Zinc finger protein 148 is dispensable for primitive and definitive hematopoiesis in mice.

Hematopoiesis is regulated by transcription factors that induce cell fate and differentiation in hematopoietic stem cells into fully differentiated hematopoietic cell types. The transcription factor zinc finger protein 148 (Zfp148) interacts with the hematopoietic transcription factor Gata1 and has been implicated to play an important role in primitive and definitive hematopoiesis in zebra fish and mouse chimeras. We have recently created a gene-trap knockout mouse model deficient for Zfp148, opening up for analyses of hematopoiesis in a conventional loss-of-function model in vivo. Here, we show that Zfp148-deficient neonatal and adult mice have normal or slightly increased levels of hemoglobin, hematocrit, platelets and white blood cells, compared to wild type controls. Hematopoietic lineages in bone marrow, thymus and spleen from Zfp148 (gt/gt) mice were further investigated by flow cytometry. There were no differences in T-cells (CD4 and CD8 single positive cells, CD4 and CD8 double negative/positive cells) in either organ. However, the fraction of CD69- and B220-positive cells among lymphocytes in spleen was slightly lower at postnatal day 14 in Zfp148 (gt/gt) mice compared to wild type mice. Our results demonstrate that Zfp148-deficient mice generate normal mature hematopoietic populations thus challenging earlier studies indicating that Zfp148 plays a critical role during hematopoietic development.

The autocrine motility factor receptor is overexpressed on the surface of B cells in Binet C chronic lymphocytic leukemia.

Chronic lymphocytic leukemia (CLL) is a heterogeneous disease with a clinical spectrum reaching from discrete lymphocytosis to extensive enlargement of lymph nodes, spleen and liver, and bone marrow failure. The aim of this study was to identify genes that differentiate between patients with disease stage A vs. C according to Binet in order to better understand the disease. To achieve this, we performed DNA microarray analysis on B cells from CLL patients with stage A and C according to Binet and matched controls. Between CLL patients and controls, there were 1,528 differentially expressed genes and 360 genes were differentially expressed between Binet A and C patients. Due to the sheer number of regulated genes, we focused on the autocrine motility factor receptor (AMFR). AMFR has not previously been investigated in hematological disorders, but high expression of AMFR correlates with a more advanced stage and invasive potential in several human tumors. AMFR mRNA expression was higher in Binet A compared with Binet C patients (P=0.0053) and healthy controls (P=0.0051). Total AMFR protein was higher in Binet A patients compared to Binet C as analyzed by intracellular flow cytometry. However, AMFR exist both in the ER involved in protein degradation and on the cell surface involved in metastasis and cell motility. Cell surface AMFR was increased in Binet C compared with Binet A+B (P=0.016). In conclusion, the mRNA levels reflect the total amount of AMFR, whereas cell surface expression is associated with progression in CLL.

Nf1 deficiency cooperates with oncogenic K-RAS to induce acute myeloid leukemia in mice.

Hyperactive RAS signaling is caused by mutations in RAS genes or a deficiency of the neurofibromatosis gene (NF1) and is common in myeloid malignancies. In mice, expression of oncogenic K-RAS or inactivation of Nf1 in hematopoietic cells results in myeloproliferative disorders (MPDs) that do not progress to acute myeloid leukemia (AML). Because NF1 is a RAS-GTPase-activating protein it has been proposed that NF1 deficiency is functionally equivalent to an oncogenic RAS. It is not clear, however, whether Nf1 deficiency would be redundant in K-RAS-induced MPD development or whether the 2 mutations would cooperate in leukemogenesis. Here, we show that the simultaneous inactivation of Nf1 and expression of K-RAS(G12D) in mouse hematopoietic cells results in AML that was fatal in primary mice within 4 weeks and transplantable to sublethally irradiated secondary recipients. The data point to a strong cooperation between Nf1 deficiency and oncogenic K-RAS.

Therapy-related patterns of cytogenetic abnormalities in acute myeloid leukemia and myelodysplastic syndrome post polycythemia vera: single center experience and review of literature.

A minor fraction of patients with polycythemia vera (PV) develop a terminal acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS). Analysis of the cytogenetic abnormalities during AML or MDS may help in understanding if this development is part of the natural course of the disease or induced by myelosuppressive therapy. Thirty-six cases with AML or MDS post PV, collected in a single Swedish institution during a 33-year period, are described with special regard to time to development of AML or MDS, therapy given during active PV, and cytogenetic findings during AML or MDS. A further 118 cases of AML or MDS post PV, in whom type of therapy during active PV and cytogenetic findings during AML or MDS were reported, were collected from the literature. AML or MDS developed in our own series after 1-30 years with a fairly constant rate (two cases per year). The most frequent cytogenetic abnormalities were +1q, -5, 5q-, -7, 7q-, +8, +9, 11q-, 13q-, and 20q-. When patients in the total material (n = 154) were divided with regard to treatment during active PV, marked differences were observed. The highest frequency of abnormalities was found in patients given multiple lines of therapy (n = 61), dominating features being -5/5q- in 28 patients (46%), -7/7q- in 19 patients (31%), numerous translocations in 24 patients (39%), and unidentified markers in 22 patients (36%). Half of the patients treated with hydroxyurea alone showed a -5 or 5q- abnormality. In patients treated with phlebotomy alone, +8 and +9 were the most frequent findings. The type of therapy given during active PV influences the type of chromosome abnormalities present during terminal AML or MDS and can also be instrumental in the development of leukemia.

Outcome of ETV6/RUNX1-positive childhood acute lymphoblastic leukaemia in the NOPHO-ALL-1992 protocol: frequent late relapses but good overall survival.

The prognostic impact of t(12;21)(p13;q22) [ETV6/RUNX1 fusion] in paediatric acute lymphoblastic leukaemia (ALL) has been extensively debated, particularly with regard to the frequency of late relapses and appropriate treatment regimens. We have retrospectively collected 679 ALLs with known ETV6/RUNX1 status, as ascertained by fluorescence in situ hybridization or reverse-transcription polymerase chain reaction, treated according to the Nordic Society of Paediatric Haematology and Oncology -ALL-1992 protocol. The assigned risk groups/treatment modalities for the 171 (25%) patients with t(12;21)-positive ALLs were 74 (43%) standard risk, 71 (42%) intermediate risk and 26 (15%) high risk. The 5- and 10-year event-free survival (EFS) of the 171 patients was 80% and 75% respectively, with no significant differences among the three risk groups. Most of the relapses occurred in boys and were late, with almost 50% of all relapses occurring > or = 5 years after diagnosis. Of all relapses after 6 years, 80% occurred in the t(12;21)-positive group. The overall survival was 94% at 5 years and 88% at 10 years; thus, the treatment of patients in second or later remission is usually successful. As yet, there is no reliable plateau in the EFS curve, a fact that raises the question as to when the prognostic ramifications of ALLs harbouring ETV6/RUNX1 should be evaluated.

Clinical and cytogenetic features of pediatric dic(9;20)(p13.2;q11.2)-positive B-cell precursor acute lymphoblastic leukemias: a Nordic series of 24 cases and review of the literature.

Although dic(9;20)(p13.2;q11.2) is a characteristic abnormality in childhood B-cell precursor acute lymphoblastic leukemias (BCP ALL), little is known about its clinical impact or the type and frequency of additional aberrations it may occur together with. We here review the clinical and cytogenetic features of a Nordic pediatric series of 24 patients with dic(9;20)-positive BCP ALL diagnosed 1996-2006, constituting 1.3% of the BCP ALL, as well as 47 childhood cases from the literature. Consistent immunophenotypic features of the Nordic cases included positivity for HLA-DR, CD10, CD19, CD20, and CD22 and negativity for T-cell and myeloid markers; no detailed immunophenotypes were reported for the previously published cases. In the entire cohort of 71 cases, the modal chromosome distribution was 45 (62%), 46 (21%), 47 (7%), 48 (4%), 49 (3%), 44 (1%), and 50 (1%). Additional changes were present in 63%, the most frequent of which were homozygous loss of CDKN2A (33%) and gains of chromosomes 21 (28%) and X (10%). The median patient age was 3 years, the female/male ratio was 2.0, the median white blood cell count was 24 x 10(9)/l, 11% had central nervous system involvement, and 5% had a mediastinal mass at diagnosis. Risk group stratification was nonstandard risk in 79%. The event-free survival and overall survival at 5 years for the 24 Nordic cases was 0.62 and 0.82, respectively. Thus, although relapses are quite common, postrelapse treatment of many patients is successful.

GGTase-I deficiency reduces tumor formation and improves survival in mice with K-RAS-induced lung cancer.

Protein geranylgeranyltransferase type I (GGTase-I) is responsible for the posttranslational lipidation of CAAX proteins such as RHOA, RAC1, and cell division cycle 42 (CDC42). Inhibition of GGTase-I has been suggested as a strategy to treat cancer and a host of other diseases. Although several GGTase-I inhibitors (GGTIs) have been synthesized, they have very different properties, and the effects of GGTIs and GGTase-I deficiency are unclear. One concern is that inhibiting GGTase-I might lead to severe toxicity. In this study, we determined the effects of GGTase-I deficiency on cell viability and K-RAS-induced cancer development in mice. Inactivating the gene for the critical beta subunit of GGTase-I eliminated GGTase-I activity, disrupted the actin cytoskeleton, reduced cell migration, and blocked the proliferation of fibroblasts expressing oncogenic K-RAS. Moreover, the absence of GGTase-I activity reduced lung tumor formation, eliminated myeloproliferative phenotypes, and increased survival of mice in which expression of oncogenic K-RAS was switched on in lung cells and myeloid cells. Interestingly, several cell types remained viable in the absence of GGTase-I, and myelopoiesis appeared to function normally. These findings suggest that inhibiting GGTase-I may be a useful strategy to treat K-RAS-induced malignancies.

Cytogenetic patterns in ETV6/RUNX1-positive pediatric B-cell precursor acute lymphoblastic leukemia: A Nordic series of 245 cases and review of the literature.

Between 1992 and 2004, 1,140 children (1 to<15 years) were diagnosed with B-cell precursor acute lymphoblastic leukemia (ALL) in the Nordic countries. Of these, 288 (25%) were positive for t(12;21)(p13;q22) [ETV6/RUNX1]. G-banding analyses were successful in 245 (85%); 43 (15%) were karyotypic failures. The modal chromosome numbers, incidence, types, and numbers of additional abnormalities, genomic imbalances, and chromosomal breakpoints in the 245 karyotypically informative cases, as well as in 152 previously reported cytogenetically characterized t(12;21)-positive ALLs in the same age group, were ascertained. The most common modal numbers among the 397 cases were 46 (67%), 47 (16%), 48 (6%), and 45 (5%). High-hyperdiploidy, triploidy, and tetraploidy were each found in approximately 1%; none had less than 40 chromosomes. Secondary chromosomal abnormalities were identified by chromosome banding in 248 (62%) of the 397 ALLs. Of these, 172 (69%) displayed only unbalanced changes, 14 (6%) only balanced aberrations, and 26 (10%) harbored both unbalanced and balanced abnormalities; 36 (15%) were uninformative because of incomplete karyotypes. The numbers of secondary changes varied between 1 and 19, with a median of 2 additional aberrations per cytogenetically abnormal case. The most frequent genomic imbalances were deletions of 6q21-27 (18%), 8p11-23 (6%), 9p13-24 (7%), 11q23-25 (6%), 12p11-13 (27%), 13q14-34 (7%), loss of the X chromosome (8%), and gains of 10 (9%), 16 (6%), and 21 (29%); no frequent partial gains were noted. The chromosome bands most often involved in structural rearrangements were 3p21 (2%), 5q13 (2%), 6q12 (2%), 6q14 (2%), 6q16 (2%), 6q21 (10%), 6q23 (6%), 6q25 (3%), 9p13 (2%), 11q13 (2%), 11q23 (2%), 12p11 (6%), 12p12 (7%), 12p13 (25%), 21q10 (6%), and 21q22 (6%). Considering that the t(12;21) is known to arise in utero and that the postnatal latency period is protracted, additional mutations are most likely necessary for overt ALL. The frequently rearranged chromosome regions may harbor genes of importance for the transformation and/or progression of an initial preleukemic t(12;21)-positive clone.

Rce1 deficiency accelerates the development of K-RAS-induced myeloproliferative disease.

The RAS proteins undergo farnesylation of a carboxyl-terminal cysteine (the "C" of the carboxyl-terminal CaaX motif). After farnesylation, the 3 amino acids downstream from the farnesyl cysteine (the -aaX of the CaaX motif) are released by RAS-converting enzyme 1 (RCE1). We previously showed that inactivation of Rce1 in mouse fibroblasts mislocalizes RAS proteins away from the plasma membrane and inhibits RAS transformation. Therefore, we hypothesized that the inactivation of Rce1 might inhibit RAS transformation in vivo. To test this hypothesis, we used Cre/loxP recombination techniques to simultaneously inactivate Rce1 and activate a latent oncogenic K-RAS allele in hematopoietic cells in mice. Normally, activation of the oncogenic K-RAS allele in hematopoietic cells leads to rapidly progressing and lethal myeloproliferative disease. Contrary to our hypothesis, the inactivation of Rce1 actually increased peripheral leukocytosis, increased the release of immature hematopoietic cells into the circulation and the infiltration of cells into liver and spleen, and caused mice to die more rapidly. Moreover, in the absence of Rce1, splenocytes and bone marrow cells expressing oncogenic K-RAS yielded more and larger colonies when grown in methylcellulose. We conclude that the inactivation of Rce1 worsens the myeloproliferative disease caused by oncogenic K-RAS.

Serial monitoring of BCR-ABL transcripts in chronic myelogenous leukemia (CML) treated with imatinib mesylate.

Survival among chronic myelogenous leukemia (CML) patients can be linked to the reduction in leukemic cell burden. Treatment with imatinib mesylate results in a high frequency of complete cytogenetic response, which can be further stratified using quantitative reverse-transcriptase polymerase chain reaction (qRT-PCR). We have serially monitored peripheral blood and bone marrow BCR-ABL transcripts using qRT-PCR in CML patients commencing imatinib therapy, and compared the results with bone marrow cytogenetics. Seventeen patients (aged 25-74 yr) with Philadelphia chromosome positive CML in first chronic phase were treated with imatinib targeting a dose of 400 mg/d. The median follow up is 30 mo (range 9-33 mo). Every third month the product of the BCR-ABL fusion gene was evaluated in both blood and bone marrow specimens by real-time RT-PCR using the TaqMan probe system. In 113 simultaneously obtained blood and bone marrow samples, the BCR-ABL transcript values agreed well with cytogenetic data. Blood and bone marrow specimens gave comparable values for BCR-ABL transcripts. Before start of imatinib therapy there was a considerable variation in BCR-ABL transcripts among the patients, ranging approximately one log (base 10). Similarly, patients with a complete cytogenetic response following imatinib therapy had variable BCR-ABL transcript levels, ranging at least three logs (base 10). The major decline in BCR-ABL transcripts occurred within 6 mo after start of imatinib therapy. The decline in BCR-ABL transcripts, following imatinib therapy, appears to level off at 12-15 mo. Two late responders were identified with a still decreasing level in BCR-ABL transcripts after 24 mo of treatment. It is concluded that BCR-ABL mRNA quantification in peripheral blood is suitable for routine monitoring of the response to treatment and long-term disease status in CML, especially in patients who have achieved a complete cytogenetic response. A plateau in BCR-ABL transcripts seems to have been reached after 12-15 mo of imatinib treatment; however, some "late responders" are seen.

Several chromosomes involved in translocations with chromosome 5 shown with fluorescence in situ hybridization in patients with malignant myeloid disorders.

In many patients with myelodysplastic syndromes or acute myeloid leukemia, complex chromosome aberrations can be seen, among which aberrations of chromosome 5 constitute a substantial part. With conventional cytogenetic technique, these aberrations are often identified as deletions or monosomy 5. We analyzed nine patients who, under conventional cytogenetic analysis, showed deletion or monosomy 5. We used fluorescence in situ hybridization with whole-chromosome painting probes to identify the counterpart chromosome and locus-specific identifiers for 5q31 and 5q33 approximately q34. A deletion of 5q was found concomitant with unbalanced translocations. Our results and cases from the literature showed that material from chromosome 5 could be translocated to almost all chromosomes. All patients but one had short survival; this one patient had a preserved 5q31 and 5q33 approximately q34 but a deletion of the q-arm more centromeric than these bands. In eight of the nine patients, further 14 translocations were revealed, not involving chromosome 5.

Philadelphia (Ph) chromosome-positive thrombocythemia without features of chronic myeloid leukemia in peripheral blood: natural history and diagnostic differentiation from Ph-negative essential thrombocythemia.

We have evaluated the clinical symptoms, hematological features, and natural history of 3 cases and 20 reported cases described as Philadelphia chromosome-positive (Ph+) essential thrombocythemia (ET). The presence of increased small mononuclear megakaryocytes in bone marrow smears and biopsy material in patients with pronounced thrombocytosis and no evidence of chronic myeloid leukemia (CML) in peripheral blood appeared to be a diagnostic clue to the diagnosis of Ph+ (essential) thrombocythemia. As compared to cases of reactive thrombocytosis, the megakaryocytes in Ph+ thrombocythemia are smaller than normal ones and typically have hypolobulated round nuclei. This contrasts with the finding of clustered mature and enlarged megakaryocytes in Ph-negative true ET. Patients diagnosed as Ph+ ET may progress to CML and show a high tendency to myelofibrosis and blastic transformation. These observations indicate that both Ph+ ET and Ph+ thrombocythemia associated with CML can be regarded as early manifestations of the chronic stable phase of CML.

Cytogenetic abnormalities in childhood acute myeloid leukaemia: a Nordic series comprising all children enrolled in the NOPHO-93-AML trial between 1993 and 2001.

Between 1993 and 2001, 318 children were diagnosed with acute myeloid leukaemia (AML) in the Nordic countries. The patient group comprised 237 children < 15 years of age with de novo AML, 42 children < 15 years with Down syndrome (DS) and de novo AML, 18 adolescents 15-18 years of age with de novo AML, and 21 children < 15 years with treatment-related AML (t-AML). The first group was all-inclusive, yielding an annual childhood de novo AML incidence of 0.7/100 000. Cytogenetic analyses were successful in 288 cases (91%), and clonal chromosomal abnormalities were detected in 211 (73%). The distribution of ploidy levels were pseudodiploidy (55%), hyperdiploidy (34%) and hypodiploidy (11%). The most common aberrations (> 2%) were + 8 (23%) (as a sole change in 6.2%), 11q23-translocations, including cryptic MLL rearrangements (22%) [t(9;11)(p21-22;q23) in 11%], t(8;21)(q22;q22) (9.0%), inv(16)(p13q22) (6.2%), -7/7q- (5.2%), and t(15;17)(q22;q12) (3.8%). Except for +8, these abnormalities were rare in group 2; only one DS patient had a t(8;21) and none had 11q23-translocations, t(15;17) or inv(16). In the t-AML group, three cases displayed 11q23-rearrangements, all t(9;11); and there were no t(8;21), t(15;17) or inv(16). Overall, the observed frequencies of t(8;21) and t(15;17) were lower, and frequencies of trisomy 8 and 11q23-translocations higher, than in previous studies. Furthermore, seven abnormalities that were previously reported as only single AML cases were also seen, meaning that der(4)t(4;11)(q26-27;q23), der(6)t(1;6)(q24-25;q27), der(7)t(7;11)(p22;q13), inv(8)(p23q11-12), t(11;17)(p15;q21), der(16)t(10;16)(q22;p13) and der(22)t(1;22)(q21;q13) are now classified as recurrent abnormalities in AML. In addition, 37 novel aberrations were observed, 11 of which were sole anomalies.

Is monosomy 5 an uncommon aberration? Fluorescence in situ hybridization reveals translocations and deletions in myelodysplastic syndromes or acute myelocytic leukemia.

Acquired loss of material from chromosome 5 in bone marrow cells is common in myelodysplastic syndromes (MDS) and acute myelocytic leukemia (AML). In this study, we have applied fluorescence in situ hybridization (FISH) analyses with probes for the three regions 5p15.2, 5q31, 5q33-q34, and whole chromosome 5 painting probes (WCP 5) to investigate what further information could be gained regarding the cytogenetic abnormalities of chromosome 5 in 35 patients with MDS or AML. With FISH, a del(5q) was found in all patients except for two. Translocations of material from chromosome 5 were found in 10 patients. Among 16 patients with clones of monosomy 5 seen by cytogenetics, 14 had deletions or translocations. Different breakpoints on chromosome 5 were observed. In conclusion, the extended FISH analyses yielded additional information about chromosome 5 abnormalities in 60% of the patients. Of interest is the finding of a high proportion of translocations and that monosomy 5 occurs less often than is generally believed.

Trisomies 8 and 9 not detected with fish in patients with mastocytosis.

Mastocytosis is a rare disease characterized by proliferation of mast cells in one or several organs. With conventional cytogenetics about 35% of the patients have chromosomally abnormal clones in bone marrow cells. It has been proposed to include the mast cell disease among the myeloproliferative disorders, in which trisomy for chromosome 8 and 9 can appear in the bone marrow cells. In this study bone marrow cells from eight patients with mastocytosis, two had as well an associated hematological disease, have been examined with fluorescence in situ hybridization (FISH) for enumeration of chromosome no8 and 9. In conventional cytogenetics two patients had clones with del(20) and 47,XY,+14/45, X,-Y, respectively. None of the patients with mastocytosis had clones with trisomy 8 or 9 with either cytogenetics or FISH.

Patients with idiopathic myelofibrosis show increased CD34+ cell concentrations in peripheral blood compared to patients with polycythaemia vera and essential thrombocythaemia.

The aim of the present work was to compare the results for some haematological variables in 12 patients with idiopathic myelofibrosis (IMF) with those of 21 patients with polycythaemia vera (PV), 22 patients with essential thrombocythaemia (ET) and 10 healthy control subjects. In each patient and control subject peripheral blood was used for analysis of flow cytometric measurement of CD34-positive (CD34+) cells, in vitro colony growth and plasma erythropoietin (EPO) concentration. The mean concentration of CD34+ cells in the IMF group was 568 +/- 686 x 10(3) mL, which significantly (P<0.001 for each group) exceeded the means in PV, ET and control groups ((10.2 +/- 32.0) x 10(3) mL, 3.0 +/- 3.7 x 10(3) mL and 1.9 +/- 0.8 x 10(3) mL, respectively). The mean number of EPO-independent erythroid colonies (EEC) was 110 +/- 215 colonies per 10(5) cells for the IMF patients. In comparison with the means for PV and ET patients (40 +/- 140 and 12 +/- 27 colonies per 10(5) cells, respectively) the difference did not reach statistical significance. The mean EEC for IMF patients was, however, significantly higher compared with the mean for the control subjects (P<0.05). The means for total erythroid colony growth with EPO added to the growth medium as well as for granulocyte-macrophage colony-forming units were significantly higher in the IMF group compared with the means for PV, ET and control groups (P<0.001 for each group). The mean plasma EPO was 204 +/- 290 IU/L in the patients with IMF compared with 6.6 +/- 7.9 IU/L in PV patients, 19.1 +/- 23.2 IU/L in ET and 10.3 +/- 7.8 IU/L in the control subjects. Due to considerable differences in haemoglobin concentrations no relevant conclusions could be drawn from the results for plasma EPO concentrations. Indeed, the majority of patients with IMF, PV and ET were on myelosuppressive treatment; additionally most PV patients received phlebotomy therapy. The results of the present study suggest that the circulating pool of stem cells and progenitor cells in peripheral blood is significantly increased in IMF patients compared with PV and ET patients as well as healthy control subjects. The most likely source for the elevated CD34+ cell concentration in peripheral blood is progenitor cells of extra-medullar origin.

Autologous blood transfusion in radical hysterectomy with and without erythropoietin therapy.

OBJECTIVE: To investigate whether preoperative treatment with erythropoietin facilitates the collection of a sufficient amount of autologous blood in a short period of time. METHODS: Forty-one women scheduled for radical hysterectomy were randomized to preoperative autologous blood donation with or without preoperative recombinant human erythropoietin therapy. All patients were scheduled to deposit three units of blood within 2 weeks before surgery. Hemoglobin, erythrocyte volume fraction, blood cells, iron status, and hemolysis were analyzed before and after surgery. RESULTS: Hemoglobin levels decreased continuously in both groups after the first autologous donation until day 1 postoperatively. With erythropoietin therapy, the erythrocyte volume fraction and hemoglobin levels were significantly higher during precollection and day 1 after surgery. Preoperatively, the drop was 12 g/L less in the erythropoietin-treated group. The additional use of erythropoietin therapy reduced the inability of patients to predeposit blood from 17.8% to 3.4%. CONCLUSION: Most women can predeposit three units of whole blood in only 2 weeks without obtaining severe anemia. By treating women with erythropoietin, one out of seven can be prevented from a hemoglobin level below the 100 g/L limit for donation.