Steensma, David P.

Abstract BACKGROUND: Acquired hemoglobin disorders including α thalassemia (hemoglobin H (HbH) disease) can complicate neoplastic myeloid disorders, especially myelodysplastic syndrome (MDS). Acquired α thalassemia in MDS has recently been linked to mutations in the trans-acting ATRX at Xq13, a gene previously associated with inherited ATR-X (α thalassemia-mental retardation-X linked) syndrome. In the 3 instances where acquired ATRX point mutations predicted to affect splicing have been encountered, mRNA has been unavailable for assessment or unexpressed; in the 3 cases where ATRX splicing changes were described in MDS, the responsible genomic DNA mutation could not be located. Here we report a novel genomic point mutation in an ATRX splice donor motif and define the mutation’s splicing and expression consequences. METHODS AND RESULTS: A 59 year old anemic Swiss man with chronic myelomonocytic leukemia (CMML) and multiple autoimmune complications including granulomatous uveitis, diabetes insipidus, and pleuropericarditis, developed anemia (Hb 10 g/dL) with MCH of 16.7 pg and MCV of 54 fL. We detected 50% HbH-containing cells by supravital staining and 7.5% HbH by weak cation chromatography, confirmed as HbH by isoelectric focusing. Peripheral blood granulocytes, mononuclear cells and bone marrow mononuclear cells were isolated using dual density Ficoll-Hypaque, and DNA and RNA extracted for mutation screening. ATRX coding region (35 exons) and splice site DNA from mononuclear cells was amplified and subjected to denaturing high performance liquid chromatography (DHPLC). A heteroduplex peak was encountered in the amplicon centered on exon 4. Sequencing demonstrated thymidine to cytidine point mutation with high proportional clonality in the canonical GT splice donor motif 3′ to ATRX exon 4 (IVS 4+2 T>C), also detectable in patient granulocytes. Electrophoresis of raw PCR product from the 5′ end of ATRX cDNA demonstrated multiple abnormal bands. Subcloning and sequencing of cDNA demonstrated 2 splicing consequences of this mutation: exon 4 skipping (i.e., exon 3 was spliced directly to exon 5), and a second isoform with intact exons 3, 4, and 5 but retaining 43 base pairs of intron 4 (including the T>C point mutation in the splice donor site), activating a previously unrecognized cryptic splice site. Since exon 4 is 53 base pairs in length, both aberrant splicing consequences result in a frameshift and generation of a premature stop codon in the 5′ end of exon 5. Exon 6 is normally alternatively spliced, and this variation was retained, so there were two subforms of each aberrant spliceoform. Expression of ATRX mRNA in the patient’s monocytes, measured by RT-PCR probing the 3′ helicase domain, was 11.1% of healthy controls; in the granulocytes, expression was 15.3% of normal. CONCLUSION: Pre-mRNA splicing abnormalities such as that described here are increasingly recognized in association with human disease and particularly neoplasia, and may represent a potential therapeutic target. Point mutations in general are sought in myeloid disorders as they can provide proof of clonality in ambiguous cases and allow tracking of clonality with therapy and disease evolution. Germline 5′ ATRX splicing abnormalities are only rarely associated with inherited ATR-X syndrome, presumably because of the devastating consequences on gene expression and fetal development, but may be more common in acquired thalassemia. In future analyses of acquired HbH disease, when seeking ATRX mutations, splicing abnormalities should be routinely considered.

Abstract BACKGROUND: The molecular etiology of idiopathic acquired sideroblastic anemia (IASA), now considered a form of myelodysplastic syndrome, is currently unknown. Romslo et al (Blood 1982) reported a patient with IASA who had moderately elevated free erythrocytic protoporphyrin (FEP); this observation is now frequently made in IASA, helping distinguish IASA (high FEP) from inherited sideroblastosis (low FEP). In addition, rare patients with erythropoietic protoporphyria (EPP)—defined by cutaneous photosensitivity, dramatically elevated FEP levels, and germline point mutations in the ferrochelatase (FECH) gene at 18q21.3—have ringed sideroblasts in their marrow. We hypothesized that patients with IASA might have acquired somatic FECH mutations. METHODS AND RESULTS: We designed a denaturing high performance liquid chromatography (DHPLC) assay to explore this possibility and to avoid problems related to mutation screening in the setting of mixed clonality. To validate the DHPLC assay, the coding region of the FECH gene from 2 molecularly undiagnosed EPP patients without liver disease was analyzed. In one patient, both a heterozygous 69delG mutation (GenBank Accession NM_000140) and heterozygosity for the FECH expression-modulating IVS3-48C/T polymorphism were detected. In the other patient, no FECH coding mutations or polymorphisms were detected, either by DHPLC or conventional dye sequencing. FEP measurements were then obtained on 2 IASA patients and were elevated (65 and 115 mcg/dL; normal 1–10 mcg/dL) with normal urine and fecal porphyrins. Genomic DNA obtained from these 2 patients as well as archival DNA from 30 other patients with IASA was amplified and tested for FECH mutations. No coding region mutations were detected. Synonymous polymorphisms in exon 7 (rs536765) and 9 (rs536560) were found in 6/32 (19%; normal heterozygosity 0.398) and 14/32 (44%; normal heterozygosity 0.402) samples, respectively. The IVS3-48C/T polymorphism (rs2272783) was found in 3/32 (9%) (prevalence in the general population is 11%, Gouya Nat Genet 2002). In addition, 3 intronic polymorphisms not in the refSNP database were detected: IVS8+34 C/T (2/32), IVS8-61delG (5/32), and IVS9-59delA (2/32). CONCLUSION: IASA is not associated with coding mutations in FECH. Elevation of FEP in ASA must instead be due to the lack of mitochondrial iron in the proper form for incorporation into the porphyrin ring; attention should instead focus on factors responsible for maintaining the appropriate redox state and compartmentalization of iron. DHPLC can detect FECH mutations and polymorphisms, but because of the high frequency of the latter and the consequent need to sequence multiple exons, it is not a practical screening tool for studying undiagnosed EPP patients.

Abstract Background: The occurrence of an activating JAK2 mutation (JAK2 V617F) has been reported in a variable proportion of patients with bcr/abl-negative myeloproliferative disorders including myelofibrosis with myeloid metaplasia (MMM). In the current study, we report mutational frequencies in each of the 3 MMM variants as well as examine the prognostic relevance of the mutation in a large group of patients from a single institution. Methods: Mutation analysis for JAK2 V617F was performed in DNA derived from peripheral blood mononuclear cells (PBMC) from all study patients (n=157). Concomitant analysis was also performed in DNA derived from purified granulocytes (n=57), CD34-positive cells (n=25), and T cells (n=19). Genomic DNA was amplified by PCR and fluorescent dye chemistry sequencing was performed using the same primers used for amplification. Results: The study cohort included 157 consecutive patients with MMM including 117 with agnogenic (AMM), 22 with post polycythemic (PPMM), and 18 with post thrombocythemic (PTMM) myeloid metaplasia. The overall detection rate for JAK2 V617F in PBMC was 51% (homozygous in 5.7%) and was significantly (p=0.0002) higher in PPMM (91%; homozygous in 18%) compared to either AMM (45.3%; homozygous in 2.6%) or PTMM (38.9%; homozygous in 11.1%). In AMM (n=117), the presence of JAK2 V617F was significantly associated with an older age at diagnosis (p=0.03), and a history of both thrombosis (p=0.01) and pruritus (p=0.005). No other statistical associations were evident and the presence of the mutation did not predict leukemic transformation. Furthermore, multivariate analysis for overall survival of patients with AMM identified only age and the Dupriez prognostic score as independent prognostic factors; JAK2 V617F had no prognostic significance. Finally, the mutation was not found in T cells and its detection rate in either granulocytes or CD34-positive cells was similar to that in PBMC. Conclusion: JAK2 V617F is a myeloid lineage-specific event, its incidence in MMM is significantly higher with an antecedent history of polycythemia vera, and its presence in AMM does not affect prognosis but is associated with PV-characteristic clinical features.

Abstract Background: Most reports of the causes of extreme leukocytosis (i.e., total white blood count (WBC) >50 × 109/L) have included relatively small numbers of cases or data from the era prior to FDA approval of filgrastim (recombinant G-CSF; February 1991), sargramostim (GM-CSF; March 1991), and pegfilgrastim (G-CSF conjugated to polyethylene glycol; January 2002). The contribution of myeloid growth factors to extreme leukocytosis in contemporary medical practice is unclear. Methods: We evaluated the medical records of all patients with extreme leukocytosis evaluated in the clinical laboratory of the University of Missouri-Columbia (UMC) Medical Center during the years 2000 to 2007. We collected complete blood count data, information on the etiology of the leukocytosis, demographic factors, and clinical complications. Results: We identified 371 unique patients who had at least one WBC >50 × 109/L during this 8-year period (43.7% females; median age 56 (range 0–92)). Review of the primary cause of the leukocytosis in each case revealed that, while the various subtypes of leukemia (n=165, 44.5% of cases) and leukemoid reactions to general medical disorders (n=155, 41.8%) still represent the most common etiologies of extreme WBC elevation, growth factor-associated leukocytosis (GFL) is common, accounting for 13.7% (n=51) of patients. Among the subset of patients with GFL, the median peak WBC count was 60.8 × 109/L (range, 50–114.7); 51% were female, and the median age was 52 years (range 0–85). The median absolute WBC differential was as follows: neutrophils 40.8 × 109/L, granulocytes 55.7, bands 7.1, lymphocytes 2.4, monocytes 1.3, eosinophils 0, and basophils 0. Most (84%) of the patients receiving growth factors were being treated for cancer, the most common being non-Hodgkin lymphoma (NHL, 19.6%), Hodgkin lymphoma (HL, 11.8%), breast cancer (9.8%), ovarian cancer (5.9%), melanoma (5.9%), and endometrial cancer (5.9%). The non-malignant diagnoses (n = 8) included autoimmune neutropenia, bruton’s agammagloblinemia with pneumonia, leukopenia of undefined etiology, preterm neutropenia/sepsis (n = 4), and neutropenia associated with HIV. 70.6% of patients with GFL were receiving filgrastim, 29.4% pegfilgrastim, and 0% sargramostim. 0% of patients were undergoing conditioning for stem cell transplant. While myeloid growth factors are generally well tolerated, extreme leukocytosis is concerning because it may be associated with serious adverse events such as spontaneous splenic rupture, leukostasis/hyperviscosity, and hemorrhage. Other diagnoses continue to be important contributors to extreme leukocytosis. Of the 165 patients with leukemia, 43% were diagnosed with CLL, 15.8% CML, 15.2% AML, 13.3% ALL, 3.6% CMML, 3.6% BCR/ABL-negative myeloproliferative disorders, and 3.6% NHL. Other diagnoses included plasma cell leukemia, undifferentiated leukemia, and an undefined lymphoproliferative disease with macroglobulinemia variant. The causes of non-clonal leukocytosis (leukemoid reactions) could be divided into four broad categories: infection (n=109, 70.3%), reaction to hemorrhage (n=15, 9.7%), paraneoplastic (n=11, 7.1%), and other miscellaneous causes (20, 12.9%). The most common infections causing extreme leukocytosis include sepsis (59, 54.1%), pneumonia (19, 17.4%), and Clostridium difficile-associated colitis (5, 4.6%). Paraneoplastic causes of leukocytosis were diverse and associated neoplasms included cancer of unknown primary (n = 2), cholangiocarcinoma, extranodal marginal zone B-cell lymphoma, HL, SCLC, NSCLC, lymphoplasmacytic lymphoma, melanoma, peripheral T-cell lymphoma, and squamous cell carcinoma of the penis. Conclusion: In the current era, extreme leukocytosis is often iatrogenic and associated with filgrastim or pegfilgrastim usage. WBC counts should be routinely monitored in patients receiving pegfilgrastim or filgrastim and the medication discontinued with evolving leukocytosis.

Abstract Congenital erythropoietic porphyria (CEP), a rare photomutilating disorder, is due to homozygosity or compound heterozygosity for mutant alleles encoding uroporphyrinogen III synthase (U3S), a cytosolic enzyme that converts hydroxymethyl bilane to uroporphyrinogen III. Deficiency of U3S results in an excess of biologically useless uroporphyrin I (URO I), which causes the photomutilation. Clinical expression of CEP is usually associated with co-existence of congenital dyserythropoietic anemia type I, also an autosomal recessive trait. We evaluated a 3-year-old boy of English-French extraction with a photosensitive bullous dermatosis; CEP was suspected. The boy had a hypochromic, microcytic anemia and thrombocytopenia from birth with hemoglobin values averaging 7.0 g/dL, MCV’s averaging 67 fL and platelet counts averaging 70,000/μL. A physical exam revealed scars on the face, hands and forearms, generalized hirsutism and splenomegaly. The diaper fluoresced under a Wood’s lamp. Fifty mL of urine contained 2,003 μg of URO (normal=trace), 92% of which was URO I. Erythrocyte U3S activity was 21% of controls, confirming the diagnosis of CEP. Red cell morphology was compatible with thalassemia and fluorescent red cells were apparent. Hb electrophoresis revealed 36.3% Hb A, 2.4% Hb A2, 59.5% Hb F and 1.8% of a fast moving peak, likely Hb Bart’s. Both parents were hematologically normal, had no evidence of porphyria, and had normal Hb electrophoreses. Genomic U3S was characterized (including the promoter), no mutations or deletions were found in the child or the parents. Characterization of the α and β-globin loci also revealed no mutations or deletions. Lack of a molecular explanation for either the low U3S activity or the β-thalassemia/thrombocytopenia phenotype led us to sequence GATA1, an X-linked transcription factor common to globin genes and the first 4 steps of the heme biosynthetic pathway in erythrocytes. A mutation was found at codon 216 in the child and on one allele of his mother, changing arginine to tryptophan (R216W). A more conservative arginine to glutamine mutation (R216Q) previously described by Yu et al. (Blood2002; 100:2040–5) is the only known human mutation of the DNA-binding domain of the N-finger of GATA1, and did not result in porphyria. Here, in contrast to the reduced activity of U3S, erythrocyte activities of the proximal heme biosynthetic enzymes were normal. To explore the mechanism responsible for the extremely high Hb F levels we sequenced the proximal promoter regions of the γ globin loci. The Xmn I polymorphism (−158 C to T), associated with increased production of G-γ chains, was identified in the promoter region of one G-γ allele in the child and in one allele of his father. The G-γ to A-γ ratio in the child’s Hb F was 2:1, a moderate increase in the expected ratio. Xmn I, when present with the Hb S mutation, results in variable Hb F elevations but almost always <20% (Miller et al Blood1987; 70:716–20), far below the 59.5% level observed. This is the first report of a human porphyria due to a mutation in a trans-acting factor, and the first association of CEP with thalassemia and thrombocytopenia. The Hb F level of 59.5% is not fully explained but a role for GATA1 R216W in altering globin switching is possible. The selective effect of the R216W mutation on U3S suggests that cis-elements vary in their affinity for GATA1. A bone marrow allograft corrected both the porphyria and the thalassemia.

Abstract Germline mutations in the X-linked hematopoietic transcription factor, GATA-1, have been associated with dyserythropoietic anemia (DEA), macrothrombocytopenia (MTP), and congenital erythropoietic porphyria. Recently, morphological features suggestive of the Gray Platelet Syndrome (GPS) were described in a pedigree with a germline R216Q GATA-1 mutation (Blood106:6a, 2005). The present study has evaluated platelets (Pl) in the electron microscope from members of a previously described pedigree (Blood98:2681–2688, 2001) with GATA-1 G208S, MTP, and dyserythropoiesis without anemia, for whom detailed platelet morphology studies have not been reported. The presence of a hemizygous GATA-1 mutation was confirmed by conventional fluorescent dye chemistry sequencing of DNA from patient peripheral blood mononuclear cells. Their Pl revealed wide variations in size, shape, internal and external structure. Some normal sized and giant Pl contained usual numbers of alpha granules and dense bodies, while most of the large cells were hypogranular. Some contained no membrane systems or organelles. Others were filled with masses of dense tubular system (DTS) channels, and others contained the tubular inclusions found in the Medich Giant Platelets Disorder (Platelets15:345–353, 2004). Many of the hypogranular cells contained small vacuoles that may have been enclosing membranes of alpha granules. Except for the tubular inclusions, all of these structural variations are observed in GPS Pl (Am. J. Pathol.95:445–462, 1979). However, additional striking abnormalities were seen in GATA-1 Pl that are not characteristic of GPS Pl. Many GATA-1 Pl contained unusual, closely associated dense double membranes differing from normal elements of rough endoplasmic reticulum, smooth endoplasmic reticulum, DTS or surface connected open canalicular system (OCS), but resembling channels of the OCS after exposure to EDTA. They were observed previously only in megakaryocytes (Mk) from one family with GATA-1 DEA-MTP (Nat. Genet, 24:266–270, 2000). The dense double membranes were often found in parallel association and, in some examples, isolating areas of cytoplasm. However, the isolated areas were not undergoing autophagic degradation as such areas enclosed by elements of the DTS do in White Platelet Syndrome Pl (Platelets 15:173,2004). Instead, they proved to be another unique feature of GATA-1 platelets, the presence of Pl within Pl. The sequestered cells were usually mature in appearance and often discoid in form with circumferential coils of microtubules. In some GATA-1 Pl there were two Pl within the same cell, and on rare occasions, a Pl within a Pl within a Pl. Pl within Pl have never been observed previously in any human platelet disorder. Another unique feature, related to Pl within Pl, was the frequent Pl to Pl surface attachment of non-activated cells forming large macrothrombocytes with no similarity to aggregates of stimulated normal platelets. Conclusion: The substructural abnormalities of Pl from patients with germline GATA-1 mutations share some overlap with those seen in GPS Pl, but are distinct and unique. The MTP, Pl within Pl and surface Pl to Pl attachments suggest a major defect in the formation and separation of GATA-1 Pl from pro-Pl of the parent Mk, resulting in the MTP characteristic of the disorder.

Abstract Background: Coagulation factor V (FV) is an essential cofactor of the prothrombinase complex, catalyzing conversion of prothrombin to thrombin. Severe FV deficiency is a rare coagulation defect with a prevalence in the general population of approximately 1 in 1,000,000 and a variable bleeding risk (Asselta R et al. J Thromb Haemostas 2006). Genetic mutations underlying a FV-deficient hemorrhagic diathesis have been described in a limited number of cases, display considerable allelic heterogeneity, and have been due to point mutations in the FV gene located at chromosome 1q23. Here, we report a 15 year-old girl with psychomotor delay, CNS abnormalities, and a solitary left kidney - features consistent with a congenital 1q deletion including band 1q23 - who was found to have severely reduced plasma FV activity after a routine preoperative laboratory evaluation prior to an elective spinal stabilization surgery revealed a mildly prolonged prothrombin (PT, 13.2 seconds; normal 8.4 – 12 seconds) and activated partial thromboplastin (aPTT, and 38 seconds; normal 21 – 33 seconds) time. She had undergone minor surgical procedures in the past without reports of significant bleeding, and her family members were also clinically unaffected. Methods: FV procoagulant activity level was measured in the patient with an MDA-180 automated coagulation analyser (Trinity Biotech, Ireland). The patient’s peripheral blood leukocytes were separated by density centrifugation and fixed in methanol:glacial acetic acid (2:1). We then performed interphase fluorescent in-situ (FISH) analysis using probes covering the FV gene and the more telomeric HPC1 gene as a control. Genomic DNA was also isolated from the patient, and the coding region (25 exons) of the FV gene amplified by polymerase chain reaction (PCR) and sequenced. Results: FV activity assays revealed a severely reduced level of 9% (normal range: 60–130%). Factor V activity levels in the patient’s family members were not assessed. Peripheral blood karyotyping revealed an interstitial deletion of 1q: 46,XX,del(1)(q24.2q25.3 or q23.3q25.1). Interphase FISH analysis was performed. Red signals corresponded to the FV probe and green signals corresponded to HPC1, which is approximately 13 MB telomeric to FV. In contrast to healthy controls, the patient had a signal pattern of only 1 green HPC1 signal and 1 red FV signal, corresponding to heterozygous loss of the FV gene and adjacent chromosomal material. Sequencing of the retained allele revealed a novel hemizygous S234W mutation (encoded by exon 6). The lost serine is part of a conserved “SGP” peptide segment predicted to be a solvent-exposed region of the heavy chain of the FV protein. Conclusion: Most mutations causing FV deficiency are autosomal recessive and introduce frame shifts resulting in the generation of a premature termination codon. Karyotypic and FISH analysis of this patient revealed a novel mechanism for FV deficiency: loss of an entire FV allele as part of a chromosomal deletion, coupled with a point mutation in the other allele. Structural modeling of the consequences of the S234W point mutation, which is predicted to alter the hydrophobicity of the protein, is ongoing. Syndromes involving an interstitial deletion of 1q have been described in association with ATIII deficiency (Pallotta, Dalpra et al. 2001), but to our knowledge, this is the first description of a 1q deletion with FV deficiency.

Abstract Background: It is widely accepted that an elevated serum endogenous erythropoietin (EPO) level in a patient presenting with erythrocytosis and a genuinely increased red cell mass makes a diagnosis of polycythemia vera (PV) extremely unlikely (Mossuz et al Haematologica2004:1194; Tefferi and Gilliland Mayo Clin Proc2005:947.) However, in the absence of a definitive molecular marker for PV, there is often uncertainty, especially when PV-associated clinical features are present. Here we describe 4 patients presenting with Budd-Chiari syndrome (BCS) and erythrocytosis who had EPO levels well above the normal reference range (Levy et al Hepatology1985:858), yet all had the newly described JAK2 V617F point mutation in myeloid cells, confirming PV. Methods and Results: A 30-year-old man presented in March 2005 with ascites, mild splenomegaly, abnormal liver tests, and erythrocytosis (hemoglobin (Hb) 18.6 g/dL) with a normal platelet count (187 x 10(9)/L). BCS was diagnosed by Doppler ultrasonography, transjugular venogram, and liver biopsy. While bone marrow examination showed hypercellularity and megakaryocyte clustering consistent with chronic myeloproliferative disorder (CMPD), his EPO level was elevated to 32 U/L (normal 4–16 U/L), suggesting secondary erythrocytosis. Arterial oxygen saturation, Hb oxygen-dissociation curves (P50), molecular studies of the globin loci, Hb chromatography, and CT scans of the body were all normal. Granulocyte DNA demonstrated the JAK2 V617F mutation (heterozygous/mixed clonality), absent in buccal cells, confirming PV. Karyotype was normal and BCR/ABL was not present. The patient was treated with phlebotomy, warfarin, and hydroxyurea. One month later, his EPO level was markedly reduced at 13 U/L. We hypothesized that since hepatocytes express low levels of EPO, a remnant of fetal erythropoiesis, acute necrosis from BCS could lead to transient EPO surge even in the presence of chronic EPO suppression from PV. We then reviewed clinical data from all 33 patients at our institution since 1995 with BCS that was ultimately believed to be associated with PV. We identified 13 such patients in whom endogenous EPO levels were measured. Most had a low EPO level at the time of diagnosis (range of <1.5 to 9 U/L), but 3 of these patients were found to have an elevated EPO level. DNA was extracted from archival samples from these 3 patients; JAK2 V617F mutation was detected in all 3. Clinical presentations were as follows: Patient 1: 21-year-old woman with BCS and pulmonary embolus, Hb 15.9 mg/dL, EPO 28 U/L. Bone marrow biopsy was consistent with PV. Subsequent EPO level was 19U/L in the setting of progressive liver failure. Patient 2: 48-year-old woman with known PV receiving intermittent phlebotomy presented with BCS, Hb 17.2 mg/dL, EPO 25 U/L. Marrow was consistent with CMPD. Patient 3: 33-year-old woman presented with BCS, Hb 18.8 mg/dL, EPO 53 U/L. Bone marrow biopsy showed early CMPD. Conclusion: Elevated EPO levels do not exclude the diagnosis of PV, especially in the presence of BCS with hepatic necrosis. These cases highlight the continuing challenges in accurate PV and CMPD diagnosis and emphasize the importance of complementing clinical assessment with molecular genetic testing, including JAK2 mutation analysis.

Abstract Background: Erythropoiesis stimulating agents (ESAs) consistently have been shown to decrease transfusions in anemic oncology patients. However, whether they increase mortality in cancer patients is under debate. Results from individual studies conflict, and results from literature based meta-analyses are inconclusive. We conducted a meta-analysis based on individual patient data (IPD) from all available randomized controlled trials (RCTs). Meta-analyses with individual patient data offer several advantages over study-level analysis, including the ability to gain statistical power and increase validity using time-to-event analyses, to adjust for prognostic variables that may have confounded the original treatment comparisons and to investigate subgroups in which treatment may be either more or less effective or harmful. Methods: An international collaboration conducted an individual patent data meta-analysis of ESA effects on mortality in cancer patients. With guidance from an independent steering committee of international experts in hematology, oncology, radiotherapy, epidemiology, medical statistics and a consumer representative, we developed a detailed protocol and statistical analysis plan. Independent RCT investigators and representatives from pharmaceutical companies who submitted data provided additional input through the project’s advisory board. IPD from RCTs of ESA plus red blood cell transfusion (RBCT) (as needed) versus placebo or no ESA plus RBCT (as needed), for prophylaxis or treatment of anemia in cancer patients with or without concurrent antineoplastic therapy, were included. Hazard ratios and 95% confidence intervals (CIs) were calculated per study and meta-analyzed with fixed-effects and random-effects models. Primary endpoints were overall survival (during active study phase and for the longest follow-up available) for patients receiving chemotherapy, and for all cancer patients regardless of anticancer treatment. Stratified multivariable Cox-regression analyses were conducted to assess the impact of baseline imbalances and to identify potential effect modifiers. Duplicate main statistical analyses were conducted independently at two academic statistical departments. Results: Data on 13933 patients enrolled in 53 studies were included in the analysis. Data were provided by the companies Amgen Inc., Johnson & Johnson Pharmaceutical Research & Development, L.L.C., and F. Hoffmann-La Roche Ltd.; and by five independent investigators. Results are currently undergoing internal verification and final evaluation and will be presented at the meeting. Conclusion: Final conclusions will be presented at the meeting. Future analyses using IPD will be conducted to estimate the risks (clots, tumor progression) and potential benefits (transfusion needs and quality of life/fatigue) from other outcomes.

Abstract INTRODUCTION: The majority of patients (pts) with MDS require red blood cell transfusions, which can result in iron overload and its clinical sequelae. The US03 trial is designed to evaluate the long-term efficacy and safety of the once-daily, oral iron chelator, deferasirox (DFX), in pts with lower-risk MDS. In this ongoing study 93/176 pts have now completed 12 months of treatment. METHODS: US03 is a Phase II, open-label, 3-year trial in pts with Low- or Int-1 IPSS-risk MDS and transfusional iron overload (serum ferritin [SF] 1000 μg/L and >20 units RBC transfusions), with serum creatinine (SCr) 2-fold the upper limit of normal (ULN). Initial DFX dose was 20 mg/kg/day and could be increased to 40 mg/kg/day based on tolerability and response. SF was monitored monthly; labile plasma iron (LPI), the reactive species of non-transferrin-bound iron, was assessed quarterly. BASELINE FEATURES: 176 pts were enrolled at 45 centers. Mean age: 70 years (range 21–90); 105 men and 71 women; IPSS Low risk (n=47; 27%); Int-1 (n=126; 72%); other (n=3; 2%). Mean baseline iron status was: SF, 3397 μg/L (863–36,280); LPI, 0.4 μmol/L (0.0–3.6). Forty-one percent of pts had elevated LPI at baseline (≥0.5 μmol/L). Mean number of lifetime prior transfusions: 63; mean duration of transfusions: 3.5 years (0–34). MDS therapy at study entry included azacitidine, hydroxyurea, lenalidomide, thalidomide or decitabine in 25 pts, and growth factors in 48 pts. Calculated creatinine clearance was normal (>80 mL/min) in 77 pts and abnormal (abn) in 99: mildly abn (51– 80 mL/min) in 71; moderately abn (30–50 mL/min) in 25; severely abn (<30 mL/min) in 3. RESULTS: Over 12 months, the mean dose of DFX was 21 mg/kg/day, and the mean transfusion rate was 3.4 units/month. Mean SF±SEM (μg/L) values at baseline, 3, 6, 9 and 12 months were 3397±233 (n=176), 3057±144 (n=143), 2802±128 (n=126), 2635±148 (n=109) and 2501±139 (n=93), respectively. In pts with elevated baseline LPI, sustained suppression of mean LPI to the normal range was achieved after 3 months of treatment. The figures show reductions (±SEM) in SF, and in LPI in pts with baseline LPI ≥0.5 μmol/L. Hematological improvement by IWG 2000 criteria was achieved in 8 pts (5%): erythroid response in 5 (major 3; minor 2); platelet in 1 (major); neutrophil in 1 (major); and combined platelet and neutrophil in 1. SAFETY: Of 173 pts, 18 (10%) discontinued DFX because of suspected adverse events (AEs), and an additional 5 (3%) because of serious AEs (SAEs). The most common AEs were diarrhea (n=9), rash (n=3) and nausea (n=2), while the most common SAEs were rash (n=2) and diarrhea (n=1). Of 147 pts with normal baseline SCr, 26 (18%) increased >ULN on at least two occasions (3.0 mg/dL max SCr). There were 9/172 (5%) and 22/168 (13%) new onset cases of grade 3–4 thrombocytopenia and neutropenia, respectively, none suspected to be related to DFX; 7 and 11 of these pts, respectively, were receiving other MDS treatment therapies, including lenalidomide, decitabine, hydroxyurea and azacitidine. There were 17 deaths (10%), due to sepsis/infection (n=8), disease progression (n=4), intracranial bleed (n=2), cardio-respiratory arrest (n=2) and renal failure (n=1), all thought to be unrelated to DFX. CONCLUSIONS: In these heavily iron-overloaded pts, DFX was generally well tolerated. New onset cytopenias were consistent with the underlying disease. A 2-year extension phase of this study will assess the long-term safety and efficacy of DFX as well as the clinical impact on cardiac, hepatic and endocrine function. Figure Figure