Correlation of dihydrofolate reductase elevation with gene amplification in a homogeneously staining chromosomal region in L5178Y cells.

A methotrexate (MTX)-resistant murine lymphoblastoid cell line has been obtained by serial passage in increasing concentrations of MTX which is greater than 100,000-fold resistant to MTX (L5178YR) and has dihydrofolate reductase (DHFR) levels 300-fold higher than the parental line. The L5178YR cell line synthesizes approximately 10-11% of its total soluble cell protein as DHFR regardless of growth phase, as measured by direct immunoprecipitation with a monospecific antiserum. Molecular hybridization of a purified [3H]DNA probe complimentary to DHFR specific mRNA with cellular DNA and RNA indicates that DHFR coding sequences are elevated several hundred fold in both nucleic acid species in the mutant cell line. Giemsa-banding studies of the diploid mutant line indicate the presence of a large homogeneously staining region on chromosome No. 2. In situ molecular hybridization studies indicate that the DHFR genes are localized in this homogeneously staining region. The homogeneously staining region probably consists of tandom repeats of a basic segment approximately 800 kilo base pairs long.

Antigen CD34+ marrow cells engraft lethally irradiated baboons.

The CD34 antigen is present on 1-4% of human marrow cells including virtually all hematopoietic progenitors detected by in vitro assays. Since the anti-CD34 monoclonal antibody 12-8 reacts with a similar marrow population in baboons, it was possible to test whether this antigen is expressed by stem cells responsible for hematopoietic reconstitution in vivo. CD34+ cells were enriched from marrows of five baboons using avidin-biotin immunoadsorption. After lethal irradiation, the five animals were given 15-27 X 10(6) autologous marrow cells (3.2-4.4 X 10(6) cells/kg) containing 65-91% CD34+ cells. All animals achieved granulocyte counts greater than 1,000/mm3 and platelet counts greater than 20 X 10(3)/mm3 by 13-24 d posttransplant and subsequently developed normal peripheral blood counts. Two additional animals received 184 and 285 X 10(6) marrow cells/kg depleted of CD34+ cells. One animal died at day 29 without engraftment, while the other had pancytopenia for greater than 100 d posttransplant. The data suggest that stem cells responsible for hematopoietic reconstitution are CD34+.

Multiple myeloma clones are derived from a cell late in B lymphoid development.

We have previously demonstrated that the immunoglobulin (Ig) heavy chain variable region (VH) sequences expressed by the malignant clone in multiple myeloma (MM) contain a high degree of somatic mutation without clonal diversity. This sequence can be used to identify all members of the malignant clone in this B cell malignancy. We sequenced the variable regions expressed by patients with MM and generated primers from the complementarity determining region (CDR) sequences specific for each patient's tumor. Using these primers, we performed PCR amplification on highly purified subpopulations of cells separated by expression of CD10, CD34 and CD38. The results of these experiments demonstrate: 1) there is a small fraction of CD10-expressing tumor cells in MM patients, 2) CD34-bearing malignant cells do not exist in MM, and 3) although the vast amount of tumor is in the CD38-expressing cells, a small amount of tumor is in the CD38-negative population. We also used these primers to determine whether pre-class switch (i.e., Cmu-expressing lymphocytes) clonal cells exist in these patients. After PCR amplification with CDR1 and Cmu primers, colony hybridization was performed using both framework 3 (FR3) and CDR3 probes. Out of > 200 FR3-hybridizing colonies, < or = 5 colonies also hybridized with the CDR3 probe. Colonies which hybridized with both these probes were sequenced, and none of these sequences matched even closely the CDR3 expressed by the malignant clone. These results make the existence of a pre-class switch malignant cell unlikely in MM. Overall, these results suggest that the malignant clone in MM derives from a cell late in B lymphocyte development.

Positive selection of viable cell populations using avidin-biotin immunoadsorption.

We have developed a new method for the selective enrichment of lymphoid subpopulations from dog and human bone marrow and peripheral blood. A mononuclear cell preparation was treated successively with monoclonal antibody, biotinylated goat anti-mouse immunoglobulin and passed over a column containing avidin linked to polyacrylamide or Sepharose beads. Adherent cells were recovered by mechanical agitation and analyzed by immunofluorescence staining and fluorescence-activated cell sorter analysis. Dog bone marrow mononuclear cells were treated successively with the antibody 7.2, which recognizes the Ia-antigen, 1:500 dilution of biotinylated goat anti-mouse immunoglobulin and passed over avidin-Biogel (1 mg/ml) at a flow rate of 3.0 ml/min. Enrichment from a starting population that was 24.4 +/- 9.1% 7.2-positive to 78.3 +/- 6.8% 7.2-positive adherent cell population was observed with 47.7 +/- 7.8% recovery of 7.2-positive cells. Human bone marrow mononuclear cells were treated successively with the T cell antibody Leu-4 followed by 1:500 dilution of B-GAMIg and passed over a column of avidin-Biogel (1 mg/ml) at a flow rate of 1.5 ml/min. Enrichment from 7.2 +/- 3.3% Leu-4-positive cells in the starting cell population to 73.1 +/- 6.8% Leu-4-positive cells in the adherent cell population with total recovery of Leu-4-positive cells averaging 64.0 +/- 12.7%. Human bone marrow mononuclear cells positively selected with antibody Leu-4 or another T cell antibody, Leu-5 had a markedly enhanced response to the T cell mitogen, phytohemagglutinin compared to untreated bone marrow. Enrichment of a subpopulation of lymphocytes from dog peripheral blood mononuclear cells has been accomplished using antibody DT2, which reacts with a broad spectrum of dog lymphocytes. Nonspecific cell binding is primarily limited to granulocytes and monocytes. Future work is being directed at improving recovery of positively selected cells, reducing nonspecific cell binding and applying the technique to the selective enrichment of hematopoietic stem cells from bone marrow.

Cellular immunoabsorption using monoclonal antibodies. Selective removal of T cells from peripheral blood and bone marrow.

T cells can be selectively removed from human peripheral blood and bone marrow by passage over a column containing monoclonal anti-T-cell antibodies covalently attached to Sepharose 6MB gel. Effective depletion of T cells from peripheral blood mononuclear cells (PBMC) resulted in the appearance of Leu-2-positive cells, most of which do not express Leu-1 or Leu-4 antigens. Using a column containing anti-Leu-1 or anti-Leu-4 attached to Sepharose 6MB gel, depletion of 98.3% and 99% of T cells from bone marrow mononuclear cells (BMMC), respectively, was demonstrated with recovery of approximately 75% of non-T cells. These columns removed 92.3-98.4% T cells from PBMC with 43.5-74.8% recovery of non-T cells. Combining anti-Leu-1 and anti-Leu-4 antibodies on the same gel removed all detectable T cells from PBMC and BMMC. Proliferative responses to the T cell mitogen, phytohemagglutinin, were abolished from both PBMC and BMMC after column treatment. Preservation of hematopoietic progenitors was observed after treatment of bone marrow, with stem cell recovery averaging 83 +/- 26% for burst-forming units (erythroid), 86 +/- 14% for granulocyte-macrophage progenitors and 94 +/- 16% for granulocyte, erythroid macrophage, and megakaryocitic elements. These results suggest that clinical application of cellular immunoabsorption techniques using monoclonal antibodies will be useful in bone marrow transplantation.

Positively selected autologous blood CD34+ cells and unseparated peripheral blood progenitor cells mediate identical hematopoietic engraftment after high-dose VP16, ifosfamide, carboplatin, and epirubicin.

To investigate the feasibility of peripheral blood CD34+ cell selection and to analyze CD34+ cell-mediated engraftment after high-dose chemotherapy, we performed a phase I/II trial in 21 patients with advanced malignancies. The rationale for the selection of CD34+ cells from peripheral blood progenitor cell (PBPC) collections is based on the observation that contaminating tumor cells can be depleted approximately 3 logs using this procedure. CD34+ cells from chemotherapy+granulocyte colony-stimulating factor-mobilized PBPCs were positively selected with an avidin-biotin immunoadsorption column (CEPRATE SC system). One leukapheresis product with a median number of 2.8 x 10(6) CD34+ cells/kg was labeled with a biotinylated anti-CD34 monoclonal antibody and subsequently processed over the column. The yield of selected CD34+ cells was 73% +/- 24.6%. The purity of the CD34+ cell fraction was 61.4% +/- 19.7%. CD34+ cells were shown to represent predominantly committed progenitors coexpressing CD33, CD38, and HLA-DR molecules (lin+). They gave rise to myeloid as well as erythroid and multilineage colonies in vitro. In addition, positively selected CD34+ cells also comprised early hematopoietic progenitor cells, as shown by the presence of CD34+/lin- cells. Transfusion of positively selected CD34+ cells (2.5 x 10(6) CD34+/kg; range, 0.45 to 5.1) after high-dose VP16 (1,500 mg/m2), ifosfamide (12 g/m2), carboplatin (750 mg/m2), and epirubicin (150 mg/m2) (VIC-E) in 15 patients resulted in a rapid and stable engraftment of hematopoiesis without any adverse events. As compared with 13 historical control patients reconstituted with a comparable number of unseparated PBPCs, time to neutrophil and platelet recovery was identical in both groups (absolute neutrophil count > 500/microL, day + 12; platelet count > 50,000/microL, day + 15). These data indicate that autologous peripheral blood CD34+ cells and unseparated PBPCs mediate identical reconstitution of hematopoiesis after high-dose VIC-E chemotherapy. Because positive selection of CD34+ cells from mobilized blood results in a median 403-fold depletion of T cells, allogeneic CD34+ cells from mobilized blood should be investigated as an alternative to bone marrow cells for allotransplantation.

Ex vivo expansion of enriched peripheral blood CD34+ progenitor cells by stem cell factor, interleukin-1 beta (IL-1 beta), IL-6, IL-3, interferon-gamma, and erythropoietin.

To provide sufficient numbers of peripheral blood progenitor cells (PBPCs) for repetitive use after high-dose chemotherapy, we investigated the ability of hematopoietic growth factor combinations to expand the number of clonogenic PBPCs ex vivo. Chemotherapy plus granulocyte colony-stimulating factor (G-CSF) mobilized CD34+ cells from 18 patients with metastatic solid tumors or refractory lymphomas were cultured for up to 28 days in a liquid culture system. The effects of interleukin-1 beta (IL-1), IL-3, IL-6, granulocyte-macrophage-CSF (GM-CSF), G-CSF, macrophage-CSF (M-CSF), stem cell factor (SCF), erythropoietin (EPO), leukemia inhibitory factor (LIF), and interferon-gamma, as well as 36 combinations of these factors were tested. A combination of five hematopoietic growth factors, including SCF, EPO, IL-1, IL-3, and IL-6, was identified as the optimal combination of growth factors for both the expansion of total nucleated cells as well as the expansion of clonogenic progenitor cells. Proliferation peaked at days 12 to 14, with a median 190-fold increase (range, 46- to 930-fold) of total clonogenic progenitor cells. Expanded progenitor cells generated myeloid (colony-forming unit-granulocyte-macrophage), erythroid (burst-forming unit-erythroid), as well as multilineage (colony-forming unit-granulocyte, erythrocyte, monocyte, megakaryocyte) colony-forming units. The number of multilineage colonies increased 250-fold (range, 33- to 589-fold) as compared with pre-expansion values. Moreover, the absolute number of early hematopoietic progenitor cells (CD34+/HLA-DR-; CD34+/CD38-), as well as the number of 4-HC-resistant progenitors within expanded cells increased significantly. Interferon-gamma was shown to synergize with the 5-factor combination, whereas the addition of GM-CSF significantly decreased the number of total clonogenic progenitor cells. Large-scale expansion of PB CD34+ cells (starting cell number, 1.5 x 10(6) CD34+ cells) in autologous plasma supplemented with the same 5-factor combination resulted in an equivalent expansion of progenitor cells as compared with the microculture system. In summary, our data indicate that chemotherapy plus G-CSF-mobilized PBPCs from cancer patients can be effectively expanded ex vivo. Moreover, our data suggest the feasibility of large-scale expansion of PBPCs, starting from small numbers of PB CD34+ cells. The number of cells expanded ex vivo might be sufficient for repetitive use after high-dose chemotherapy and might be candidate cells for therapeutic gene transfer.

Reconstitution of hematopoiesis after high-dose chemotherapy by autologous progenitor cells generated ex vivo.

BACKGROUND: Autologous peripheral-blood progenitor cells can restore hematopoiesis after high-dose chemotherapy in patients with solid tumors or hematologic cancers. We investigated the ability of peripheral-blood progenitor cells generated ex vivo to restore hematopoiesis in patients with cancer who have undergone high-dose chemotherapy. METHODS: Ten patients who had received high-dose chemotherapy were given transplants of autologous progenitor cells that had been generated ex vivo. We used 11 million CD34+ hematopoietic progenitor cells as the starting population for the cell growth. This number corresponds to less than 10 percent of the usual preparation of peripheral-blood CD34+ mononuclear cells used in leukapheresis. The CD34+ cells were grown in medium containing autologous plasma, recombinant human stem-cell factor, interleukin-1 beta, interleukin-3, interleukin-6, and erythropoietin. RESULTS: No toxic effects were observed with the infusion of the generated cells. The cells promoted a rapid and sustained hemopoietic recovery when transplanted after treatment with high-dose etoposide (1500 mg per square meter of body-surface area), ifosfamide (12 g per square meter), carboplatin (750 mg per square meter), and epirubicin (150 mg per square meter). The pattern of hematopoietic reconstitution was identical to that in historical controls treated with unseparated mononuclear cells or positively selected CD34+ cells. CONCLUSIONS: A small number of peripheral-blood CD34+ cells, when grown ex vivo, can supply a population of hematopoietic precursors that have the ability to restore blood formation in patients treated with high doses of chemotherapy. This method, which requires only a small volume of the patient's blood, may reduce the risk of tumor-cell contamination, circumvent the need for leukapheresis, and allow repeated cycles of high-dose chemotherapy.

Elimination of Daudi lymphoblasts from human bone marrow using avidin-biotin immunoadsorption.

Biotinylated antibodies and an avidin-Sepharose 6MB column were utilized in a novel approach to deplete selected cell populations from human bone marrow. Fluorescein-labeled Daudi lymphoblasts were mixed with bone marrow mononuclear cells in a model system, and removal of Daudi cells was quantitatively assessed with an inverted fluorescent microscope. Treatment using the biotinylated monoclonal antibody 2H7 reactive with Bp32 antigen (expressed on Daudi cells) followed by passage over avidin-Sepharose produced greater than two logs of Daudi cell removal from bone marrow. An alternative method was tested by treating cells successively with nonbiotinylated monoclonal antibody and biotinylated goat antimouse immunoglobulin followed by passage over avidin-Sepharose. Up to three logs of Daudi cells could be eliminated from bone marrow with quantitative recovery of hematopoietic progenitors. The use of biotinylated goat antimouse immunoglobulin eliminates the need to prepare a biotin conjugate of each individual monoclonal antibody. The clinical application of cellular immunoadsorption using the avidin-biotin system may prove useful in bone marrow transplantation.

Primary osteogenic sarcoma of the bladder. Case report and review of the literature.

Primary osteogenic sarcoma of the bladder is an extremely rare disease. Fewer than 30 cases of tumors of the bladder containing bone or cartilage have been reported. Only 14 cases of well-documented primary osteogenic sarcoma of the bladder appear in the world's literature. The authors describe the first detailed description of such a patient treated with radiotherapy and chemotherapy as well as provide a review of this rare and interesting entity.

Engraftment of dogs with Ia-positive marrow cells isolated by avidin-biotin immunoadsorption.

Previous work has shown failure of engraftment in lethally irradiated dogs when autologous marrow was depleted of Ia-positive cells with an anti-Ia antibody and complement before infusion. In the current study, we have utilized an avidin-biotin immunoadsorption procedure to obtain a population of highly enriched Ia-positive cells for autologous bone marrow transplantation in dogs given lethal irradiation. Dog marrow cells (2.4 to 7.0 X 10(9) cells) that contained 8.6% to 19.9% Ia-positive cells were treated successively with monoclonal antibody 7.2, which reacts with a framework determinant of Ia-antigen, and biotin-conjugated goat antimouse immunoglobulin. These treated cells were passed over a column of avidin-Biogel (polyacrylamide) and the adherent cells removed by mechanical agitation. Seven lethally irradiated dogs were transplanted with 5.9 to 33.4 X 10(6) recovered adherent cells per kilogram of which 69.0% to 88.0% were Ia-positive. All dogs had hematologic recovery; six are alive and well with durable engraftment and one died on day 15 posttransplant. They are immunologically normal as determined by lymph node and bone marrow biopsies, lymphocyte function, and immunophenotyping of peripheral blood and bone marrow cells. These data provide further evidence that canine hematopoietic stem cells express Ia-like antigens and that these cells are capable of complete hematopoietic and immunologic reconstitution in an autologous model.

Transplantation of CD34+ hematopoietic progenitor cells.

We have developed an avidin-biotin immunoadsorption technique in conjunction with a monoclonal anti-CD34 antibody that is capable of selecting CD34+ progenitor cells from marrow and mobilized peripheral blood. Clinical studies with these CD34+ selected cells have shown that the cells are capable of rapid and durable engraftment. In addition, there is significantly less infusional toxicity to the patient because the volume in which the CD34+ selected cells are contained is much less than that of a typical marrow or apheresis buffy coat. Selection of CD34+ progenitor cells also offers other potential advantages, including T-cell depletion of allografts and tumor cell depletion of autografts. CD34+ selection can also be used to facilitate other manipulations of marrow and peripheral blood, including gene transfection, ex vivo stem cell expansion, tumor purging, and progenitor cell banking. Future graft engineering studies are expected to clarify these relationships and enable refinement of the graft to the point at which GVHD can be minimized, graft survival maximized, and relapse-free survival prolonged.

Isolation and ex vivo expansion of CD34+ cells from cord blood using dextran sedimentation and avidin column selection.

Human umbilical cord bloods were fractionated by unit gravity sedimentation in 1% (v/v) dextran, followed by immunoaffinity selection for CD34+ stem and progenitor cells. Dextran sedimentation alone enabled recovery of more than 80% of the nucleated cells present and 90% of the CD34+ cells, as determined by flow cytometry. The addition of an immunoaffinity selection step for CD34+ cells resulted in a 134-fold enrichment for CD34+ cells, with a mean yield of 64 +/- 15%. The resultant CD34+ population contained almost half the CFU-GM activity initially present in the cord bloods and could be expanded ex vivo in liquid culture.

The hematopoietic stem cell antigen, CD34, is not expressed on the malignant cells in multiple myeloma.

Autologous stem cell transplantation has become an important therapy in multiple myeloma (MM). To develop adequate autograft purging methods, it is necessary to determine whether antigens expressed on early hematopoietic progenitors exist on malignant cells. The Ig heavy chain produced by the MM cells shows evidence of prior somatic mutation without intraclonal diversity. As a result, this sequence can be used as a specific marker to detect all members of the malignant clone. The Ig heavy chain sequence expressed by the MM cells was obtained in five patients with advanced disease. Patient specific oligonucleotide primers were designed based on the complementarity determining regions (CDR) of each MM Ig sequence and used to amplify DNA by polymerase chain reaction for the detection of malignant cells. A highly purified collection of CD34+ cells was obtained after passage of the initial bone marrow cells through an immunoadsorption column and fluorescence-activated cell sorting. Despite an assay sensitivity of 1 tumor cell in 2,500 to 44,000 normal cells, none of the CD34+ samples showed product with the myeloma-specific CDR primers. Therefore, positive selection for cells bearing this antigen should yield a tumor-free autograft capable of providing hematopoietic recovery after myeloablative chemotherapy.

Transplantation of CD34+ hematopoietic progenitor cells.

Sixty-six stage IV breast cancer patients received high dose chemotherapy followed by autologous transplantation of CD34-positive(+) cells obtained from the bone marrow and/or granulocyte colony stimulating factor (G-CSF)-mobilized peripheral blood. Grafts were examined for the presence of tumor using conventional histology and immunocytochemical staining. Patients achieved a granulocyte count of 500 x 10(9)/liter 10-12 days posttransplant, with a platelet count of > 20 x 10(9)/liter in 14-15 days. Enrichment of CD34+ cells from the peripheral blood progenitor cell (PBPC) collections resulted in a 1.3 to 4.0 log depletion of breast cancer cells from the graft.

Preparation and successful engraftment of purified CD34+ bone marrow progenitor cells in patients with non-Hodgkin's lymphoma.

From September 1992 to January 1994, we evaluated the use of the CEPRATE SC stem cell concentrator (CellPro, Inc, Bothell, WA) to select CD34+ cells from the bone marrow (BM) of 25 patients with non-Hodgkin's lymphoma in complete remission. This system uses the biotinylated 12.8 IgM MoAb to select CD34+ cells. Cells are retained on an avidin column and detached by agitation. Fifteen patients have been transplanted with the CD34+ purified fraction. The CD34+ purified fraction of the 25 processed BMs contained a median of 0.54% of the original nucleated cells in a volume of 5 to 10 mL. The median concentration of CD34+ cells was 49% (range, 12% to 80%), and the median enrichment of CD34+ cells was 33-fold (range, 9- to 85-fold). This selected CD34+ fraction retained 60% (range, 15% to 95%) of late granulocyte-macrophage colony-forming units (CFU-GM), 55% (range, 12% to 99%) of early CFU-GM, and 31% (range, 2% to 100%) erythroid burst-forming units (BFU-E) corresponding to median enrichments of 22-fold (range, 1- to 71-fold), 19-fold (range, 2- to 58-fold), and 14-fold (range, 2- to 200-fold), respectively. There was a correlation between immune phenotypes and progenitor cells. In the initial buffy-coat fractions, the percentage of CD34+ cells was correlated to the cloning efficiency of both late CFU-GM (P < .05) and early CFU-GM (P < .001). In the final selected fraction, there was a correlation between the percentage of CD34+/CD33- and the cloning efficiency of early CFU-GM (P < .05) and between the percentage of CD34+/CD33+ and the cloning efficiency of late CFU-GM (P < .05). Lymphoma cells positive for t(14; 18) were found by polymerase chain reaction in 9 of 14 buffy coats tested before CD34+ cell purification. In 8 cases, the CD34(+)-selected fraction was found to be negative, and the CD34- fraction was found to be positive. After cryopreservation, the recoveries of progenitor cells in the CD34(+)-purified fraction were 79% for late CFU-GM, 71% for early CFU-GM, and 73% for BFU-E. The 15 patients transplanted with the concentrated CD34+ fraction received a median dose of 1 x 10(6) CD34+ cells/kg (range, 0.3 to 2.96) and 10.62 x 10(4) early CFU-GM/kg (range, 0.92 to 25.55). Median days to recovery to 0.5 x 10(9)/L neutrophils and 50 x 10(9)/L platelets were days 15 (range, 10 to 33) and 23 (range, 11 to 68), respectively.(ABSTRACT TRUNCATED AT 400 WORDS)

Stem cell selection--clinical experience.

The ability to isolate large numbers of hematopoietic progenitors will facilitate an understanding of the growth and differentiation of bone marrow. Furthermore, isolating hematopoietic progenitors will have widespread clinical applications to autologous marrow transplantation, allogeneic marrow transplantation, gene therapy, and in vitro marrow expansion. With the development of avidin-biotin immunoadsorption, it is now feasible to isolate large numbers of these progenitor cells for clinical purposes. Successful hematopoietic reconstitution has been demonstrated in lethally irradiated baboons transplanted with CD34+ cells isolated by immunoadsorption with the anti-CD34 antibody 12-8. Recent studies have shown that CD34+ cells enriched from the marrow of patients with metastatic breast cancer can be used for autologous marrow transplantation.

Effect of CD34+ selection and various schedules of stem cell reinfusion and granulocyte colony-stimulating factor priming on hematopoietic recovery after high-dose chemotherapy for breast cancer.

We evaluated the effects of various schedules of peripheral blood stem cell (PBSC) reinfusion, granulocyte colony-stimulating factor (G-CSF) priming, and CD34+ enrichment on hematopoietic recovery in 88 patients with advanced breast cancer treated with high-dose chemotherapy, consisting of cisplatin 250 mg/m2, etoposide 60 mg/kg, and cyclophosphamide 100 mg/kg. PBSC (> or = 7.5 x 10(8) nucleated cells/kg) were collected following priming with G-CSF and were either immediately cryopreserved (48 patients; cohorts A and B) or were first processed for CD34+ enrichment (40 patients; cohorts C and D). Patients in cohorts A and C received PBSC on day 0; patients in cohorts B and D received 25% of their nucleated cells on day -2 and 75% on day 0 (split reinfusion). Patients in cohorts A, B, and C were primed with G-CSF 10 micrograms/kg, subcutaneously (SC), once a day; patients in cohort D were primed with 5 micrograms/kg G-CSF, SC, twice daily (bid). Bid administration of G-CSF yielded 2.3 to 4.7 x higher numbers of CD34+ cells in the PBSC product than the same total dose given once a day (P = .002). Reinfusion of 25% of unselected PBSC on day -2 (median, 2.26 x 10(8)/kg nucleated cells [range, 1.7 to 3.3 x 10(8)/kg]) with the remaining cells reinfused on day 0 resulted in earlier granulocyte recovery to > or = 500/microL when compared with reinfusion of all stem cells on day 0 (group B, median of 8 days [range, 7 to 11] v group A, 10 days [range, 8 to 11], P = .0003); no schedule-dependent difference was noted in reaching platelet independence (group B, 11.5 days [range, 5 to 21]; group A, 12 days [range, 8 to 24], P = not significant). Split schedule reinfusion of CD34(+)-selected PBSC did not accelerate granulocyte recovery. In groups D and C, the median number of days to granulocyte recovery was 12 (range, 8 to 22) and 11.5 (range, 9 to 13); patients became platelet independent by day 15 (range, 6 to 22) and 14 (range, 12 to 23), respectively. CD34(+)-selected PBSC rescue decreased the incidence of postreinfusion nausea, emesis, and oxygen desaturation in comparison to unselected PBSC reinfusion (P < or = .005 for each). Hematopoietic recovery may be accelerated by earlier reinfusion of approximately 2.26 x 10(8)/kg unselected nucleated cells. Earlier recovery may be triggered by components other than the progenitors included in the CD34+ cell population. Sustained hematopoietic recovery can also be achieved with CD34(+)-selected PBSC alone. Dosing of G-CSF on a bid schedule generates higher CD34+ cell yield in the leukapheresis product. Whether even earlier "sacrificial" reinfusion of approximately 2 x 10(8)/kg unselected nucleated cells concomitant with the administration of high-dose chemotherapy would reduce the duration of absolute granulocytopenia further while initiating sustained long-term hematopoietic recovery will require further investigation.