Search

Background Intramyocardial injection of autologous bone marrow (ABM) may induce angiogenesis. We tested the safety and feasibility of catheter-based direct percutaneous intramyocardial delivery of ABM in patients with refractory angina pectoris. Methods Ten patients (9 men, 67 ± 8 years) with refractory angina (Canadian Cardiovascular Society class III-IV) and documented myocardial ischemia were enrolled. After left ventricular electromechanical mapping, freshly aspirated and filtered ABM was percutaneously injected into target myocardial ischemic areas. Clinical symptoms (as assessed according to the Canadian Cardiovascular Society class), quality of life, and myocardial perfusion were evaluated before the procedure and through the follow-up. Results In all patients, ABM was successfully injected into the target regions. No periprocedural complications occurred. At 12 months, no major cardiac events (death, acute myocardial infarction, stroke, and malignant ventricular arrhythmias) occurred. Severity of angina improved of ≥2 classes in 3 patients. Quality of life showed a significant improvement in all patients. Myocardial perfusion in the target regions improved in 4 of 8 patients. Conclusions Direct percutaneous intramyocardial delivery of ABM appears feasible and safe. Further evaluation is warranted to test its clinical efficacy.

The aim of the study was to determine the safety and efficacy of the combination of fludarabine (FLU), cyclophosphamide (CY) and mitoxantrone (FLU/CY/MITO) in untreated follicular lymphomas (FL), Sixty patients with newly diagnosed stage II bulky to IV FL, median age 59 years (range 36-70), received FLU/CY/MITO regimen (FLU 25 mg/m2 days 1-3, CY 300 mg/m2 days 1-3, Mito 10 mg/m2 day 1). Patients received antibiotic oral prophylaxis during all treatments, and growth factors (G-CSF) when grade III granulocytopenia (WHO) occurred. The overall response rate was 87%: 46 patients achieved complete response (CR) (77%), 6 a partial response (10%) and 8 were non-responders. Fifty patients are surviving with a median observation time of 31 months. The 4-year estimated probability of overall survival and failure-free survival were 78.2% and 45% respectively. Thirty-five patients (58%) are still in CR. Sixty percent of patients experienced grade III-IV granulocytopenia. Two patients suffered grade III pulmonary infection and one grade III liver toxicity. In a subset of 46 patients, bcl-2 translocation was positive in bone marrow (BM) and/or peripheral blood (PB) of 36 patients. At the end of treatment, 25 of these patients had CR and 19 (76%) converted to polymerase chain reaction (PCR) negativity. FLU/CY/MITO regimen showed a high level of activity in follicular lymphoma. Toxicity, mainly hematological, was acceptable and the treatment was made feasible by the use of antibiotic prophylaxis and G-CSF. Significant non-hematological toxicities were seen, but no patients died. The conversion of bcl-2 from positive to negative by PCR in BM and/or PB suggests a possible role for this treatment in clearing minimal residual disease and improving patients' outcome.

Here we report the results of a randomised multicenter phase III clinical trial which assesses the therapeutic efficacy and tolerability of a chemotherapy protocol CEMP (cyclophosphamide, etoposide, mitoxantrone and prednisone) in adult and elderly patients with advanced intermediate and high-grade NHL. Between October 1991 and October 1995, 139 patients, aged 55 to 79 years, with diffuse intermediate and high-grade lymphoma, were enrolled. A considerable percentage of patients had clinically aggressive disease: 32.4% had systemic symptoms, 79% had stage III or IV disease, 33.8% had bone marrow involvement, 46% had splenic involvement and 42.5% had increased values of serum lactate dehydrogenate. Complete remission was achieved in 70 of the 139 patients (51.9%) and PR in 12 (16.6%) with an overall response of 68.5%. The overall response survival rate at 6 years was 39%, whereas DFS rate was 48.7% and PFS rate was 28.5%. At four years 49% of the patients were still in CR. Dividing the patients in two groups, under and over 65 years of age, we obtained the same results as far as overall response is concerned. No toxic deaths occurred, neither cardiac, renal nor liver complications happened. CEMP regimen is an effective and safe protocol with good results in elderly people, well comparable to those achieved in younger ones.

Adult T-lymphoblastic lymphoma is rare and has a poor prognosis. In the 80s, following the introduction of sequential, intensified chemotherapy, complete remissions in the order of 75%-95% of treated patients, were achieved. However, several patients, namely those with advanced disease, continued to relapse either in remission or during maintenance therapy. Moreover, all these early studies were not able to detect any valuable prognostic index to predict the outcome. In an attempt to reduce the relapse rate, upfront autologous stem cell transplantation in patients in complete remission was introduced. The results obtained with this approach were quite homogeneous, indicating a probability of disease-free survival of about 65%-75% and an overall survival rate of 60%. Successive therapies designed since 2000 were able to obtain complete remissions of above 90%, with a relapse rate in the order of 30% and an overall survival comparable to that obtained with the transplant procedure. Yet, these studies were also unable to detect valuable prognostic factors predictive of the outcome. Moreover, no study on the biologic profile of the disease has been developed. To improve the prognosis of T-lymphoblastic lymphoma it seems necessary to create national registries to collect both clinical and biological data of all lymphoblastic lymphoma patients. In this way it will be possible to reach critical numbers of data with which valid statistical analysis may be performed that is able to detect factors influencing the outcome. Moreover, subsets of patients needing intensified procedures such as stem cell transplant may be detected at diagnosis.

Purpose: To determine, in a randomized clinical trial, whether high-dose therapy (HDT) followed by autologous stem-cell transplantation is more effective than standard treatment with regard to progression-free survival (PFS) and overall survival (OS) in patients with relapsed follicular non-Hodgkin’s lymphoma; and to assess the additional value of B-cell purging of the stem-cell graft with regards to PFS and OS. Patients and Methods: Patients received three cycles of chemotherapy. Responding patients with limited bone marrow infiltration were eligible for random assignment to three further cycles of chemotherapy (C), unpurged HDT (U), or purged HDT (P). Results: Between August 1993 and April 1997, 140 patients were registered from 36 centers internationally, and 89 were randomly assigned. Reasons for not randomizing included patient refusal, early progression, or death on induction therapy. With a 69-month median follow-up, the log-rank P value for PFS and OS were .0037 and .079, respectively. For PFS, the hazard ratios (95% CIs) for U versus C, P versus C, and P versus U were 0.33 (0.16 to 0.70), 0.38 (0.19 to 0.79), and 1.02 (0.51 to 2.05), respectively. The hazard ratio (95% CI) for C versus U + P was 0.30 (0.15 to 0.61). Hazard ratios (95% CIs) for OS were 0.43 (0.18 to 1.06), 0.43 (0.18 to 1.02), and 0.72 (0.32 to 1.63). For C versus U + P, the hazard ratio (95% CI) was 0.40 (0.18 to 0.89). Kaplan-Meier estimates (95% CIs) of 2-year PFS for C, U, and P were 26% (8% to 44%), 58% (37% to 79%), and 55% (34% to 75%), respectively. OS at 4 years for C, U, and P are 46% (25% to 67%), 71% (52% to 91%), and 77% (60% to 95%) respectively. Conclusion: HDT significantly improves PFS and OS. There is no clear evidence of benefit through purging.

Alpha interferon has shown initial promise in the treatment of low-grade non-Hodgkin's lymphoma (NHL), especially with the nodular form of the disease. The present study enrolled 70 NHL patients who received either chlorambucil (CB; 10 mg/day) or CB plus interferon alfa-2b (5 million units (MU)/m2 subcutaneously three times a week). Among 63 evaluable patients, similar response rates (62.1% and 64.7% respectively) were recorded for the treatment arms. In patients receiving no maintenance therapy, those who received interferon alfa-2b during the induction phase showed a favourable trend in terms of incidence of relapse compared to those who had received chlorambucil alone. During maintenance therapy with interferon alfa-2b, no significant differences in the occurrence of relapse have yet been seen compared to patients on no maintenance therapy. A longer observation period is needed to make a definitive conclusion about the usefulness of interferon maintenance therapy and to evaluate further the effects of the combined schedule of chlorambucil and interferon induction on the duration of remission.

Translocation t(9;22) or Philadelphia chromosome (Ph)/BCR-ABL rearrangement positive acute lymphoblastic leukemia (Ph/BCR+ ALL) is associated with a very short survival of about one year in most patients.We analyzed long-term outcome of 76 adults with Ph/BCR+ ALL, in order to detect which factors were associated with longer survival. Modifiable prognostic factors included type of treatment, allogeneic marrow transplant (allo-BMT), and early anthracycline dose intensity (high=H/A, low=L/A); unmodifiable factors were age, gender, FAB morphology, phenotype, blast count, P190/210 transcript, hepato-spleno-lymphadenopathy, LDH level. Median patient age was 43 years (range 15–71). Four favorable prognostic factors (FPF) were found associated with greater likelihood of complete remission (blast count

Thirty-six successive adult patients with lymphoblastic lymphoma entered a study of sequential chemotherapy consisting of an intensive LSA2-L2-type protocol to induce first complete remission. Eighteen patients in first CR (median age 22 years, range 15–51), underwent autologous bone marrow transplantation after receiving a conditioning regimen consisting of cyclophosphamide and total body irradiation. Of these 18 patients, 2 were in stage III and 16 in stage IV; 15 showed mediastinal and 9 bone marrow involvement at diagnosis. The transplant procedure was well tolerated and no treatment-induced deaths occurred. At this time, 14 out of 18 patients are alive and well between 1 and 60 months post transplant (median follow-up time 46 months) with an actuarial disease-free survival of 74%. This phase II study suggests that high-dose chemo-radiotherapy followed by autologous bone marrow transplantation may improve long-term disease-free survival in advanced stage adult lymphoblastic lymphoma.

The CFU-Gm content of liver, spleen and bone marrow of human fetuses was determined from the 8th to 28th week of gestational age. The study of progenitor cells from 43 fetal livers revealed that the CFU-Gm increased from 8 to 12 weeks, was relatively stable through 16 weeks and thereafter declined by 22 to 24 weeks. The total number of liver CFU-Gm rose from 10(4) to greater than 10(6) CFU-Gm by the 16th week of gestation and declined to approximately 4 X 10(5) progenitor cells by 22 to 24 weeks. Peak values of bone marrow CFU-Gm were noted at 10 to 14 weeks of gestation. Generally, the numbers of these progenitor cells declined by 16 to 18 weeks; however, several high values were obtained at later times. Splenic CFU-Gm were relatively constant in number between 12 and 28 weeks of gestation with no correlation with fetal age. Thus, these studies indicate that certain fetal tissues contain substantial numbers of hemopoietic progenitor cells. Based on the quantity of CFU-Gm obtained, it appears feasible to use fetal liver cells as a source of progenitor cells for transplantation.

Binding of radiolabeled L-cell colony-stimulating factor (CSF) was studied using murine bone marrow and fetal liver cells. With 10(7) cells, saturation of binding was seen with approximately 500,000 cpm of 125I-CSF. Minimal binding was detected after one hour incubation with tracer at 37 degrees C; however, marked cellular uptake of radioactivity was noted after 24-hr exposure to CSF. As judged by autoradiographs, small numbers of myeloblasts, promyelocytes, and large mononuclear cells were labeled with 1-hr exposure to tracer. By 6 hr of incubation, 50%-70% of myeloblasts and promyelocytes and small numbers of late granulocytic cells were labeled. Virtually all myeloblasts and promyelocytes and approximately 50% of myelocytes, metamyelocytes, polymorphonuclear granulocytes, and monocytes were labeled after 24-hr exposure to the radioiodinated CSF. Label was not detected on erythroblasts, eosinophils, or megakaryocytes. Suspensions of fetal liver cells had lower uptake of radioactivity than bone marrow cells. This appeared to result from a lesser concentration of granulocytic cells in fetal liver, as labeling of individual cells was similar with both tissues. In additional experiments, CSF binding to marrow cells was assessed after 30-min exposure to tracer at 0 degrees C. Uptake of 125I-CSF exceeded that observed after 24-hr incubation at 37 degrees C. With this technique, cellular label was also confined to granulocytic and monocytic cells. These findings suggest that purified CSF reacts with and may stimulate immature and mature cells of the granulocytic and monocytic lineages.

There are many case reports of secondary neoplasms occurring after treatment with alkylating agents. A case of malignant melanoma of the glans penis in a chronic myeloid leukemia (CML) Ph'-positive patient after 13 years on busulfan treatment is described. Since neither impairment of immune status nor increased incidence of secondary neoplasm have hitherto been reported in CML, the suggestion that busulfan has a carcinogenetic effect is discussed.

To stimulate a leukemia remission marrow, cell suspensions of normal human bone marrow were mixed with human acute lymphoblastic or myelogenous leukemic cells of the CCRF-SF, Nalm-6, and K-562 lines. The cell mixtures were incubated in vitro with mafosfamide (AZ) or with the photoreactive dye merocyanine 540 (MC-540). A quantity of 10(4) cells of the treated suspensions was dispensed into microculture plates, and graded cell numbers of the line used to contaminate the normal marrow were added. Limiting-dilution analysis was used to estimate the frequency of leukemia cells persisting after treatment with the decontaminating agents. Treatment with AZ or MC-540 produced a total elimination (ie, 6 logs or 5.3 logs respectively) of B cell acute leukemia cells (CCRF-SB), whereas nearly 1.7 logs and 2 logs of K-562 acute myelogenous blasts were still present in the cell mixtures after treatment with MC-540 and AZ, respectively. Treatment of the Nalm-6- contaminated cell mixtures with AZ resulted in 100% elimination of clonogenic cells, whereas nearly 80% decontamination was obtained with MC-540. Our results suggest that treatment with AZ could be an effective method of eliminating clonogenic tumor cells from human bone marrow. MC-540, shown by previous studies to spare sufficient pluripotential stem cells to ensure hemopoietic reconstitution in the murine model and in clinical application, has comparable effects and merits trials for possible clinical use in autologous bone marrow transplantation.

Pluripotential cells derived from fetal liver had a lower plating efficiency than adult marrow cells, but estimates of the generation time derived from the growth curve are significantly shorter and may account for the earlier erythroid population.

After exposure to hypoxia there was only a moderate stimulation of erythropoiesis in a strain of mouse which produces minimal amounts of erythropoietin in response to such a stimulus. This was associated with minimal changes in bone marrow CFU’s, but there was a striking decrease in splenic CFU’s. This splenic depopulation seemed related to a specific effect of hypoxia rather than stem cell depletion secondary to erythroid differentiation. On return to ambient pressure there was a prompt onset of splenic red cell aplasia, but only a gradual decrease in marrow erythropoiesis. Despite the prompt splenic red cell aplasia, there was only a slow return to splenic CFU’s to normal levels during the first 4-5 days. Thereafter a slight rebound or overshoot was noted. Although differentiated erythroid elements were still present in the marrow on the fourth post-hypoxic day, a moderate rebound in marrow CFU’s to 150 per cent of control was evident. These findings might best be explained by a migration of stem cells during the post-hypoxic interval. Marrow lymphocytes decreased during hypoxia and rose gradually there-after following, rather than preceeding, the increase in bone marrow CFU’s.

An antiserum was produced in rabbits to acute undifferentiated leukemia (AUL) antigens. Partial absorption of the antiserum allowed the removal of the cytotoxic activity against normal mononuclear cells of peripheral blood and bone marrow, in contrast to blast cells from acute and chronic leukemias, which were killed by the antiserum in a standard NIH cytotoxicity test. These observations suggest that antigens present on blast cells of AUL are also expressed on some acute and chronic leukemic cells and that these antigens could be interpreted as B or leukemia-associated antigens because of the disappearance of all cytotoxic activity of the immune serum after extensive absorption with B cells from chronic lymphoid leukemia.

SummaryLimulus lysate assays showed that several samples of purified L-cell CSF were markedly contaminated with endotoxin. Several pools of material contained 1 to 10 μg/ml of this bacterial lipopoly-saccharide. Ultracentrifugation in sucrose density gradients led to a 95% separation of CSF and endotoxin activities. In four studies, endotoxin content of the CSF preparations was reduced to levels varying from 0.05 to 0.0001 μg/ml.In vivo studies with known quantities of E. coli endotoxin were conducted in mice. Acute and repeated injections of 0.1 μg of endotoxin led to increased serum CSF activity and increased number of splenic CFU-C and CFU-E. In contrast, similar injections of 0.01 μg of endotoxin produced virtually no effect on these parameters of hemopoiesis. Thus, the removal of endotoxin from CSF as outlined herein will permit in vivo studies of the effects of CSF on hemopoiesis, without concern for the complicating effects of endotoxin.The authors gratefully appreciate the excellent technical assi...

Purified L-cell colony-stimulating factor (CSF) was iodinated by several techniques and used to determine the physiologic distribution in vivo. Mice were injected i.v. or i.p. with 100,000 cpm; plasma and tissue radioactivity were monitored at intervals up to 6 hr. A complex plasma disappearance curve was observed after i.v. injection. There was a 80%-90% reduction in circulating radioactivity at 4 hr. This resulted from a biphasic pattern, in which the t½ was 24-40 min and 2-2.5 hr in the first and second phases, respectively. Plasma levels of radioactivity were lower following i.p. administration, with peak activity of 10%; this was observed 1 hr after injection. Thereafter, plasma radioactivity declined with a half-life (½) of 3.5 hr. In parallel experiments, clearance of unla-beled partially purified CSF was monitored by radioimmunoassay (RIA). A nearly linear serum disappearance was observed (t½ of 1.5 hr), which approximated the major curve noted with the 125I-CSF. In 4 studies, 6-hr urinary excretion of radioactivity ranged from 14% to 32%. Bilateral nephrectomy did not alter the initial decline in plasma activity but markedly prolonged the second phase (t½ of 5 hr). Despite substantial amounts of radioactivity in the urine, virtually no CSF could be detected by RIA. This suggests that the kidneys may serve a key role in the metabolism and excretion of CSF. These studies show that CSF is rapidly cleared and metabolized. Based on a t½ of 1.5-2.5 hr, in vivo studies will require large repetitive doses of this presumptive granulopoietin.

Human acute myelogenous or lymphoblastic leukemia cells of the K-562 and CCRF-SB lines were mixed with an excess of normal human bone marrow cells to simulate a leukemia remission marrow. The cell mixtures were then incubated in vitro with mafosfamide (AZ) followed by the photoreactive dye merocyanine-540 (MC). Treated cells (1 x 10(4] were seeded in microwell plates, and increasing numbers of the line used to contaminate the normal marrow were added. Treatment with AZ alone produced total elimination (i.e., 6 logs) of CCRF-SB cells, while addition of merocyanine-540 increased the cloning efficiency from 22% to 24.4%. After treatment of the K-562-contaminated cell mixtures with AZ, nearly 1.6 logs of K-562 acute myelogenous blasts were still present, whereas AZ purging followed by MC-mediated photosensitization resulted in 100% elimination of clonogenic cells. Moreover, the combined treatment caused an increase of the cloning efficiency from 37.3% to 62%, clearly indicating that cleansing by the two agents combined was more effective than treatment with one agent alone.

Bone marrow cells collected from patients with hematologic malignancies were cryopreserved using DMSO as a cryoprotective agent. The growth kinetics of hemopoietic stem cells frozen to −196 °C was monitored immediately after thawing by the semisolid agar CFU-C assay and two different methods of cell reconstitution were compared. In the first procedure, thawed cells were plated after the removal of DMSO by washing the cell suspension; in the second, cell suspensions were cultured after a simple 1:1 dilution of DMSO with medium. The numbers of CFU-C per 2 × 105 cells plated was higher by washing out the DMSO in all the groups studied. However, the absolute numbers of CFU-C contained in the whole ampoules after the freezing procedures was approximately the same using both methods. It is concluded that washing the cells only apparently yielded a better cloning efficiency, suggesting that such a procedure led to a higher mature nucleated cell loss with the consequence of a CFU-C concentration. This trend seems particularly evident in cells from the AML and CML patients.

Normal and pathological lymphoid cells, collected from different sources, were cryopreserved using a programmed freezing procedure. With this cryopreservation technique, the percentage of T and B cell surface markers and the proliferative response to mitogens were not influenced by 21 and 35 days of storage in liquid nitrogen. The recovery percentage of the lymphoid cells was satisfactory when fetal calf serum was added, as a protein source, to the medium during freezing and thawing phases, while a very low percentage of cells was recovered if the fetal calf serum was omitted.

The major stumbling-block in ABMT is the high relapse rate compared to that observed in both allogeneic and syngeneic marrow transplant (1). This is possibly due to occult leukemic cells contaminating the remission marrow. Although the efficacy of the ex-vivo purging procedure has yet to be definitely established in clinical practice, there are indications that cleansing the marrow is associated with improved survival in some cases (1, 2). To meet the obvious need for methods capable of eliminating these cells from autologous marrow, several techniques have been investigated (3–7).

SummaryThe recovery of erythropoiesis in heavily irradiated mice receiving fetal liver was compared with that of mice receiving adult bone marrow. The number of cells was adjusted so that each group received a comparable number of cfu. Red cell production as judged by both iron incorporation and peripheral reticulocyte values began earlier in the fetal liver transplant than after adult bone marrow. Further, the extent of erythroid colonization of the bone marrow appeared greater in recipients of fetal tissue than adult. These differences in the recovery pattern suggest that stem cell regulation of fetal erythropoiesis differs from that of the adult.

SummaryNine units of erythropoietin given in equally divided doses at 8-hour intervals for 23 days to the CAF1 mouse produced a significantly greater polycythemia than had been observed in this strain when exposed to 23,000 feet for a similar period of time. The splenic CFU began to increase within 24 hours after the beginning of treatment and reached levels of 5-fold those of controls between days 8 and 16. The extent to which migration of CFU from the marrow and or replication of CFU in situ is responsible for the increase in splenic CFU is discussed. It is concluded that the stem cell participates in the physiologic response to erythropoietin but without depletion of this compartment. The possibility that the previously reported decrease in CFU in hypoxic animals represents a secondary perhaps “toxic” effect of hypoxia unrelated to erythropoietin is considered.

The aim of this study was to verify, and possibly improve, culture conditions to expand human mobilized peripheral blood stem cells (PBSCs). We investigated the role of three parameters: A) the culture medium (serum-free versus serum-dependent); B) the initial cell population (Ficoll-separated mononucleated cells versus CD34(+)-selected cells), and C) the low concentration of recombinant cytokines, flt3 ligand, and thrombopoietin in association with a basic cocktail of stem cell factor, interleukin (IL)-6, IL-3, GM-CSF, and erythropoietin. Eighteen leukapheresis samples were monitored in static culture for 15 days. The expansion potential was assessed at day 10 and 15 by total nuclear cells, colony-forming-units (CFUs) (burst-forming units-erythroid [BFU-E], colony-forming units-granulocyte-macrophage [CFU-GM], and colony-forming units-granulocyte-erythroid-macrophage-megakaryocyte [CFU-GEMM]), and flow cytometry immunophenotyping (CD34(+)/CD38(-), CD38(+), CD33(+), CD41(+), GlyA(+) progenitor cells). The results, evaluated by multivariate analysis of variance, emphasize that some variables affected the outcome of stem and progenitor cell expansion. CD34(+) enrichment increased expansion of total nuclear cells, number of CD38(+) and CD33(+) late precursors, and number of the CFU-GM compartment. Interestingly, however, quantitative expansion of GlyA(+) and the early progenitor cells (CD34(+)/CD38(-), CFU-GEMM, BFU-E) are favored by the use of unselected mononucleated cells. Regarding the role of serum, no significant difference was observed except for expansion of total nuclear cells, CFU-GM, and BFU-E. Cytokine combinations, in particular the use of flt3 ligand, stimulated expansion of almost all the cellular subsets, reaching a statistical significance for total nuclear cells and CFU-GM. Our study indicates that progenitor and late precursor multilineage cell compartments of mobilized PBSCs may be significantly expanded in short-term cultures by well-defined experimental conditions. Furthermore, these data might be useful when evaluating ex vivo expansion of hematopoietic cells for clinical purposes.

2 patients with acute leukaemia in relapse were transplanted with fetal liver cells following a conditioning regimen of cyclophosphamide (120 mg/kg) and total body irradiation (1000 r). Each patient achieved a remission with haematopoietic recovery that was rapid in one case and delayed in the other. In one case there was evidence of chimerism as demonstrated by the presence of the XYY karyotype of the donor fetus in 20% of marrow metaphases, by the presence of double Y bodies in the peripheral blood, by the appearance of new HLA-antigens, and by red cell isoenzyme phenotypes of donor origin. In the second case there was prompt haemotopoietic recovery and the appearance of red cell isoenzyme phenotypes of donor origin. Survival was 153 and 30 d, respectively, and both patients died of interstitial pneumonia without evidence of graft versus host disease.

The erythroid and stem cell (colony-forming unit) responses were observed in the mouse after transfusion-induced polycythemia. Although there was a 40% decrease in erythroid differentiation 24 hours after transfusion, an increase in splenic CFU was not evident until 48 hours. This reached a level of approximately 200% by the seventh day, a time when maximum suppression of erythropoiesis was observed. With the administration of erythropoietin, a marked increase in splenic erythropoiesis was noted within 24 hours but the number of splenic CFU was not significantly changed from control values. Thereafter splenic CFU increased, following the erythroid response by approximately 24 hours. Bone marrow erythropoiesis gradually returned to the normal levels; however, quantitative estimates indicated at 15-20% decrease in marrow CFU. These responses to erythropoietin are similar to what has been observed in nonpolycythemic animals and suggest stem cell migration from the marrow to the spleen. The delays noted in st...