The Results of Modern Surgical Therapy for Multiple Primary Lung Cancers*

Transcription

1 The Results of Modern Surgical Therapy for Multiple Primary Lung Cancers* Samuel A Adebonojo, MD, FCCP; Dennis M. Moritz, MD, FCCP; and Christopher A Danby, MD Study objectives: The purpose of this report is to review our experience with multiple primary lung cancers (MPLC) at the Walter Reed Army Medical Center, Washington, DC, and to determine the outcome of our surgical management of this complex problem. Patients and methods: The data from the Lung Cancer Registry on patients with MPLC from January 1984 to December 1995 were reviewed. We used the criteria of Martini and Melamed modified by Antakli for the diagnosis of synchronous and metachronous MPLC. Survival probabilities were calculated by the Kaplan-Meier actuarial method with the dates of resection as the starting point and included deaths from all causes. The log rank test was used to compare survival rates between groups and Wilcoxon rank sum test was used to compare the intervals between the first and the second metachronous cancers. A p value of 0.05 was considered statistically significant. Results: Fifty-two patients, consisting of 51 patients who had "curative" pulmonary resections and 1 patient who had radiation therapy for previous primary lung cancer, developed second or third primary lung cancers. Thirty-seven patients developed metachronous cancers within 1 to 15 years of the first operation (median, 24 months) while 15 patients had synchronous cancers (10 unilateral, 5 bilateral). The probability of cancer-free interval among patients with metachronous cancers was 41% at 3 years, 16% at 5 years, and 3% at 10 years. Two of the 36 patients who had pulmonary resection for the second metachronous cancer died in the perioperative period (operative mortality, 5.6%), and one patient had radiation therapy for the second metachronous cancer. There were no deaths among patients with synchronous cancers. The actuarial 5-year survival for second metachronous cancers was 37% and for synchronous cancers was 0%. Conclusions: We conclude that an aggressive surgical approach is safe and justified in most patients with MPLC, especially patients with metachronous cancers, while patients with synchronous lung cancers have poorer prognosis. The operative morbidity and mortality are acceptable and long-term survival is possible in many patients with metachronous lung cancer. (CHEST 1997; 112: } Key words: lung neoplasm; metachronous lung cancer; multiple lung cancers; second primary lung cancer; synchronous lung cancer Abbreviations: MPLC=multiple primary lung cancers; SPLC =second primary lung cancer; WRAMC=Walter Reed Army Medical Center The simultaneous discovery of two pulmonary nodules in different lobes or lungs raises the clinical dilemma of whether these lesions represent metastases or primary synchronous lung cancers. However, when a new solitary pulmonary nodule *From the Department of Cardiothoracic Surgery, Walter Reed Army Medical Center, vvashington, DC. The opinions and assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of the Army or the Department of D efense. Manuscript received October 22, 1996; revision accepted February 27, Reprint requests: Samuel A Adebonojo, MD, FCCP, Surgical Services Department, Veterans Affairs Medical Center, Dayton, OH ; s.adebonojo@mc12000.com develops 2 to 3 years after curative pulmonary resection, the new lesion may represent a recurrent cancer, a metastatic process, or a second primary lung cancer (SPLC). Differentiation of these clinical entities is important in terms of prognosis but, most importantly, it will have a great impact on the extent of surgical resection if the lesion is resectable. n two series of patients \vith stage lung cancer who were followed up after curative resection, Martini and associates 1 found that 27% of the patients developed recurrent disease (70% disseminated, 30% locoregional) while 11.5% developed an SPLC. Pairolero and his colleagues, 2 in an earlier publication in 1984, also found that 39% of patients with CHEST SEPTEMBER,

2 resected stage lung cancer subsequently developed recurrent disease and 10% had multiple primary lung cancers (MPLC). These findings provide compelling evidence for frequent, early, and sustained long-term surveillance of all patients undergoing resection for lung cancer. MPLC represents an interesting subgroup of cancer cases that may be encountered after curative resection of bronchogenic carcinoma. The actual frequency of MPLC is unknown, but varies from 0.8 to 14.5% 1-14 depending on whether it was calculated from cancer registry, autopsy series, 4 or from surgical series.l-3,5-14 This may be due to the difficulty in differentiating a true SPLC from metastatic or recurrent disease, especially in patients with synchronous lung cancers. Patient selection also plays a major role as most surgical series tend to focus on stage lung cancer patients, reporting a frequency of 10 to 11.5% of MPLC. 1 2 n many series, metachronous lung cancers constitute 55 to 65% of all MPLC; these are patients who are most likely to have been cured of their index cancer and more likely to survive long enough to develop second and third primary lung cancers. 1 4 The purpose of this report is to review our experience with MPLC at a tertiary military health facility with a broad referral base and to determine the outcome of our aggressive surgical approach to this complex problem. MATERALS AND METHODS Between January 1984 and December 1995, 1,325 patients were registered in the Lung Cancer Registry at Walter Reed Army Medical Center (WRAMC). The data of all patients with MLPC during this period were reviewed. We used the criteria of Martini and Melamed 5 with the recent modifications by Antakli and associates 6 for the diagnosis of synchronous and metachronous primary lung cancers (Table 1). Sixty-eight of the 1,325 patients with lung cancer were found to have MPLC for a frequency of 5.1% among patients vvith lung cancer at WRAM C. Sixteen patients were excluded from further analysis because of incomplete data or because they did not receive optimal "curative" therapy for the index cancer. Because of the propensity toward multicentric disease, patients with bronchoalveolar carcinoma were also excluded. During this period, 576 of the 1,325 patients with primary lung cancer had "curative" pulmonary resections (43.4% resection rate). Fifty-one of these 576 patients (8.9%) and 1 patient who was initially treated by radiation therapy for a primary lung cancer were documented to have MPLC. A review of the medical records of these 52 patients forms the basis of this report. All patients who underwent resection for lung cancer at WRAM C were seen in the outpatient clinic quarterly for the first 2 years, semiannually for the next 3 years, and annually thereafter for life. History and physical examination along with chest radiographs in the posteroanterior and lateral projections were obtained at each clinic visit. Patients with abnormal chest radiographs, symptoms, or positive physical findings were further investigated by CT of the chest and upper abdomen. Our diagnostic protocol for patients with suspected metachronous or 694 Table!-Criteria for Diagnosis of MPLC Martini and Melamed, Metachronous tumors. Histologic type different. Histologic type the same, if: A. Free interval between cancers is at least 2 yr, or B. Origin from carcinoma in situ, or C. Second cancer in different lobe or lung, but: l. No carcinoma in lymphatics common to both 2. No extrapulmonary metastases at time of diagnosis Synchronous tumors. Tumors physically distinct and separate. Histologic type: A. Different B. Same, but in different segment, lobe, or lung, if: l. Origin from carcinoma in situ 2. No carcinoma in lymphatics common to both 3. No extrapulmonary metastases at time of diagnosis Antakli et al, A. Different histologic condition B. Same histologic condition with two or more of the following: l. Anatomically distinct 2. Associated premalignant lesion 3. No systemic metastases 4. No mediastinal spread 5. Different DNA ploidy synchronous cancers includes sputum cytologic testing, flexible bronchoscopy with or without transbronchial biopsy, bone scintigraphy, CT-guided thransthoracic biopsy, MR of the chest and brain, and mediastinoscopy. n addition, anterior mediastinal surgical exploration or video-assisted thoracoscopy was performed when indicated. Other appropriate studies were performed to exclude extrapulmonary primary cancer which may affect diagnosis and treatment options. When resection was contemplated, full biochemical analysis, ECG, and pulmona1y function tests were obtained. Exercise testing in the form of stair climbing was used during the latter half of the review period. n addition, all patients evaluated for operation of a second metachronous cancer had quantitative ventilation/perfusion scintigraphs with estimation of the postoperative FEY 1. Maximal oxygen consumption with exercise was obtained in two patients with marginal pulmonary function test results. Right heart catheterization with temporary occlusion of the involved pulmonary arte1y was not performed in any of the patients undergoing resection of a second metachronous cancer. Similarly, we did not use DNA flow cytometry for histologic differentiation as suggested by chinose and associates 15.l 6 in any of these patients. All lesions were staged according to the new nternational Staging System of Lung Cancer as proposed by Mountain 17 and the American Joint Committee on Cancer 18 Operative mortality included deaths from all causes occurring within 30 days of surgery or beyond 30 days during the same hospitalization. The interval between metachronous cancers was calculated from the date of resection or radiation therapy of the index cancer to the date of pathologic diagnosis of the second cancer. Survival interval was calculated from the date of the first or second operation to the date of last follow-up or death. Survival probability was calculated by the Kaplan-Meier actuarial method, 19 vvith the date of resection as the starting point, and included Clinical nvestigations

3 deaths from all causes. The log rank test was used to compare survival rates between groups and \Vilcoxon rank sum test was used to compare the intervals between the first and second metachronous cancers. A p value of 0.05 was considered statistically significant. RESULTS There were 52 patients with MPLC, consisting of 38 men and 14 women with ages at the time of initial diagnosis ranging from 41 to 76 years (mean age ±SO, 62.9:±:8.1 years; median age, 64 years). All patients were smoking at the time of diagnosis of the synchronous cancers or the detection of the second metachronous cancer. Thi1ty-seven patients ( 71%) developed a second cancer within 12 to 175 months of the first operation (median, 24 months) while two patients developed a third cancer 27 and 69 months after the second cancer operation. The cumulative probability of cancer-free interval for metachronous cancers was 41% at 3 years, 16% at 5 years, and 3% at 10 years. There was no difference in the intervals between the first and second metachronous cancers based on stage or histologic features of the first cancer. The median interval for stage was 24 months (range, 12 to 175 months) and for stages and lia, it was 35 months (range, 13 to 115 months) (p=0.54). The median interval for adenocarcinoma was 28 months (range, 12 to 175 months) and for squamous cell carcinoma, it was 22 months (range, 13 to 66 months) (p=0.80). Fifteen patients (29%) had multiple synchronous cancers at the time of presentation; 10 were unilateral and 5 were bilateral. The median length of follow-up after the second operation was 48 months (range, 1 to 136 months) among patients with metachronous cancers and 43 months (range, 6 to 58 months) among patients with synchronous cancers. Criteria for Diagnosis of MPLC Thirty-seven patients with metachronous and 15 patients with synchronous cancers met the criteria for diagnosis of MPLC. Seventeen of the 37 patients ( 46%) with metachronous cancers had different histologic features, of which 12 were contralateral and 5 were ipsilateral. Sixteen patients had second metachronous cancers of similar histologic condition but were located in the contralateral lung. The remaining four patients developed second cancers of similar histologic condition >3 years after their first cancer operation. Twelve of the 15 synchronous cancers (80%) were of different histologic features and 3 were of similar histologic features, of which 2 were in different lungs (bilateral) and 1 in the same lung but in different lobes. Methods of Diagnosis All the second metachronous lesions were discovered by chest radiographs during routine follow-up and confirmed by CT. Ten of the 15 synchronous cancers were detected by chest radiographs; in four other patients, the second synchronous nodules were not evident on chest radiographs but were subsequently discovered by chest CT. n one patient, a contralateral synchronous cancer was found incidentally at bronchoscopy. Location of Cancers Metachronous lesions presented in the contralateral lung in 27 of the 37 patients (73%) with a higher prevalence for the upper lobes, while 10 of the 15 ( 66%) synchronous lesions were unilateral with a higher prevalence for the upper lobes. Histopathology n patients with metachronous cancers, the index cancers were adenocarcinoma in 23 (62%) and squamous cell carcinoma in 14 (38%). The second cancers were adenocarcinoma in 22 (60%) and squamous cell carcinoma in 15 (40%). However, the histologic condition of the second cancer was different from that of the first cancer in 17 patients ( 46%). Nine of these had initial adenocarcinoma followed by a squamous cell carcinoma; eight had an initial squamous cell carcinoma followed by adenocarcinoma. n patients with synchronous cancers, 12 (80%) had an adenocarcinoma-squamous cell carcinoma combination of which 9 were in the same lung and 3 were in different lungs. The remaining three patients had similar histologic condition of squamous-squamous or adenocarcinoma-adenocarcinoma combination in the same lung in one patient and in different lungs in two patients. The two patients with a third metachronous cancer had squamous cell carcinoma and small cell carcinoma, respectively. Stage Classification The top part of Table 2 shows the stage classification of the index and second metachronous cancers while the bottom of Table 2 shows the stage classification of the second cancers with respect to those of the index cancers. Twenty-nine of the 37 (78%) second metachronous cancers were stage, 6 were stage, and 2 patients presented with stage lia disease. Similarly, 27 of the 37 index metachronous cancers were stage, 5 were stage, while the remaining 5 patients presented with stage lia disease. Fourteen of the 15 index and 13 of the 15 second synchronous cancers were stage. CHEST SEPTEMBER,

4 Table 2-Stage Classification of Metachronous Lung Cancer (n=37) (Top) and Stage of Second Metachronous Cancer With Reference to First Cancer (Bottom) ndex Cancer Second Cancer Stage (78%) Stage 5 6 (16%) Stage lia 5 2(5%) Stage 27 Stage 22 Stage 5 Stage 5 Stage 4 Stage 1 Stage lia 5 Stage 3 Stage lia 2 Management Table 3 shows the treatment options for the second cancer operation based on the procedure performed for the index cancer. n patients with metachronous cancers, the index cancers were treated by lobectomy in 35, pneumonectomy in one, and radiation therapy in one patient. The second metachronous cancers were treated by lobectomy in 22, completion pneumonectomy in six, wedge resection in eight, and radiation therapy in the patient with previous pneumonectomy. Four of the 10 patients with unilateral synchronous cancers were treated by bilobectomy, two patients by pneumonectomy, one patient by a combination of lobectomy plus wedge resection, and three patients by multiple wedge resections. The five patients with bilateral synchronous cancers were offered staged thoracotomies performed at 4- to 6-week intervals. Four of the five index cancers were treated initially by lobectomy followed later by contralateral lobectomy in two patients and wedge resection in two other patients. One patient who had wedge resection for the index cancer later had a contralateral lobectomy for the second cancer. Results of Treatment Fourteen of the 37 patients (38%) with metachronous cancer were alive at the end of the review period. Twelve were free of cancer while the remaining two patients with stage lia disease were receiving adjuvant radiation therapy for recurrent disease. Twenty-three patients died within 1 to 136 months (median, 34 months; mean±sd, 36.6±11.4 months) of the second operation, 10 of recurrent cancer (8 after wedge resection and 2 after lobectomy) within 13 to 53 months of the second operation, 6 died of extrapulmonary cancers (4 colon; 2 esophagus), and 5 died of myocardial infarction but free of recurrent cancer at the time of death. The remaining two patients died in the perioperative period. Twelve of the 15 patients (80%) with synchronous cancers died of recurrent disease, while the other 3 patients died of progressive emphysematous lung disease but were free of cancer at the time of death. There were no perioperative deaths among the 15 patients with synchronous cancers, but 2 of the 36 patients (5.6%) who were treated by pulmonary resection for metachronous cancer died in the perioperative period. A 68-year-old woman who had a Table 3-Treatment Options for Second Cancer With Reference to First Procedure Metachronous cancers First Operation N Second Operation N Lobectomy Pneumonectomy Radiation therapy Lobectomy (Contralateral 20) (psilateral l ) Wedge resection (Contralateral 6) (psilateral 2) Completion pneumonectomy Radiation therapy Lobectomy l Synchronous cancers Unilateral (10) First and Second Cancers N Bilateral (5) ndex Cancer N Second Cancer N Bilobectomy (include RML*) Wedge +wedge Pneumonectomy Lobectomy + wedge Lobectomy Wedge 4 Lobectomy 2 Wedge 2 1 Lobectomy *HML=right middle lobe. 696 Clinical nvestigations

5 right lower lobectomy for a stage (T2NOMO) squamous cell carcinoma in July 1993 presented with a new stage (TlNOMO) left lower lobe adenocarcinoma in October She died of ARDS and sepsis on the 45th postoperative day. The second death occurred in a 61-year-old man who had a left lower lobectomy for stage (T2NOMO) adenocarcinoma in July He presented with a new stage (T2NOMO) right upper lobe squamous cell carcinoma a year later and died of ARDS on the 14th postoperative day. The overall actuarial 5-year survival for all patients was 32% with a median survival of 43 months (range, l to 136 months). The actuarial 5-year survival for second metachronous cancers was 37%, median survival of 48 months (range, l to 136 months), while the actuarial 5-year survival for synchronous cancers was 0% with a median survival of 43 months (range, 6 to 58 months). The survival curves of both second metachronous and synchronous cancers were identical up to the fourth year; thereafter no patient with synchronous cancer survived beyond the fifth year. When the actuarial 5-year survival rates for second metachronous cancers were calculated on the basis of stage of the disease, the 5-year survival for stage was 39% (median, 36 months; range, l to 136 months) while that of stages and lia was 38% (median, 53 months; range, 2 to 93 months) (p=0.2l). Also, there was no survival advantage based on histologic condition of the second metachronous cancers. The 5-year survival for squamous cell carcinoma was 46% (median, 53 months; range, l to 136 months) and that for adenocarcinoma was 28% (median, 36 months; range, l to 78 months) (p=0.26). The actuarial 5-year survival for second metachronous cancer in patients with different histologic features was 36.9% (median, 53 months; range, l to 75 months) as compared with 36.5% (median, 36 months; range, l to 136 months) for patients with similar histologic conditions (p=0.79) (Table 4; Figs l-6). DSCUSSON Very little was written about MPLC until 1924 when Beyreuther 20 identified two separate primary lung cancers in a patient with pulmonary tuberculosis. n 1932, Warren and Gates 21 established the criteria for the diagnosis of multiple primary malignant tumors while Auerbach and associates 4 reported a frequency of 3.5 to 14.5% of MPLC in However, it was not until 1975 that Martini and Melamed 5 provided the criteria that have now become the gold standard for the diagnosis of MPLC. Although the actual incidence of MPLC is unknown, its frequency is probably underestimated in the literature with reported frequency ranging from 0.8 to 14.5%_1-3,5-14 The diagnosis of synchronous or metachronous MPLC rests on certain criteria as outlined by Martini and Melamed 5 with recent modification by Antakli and associates. 6 A second lung cancer with different histologic features or arising from carcinoma in situ or located in the contralateral lung should be considered as synchronous MPLC and should be staged separately; however, the higher stage should be recorded for the patient. However, when two synchronous lesions of similar histologic condition are located in different lobes of the same lung, the diagnosis of synchronous MPLC is difficult. Also, it is very difficult to distinguish between poorly differentiated adenocarcinoma, squamous cell carcinoma, and large cell carcinoma by conventional pathologic Table 4-Analysis of Actuarial Survival of Patients With MPLC* 3-yr 5-yr Median Range N Surv,% Surv,% Surv, mo Surv, mo All patients Synchronous Ca Metachronous Ca First cancer Second cancer Squamous ce ll Adenocarcinoma Stage Stages and A Pneumonectomy Lobectomy Wedge resection Different histologic condition Same histologic condition *Surv=survival; Ca=cancer. CHEST SEPTEMBER,

6 g :c jg 6, e Cl. Sj.4 J O.OJ:-- 0-1, ; 2: c 2 o -: 4- : 3-6: -,- - 4, 8: :- : 6-0: " " ,. 10, FGURE 1. The overall actua1ial survival of 52 patients with metachronous and synchronous multiple p1imary lung cancers. The actuarial 5-year survival was 32%, median survival was 43 months (mean±sd, 52± 7 months; range, 1 to 136 months). studies. n fact, it has been reported that most poorly differentiated adenocarcinomas contained squamous cell components when analyzed by electron microscopy.22 chinose and his colleagues have shown that DNA flow cytometry can be used to differentiate between two synchronous cancers of the same histologic condition or between the index and the second metachronous cancer of similar histologic condition by analysis of the DNA ploidy of the first and second tumors using DNA flow cytometry. The tumors are considered to be independent of each other if one tumor showed diploidy and the other showed aneuploidy or when the two tumors showed aneuploidy but with different DNA index of abnormal clones. The tumors are considered to be related to each other if they showed diploidy or when at least. 8 g.7.6 [. 5 ro -.4 1:.3.1 ; ;.... 1stMetachronous(n,.37) 2nd Metachronous (n-37) Synchronous (n 15j ' 't..., O. OO!- - 1" 2' -" -2' 4" ' -3 6 x-c - 84= '" 0" ' -7 " " FGURE 2. The actuarial survival of the first and second metachronous lung cancers compared with that of synchronous cancer. The 5-year actuarial survival after the first metachronous cancer operation was 75%. The survival curve of the second metachronous cancer is almost identical to that of synchronous cancer up to the fourth year but reaching statistical significance at 5 y ears (p < 0.001). The actuarial 5-year survival for second metachronous cancer was 37% ( median, 48 months; mean±sd, 54±8 months) a s compared with 0% (median, 43 months; mean ±SD, 40± 5 months) for synchronous cancer. g.7 :c jg : $.3 i.2 ' 1i,stage! (n 29) Staguii&A(n"'8) o.o o! -1" 2 '- - 2"4 - c FGURE 3. Actuarial5-year survival of stage v stages and lia following operation for the second metachronous cancer. The actuarial5-year survival for stage was 39% (median, 36 months; mean±sd, 52±10 months) a s compared with 38% ( median, 53 months; mean±sd, 50± ll months) for stages and A (p = 0.21 ). one DNA index of abnormal clones between two aneuploidy tumors was the same or almost identical. This is a promising tool in the diagnostic armamentarium of lung cancer which, in combination with molecular markers, will help with better stage classification, determination of prognosis, and our understanding of the biological b ehavior of lung cancer in various individuals. The mainstay of diagnosis of metachronous MPLC is careful follow-up of patients after cancer surgery As a result of our aggressive follow-up policy, 29 of the 37 (78%) second metachronous cancers were discovered at stage. Without such policy, it is possible that most of these patients would have been discovered with more advanced disease. Despite the vvidespread education on smoking and lung cancer, it was disturbing to note that all the 37 :c : Ll - - -,., SquamousCell(n 15) L Adenocarcinoma (n 22) 1 L O.OJ: ,1:-::- 2--:2,-; 4--:3:-: :-;: 8---,6o:;:0--:7o:: 2- -:8,-; 4--:9:-:-6--:1-,c08,---,-, 120 FGURE 4. Actuarial 5-year survival of squamous cell carcinoma and adenocarcinoma following operation for the second metachronous cancer. The actuarial 5-year survival for squamous cell carcinoma was 46% (median, 53 months; mean±sd, 60±15 months) as compared with 28% ( median, 36 months; mean±sd, 42±6 months) for adenocarcinoma ( p = 0.26). 698 Clinical nvestigations

7 !:-::: - Oirferent Histology (n 17).1 same Histology (n 20) o.o±-o --1"'"2 --=2' : ---;,4S;;------;:';6o, ,;;-'------;a;;-;4---;9,-6 : : - ; 1 o a - - : ; : : FGURE 5. Actuarial 5-year survival following second operation for metachronous cancer based on change of histologic condition. There was no survival advantage between both groups (37%). However, there was a significant difference in the median survival. Patients with different histologic conditions had a median survival of 53 months (mean±sd, 46::'::7 months) as compared to a median survival of 36 months (mean::'::sd, 58::+::4 months) for patients \Vith similar histologic conditions. patients with metachronous cancers were still smoking at the time of discovery of the second cancer while 15 continued to smoke after their second operation. Richardson and associates 26 noted fewer smoking-related second primary cancers in patients who stopped smoking after their curative lung cancer surgery and emphasized the need for continuous and concerted education of patients in smoking cessation. From our findings, it appears that continuation of smoking after lung cancer resection increases the risk of a new primary lung cancer irrespective of the stage of the index cancer. Our results also show that :c e U:.:.-:..:i l c...5 (ij ].1 - = : L.-, c l - 1, : '-=+-;,,_ : Pneumonectomy (n S) -----_Lobectomy (n-22) --Wedge Resection (n-3) 2 4c - o.o,_o - -,1""2 - -, , , -... l : = , : i -- 4 a, o : t : : - ; ; 0 - -, 7 c : : : ' 2 - = a " " 4 --=oo-= ,-,1o"'"a - -:o. FGURE 6. Actuarial 5-year survival based on the type of pulmonary resection for second metachronous cancer. The actuarial 5-year survival for pneumonectomy was 63% (median, 75 months; mean, 74 ::':: 26 months). This apparent survival advantage of pneumonectomy over lobectomy and wedge resection may be due to the small number of patients who had pneumonectomy. However, the survival curve for lobectomy is similar to that of wedge resection. The actuarial 5-year survival for lobectomy was 33% (median, 35 months; mean::'::sd, 45::'::7 months) as compared with 29% (median, 38 months; mean::'::sd, 42::'::10 months) for wedge resection (p=0.3l). 120 the prognosis of metachronous cancers is poorer than that of primary lung cancer when compared stage for stage. The actuarial 5-year survival of 37% for stage second metachronous cancer is very disconcerting when compared with the actuarial 5-year survival of 75% for primary stage disease in our institution. The diagnosis of synchronous MPLC is commonly made from chest radiographs and CT. Some are incidental findings at bronchoscopy or at surgery and others are less commonly found by pathologists during examination of resected specimens. n a large number of cases, the diagnosis is never made but discovered at autopsy as shown by Auerbach et al 4 and other authors t is our policy to perform bronchoscopy on all patients undergoing pulmonary resection preoperatively or intraoperatively to assess bronchial anatomy, determine the extent of disease, and rule out unsuspected endobronchial lesions not detected by chest radiographs or CT. The surgical options for second metachronous cancers depend on the extent of the disease, the initial surgical procedure, and the patient's pulmonary reserve. n general, limited resection is favored for a second peiipheral cancer However, if the patient can tolerate another lobectomy, this should be the procedure of choice. n a recent randomized trial of lobectomy vs limited resection for T1NO non-small cell lung cancer, Ginsberg and Rubinstein29 found that limited resection does not confer improved perioperative morbidity, mortality, or late postoperative pulmonary function when compared with a lobectomy. Limited resection was associated with a 75% increase in recurrence rates attributable to tripling of the local recurrence rate, and a 30% increase in overall death rate when compared with patients undergoing lobectomy. However, if a lobectomy or segmentectomy had been performed initially, a completion pneumonectomy is feasible for an ipsilateral cancer; a lobectomy or limited resection for a contralateral cancer; a middle lobectomy on the same side or a wedge resection for an ipsilateral or contralateral cancer n few selected patients with adequate pulmonary reserve, a left pneumonectomy may be feasible after an initial right upper or middle lobectomy. n patients with deep-seated lesions in which the anatomic locations of the lesions make wedge resection difficult or impossible, the "precision cautery excision" technique first described by Perelman33 in 1983 and later popularized by Cooper and associates 34 in 1986 will provide adequate resection of a second metachronous cancer with little morbidity and mortality while preserving adequate lung tissue. Although a previous pneumonectomy is generally CHEST /112 /3/ SEPTEMBER,

8 considered a contraindication to further pulmona1y resection, with careful selection, some of these patients may be good candidates for limited resection of a second primary lung cancer in the remaining lung. Such patients, however, should undergo full metastatic workup, quantitative ventilation/perfusion scintigraphs with estimation of the postoperative FEY v and measurement of maximal oxygen consumption with exercise. Patients with borderline cardiopulmonary reserve may need to undergo right heart catheterization with balloon occlusion of the lobe to be resected in order to quantitate the expected postoperative pulmonary artery pressure Patients with previous left pneumonectomy may tolerate wedge resection or a right upper or middle lobectomy if pulmonary function tests justify such resection. However, patients with a previous right pneumonectomy will tolerate only very limited resection of the left lung When a patient presents with ipsilateral synchronous cancers, the two cancers can be resected by bilobectomy, pneumonectomy, multiple wedge resections, or a combination of lobectomy and wedge resection. Patients with bilateral synchronous cancers should be offered staged bilateral thoracotomies at an interval of 4 to 6 weeks starting with the side with the higher stage of disease Median sternotomy may be used for wedge resections of small peripheral upper lobe cancers, but this approach is not recommended for lower lobe cancers.4l Our overall operative mortality for all patients with MPLC was 3.9%. Our operative mortality of 5.6% for metachronous cancers and 0% for synchronous cancers compares favorably with reported operative mortality of 0 to 20% for metachronous cancers and 4.5 to 20% for synchronous cancers Also, our overall actuarial 5-year survival of 32% for all patients with MPLC is in consonant with 18 to 36% actuarial 5-year survival reported in the literature for this group of patients Patients with metachronous cancers had a 37% actuarial 5-year survival while no patient with synchronous cancers had survived 5 years so far. This is also similar to other reports of ll to 36% 5-year survival for metachronous cancers and 0 to 28% 5-year survival for synchronous cancers quoted in the literature. 3,5-7,9-J 3,45-48 CONCLUSONS Multiple pulmonary nodules occuning in different lobes or lungs may represent synchronous lung cancers and should not be presumed to be metastatic disease. A careful search for extrapulmonary primary cancer is mandatory. n the absence of any positive findings, such lesions should be considered synchro- 700 nous MPLC and given surgical considerations. Similarly, when a new solitary pulmona1y nodule develops 2 or more years after curative resection for lung cancer, the patient should be offered surgical therapy if the lesion is resectable since this offers the best chance for prolonged survival. We believe that all patients who have undergone curative resection for lung cancer should be followed up for life since a second cancer appears only in long-term survivors and appears within the third and fifth year and possibly as late as 15 to 25 years after initial resection. We conclude that an aggressive surgical approach is safe and justified in most patients with metachronous lung cancers, the operative morbidity and mortality are acceptable, and long-term survival is possible in many of these patients. ACKNOWLEDGMENTS: The authors wish to express their sincere thanks and appreciation to Gayle Taylor for her valuable help in extracting the data from the Cancer Registry, Robin Plauche for her secretarial assistance, Danielle Howard for her assistance with data ently, and Professor Oluwatope Abimbola Mabogunje for his editorial assistance. REFERENCES 1 Martini N, Bains MS, Burt ME, et a!. ncidence of local recurrence and second primary tumors in resected stage lung cancer. J Thorac Cardiovasc Surg 1995; 109: Pairolero PC, Williams DE, Bergstralh EJ, eta!. Postsurgical stage bronchogenic carcinoma: morbid implications of recurrent disease. Ann Thorac Surg 1984; 38: Deschamps C, Pairolero P, Trastek V, et al. Multiple primary lung cancers. J Thorac Cardiovasc Surg 1990; 99: Auerbach 0, Stout AP, Hammond EC, et al. Multiple primary bronchial carcinomas. Cancer 1967; 20: Martini N, Melamed MR. Multiple primary lung cancers. J Thorac Cardiovasc Surg 1975; 70: Antakli T, Schaefer RF, Rutherford JE, et a!. Second primary lung cancer. Ann Thorac Surg 1995; 59: Jensik RJ, Faber LP, Kittle CF, et a!. Survival following resection for second primary bronchogenic carcinoma. J Thorae Cardiovasc Surg 1981; 82: Kono M, Fujii M, Adachi S, et a!. Multiple primary lung cancers: radiographic and bronchoscopic diagnosis. J Thorac maging 1993; 8: Mathisen DJ, Jensik RJ, Faber LP, et a!. Survival following resection for second and third lung cancers. J Thorac Cardiavase Surg 1984; 88: Mousset X, Deslauriers J, Beaulieu M, eta!. Multiple primaty lung cancers: the importance of early detection and survival after a new surgical excision. Ann Chir 1989; 43: Ribet M, Dambron P. Multiple primary lung cancers. Eur J Cardiothorac Surg 1995; 9: Verhagen AF, van de Wal HJ, Cox AL, et a!. Surgical treatment of multiple primary lung cancers. Thorac Cardiavase Surg 1989; 37:107-ll 13 Verhagen AF, Tavilla G, van de Wal HJ, e t a!. Multiple primary lung cancers. Thorac Cardiovasc Surg 1994; 42: Wu SC, Lin ZQ, Xu CW, eta!. Multiple primary lung cancers. Chest 1987; 92: chinose Y, Hara N, Ohta M. Synchronous lung cancers defined by deoxyribonucleic acid flow cytometry. J Thorac Cardiovasc Surg 1991; 102: Clinical nvestigations

9 16 chinose Y, Hara N, Ohta M, et a!. DNA ploidy patterns of tumors diagnosed as metachronous or recurrent lung cancers. Ann Thorac Surg 1991; 52: Moutain CF. A new international staging system for lung cancer. Chest 1986; 89:225S-33S 18 Beahrs OH, Henson DE, Hunter RV, eta!. American Joint Committee on Cancer: manual for staging of cancer. 4th ed. Philadelphia: JB Lippincott, 1992; Kaplan EL, Meier P. Nonparametric estimation from incomplete observation. J Am Stat Assoc 1958; 53: Beyreuther H. Multiplicitat von carcinomen bei ein fall von sog: 'schneeberger' lungenkrebs mit tuberculose. Virchows Arch Pathol Anat 1924; 250: Warren S, Gates 0. Multiple primary malignant tumors: a survey of the literature and a statistical study. Am J Cancer 1932; 16: McDowell EM, Harris CC, Trump BF. Histogenesis and morphogenesis of bronchial neoplasms. n: Shimosato Y, Melamed MR, Nettessheim P, eds. Morphogenesis of lung cancer. Boca Raton, Fla: CRC Press, 1982: Boucot KR, Weiss W, Cooper DA. Second pulmonary neoplasm among long-term survivors of lung cancer. Am Rev Respir Dis 1965; 92: Bower SL, Chopin H.H, Muss HB. Multiple primary bronchogenic carcinoma. AJR Am J Roentgenoll983; 140: Van Bodegom PC, Vagenaar SS, Corrin B, et al. second primary lung cancer: importance of long-term follow-up. Thorax 1989; 44: Richardson GE, Tucker MA, Venzon DJ, et al. Smoking cessation after successful treatment of small cell lung cancer is associated with fewer smoking-related second primary cancers. Ann ntern Med 1993; 119: Jensik H.J, Faber LP, Kittle CF, eta!. Segmental resection for bronchogenic carcinoma. Ann Thorac Surg 1979; 28: Shields TW, Higgins GA Jr. Minimal pulmonary resection in the treatment of carcinoma of the lung. Arch Surg 1974; 108: Ginsberg RJ, Rubinstein LV. Randomized trial of lobectomy versus limited resection for T1NO non-small cell lung cancer: Lung Cancer Study Group. Ann Thorac Surg 1995; 60: Temeck BK, Flehinger BJ, Martini N. A retrospective analysis of 10-year survivors from carcinoma of the lung. Cancer 1984; 53: McGovern EM, Trastek VF, Pairolero PC, et al. Completion pneumonectomy: indications, complications and results. Ann Thorac Surg 1988; 46: Neptune WB, Woods FM, Overholt H.H. Reoperation for bronchogenic carcinoma. J Thorac Cardiovasc Surg 1966; 52: Perelman M. Precision techniques for removal of pathological structures from the lung. Surgery 1983; 11: Cooper JD, Perelman M, Todd TRJ et a!. Precision cautery excision of pulmonary lesions. Ann Thorac Surg 1986; 41: Barker JA, Yahr WZ, Krieger BP. Right upper lobectomy 20 years after pneumonectomy: preoperative evaluation and follow-up. Chest 1990; 97: Boysen PC, Block AJ, Olsen GN, eta!. Prospective evaluation for pneumonectomy using the 99 mtechnetium quantitative perfusion lung scan. Chest 1977; 72: Williams AJ, Gayton RM, Harding LK, et al. Quantitative lung scintigrams and lung function in the selection of patients for pneumonectomy. B J Dis Chest 1984; 78: Kittle CF, Faber LP, Jensik RJ, et al. Pulmonary resection in patients after pneumonectomy. Ann Thorac Surg 1985; 40: Westermann CJJ, van Swieten HA, de a H.iviere AB, eta!. Pulmonary resection after pneumonectomy in patients with bronchogenic carcinoma. J Thorac Cardiovasc Surg 1993; 106: Chaudri MR. ndependent bilateral primary bronchial carcinomas. Thorax 1971; 26: Clifton EE, Gupta TD, Pool JL. Bilateral pulmonary resection for primary or metastatic lung cancer. Cancer 1964; 17: Ferguson MK, De Meester TR, Deslauriers J, et al. Diagnosis and management of synchronous lung cancers. J Thorac Cardiovasc Surg 1985; 89: Ferguson MK. Synchronous primary lung cancers. Chest 1993; 103:398S-400S 44 Payne WS, Clagett OT, Harrison EG. Surgical management of bilateral malignant lesions of the lung. J Thorac Cardiovasc Surg 1962; 43: Salerno T, Munro D, Blundell P, et a!. Second primary bronchogenic carcinoma: life-table analysis of surgical treatment. Ann Thorac Surg 1979; 27: Benfield JR. Multiple primary lung cancers. J Thorac Cardiovasc Surg 1991; 101: H.osengart T, Martini N, Chosen P, et a!. Multiple primary lung carcinoma. Ann Thorac Surg 1991; 52: 'Neiss W. Prognosis after resection of metachronous multiple bronchogenic carcinoma. Am J Surg 1979; 138: CHEST SEPTEMBER,