One advantage of the use of peripheral blood stem cells (PBSCs) over autologous bone marrow would be a reduced risk of tumor cell contamination. However, the level of neoplastic cells in the PB of multiple myeloma (MM) patients after mobilization protocols is poorly investigated. In this study, we evaluated PB samples from 27 pretreated MM patients after the administration of high dose cyclophosphamide (7 g/m2 or 4 g/m2) and granulocyte-colony stimulating factor for the detection of myeloma cells as well as hematopoietic progenitors. Plasma cells containing intracytoplasmic lg were counted by microscope immunofluorescence after incubation with appropriate antisera directed against light- and heavy-chain lg. Moreover, flow cytometry studies were performed to determine the presence of malignant B-lineage elements by using monoclonal antibodies against the CD19 antigen and the monotypic light chain. Before initiation of PBSC mobilization, circulating plasma cells were detected in all MM patients in a percentage ranging from 0.1% to 1.8% of the mononuclear cell fraction (mean value, 0.7% +/- 0.4% SD). In these patients, a higher absolute number of PB neoplastic cells was detected after chemotherapy and granulocyte colony-stimulating factor. Kinetic analysis showed a pattern of tumor cell mobilization similar to that of normal hematopoietic progenitors with a maximum peak falling within the optimal time period for the collection of PBSCs. The absolute number of plasma cells showed a 10 to 50-fold increase as compared with the baseline value. Apheresis products contained 0.7% +/- 0.2% SD of myeloma cells (range, 0.2% to 2.7%). Twenty-three MM patients were submitted to PBSC collection. In 10 patients, circulating hematopoietic CD34+ cells were highly enriched by avidin-biotin immunoabsorption, were cryopreserved, and used to reconstitute bone marrow function after myeloablative therapy. The median purity of the enriched CD34+ cell population was 89.5% (range, 51% to 94%), with a 75-fold increase as compared with the pretreatment samples. The median overall recovery of CD34+ cells and colony-forming unit-granulocyte-macrophage was 58% (range, 33% to 95%) and 45% (range, 7% to 100%), respectively. Positive selection of CD34+ cells resulted in 2.5- to 3-log depletion of plasma cells and CD19+ B-lineage cells as determined by immunofluorescence studies, although DNA analysis of CDR III region of IgH gene showed the persistence of minimal residual disease in 5 of 6 patient samples studied. Myeloma patients were reinfused with enriched CD34+ cells after myeloablative therapy consisting of total body irradiation (1,000 cGy) and highdose melphalan (140 mg/m2). They received a median of 4 x 10(6) CD34+ cells/kg and showed a rapid reconstitution of hematopoiesis; the median time to 0.5 x 10(9) neutrophils and to 20 and 50 x 10(9) platelets per liter of PB was 10, 11, and 12 days, respectively. These results, as well as other clinically significant parameters, did not significantly differ from those of patients (n = 13) receiving unmanipulated PBSCs after the same pretransplant conditioning regimen. In summary, our data show the concomitant mobilization of tumor cells and hematopoietic progenitors in the PB of MM patients. Positive selection of CD34+ cells reduces the contamination of myeloma cells from the apheresis products up to 3-log and provides a cell suspension capable of restoring a normal hematopoiesis after a total body irradiation-containing conditioning regimen.