Despite their high response rate to treatment, virtually all patients with small-cell lung cancer (SCLC) eventually die. More accurate staging procedures may identify patients who might perhaps benefit from novel treatment strategies.Monoclonal antibodies (MAbs) recognize antigens expressed by tumor cells and have thie potential to detect small numbers of infiltrating tumor cells ;at metastatic sites. Many of the antibodies that react with SCLC have been grouped into clusters based on their pattern of reactivity with tumors and normal tissues (Souhami et al., 1987(Souhami et al., , 1991. Since SCLC frequently metastatises to bone marrow, these antibodies could be used to identify tumor cells in bone-marrow samples of patients treated for SCLC.We have used MAbs from 2 clusters (Souhami et al., 1991) to identify tumor cells in bone-marrow aspirates of patients with SCLC and we have compared the results with conventional histomorphology.
PATIENTS AND METHODS
Collection of .sumplesBone-marrow aspirates were taken from the posterior iliac crcst of 52 patients with histological or cytological diagnosis of SCLC. A total of 108 bone-marrow aspirates were taken: bilaterally in 48 patients (4 of thlese had also the marrow aspirate repeated) and unilaterally in 4; 84 samples were collected at diagnosis from 44 patients and 24 during different phases of the disease (Table I).Bone-marrow aspirates (3 to 5 ml) were collected in heparinized syringes, diluted with an equivalent volume of RPMI 1640 with 5% FCS (MA Bioproducts, Walkersville, MD), layered ovcr the same amount of Ficoll-Paque and centrifuged at 400g for 30 rnin at 19°C. The cell layer was diluted in 10 ml of RPMI 1640 with 5 % FCS and centrifuged at 350 g for 7 rnin at 4°C. The cell pellet was washed twice in 10 ml of RPMI 1640 with 5% FCS. Cytospin slides (500 rpm lor 4 min), pre-treated with poly-L-lysine, were then prepared for immunostaining using 200 11. 1 of a cell suspension containing about lo6 ml-l nucleated cells. For each antibody, 2 to 4 slides were prepared, and fixed in cold acetone for 10 min as wet preparations. After fixation, the cytospin slides were air-dried for 1 hr, then placed in 10% human serum (Sigma, St. Louis, MO) and immediately immunostained or stored at -20°C until required. As an external control, we stained cytospin preparations of N592 cells, of bone marrow of patients with non-neoplastic diseases and of peripheral-blood lymphocytes from healthy volunteers.Eighty-three bone-marrow biopsies were obtained, by means of the Jamshidi needle, from 50 of these patients. In 27 patients, the samples were collected bilaterally (in 3 they were also later obtained bilaterally during different phases of the disease) and in 23 unilaterally; 65 specimens were collected at diagnosis (Table 11) from 41 patients. Bone-marrow biopsies were fixed in 10% buffered formaldehyde solution for 12 hr, and then processed routinely for paraffin embedding. Histologic sections 3-11. thick were evaluated morphologically by hematoxylin-eosin stain, Giemsa stain, Gor...