Proposals for revised IWG 2018 hematological response criteria in patients with MDS included in clinical trials.

The heterogeneity of myelodysplastic syndromes (MDSs) has made evaluating patient response to treatment challenging. In 2006, the International Working Group (IWG) proposed a revision to previously published standardized response criteria (IWG 2000) for uniformly evaluating clinical responses in MDSs. These IWG 2006 criteria have been used prospectively in many clinical trials in MDSs, but proved challenging in several of them, especially for the evaluation of erythroid response. In this report, we provide rationale for modifications (IWG 2018) of these recommendations, mainly for "hematological improvement" criteria used for lower-risk MDSs, based on recent practical and reported experience in clinical trials. Most suggestions relate to erythroid response assessment, which are refined in an overall more stringent manner. Two major proposed changes are the differentiation between "procedures" and "criteria" for hematologic improvement-erythroid assessment and a new categorization of transfusion-burden subgroups.

Erythropoiesis-stimulating agents significantly delay the onset of a regular transfusion need in nontransfused patients with lower-risk myelodysplastic syndrome.

BACKGROUND: The EUMDS registry is an unique prospective, longitudinal observational registry enrolling newly diagnosed patients with lower-risk myelodysplastic syndrome (MDS) from 17 European countries from both university hospitals and smaller regional hospitals. OBJECTIVE: The aim of this study was to describe the usage and clinical impact of erythropoiesis-stimulating agents (ESAs) in 1696 patients enrolled between 2008 and 2014. METHODS: The effects of ESAs on outcomes were assessed using proportional hazards models weighting observations by propensity to receive ESA treatment within a subset of anaemic patients with or without a regular transfusion need. RESULTS: ESA treatment (median duration of 27.5 months, range 0-77 months) was administered to 773 patients (45.6%). Outcomes were assessed in 897 patients (484 ESA treated and 413 untreated). ESA treatment was associated with a nonsignificant survival benefit (HR 0.82, 95% CI: 0.65-1.04, P = 0.09); this benefit was larger amongst patients without prior transfusions (P = 0.07). Amongst 539 patients for whom response to ESA treatment could be defined, median time to first post-ESA treatment transfusion was 6.1 months (IQR: 4.3-15.9 months) in those transfused before ESA treatment compared to 23.3 months (IQR: 7.0-47.8 months) in patients without prior transfusions (HR 2.4, 95% CI: 1.7-3.3, P < 0.0001). Responding patients had a better prognosis in terms of a lower risk of death (HR 0.65, 95% CI: 0.45-0.893, P = 0.018), whereas there was no significant effect on the risk of progression to acute myeloid leukaemia (HR 0.71, 95% CI: 0.39-1.29, P = 0.27). CONCLUSION: Appropriate use of ESAs can significantly delay the onset of a regular transfusion need in patients with lower-risk MDS.

Somatic SF3B1 mutation in myelodysplasia with ring sideroblasts.

BACKGROUND: Myelodysplastic syndromes are a diverse and common group of chronic hematologic cancers. The identification of new genetic lesions could facilitate new diagnostic and therapeutic strategies. METHODS: We used massively parallel sequencing technology to identify somatically acquired point mutations across all protein-coding exons in the genome in 9 patients with low-grade myelodysplasia. Targeted resequencing of the gene encoding RNA splicing factor 3B, subunit 1 (SF3B1), was also performed in a cohort of 2087 patients with myeloid or other cancers. RESULTS: We identified 64 point mutations in the 9 patients. Recurrent somatically acquired mutations were identified in SF3B1. Follow-up revealed SF3B1 mutations in 72 of 354 patients (20%) with myelodysplastic syndromes, with particularly high frequency among patients whose disease was characterized by ring sideroblasts (53 of 82 [65%]). The gene was also mutated in 1 to 5% of patients with a variety of other tumor types. The observed mutations were less deleterious than was expected on the basis of chance, suggesting that the mutated protein retains structural integrity with altered function. SF3B1 mutations were associated with down-regulation of key gene networks, including core mitochondrial pathways. Clinically, patients with SF3B1 mutations had fewer cytopenias and longer event-free survival than patients without SF3B1 mutations. CONCLUSIONS: Mutations in SF3B1 implicate abnormalities of messenger RNA splicing in the pathogenesis of myelodysplastic syndromes. (Funded by the Wellcome Trust and others.).

Myelodysplastic syndromes: biology and treatment.

Optimal management of patients with myelodysplastic syndromes (MDS) requires an insight into the biology of the disease and the mechanisms of action of the available therapies. This review focuses on low-risk MDS, for which chronic anaemia and eventual progression to acute myeloid leukaemia are the main concerns. We cover the updated World Health Organization classification, the latest prognostic scoring system, and describe novel findings in the pathogenesis of 5q- syndrome. We perform in depth analyses of two of the most widely used treatments, erythropoietin and lenalidomide, discussing mechanisms of action, reasons for treatment failure and influence on survival.

Early mitochondrial alterations in ATRA-induced cell death.

All-trans retinoic acid (ATRA) induces differentiation and subsequent apoptosis in a variety of cell lines. Using the myeloid cell line P39, we show that ATRA disturbs mitochondrial functional activity long before any detectable signs of apoptosis occur. These early changes include diminished mitochondrial oxygen consumption, decreased calcium uptake by mitochondria and as a result, a lower mitochondrial matrix calcium concentration. Granulocyte colony-stimulating factor (G-CSF) increases mitochondrial respiration and calcium accumulation capacity and subsequently blocks ATRA-induced apoptosis. Nifedipine, a plasma membrane calcium channel blocker, inhibits apoptosis-related changes, such as the loss of the mitochondrial membrane potential and activation of caspases. Thus, the properties of ATRA and G-CSF to modulate mitochondrial respiration and intracellular calcium control are novel findings, which give insight into their precise molecular mode of action.

Attitude towards remission induction for elderly patients with acute myeloid leukemia influences survival.

Combination chemotherapy may induce remission from acute myeloid leukemia (AML), but validated criteria for treatment of elderly are lacking. The remission intention (RI) rate for elderly patients, as reported to the Swedish Leukemia Registry, was known to be different when comparing the six health care regions, but the consequences of different management are unknown. The Leukemia Registry, containing 1672 AML patients diagnosed between 1997 and 2001, with 98% coverage and a median follow-up of 4 years, was completed with data from the compulsory cancer and population registries. Among 506 treated and untreated patients aged 70-79 years with AML (non-APL), there was a direct correlation between the RI rate in each health region (range 36-76%) and the two-year overall survival, with no censored observations (6-21%) (chi-squared for trend=11.3, P<0.001; r2=0.86, P<0.02, nonparametric). A 1-month landmark analysis showed significantly better survival in regions with higher RI rates (P=0.003). Differences could not be explained by demographics, and was found in both de novo and secondary leukemias. The 5-year survival of the overall population aged 70-79 years was similar between the regions. Survival of 70-79-year-old AML patients is better in regions where more elderly patients are judged eligible for remission induction.

Stem cell transplantation from identical twins in patients with myelodysplastic syndromes.

In a multicentre retrospective EBMT database study, we analysed factors influencing outcome in 38 patients with MDS/sAML who were transplanted with stem cells from their syngeneic twin and compared those to 1444 patients who were transplanted from an HLA-identical sibling. The median time to leukocyte and platelet engraftment was faster in the twin group: 14 vs 17 (P=0.02) and 16 vs 26 days (P=0.09), respectively. The 5 years cumulative incidence of treatment-related mortality (TRM) was higher in the sibling than in the twin group (38 vs 27%; P=0.05). The 5 year cumulative incidence of relapse was 32% (95% CI: 29-35%) for the siblings and 39% (95% CI: 26-60%; P=0.6) for the twins. A trend for better 5-years disease-free and overall survival was observed in the twin group: 34% (95% CI: 14-54%) vs 28% (95% CI: 25-31%; P=0.2) and 36% (95% CI: 15-57%) vs 32% (95% CI: 29-35%; P=0.09), respectively. In a multivariate analysis, stem cell transplantation from identical twins had a lower TRM: HR: 0.4 (95% CI: 0.2-0.9; P=0.03). The relapse rate was similar for both groups with a HR of 1.2 (95% CI: 0.07-2.1; P=0.5), with a better survival for the twins: HR 0.6 (95% CI: 0.4-1.0; P=0.07). We conclude that twin transplantation in MDS/sAML is associated with a similar relapse risk, a lower TRM and a trend for better overall survival in comparison to transplantation from HLA-identical siblings.

The pharmacodynamic effect of busulfan in the P39 myeloid cell line in vitro.

Busulfan is an alkylating agent currently used in the myeloablative conditioning regimen before stem cell transplantation. Its mechanism of action is not fully understood, nor the reason for its narrow therapeutic window. We studied the pharmacodynamics of busulfan in an in vitro cell line model, allowing us to evaluate the effects of various doses and exposure times on clonogeneic capacity, proliferation and apoptosis. Cells were incubated with busulfan in concentrations ranging from 10 to 100 microg/ml for 2, 4 or 8 h, then washed and cultured in busulfan-free medium for 72 h. Area under the concentration-time curve (AUC) was estimated by using the trapezoidal rule from different concentrations and times of incubation. In all assays busulfan affected the cells in an AUC-dependent manner. Induced changes in the biological parameters studied appeared at different time points after exposure to busulfan stopped. Thus, the decrease in proliferation and clonogenic capacity preceded cell cycle arrest in G2 phase and development of apoptosis, implying that apoptosis is a secondary event to interruption of vital metabolic processes. Biochemically, apoptotic changes were typical for chemotherapy-induced apoptosis with caspase activation, cleavage of Bcl-2 and PARP proteins, while cleavage of actin was not observed. Cells were rescued from apoptosis with a general caspase inhibitor ZVAD-fmk, but not with granulocyte colony-stimulating factor (G-CSF). Our results add new information about busulfan pharmacodynamics and mechanisms underlying the cytotoxic effect of the drug.

Cleavage of Bcl-2 is an early event in chemotherapy-induced apoptosis of human myeloid leukemia cells.

The proto-oncogene product Bcl-2 protects a wide variety of cell types from apoptosis via a hitherto unknown mechanism. Bcl-2 has been shown to function upstream of the death proteases (caspases) in some, but not all, occurrences of apoptotic cell death. Using the myeloid leukemic cell line P39 we report the chemotherapy-induced caspase-dependent cleavage of endogenous Bcl-2. Etoposide treatment of these cells triggered a time-dependent activation of type II and type III caspases and cleavage of Bcl-2 yielding a 23 kDa cleavage fragment. The emergence of this cleavage product was blocked by the general caspase inhibitor zVAD-fmk, as well as the type III caspase inhibitor IETD-fmk and the caspase-9-selective inhibitor LEHD-fmk, while the type II caspase inhibitor DEVD-fmk proved considerably less efficient. Bcl-2 cleavage preceded cleavage of the known caspase-3 substrate, poly(ADP-ribose) polymerase (PARP), as well as that of the caspase-6 substrate, lamin B, indicating that Bcl-2 cleavage is a relatively early event in the apoptosis cascade in this experimental model. While evidence for cleavage of Bcl-2 in several subcellular compartments of etoposide-treated cells was obtained, this cleavage was detected predominantly in the mitochondrial fraction, thus providing further support for the central role of mitochondria in apoptosis. Caspase-mediated cleavage following etoposide treatment of these myeloid leukemic cells may represent a means for the attenuation of Bcl-2 function upon apoptosis induction.

Granulocyte colony-stimulating factor inhibits Fas-triggered apoptosis in bone marrow cells isolated from patients with refractory anemia with ringed sideroblasts.

Treatment with granulocyte colony-stimulating factor (G-CSF) plus erythropoietin may synergistically improve hemoglobin levels and reduce bone marrow apoptosis in patients with refractory anemia with ringed sideroblasts (RARS). Fas-induced caspase activity is increased in RARS bone marrow cells. We showed that G-CSF significantly reduced Fas-mediated caspase-8 and caspase-3-like activity and the degree of nuclear apoptotic changes in bone marrow from nine RARS patients. A decrease in mitochondrial membrane potential and an increase in intracellular reactive oxygen species occurred in Fas-treated cells, but became significant only 24 h after changes in caspase activity and decrease in proliferation. G-CSF also reduced the magnitude of these late apoptotic changes. In CD34-selected normal cells, G-CSF induced myeloid colony growth, and an overall small decrease in the number of erythroid colonies. By contrast, G-CSF induced a 33-263% increase of erythroid colony formation in CD34+ cells from four of five RARS patients with severely reduced erythroid growth, while the normal or slightly reduced erythroid growth of three other patients was not influenced by G-CSF. This study suggests that G-CSF may reduce the pathologically increased caspase activity and concomitant apoptotic changes, and promote erythroid growth and differentiation of stem cells from RARS patients. Our data support the clinical benefit of G-CSF in this subgroup of myelodysplastic syndromes.

No benefit from adding GM-CSF to induction chemotherapy in transforming myelodysplastic syndromes: better outcome in patients with less proliferative disease.

In this prospective randomized multicenter trial 93 patients, median age 72 years, with RAEB-t (n=25) and myelodysplastic syndrome (MDS)-AML (n=68) were allocated to a standard induction chemotherapy regimen (TAD 2+7) with or without addition of granulocyte-macrophage-CSF (GM-CSF). The overall complete remission (CR) rate was 43% with no difference between the arms. Median survival times for all patients, CR patients, and non-CR patients were 280, 550, and 100 days, respectively, with no difference between the arms. Response rates were significantly better in patients with serum lactate dehydrogenase (S-LDH) levels

Two pathways of apoptosis induced with all-trans retinoic acid and etoposide in the myeloid cell line P39.

P39/Tsugane is a myelomonocytoid cell line derived from a patient with myelodysplastic syndrome (MDS). The cells readily undergo apoptosis in response to various agents, and the cell line has been suggested as a useful model to study apoptosis in MDS. The aims of the present study were to assess differentiation and apoptosis induced with all-trans retinoic acid (ATRA) and etoposide, to characterize the mode of apoptosis in these two model systems, and to assess the influence of granulocyte colony-stimulating factor (G-CSF), which in combination with erythropoietin has been shown to inhibit apoptosis in MDS. ATRA induced differentiation and apoptosis in a concentration- and time-dependent manner. Differentiated cells were partially rescued (by 50%) from apoptosis with G-CSF. Etoposide induced apoptosis in a concentration- and time-dependent manner, but no signs of preceding maturation or G-CSF rescue were detected. ATRA- and etoposide-induced apoptosis were both mediated through the caspase pathway and were partially blocked with the general caspase inhibitor zVAD-fmk. Simultaneous treatment with G-CSF and zVAD-fmk additively blocked ATRA-induced apoptosis. However, the two pathways differed in terms of substrate cleavage during apoptosis. ATRA-induced apoptosis caused actin cleavage, which was not affected by G-CSF, and Bcl-2 downregulation. Etoposide induced a caspase-dependent cleavage of Bcl-2, while actin remained intact. The Fas system did not seem to play a major role in any of these apoptotic pathways. Our results may provide new tools to study the mechanisms of apoptosis in MDS.

Freezing induces artificial cleavage of apoptosis-related proteins in human bone marrow cells.

The aim of this study was to investigate whether freeze-thawing of freshly isolated human mononuclear bone marrow cells (MNC) influences the integrity of apoptosis-related proteins as determined by immunoblot analyses. Our results show that bone marrow is more sensitive to this process than either myelomonocytoid leukemic P39 or Jurkat T-lymphocyte cell lines. Specifically, bone marrow cells displayed a high level of intrinsic proteolytic activity in response to a single freeze-thaw cycle, which led to the cleavage of various proteins involved in apoptosis cell signaling. This effect was completely blocked by the inclusion of broad-spectrum protease inhibitors in the freezing medium and subsequently thawing the cells on ice. Since differences in the freezing conditions (-80 degrees C vs. liquid nitrogen) did not alter the proteins of interest, we suggest that the thawing process is the critical point when proteolytic enzyme activity is elevated.

Morphological changes and apoptosis in bone marrow from patients with myelodysplastic syndromes treated with granulocyte-CSF and erythropoietin.

A study of bone marrow morphology and apoptosis was undertaken in 51 patients with myelodysplastic syndromes (MDS) treated with granulocyte colony-stimulating factor (G-CSF) and erythropoietin (EPO). In 19 of these patients (37%), a significant improvement in the hemoglobin level was found after treatment. Apoptosis was measured using a nick-end labeling (TUNEL) technique. Patients with MDS had a significantly higher percentage of labelled (apoptotic) cells in the bone marrow compared to healthy individuals (56.3 +/- 3.8% vs. 16.2 +/- 1.4%, p = 0.0001). Patients with RAS showed a lower percentage of apoptotic cells than patients with RA (68.5 +/- 9% vs. 46.5 +/- 4.8%, p < 0.05), while patients with RAEB did not differ significantly from either RA or RAS. In the patients who responded to treatment, the bone marrow samples displayed significant morphological changes. The percentages of erythroid precursors and myeloblasts were reduced after treatment, and patients who had ring sideroblasts before treatment also showed a reduction in the percentage of these cells. Total erythroid index also decreased in responding patients. The percentage of apoptotic cells decreased significantly in responding patients (58.8 +/- 4.8% before treatment vs. 44.5 +/- 5.5% after treatment, mean reduction 18.3%, p = 0.0003), whereas no significant change was found in non-responding patients. Our results suggest that one important mechanism behind the positive effects of treatment with G-CSF and EPO is a reduction in the degree of ineffective hematopoiesis in MDS.

Variable growth in vitro of chromosomally different stem cells in myelodysplastic syndromes.

Ten patients with myelodysplastic syndromes were investigated using in vitro colony assay of bone marrow cells and chromosome analysis of single colonies. The result was compared with conventional cytogenetic analysis of bone marrow cells. The chromosome abnormalities included were 5q-, +8, -7, 11q-, -Y and one complex karyotype. Erythroid colony formation was reduced in eight patients, while the number of granulocyte-macrophage colony-forming units was normal or increased. Cytogenetic examination of single colonies showed that both chromosomally normal and abnormal stem cells had colony forming ability. The proportion of cytogenetically abnormal colonies varied between 25 and 100% of analysed colonies. No further clones than those found in direct cytogenetic analysis were revealed after in vitro growth. One patient with a 5q- abnormality and one patient with a -7 abnormality showed a significantly lower proportion of cytogenetically abnormal colonies than the proportion of abnormal cells in the direct bone marrow chromosome preparation. One patient with a +8 abnormality showed a growth advantage of the +8 clone in comparison with the chromosomally normal clones, but this was not statistically significant. In two patients a chromosomally changed stem cell gave rise to both erythroid and myeloid colonies. The FAB-class did not seem to influence the growth of either chromosomally normal or abnormal colonies.

High-dose chemotherapy and APSCT as a potential cure for relapsing hemolysing AILD.

Angioimmunoblastic lymphadenopathy with dysproteinemia (or dysgammaglobulinemia) (AILD) is a lymphoproliferative disorder with cytogenetic and molecular abnormalities characteristic of malignant T-cell lymphoma (angioimmunoblastic T-cell lymphoma -- AITL). We report the clinical course of a 58-year-old male patient with unusually aggressive AILD, including severe hemolysis and Guillain-Barré syndrome, who entered complete remission after CHOP therapy, but had a full relapse after 2 months. At relapse, treatment with high-dose chemotherapy followed by autologous peripheral stem cell transplantation (APSCT) with CD34 selected cells was shown to be successful. The patient is alive and disease-free 3 years after diagnosis and 32 months after APSCT. Considering the poor prognosis of the majority of patients with AILD, intensive treatment followed by APSCT, may be a subject for further studies.

In vitro suspension culture reactions to 1,25 dihydroxyvitamin D3 in relation to bone marrow morphology and prognosis in patients with myelodysplastic syndromes.

Thirty-four patients with MDS or AML following MDS were studied with regard to survival, peripheral blood values and bone marrow morphology. The effects of 1,25 dihydroxyvitamin D3 (D3) on differentiation (NBT positivity) and proliferation (3H-thymidine incorporation) were studied in suspension cultures of bone marrow cells. Twelve bone marrow donors served as controls. Normal cells showed spontaneous differentiation in vitro, but only 2/12 were induced to differentiation by D3. Myelodysplastic cells did not differentiate spontaneously, but cells from 18/34 patients differentiated after incubation with D3. Normal cells showed increased proliferation, myelodysplastic cells showed a heterogeneous response and leukemic cells reacted with decreased proliferation after D3 incubation. Poor survival was associated with low platelet counts, high percentage of bone marrow blasts (BM blast %), low spontaneous in vitro proliferation and absence of hypogranulation of myeloid cells. Platelet counts and hypogranulation retained their predictive value in a multi-variate analysis. Progression to AML was predicted by a high BM blast % and low scores for erythroid and total dysplasia. In conclusion, the pattern of in vitro proliferation showed prognostic value while the pattern of vitamin D3-induced differentiation failed to correlate to other parameters. An estimation of bone marrow dysplasia can be used to predict the development of AML. Our results add to the information about the biology of MDS and may be important for the evaluation of therapeutic trials.

The effect of modulation of glutathione cellular content on busulphan-induced cytotoxicity on hematopoietic cells in vitro and in vivo.

Busulphan is used in conditioning regimens prior to SCT. A relationship between exposure to busulphan, expressed as an area under the plasma concentration time curve (AUC), and effect and/or adverse effects, such as veno-occlusive disease (VOD), was reported. Exhaustion of glutathione (GSH) contributes to VOD and modulation of intracellular levels of GSH influences bulsulphan-induced toxicity in hepatocytes. Thus, increase of GSH might serve as prophylaxis against VOD. However, it should not interfere with the myeloablative effects of busulphan. We investigated the relationship between exposure to busulphan, and its in vitro toxicity to CD34(+) hematopoietic progenitors from volunteers using clonogenic assays. Busulphan inhibited colony formation by CD34(+) cells in an AUC-dependent manner. Myeloid progenitors were more sensitive than erythroid progenitors, expressed as 100% inhibition of colony formation (68.7 +/- 7.5 microg.h/ml and 140.3 +/- 35.7, respectively). The observed exposure corresponds to the total AUC obtained in patients treated with busulphan (1 mg/kg/day) for 4 days. Secondly, we studied the effect of modulation of GSH cellular levels on busulphan-induced toxicity in vitro in CD34(+) cells from volunteers, and in vivo in bone marrow cells from Balb/c mice. The intracellular concentration of GSH was increased or decreased by treatment with N-acetylcysteine or buthionine sulfoximine, respectively. Neither in vitro nor in vivo treatment with GSH modulators affected the hematological toxicity of busulphan. Thus, N-acetylcysteine would not interfere with the myeloablative effect of busulphan and therefore it is a potential candidate for VOD prophylaxis during busulphan-based conditioning regimens.

Treatment of adult myelodysplastic syndromes.

Patients with myelodysplastic syndromes (MDS) suffer from pancytopenia and are at substantial risk for progression to acute myeloid leukemia. The principal pathogenetic features of MDS are clonal evolution, ineffective hematopoiesis (apoptosis), and reduced cellular maturation. T-cell-mediated myelosupression may add to disease development in some patients. The disease is progressive and, for the majority of patients, the aim of treatment will be to improve blood values and quality of life. A minority of patients could be cured by allogeneic bone marrow transplantation, which results in an overall disease-free survival of about 40%. The value of autologous stem-cell transplantation is currently being investigated in clinical trials; this treatment may be found to be beneficial for some MDS patients. High-dose chemotherapy may lead to significant improvement and complete remission in about 50% of treated patients, but remission duration is usually short. Low-dose chemotherapy may also be used to improve peripheral blood values. Several new low-dose treatments are being tested at present. Growth factors can be used to improve both ineffective hematopoiesis and cytopenia. The value of treatment with growth factors alone for granulocytopoiesis is uncertain, whereas treatment for impaired erythropoiesis using erythropoietin alone or erythropoietin in combination with other growth factors seems more promising. Subgroups of MDS in which T-cell-mediated myelosuppression are present may respond favorably to cyclosporin A or antithymoglobulin. Antiapoptotic agents such as amifostine may improve blood values in some patients with MDS, but the value of this treatment is not yet clear. Increasing pathogenetic knowledge and better use of predictive models have resulted in some progress in MDS treatment. In the future, more effective treatments may result from further insights into the biology of the disease, the discovery of new therapeutic approaches, and the search for better ways to use existing therapeutic options.

A combination of granulocyte colony-stimulating factor and erythropoietin may synergistically improve the anaemia in patients with myelodysplastic syndromes.

In an attempt to obtain a synergistic effect on the hemoglobin levels in anaemic patients with myelodysplastic syndromes (MDS), granulocyte colony-stimulating factor (G-CSF) and erythropoietin (epo) were combined in a clinical phase II trial. Twenty-two patients with MDS were included in the study. G-CSF was given alone for six weeks and then in combination with epo for the following twelve weeks. Eight (38%) of 21 evaluable patients showed a significant increase in hemoglobin. One patient with a previous response and subsequent failure to epo alone improved after the addition of G-CSF. Responses were more frequent in patients with less advanced pancytopenia, lower endogenous levels of serum-epo and in those with ring sideroblasts in the bone marrow. The response frequency of 38% is higher than in any study of epo as monotherapy. Moreover, patients with ring sideroblasts, who respond poorly to epo alone, showed a response rate of 60%. Our findings suggest a synergistic in vivo effect of granulocyte-CSF and erythropoietin in patients with myelodysplastic syndromes.