Tohoku
J. Exp.
Med., 1992,
167, 219-230
Cytologic Effusion
Assessment and Prognosis
Patients
Who
of Peroperative Pleural in Lung Cancer
Underwent
Resection
KATSUO USUDA, YASUKI SAITO, CHIAKI ENDO, SATOMI TAKAHASHI,KEIJI KANMA, MASAMI SATO, MOTOYASU SAGAWA, NORIYOSHINAGAMOTO, SHIGEFUMIFUJIMURAand SHIN'ICHIRoQTA* Department of Surgery, the Research Institute for Tuberculosis and Cancer, Tohoku University, Sendai 980, and *Department of Thoracic Surgery, Shizuoka Central Hospital, Shizuoka 420
USUDA,K., SAITO,Y., ENDO, C., TAKAHASHI, S., KANMA,K., SATO,M., SAGAWA, M., NAGAMOTO, N., FUJIMURA,S. and QTA,S. Cytologic Assessment of Peroperative Pleural Effusion and Prognosis in Lung Cancer Patients Who Underwent Resection. Tohoku J. Exp. Med., 1992, 167 (3), 219-230 Twenty-five of 108 lung cancer patients who underwent resection had cytologically positive pleural effusions. The rate at which cancer cells were detected was not related to the amount of the effusion. Almost one third of patients with cancer cells in effusion were alive at the end of the third postoperative year, provided that the pleura itself was free of metastasis at the time of operation. Correlation of the cytologically positive rate of pleural effusion (Y) with the degree of pleural metastasis (X,), the degree of pleural involvement (X2), or the degree of nodal involvement (X3) was analyzed using the Hayashi's quantification method type I. The multiple correlation coefficient was 0.843. Partial correlation coefficients of X1, X2i and X3 were 0.733, 0.446, and 0.653, respectively. Pleural metastasis had the strongest effect on the cytologically positive rate of pleural effusion. lung cancer ; peroperative pleural effusion ; cytology ; prognosis ; Hayashi's quantification method
Pleural effusion is frequently detected at thoracotomy for lung cancer. Up to the present, considerable attention has been directed to a relatively large amount of preoperative pleural effusion (Brinkman 1959; Jarvi et al. 1972; Decker et al. 1978; Canto et al. 1985; Monte et al. 1987; Mountain 1987, 1988), while, to the author's knowledge, there has been no study on a relatively small amount of peroperative pleural effusion in association with lung cancer. In
Received
January
22, 1992; revision accepted : Katsuo Usuda, M.D.,
Address for reprints Institute for Tuberculosis Sendai
and
Cancer,
Tohoku
980, Japan. 219
for publication Department of University,
4-1
July 22, 1992. Surgery, the Research Seiryo-machi,
Aoba-ku,
220
K. Usuda
et al.
general, lung cancer patients with cytologically positive pleural effusion have been out of consideration as regards surgical intervention (Brinkman 1959; Decker et al. 1978; Canto et al. 1985; Monte et al. 1987; Mountain 1987). Our study deals with a cytologic assessment of a relatively small amount of peroperative pleural effusion in lung cancer patients, whose prognosis after removal of cancer was discussed. MATERIALS AND METHODS
A cytological study of the peroperative pleural effusion was undertaken in a total of 108 lung cancer patients (82 males and 26 females) who underwent operation in our hospital from 1977 to 1986. The average age of the above patients was 62 years (ranging from 28 to 82 years). The pleural effusions escaped detection by imaging technique before operation, and was then discovered at thoracotomy. Specimens of pleural effusion were prepared in the following manner. Fresh fluid specimens were fixed with 50% ethanol and 2% carbowax. Each fluid sample was centrifuged at 3,000 rpm for five min. The resulting cell sediment was spread on a slide, and stained by the Papanicolaou method. All the specimens were screened by cytotechnologists and then by a cytopathologist for final diagnosis. The cytologic results were classified as "positive", "suspicious", and "negative" based on the evaluation for the presence of tumor cells. For each patient the volume and macroscopic findings of the pleural effusion were recorded. The histologic cell types and pathologic stages were documented based on the criteria of Union Internationale Contre le Cancer (UICC) classification (International Union Against Cancer 1987). Lung cancer with cytologically positive pleural effusion (malignant pleural effusion) was classified as "T4 lung cancer", and as "incomplete resection" with respect to curativity. Pleural involvement was classified into four types according to the general rules for clinical and pathological record of lung cancer of the Japan Lung Cancer Society (1987) as follows : p0 lesion - without involvement of the pleural elastic membrane ; pl lesion - an invasion beyond the pleural elastic membrane with the intact pleural surface ; p2 lesion an invasion involving the pleural surfaces ; and p3 lesion - an invasion involving the pleural surface and surrounding organs (chest wall, mediastinum, pericardium, diaphragm). Pleural metastasis was classified according to the general rules for clinical and pathological record of lung cancer of the Japan Lung Cancer Society (1987) : DO lesion - absence of
TABLE1. Relationship between histologic cell types and pathologic stages of 108 resected lung cancers with peroperative pleural effusions
Cytologic
Assessment
of Pleural
Effusion
in Lung
Cancer
221
pleural metastasis ; D1 lesion - slightly positive pleural metastasis ; and D2 lesion - strongly positive pleural metastasis. There were 47 adenocarcinomas, 43 squamous cell carcinomas, eight adenosquamous carcinomas, six small cell carcinomas, and four large cell carcinomas (Table 1). When classified by pathologic staging, 22 patients were at stage I, 6 at stage II, 41 at stage IIIA, 31 at stage IIIB and 8 at stage IV (Table 1). As regards curativity, there were 57 complete resections, and 51 incomplete resections. Correlation of the cytologically positive rate (CPR) of pleural effusion with the degree of pleural metastasis, the degree of pleural involvement, or the degree of nodal involvement was analyzed using the Hayashi's quantification method type I (Hayashi 1950, 1952). The survival rate after operation was calculated by the Kaplan-Meier method. In this calculation, deaths due to diseases other than recurrence of lung cancer were excluded. A statistical evaluation was carried out by the generalized Wilcoxon test.
RESULTS Effusion was positive for cancer cells in 25 (23%), suspicious in two, and negative in 81 of the 108 patients (Table 2). No relationship could be found between the volume of pleural effusion and the cytologically positive rate (CPR). Of 48 patients with macroscopically serosanguinous effusion, 11 (23%) had cytologically positive effusions. Of 39 patients with macroscopically serous effusion, nine (23%) had cytologically positive effusions. The CPR of serosanguinous effusion did not exceed that of serous effusion. The CPR of pleural effusion was 9% (4/43) for squamous cell carcinomas and 32% (21/65) for non-squamous cell carcinomas (Table 3). The former was significantly less than the latter, and also significantly less than 30% (14/47) for adenocarcinomas (the x 2 test, p
J. Exp.
Med., 1992,
167, 219-230
Cytologic Effusion
Assessment and Prognosis
Patients
Who
of Peroperative Pleural in Lung Cancer
Underwent
Resection
KATSUO USUDA, YASUKI SAITO, CHIAKI ENDO, SATOMI TAKAHASHI,KEIJI KANMA, MASAMI SATO, MOTOYASU SAGAWA, NORIYOSHINAGAMOTO, SHIGEFUMIFUJIMURAand SHIN'ICHIRoQTA* Department of Surgery, the Research Institute for Tuberculosis and Cancer, Tohoku University, Sendai 980, and *Department of Thoracic Surgery, Shizuoka Central Hospital, Shizuoka 420
USUDA,K., SAITO,Y., ENDO, C., TAKAHASHI, S., KANMA,K., SATO,M., SAGAWA, M., NAGAMOTO, N., FUJIMURA,S. and QTA,S. Cytologic Assessment of Peroperative Pleural Effusion and Prognosis in Lung Cancer Patients Who Underwent Resection. Tohoku J. Exp. Med., 1992, 167 (3), 219-230 Twenty-five of 108 lung cancer patients who underwent resection had cytologically positive pleural effusions. The rate at which cancer cells were detected was not related to the amount of the effusion. Almost one third of patients with cancer cells in effusion were alive at the end of the third postoperative year, provided that the pleura itself was free of metastasis at the time of operation. Correlation of the cytologically positive rate of pleural effusion (Y) with the degree of pleural metastasis (X,), the degree of pleural involvement (X2), or the degree of nodal involvement (X3) was analyzed using the Hayashi's quantification method type I. The multiple correlation coefficient was 0.843. Partial correlation coefficients of X1, X2i and X3 were 0.733, 0.446, and 0.653, respectively. Pleural metastasis had the strongest effect on the cytologically positive rate of pleural effusion. lung cancer ; peroperative pleural effusion ; cytology ; prognosis ; Hayashi's quantification method
Pleural effusion is frequently detected at thoracotomy for lung cancer. Up to the present, considerable attention has been directed to a relatively large amount of preoperative pleural effusion (Brinkman 1959; Jarvi et al. 1972; Decker et al. 1978; Canto et al. 1985; Monte et al. 1987; Mountain 1987, 1988), while, to the author's knowledge, there has been no study on a relatively small amount of peroperative pleural effusion in association with lung cancer. In
Received
January
22, 1992; revision accepted : Katsuo Usuda, M.D.,
Address for reprints Institute for Tuberculosis Sendai
and
Cancer,
Tohoku
980, Japan. 219
for publication Department of University,
4-1
July 22, 1992. Surgery, the Research Seiryo-machi,
Aoba-ku,
220
K. Usuda
et al.
general, lung cancer patients with cytologically positive pleural effusion have been out of consideration as regards surgical intervention (Brinkman 1959; Decker et al. 1978; Canto et al. 1985; Monte et al. 1987; Mountain 1987). Our study deals with a cytologic assessment of a relatively small amount of peroperative pleural effusion in lung cancer patients, whose prognosis after removal of cancer was discussed. MATERIALS AND METHODS
A cytological study of the peroperative pleural effusion was undertaken in a total of 108 lung cancer patients (82 males and 26 females) who underwent operation in our hospital from 1977 to 1986. The average age of the above patients was 62 years (ranging from 28 to 82 years). The pleural effusions escaped detection by imaging technique before operation, and was then discovered at thoracotomy. Specimens of pleural effusion were prepared in the following manner. Fresh fluid specimens were fixed with 50% ethanol and 2% carbowax. Each fluid sample was centrifuged at 3,000 rpm for five min. The resulting cell sediment was spread on a slide, and stained by the Papanicolaou method. All the specimens were screened by cytotechnologists and then by a cytopathologist for final diagnosis. The cytologic results were classified as "positive", "suspicious", and "negative" based on the evaluation for the presence of tumor cells. For each patient the volume and macroscopic findings of the pleural effusion were recorded. The histologic cell types and pathologic stages were documented based on the criteria of Union Internationale Contre le Cancer (UICC) classification (International Union Against Cancer 1987). Lung cancer with cytologically positive pleural effusion (malignant pleural effusion) was classified as "T4 lung cancer", and as "incomplete resection" with respect to curativity. Pleural involvement was classified into four types according to the general rules for clinical and pathological record of lung cancer of the Japan Lung Cancer Society (1987) as follows : p0 lesion - without involvement of the pleural elastic membrane ; pl lesion - an invasion beyond the pleural elastic membrane with the intact pleural surface ; p2 lesion an invasion involving the pleural surfaces ; and p3 lesion - an invasion involving the pleural surface and surrounding organs (chest wall, mediastinum, pericardium, diaphragm). Pleural metastasis was classified according to the general rules for clinical and pathological record of lung cancer of the Japan Lung Cancer Society (1987) : DO lesion - absence of
TABLE1. Relationship between histologic cell types and pathologic stages of 108 resected lung cancers with peroperative pleural effusions
Cytologic
Assessment
of Pleural
Effusion
in Lung
Cancer
221
pleural metastasis ; D1 lesion - slightly positive pleural metastasis ; and D2 lesion - strongly positive pleural metastasis. There were 47 adenocarcinomas, 43 squamous cell carcinomas, eight adenosquamous carcinomas, six small cell carcinomas, and four large cell carcinomas (Table 1). When classified by pathologic staging, 22 patients were at stage I, 6 at stage II, 41 at stage IIIA, 31 at stage IIIB and 8 at stage IV (Table 1). As regards curativity, there were 57 complete resections, and 51 incomplete resections. Correlation of the cytologically positive rate (CPR) of pleural effusion with the degree of pleural metastasis, the degree of pleural involvement, or the degree of nodal involvement was analyzed using the Hayashi's quantification method type I (Hayashi 1950, 1952). The survival rate after operation was calculated by the Kaplan-Meier method. In this calculation, deaths due to diseases other than recurrence of lung cancer were excluded. A statistical evaluation was carried out by the generalized Wilcoxon test.
RESULTS Effusion was positive for cancer cells in 25 (23%), suspicious in two, and negative in 81 of the 108 patients (Table 2). No relationship could be found between the volume of pleural effusion and the cytologically positive rate (CPR). Of 48 patients with macroscopically serosanguinous effusion, 11 (23%) had cytologically positive effusions. Of 39 patients with macroscopically serous effusion, nine (23%) had cytologically positive effusions. The CPR of serosanguinous effusion did not exceed that of serous effusion. The CPR of pleural effusion was 9% (4/43) for squamous cell carcinomas and 32% (21/65) for non-squamous cell carcinomas (Table 3). The former was significantly less than the latter, and also significantly less than 30% (14/47) for adenocarcinomas (the x 2 test, p
