TNM stage was not available. The 11 patients designated as having definite synchronous lung cancers fulfilled the following two criteria: (1) Their tumors were separate or were located in different lobes; (2) their tumors were of different histologic types or were shown to arise from different endobronchial lesions by bronchoscopyor from separate foci of carcinoma in situ at pathological examination. The 17 patients classified as having possible synchronous lung cancers also had lesions that were separate and distinct but were all of similar histologic type. Their tumors were not shown to arise from separate endobronchial lesions, as bronchoscopy showed no abnormalities in nine and revealed only a single lesion in five. Pathological examination was also unable to confirm the presence of separate tumor origins from carcinoma in situ in these patients. Thus, we were unable to conclusively rule out lymphatic or vascular embolic spread as a cause for the second lesion in these patients.

0 0 0

Ten patients had two separate tumors classified as T1 or T2 without evidence for mediastinal or hilar adenopathy and thus had two possible or definite Stage I synchronous lung cancers. Raw survival data for these patients are plotted in Figs. 2 and 3. Six of these patients had possible synchronous Stage I lung cancers. Their survival times ranged from 10 months to 38 months, with two patients alive at 3 and 24 months, respectively. The other four patients had definite Stage I synchronous lung cancers. Survival times ranged from 2 to 37 months, with one patient alive at 24 months. For all patients with possible or definite synchronous Stage I lung cancers, the 2 year survival rate was 60% and median survival time was 27 months. In 17 patients the more extensive tumor was classified as Stage II or III, having met one of the following criteria: (1) one of the pulmonary masses qualifying as T3; (2) abnormal hilar nodes with histologic features similar to an ipsilateral lung mass; or (3) abnormal mediastinal nodes with histologic features similar to either of the lung masses. Raw survival data for 16 patients having a Stage II or III lung cancer synchronous with a separate malignant pulmonary tumor are shown in Figs. 4 and 5. One patient was lost to follow-up. Nine patients had possible Stage II or III lung cancers synchronous with a separate tumor. Survival times ranged from 3 to 57 months, with two patients alive at 2 and 44 months, respectively. Seven patients had definite Stage II or III lung cancers synchronous with a separate mass. Survival times ranged from 4 to 48 months and no patients are presently alive. For all patients with a possible or definite Stage II or III lung cancer, the 2 year survival rate was 18.7% (3/16) and the median survival time was 11 months. Discussion The incidence of multiple lung cancers in reported series ranges from 1% to 7%. An autopsy series of patients with proved lung cancer reported by Auerbach and associates' reveals 3.5% of them to have a second bronchogenic carcinoma at the time of death. Clinical series show the incidence of dual lung carcinomas to be

Volume 89 Number 3 March, 1985

1 -

Synchronous lung cancers 3 8 1

-----, I I I

1·

I

.9

I I I

I

II .8

.7

....

~

.6

.9·

I

~

I

I !L I

I

.8·

I

+

.7.

....

~

iil '0

> :::; iii

.5

I-

~

~

~

.6-

'0 >

.5,

s a: iil !:::

::!

.4

--------1

.3

.4·

.3-

I

.2

I

I I I

.2,

II

.1

o

I lD

.1'

I

!

12

24

36

48

MONTHS

o·

0

12

24

3e

48

MONTHS

Fig. 2. Probability of survival for patients with possible synchronous Stage I lung cancers (- - -) and for patients with definite synchronous Stage I lung cancers (--). Vertical lines indicate length of follow-up for patients still alive.

Fig. 3. Probability of survival for patients with possible or definite synchronous Stage I lung cancers. Vertical lines indicate length of follow-up for patients still alive.

closer to 1%.3.4 Multiple synchronous primaries are less common than metachronous primaries, comprising roughly 40% of all multiple bronchogenic carcinomas.' Differentiating multicentric lung cancers from a single lung cancer with metastases or from pulmonary metastases from cancer originating elsewhere in the body can be difficult. The radiographic appearance of pulmonary masses is frequently sufficient to make a preliminary differentiation between primary pulmonary malignancy and metastatic disease. Nevertheless, the presence of two separate lung tumors makes systemic evaluation in these patients critical. If history or physical examination points to another organ system as a possible source of malignancy, usually breast, kidney, thyroid, or colon, a diagnostic work-up should be directed to that system until such a possibility is effectively ruled out. In the absence of such clues during preliminary evaluation, additional work-up should include a whole body scan such as a gallium scan or a bone scan. We find the gallium scan to be valuable in evaluation of possible

extra pulmonary metastases or primaries and in assessing mediastinal disease.'' Chest CT is particularly useful if multiple lung tumors are suggestive of metastatic disease, as it can define other nodules too small for an ordinary chest radiograph to reveal. If the preoperative work-up indicates no evidence for an extrapulmonary source of malignancy or for extrapulmonary metastases, the possibility that the two separate lung tumors are synchronous lung cancers should be strongly considered. Evaluation of the pulmonary masses themselves includes bronchoscopy. The value of this procedure as both a staging and diagnostic manuever cannot be overemphasized. In three of the patients included in this review, second lung cancers were detected bronchoscopically that had not been suspected radiographically. Bronchoscopy also allowed the histologic type of at least one tumor in nine of the 24 patients to be diagnosed prior to thoracotomy. The ability to obtain histologic diagnosis via needle biopsy is often dependent on the

The Journal of Thoracic and Cardiovascular Surgery

382 Ferguson et al.

I I

I I I

.9

I I I I

.9

II .8

I I I

.8

L. .7 ..J

~

s a:

.6

> ~ ::;

iii

ct

B a:

.7

I I I

..J

~

I I

~

I I

::;) (/)

'0

I I I I I I I I

.5

.4

...

'---~

::;) (/)

I I I I I

> ~ ::;

I

'0

0-

.6

.5

iii

I I I

~

I

.4

I

.3

'--

.3

--------------t----------i

.2

I I I

II

.1

0

12

24

36

48

I

MONTHS

Fig. 4. Probability of survival for patients with possibleStage II or III lung cancer and a second synchronous malignant pulmonary mass (- - -) and for patients with definite Stage II or III lung cancer and a second synchronous malignant pulmonary mass (--). Vertical lines indicate length of follow-up for patients still alive.

experience in a particular institution. Although it is convenient to have confirmation of malignancy, the amount of tissue obtained through needle biopsy is frequently not sufficient to provide differentiation between a pulmonary and a nonpulrnonary tumor, especially in the case of squamous cell carcinoma or adenocarcinoma. The fact that many tumors are pleomorphic creates additional difficulty in obtaining a definitive histologic diagnosis on the basis of such small samples." Mediastinoscopy is an essential tool used in staging these masses. It is important to exclude the mediastinal lymphatics as a possible route for ipsilateral tumor spread. The incidence of contralateral pulmonary metastases from a lung cancer, even in the presence of mediastinal disease, is low." In the absence of histologic diagnosis from bronchoscopy or needle biopsy, mediastinoscopy may also provide adequate tissue for determining cell type.

.2

.1

o

12

24

36

48

MONTHS

Fig. 5. Probability of survival for patients with possible or definite Stage II or III lung cancer accompanied by a second synchronous malignant pulmonary mass. Vertical lines indicate length of follow-up for patients still alive.

Controversy still exists in the literature regarding criteria for determining which synchronous lesions are multiple primary lung carcinomas. The most recent review determined that synchronous tumors exist if the lesions are physically distinct and separate and if their histologic features are different." If the histologic characteristics are similar but tumors are located in different lobes, it is necessary that tumors arise from separate foci of carcinoma in situ, that there is no lymphatic involvement, and that no extrapulmonary metastases are present at the time of diagnosis. In our review, the majority of patients with possible synchronous Stage I tumors have similar histologic evidence for both without proof of origin from separate foci of carcinoma in situ. Despite this fact, biologic behavior of these tumors is similar to that of definite synchronous State I cancers. On the basis of these data, patients having two separate pulmonary tumors, whether or not they are of similar histologic types, have synchronous primary lung cancers in the absence of hilar or mediastinal nodal involvement or distant metastases.

Volume 89 Number 3

Synchronous lung cancers 3 8 3

March. 1985

SYNCHRONOUS MASSES ON CHEST x- RAY

HISTORY AND PHYSICAL IMPLICATE ANOTHER ORGAN SYSTEM

I

~~

EVALUATE OTHER ORGAN SYSTEM

HISTORY AND PHYSICAL NEGATJ VE

---------r----

GA SCAN J CT SCAN J MEDIAST I NOSCOPY

!

MEDIASTINUM NEGATIVE

MEDIASTINUM POSITIVE

~

RESECTION

RT, CHEMOTHERAPY

•

~ RECURRENT DISEASE ......

~ t~

STAGE

NO RECURRENCE

j

!

I, II

RE-EVALUAT I ON

Fig. 6. Schematic diagram of diagnostic and therapeutic considerations in patients presenting with synchronous pulmonary masses. RT, Roentgen therapy. GA, Gallium. CT, Computed axial tomography.

Biologic behavior of Stage II or III primary lesions combined with a second pulmonary mass categorized as a possible synchronous primary is similar to that of definite synchronous primaries with a Stage II or III tumor. It is not possible to state whether these groups should be considered as one, for we cannot rule out the possibility that metastatic intrapulmonary spread accounts for the second pulmonary mass in the group of patients with possible synchronous cancers. The fact that there are two patients with a Stage III tumor histologically similar to a second pulmonary mass who have survived longer than 4 years further complicates the issue. Our routine for diagnostic evaluation and management of patients with two synchronous lung masses is outlined in Fig. 6. Patients with lung masses on chest x-ray films undergo a routine bronchoscopic evaluation for histologic diagnosis, followed by CT of the chest and upper part of the abdomen. The presence of multiple lesions less than 2 em in diameter, frequently undetectable on a routine chest radiograph, indicates widespread metastatic pulmonary disease. If the mediastinal nodes are normal by CT, work-up for a primary lesion elsewhere in the body is indicated, including intravenous pyelograms and gastrointestinal x-ray films. A CT scan revealing only two isolated lesions is followed by gallium scanning to evaluate a probable primary lung tumor. If the gallium scan reveals the pulmonary lesions and also shows disease in the mediastinum, Stage III disease is confirmed with mediastinoscopy. If the mediastinum is normal by gallium scanning but uptake is seen in the pulmonary lesionsand elsewhere in the body, particularly the bones, confirmation of this is sought with a biopsy

or a second scan. Despite recent reports to the contrary, II the survival of patients with mediastinal nodal metastases and synchronous tumors does not warrant resection when diseased mediastinal nodes are documented prior to thoracotomy. Stage III disease is treated with a combination of radiation therapy and multi-drug chemotherapy. If the gallium scan confirms two isolated lesions or is negative then mediastinoscopy is performed. Positive mediastinoscopy is indictive of Stage III disease, whereas negative mediastinoscopy permits progression to resection. The operative management of these patients requires careful analysis at each stage of evaluation. Preoperative pulmonary testing is of utmost importance, as many patients with two synchronous lesions require more than one resection. An overzealous initial resection may limit the surgeon's ability to perform curative resection of the second lesion because of inadequate residual lung function. In patients with synchronous unilateral lesions, which most often are located in separate lobes, pneumonectomy is the operation of choice. In unusual circumstances a lobectomy with additional wedge resection may be performed if pulmonary functions are limiting. In patients with bilateral lesions, a curative resection should be performed first on the side that appears to be most advanced in stage, without unnecessarily sacrificing lung tissue. Mediastinal nodal dissection should be performed at this time to complete staging. In the presence of Stage I or II disease at the time of initial resection in patients with bilateral disease, contralateral resection is performed within 1 to 2 months. The presence of ipsilateral mediastinal nodal disease at the time of initial resection is not a deterrent from perform-

The Journal of

384

Ferguson et al.

ing resection on the opposite side if all gross disease is removed and reevaluation at a 3 month interval shows no evidence of recurrence. The patient must be allowed the benefit of the doubt that the second lesion is another primary rather than a metastasis. If pulmonary functions are borderline, axillary thoracotomy may be considered at the time of the second resection to minimize respiratory compromise secondary to postoperative pain. In situations that appear anatomically favorable, median sternotomy should be performed, especially when there are two peripheral lesions located in the upper lobes. Median sternotomy should also be considered when pulmonary function is a limiting factor in performing staged thoracotomies.

2

3

4

5

6

7

8

9 10 II

REFERENCES Razzuk MA, Pockey M, Urschel HC Jr, Paulson DL: Dual primary bronchogenic carcinoma. Ann Thorac Surg 17:425-433, 1974 Auerbach 0, Stout AP, Hammond EC, Garfinkel L: Multiple primary bronchial carcinomas. Cancer 20:699705, 1967 Struve-Christensen E: Diagnosis and treatment of bilateral primary bronchogenic carcinoma. J THoRAc CARDIOVASC SURG 61:501-513,1971 Rohwedder JJ, Weatherbee L: Multiple primary bronchogenic carcinoma with a review of the literature. Am Rev Respir Dis 109:435-445, 1974 Bower SL, Choplin RH, Muss HB: Multiple primary bronchogenic carcinomas of the lung. AJR 140:253-258, 1983 DeMeester TR, Golomb HM, Kirchner P, Rezai-Zadeh K, Bitran J, Streeter DL, Hoffman PC, Cooper M: The role of gallium-67 scanning in the clinical staging and preoperative evaluation of patients with carcinoma of the lung. Ann Thorac Surg 28:451-463, 1979 Spies WG, Spies SM, Mintzer RA: Radionuclide imaging in diseases of the chest (part 2). Chest 83:250-255, 1983 Feinstein AR, Gelfman NA, Yesner R, Auerbach 0, Hackel DB, Pratt PC: Observer variability in the histopathologic diagnosis of lung cancer. Am Rev Respir Dis 101:671-684, 1970 Onuigbo WIB: Contralateral pulmonary metastases in lung cancer. Thorax 29: 132-133, 1974 Martini N, Melamed MR: Multiple primary lung cancers. J THoRAc CARDIOVASC SURG 70:606-611, 1975 Martini NM, F1ehinger BJ, Zaman MB, Beattie EJ Jr: Results of resection in non-oat cell carcinoma of the lung with mediastinal lymph node metastases. Ann Surg 198:386-396, 1983

Discussion DR. LUCIUS D. HILL Seattle. Wash.

This series probably is the largest series of synchronous lung cancers to be reported to date. Nael Martini reported a series

Thoracic and Cardiovascular Surgery

of 50 patients, but 3 I of them had metachronous disease. The paucity of reported cases of synchronous lung cancers tells us that we are missing a lot of them. However, I do believe that the oncologists and the chest physicians, upon seeing multiple tumors in the chest, rarely think of the surgeon. They immediately jump to the conclusion that these are metastatic, and the patient is relegated to chemo-radiotherapy. Auerbach took multiple step sections through the laryngotracheobronchial tree and found that 3.5% of patients with lung cancer have separate primaries. If one accepts the criteria that the authors brought out, the number of patients in Auerbach's series who had multiple lung cancers was 14%, which is another way of saying that the entire laryngotracheobronchial tree in patients who are heavy smokers is at risk for lung cancer. Step sections will uncover many islands of small, in situ carcinomas. Thus, it is not surprising that both synchronous and metachronous lung cancers are found. The authors have brought up a message that should be gotten across to our confreres---our medical chest men and oncologists-and that is that the survival rate in patients with synchronous lung cancers is similar to that in patients who have an isolated lesion. I would like to ask the authors whether or not they have used electron microscopy in those patients with synchronous tumors, in whom they have made the statement that the cell types are different or similar. Dr. Sam Hammer, who is chairman of the Pathology Section of the National Lung Cancer Study Group, is our pathologist and submits all our specimens to electron microscopy. It has been fascinating to see that some of the patients who were thought to have squamous carcinoma by electron microscopy have adenocarcinoma. Thus, before the cell type is reported as the same or different, electron microscopy is important. The average survival rate of 26 months in the authors' patients is excellent. They have a number of patients with Stage I disease who are alive beyond that point. As they point out, the patients with Stage II and III disease in their series have a very poor outlook. For this reason, the National Lung Cancer Study Group is embarking on programs of platinumbased chemotherapy coupled with radiation, in an effort to improve the outlook in these two groups. I would like to ask the authors if they are contemplating undertaking such a study. A patient in our series illustrates some of the points that the authors have made. Dr. Richard Anderson performed an upper lobectomy on this patient in 1974 for a peripheral lesion. Unfortunately, this patient had not had bronchoscopy, which, as the authors emphasized, is necessary. When Dr. Anderson did try upper lobectomy for what proved to be an oat cell tumor, he found a squamous carcinoma at the orifice of the lower lobe that had been undetected. These were obviously synchronous tumors. This patient is alive and well and free of disease at IO years. One wonders whether our chest physicians would not have referred this patient for operation if a needle biopsy had been done on this upper lobe lesion and oat cell carcinoma had been discovered. Also, one wonders whether the same chest physicians would have assumed that the upper lobe lesion was a metastasis if this patient had had bronchoscopy and the lesion had been discovered in the lower lobe.

Volume 89 Number 3 March, 1985

Since these patients who are not treated surgically, both in the authors' study and in our group, have a survival time of only about 6 months, it is mandatory that these patients at least be considered for operation. Some of the authors' points are underscored by the case of a patient who had a left upper lobectomy. Dr. Sam Hammer found an adenocarcinoma in the hilum of that lobe, a squamous cell carcinoma in the apex, and an oat cell carcinoma in the posterior segment. The ear, nose, and throat department had been asked to see this patient because he was hoarse. A separate squamous in situ carcinoma was found in the larynx. This case underscores the fact that the whole laryngotracheobronchial tree is at risk in these patients. In this particular patient, a recurrence developed in the larynx, but he is free of any disease in the thorax. He is alive at 31 months. A tragic case was that of a patient who had a right upper lobectomy in 1976 for synchronous tumors. This was done in a staged fashion. A left upper lobectomy was done 3 months later. The patient did fine. He returned 8 years later with a mass in the abdomen with a large gastric ulcer that was immediately assumed to be malignant. The surgeon was told to do a massive gastric resection, and this resulted in the patient's death. At postmortem examination, there was no tumor anywhere in the body. The fact that these patients have had synchronous or metachronous lung cancers does not necessarily mean that a lesion discovered later is recurrent lung cancer. The authors have brought up a very important area of lung cancer. I think we must get the message across to our medical confreres that when they see multiple lesions, they should try not to relegate the patients to radio-chemotherapy immediately, but refer them for operation. The survival rate of these patients will be as good as that of patients having solitary lesions. The survival time in our series of 16 patients was about 30 months, and there are a number of patients who set the data askew. One is alive at IOyears and several others are alive following resection for synchronous lung cancers at 2, 3, and 4 years. DR. DONALD F. ROWLES Palo Alto, Calif

Why not approach these tumors from a sternal-splitting incision and do wedge resections whenever possible?

Synchronous lung cancers

385

DR.FERGUSON(Cwsm~

I would like to thank the discussants for their commentary. I would also like to compliment Dr. Hill on discovering a patient who had more tumors synchronously that I ever thought possible. It is true that a problem exists when a patient reaches the oncologists or the medical chest physicians first and more than one lung mass is discovered. These patients are not normally referred for evaluation by a surgeon or considered for possible resection. We overcome that to some extent in our lung cancer group at the University of Chicago. The case history of every patient with lung cancer is extensively reviewed, and we require histologic proof of metastatic disease, real certainty that it has metastasized from another cancer, before we will omit the possibility of resection. We have recently reviewed our experience with electron microscopy in tumors that were very poorly differentiated and in large cell cancers. About a third of these tumors differentiated into squamous cell types, another third into adenocarcinoma, and the remaining third stayed classified as large cell carcinomas. However, the breakdown by histologic types did not predict the survival in any of these groups. It is our belief currently that, although electron microscopy may give a better histologic diagnosis, it will not affect prognosis at all. In patients who we believe have extensive disease, we are beginning a protocol involving platinum-based chemotherapy prior to reevaluation and possible resection. I would like to emphasize that bronchoscopy is essential in the staging of this disease. Twenty of our patients were found to have multiple primary tumors on x-ray films. The remaining eight patients were not known to have two tumors until bronchoscopy in three of them and thoracotomy in the remaining five. I do not think we would routinely advocate median sternotomy, for it does not permit complete staging. In certain select situations in which both tumors are quite peripheral and are located in the upper lobes, it may be considered, especially when the patient has poor pulmonary function.