Summary Cell kinetics have been shown to be an important predictor of clinical evolution of operated breast cancer. We established a method for the estimation of the proliferative activity of tumour cells obtained by fine needle sampling without aspiration (FNS), using simultaneously S-phase fractions (SPF) measured on DNA histograms and 5-bromodeoxyuridine (BrdU) labelling index (BLI) Proliferative activity has consistently been found to be an important biological predictor of the clinical evolution of breast cancer (Tubiana et al., 1984;Silvestrini et al., 1985;Meyer, 1986;Hery et al., 1987). Traditionally, it has been measured by autoradiography of a radiolabelled precursor (3H-thymidine) incorporated in tumour-cell DNA.With the introduction of DNA flow cytometry (FCM), fast cell-cycle distribution analysis of large numbers of cells has become available (Barlogie et al., 1982). Retrospective studies have shown that S-phase fraction (SPF), computed from DNA histograms, is also a significant prognostic factor (Hedley et al., 1987;Kallioniemi et al., 1988) and a good predictor of response to neo-adjuvant chemotherapy (Remvikos et al., 1989). Nevertheless, the possibility of establishing SPF's correctly has been questioned, either from poor quality-or complex histograms presenting multiple aneuploid peaks (Meyer et al., 1984; Kallioniemi et al., 1985 In a prospective study of feasibility, we developed complementary methods for assaying SPF and BrdU labelling index (BLI) in vitro on the same tumour samples. Indeed, the substitution of 5-bromodeoxyuridine (BrdU) to 3H-thymidine has been proposed as a more convenient method of measuring DNA synthesis, due to the instant immunofluorimetric detection (Dolbeare et al., 1983;Schutte et al., 1987). Tumour cells were obtained directly from the patients by fine needle sampling without aspiration. The inherent difficulties of each methodology are discussed in order to define the best approach for an adequate estimation of the proliferative activity of each breast cancer before treatment decision.
Materials and methodsOne hundred and eighty-nine consecutive patients were subjected to fine needle sampling without aspiration, using 23 or 25 gauge needles (Zajdela et al., 1987). The cytological samples were expelled in 1 ml of incubating medium: RPMI (Gibco), 10% foetal calf serum (Calbiochem), 30 pM BrdU (Sigma). The tubes were incubated at 37°C for 15 min, then 100 yl of DMSO were added and the tubes stored at -80C.FNS were processed for DNA-FCM according to a onestep protocol. The samples were thawed and rinsed with PBS at 1,500 r.p.m., 5 min. The pellets were resuspended in 600 tl of PBS, 0.2% Tween 80 (Sigma), 50 lg m -' propidium iodide (Sigma), 250 iLg ml-Ribonuclease A (B6ehringer), and left at room temperature for 20 min.For BrdU/DNA labelling, cell-rich samples were selected and fixed in 70% ethanol. In all cases, at least 50,000 cells were kept for DNA analysis as above. The fixed samples were centrifuged, digested with 3 ml of pepsin (Sigma