Myelodysplastic Syndromes

Christopher R. Cogle

Rami S. Komrokji

Alan F. List

The myelodysplastic syndromes (MDS) comprise a spectrum of neoplastic hematopoietic stem cell disorders that share the features of ineffective and dysplastic hematopoiesis accompanied by peripheral blood cytopenias. Subclassifications of MDS can be made based on differences in hematopoietic lineage affected (e.g., erythroid, myeloid, and megakaryocytic), percentage of bone marrow myeloblasts, and molecular abnormalities. The incidence of MDS in the United States is estimated at 3.3 per 100,000.¹^,² MDS most often affects older individuals. In the population >60 years of age, the incidence of MDS is estimated at 50 per 100,000. Predisposing causes of MDS include prior exposure to organic solvents such as benzenes, ionizing irradiation, chemotherapy, and tobacco smoking.³ Most cases of MDS occur de novo with a rising fraction secondary to prior therapy such as chemotherapy (particularly alkylating agents) and radiation therapy. The natural history of MDS is highly variable and likely related to the myriad of genetic, epigenetic, and microenvironmental changes that characterize each individual’s disease. In patients who develop therapy-related MDS, the disease is often characterized by adverse cytogenetic abnormalities, an aggressive clinical course, and rapid transformation into acute myeloid leukemia (AML). Approximately 30% of MDS cases progress to AML, and for that reason, this syndrome has been historically regarded as a preleukemia. However, most patients with MDS succumb to infection and other complications related to cytopenias.⁴

PATHOGENESIS

MDS arises from genetic or epigenetic changes in hematopoietic stem cells or as a result of aberrant growth of a stem cell under adverse microenvironmental pressures.⁵^,⁶^,⁷ Subsequent DNA damage events and immunologic responses that promote survival of the dysplastic stem/progenitor cell and clonal dominance may occur.⁶^,⁸ Genomic instability, because of an accumulation of genomic damage and failure to repair these sites of damage, is a hallmark feature of MDS pathology. Another factor contributing to genomic instability is age-related telomere erosion arising from senescence-related DNA replication stress and impairment in telomere repair.⁹ Early in the disease course, the MDS bone marrow reveals impaired differentiation and increased apoptosis.¹⁰ As the MDS progresses, the bone marrow shows increased cell proliferation and accumulation of immature stem/progenitor cells. Eventually, a clone of mutated stem/progenitor cells loses responsiveness to apoptosis signaling, resulting in accumulation of neoplastic myeloblasts associated with a diagnosis of leukemia.

In contrast to leukemias like chronic myeloid leukemia and acute promyelocytic leukemia (APL), where a distinguishing genetic mutation accounts for pathogenesis of the disease, MDS can present with variable and heterogeneous genetic or epigenetic changes. This variability is likely the reason for the syndrome’s heterogeneity in clinical presentation and bone marrow pathology. In MDS, certain chromosomal abnormalities present more frequently. Among the most common include structural or numerical deletions of chromosomes 5, 7, and 20, and in men, the deletion of Y chromosome. Identifying cytogenetic abnormalities in MDS is a key determinant of diagnosis and prognosis, providing genetic confirmation of diagnosis, predicting tempo of disease progression, and assisting in the selection of appropriate therapeutics. Unfortunately, because of technical limitations of commercially available metaphase cytogenetics (i.e., Giemsa band staining of whole chromosomes), approximately 30% to 60% of MDS patients have normal karyotype. This group of normal cytogenetic MDS patients is clinically heterogeneous in presentation, disease behavior, and response to therapy, which likely reflects the heterogeneous changes at the molecular level.

With the majority of MDS patients showing no cytogenetic alteration in their MDS clone, candidate gene approaches have been employed to search for recurring gene mutations. Investigations of select genes (e.g., RUNX1, NPM-1, FLT3, TP53, and rat sarcoma (Ras) pathway) have found gene mutations in MDS patients with normal karyotype; however, these select mutations were present in only a minority (<5%) of patients. By studying the whole genome of MDS cells using single-nucleotide polymorphism (SNP) array and genome sequencing, investigators found acquired deletions, and missense and nonsense mutations in TET2 in 26% of MDS patients, representing the most frequently reported gene abnormality in MDS.¹¹ In cases where a TET2 mutation was found the abnormality was found in nearly all bone marrow cells including CD34+ stem/progenitor cells. These loss of function abnormalities of TET2 are believed to be a first hit in the pathogenesis of MDS. Functionally, TET2 is involved in regulating DNA methylation.¹² Perturbations in TET2 result in dysregulated methylation (both hypomethylation and hypermethylation). Thus, it has been proposed that treatment of TET2 mutant MDS with hypomethylating agents may bring about improved outcomes. An early study corroborates this hypothesis by showing higher complete remission (CR) rates after azacitidine in MDS patients with TET2 mutation compared with those with wild type TET2 (65% vs. 30%, P = .01),¹³ although, there was no difference in survival detected.

The 5q-syndrome is a discrete clinical and pathologic subset of MDS characterized by an isolated chromosome 5q deletion, a hypoplastic macrocytic anemia, normal to elevated platelet count and pathologically characterized by the presence of hypolobulated megakaryocytes. In del(5q) MDS, there is a commonly deleted region (CDR) on chromosome 5; a 1.5 Mb interstitial segment that extends between 5q31 and 5q32 that contains 44 candidate genes.¹⁴ Using a functional RNA interference screen for each of the genes in the CDR, only knock down of the RPS14 gene yielded an erythroid phenotype similar to the 5q-syndrome. RPS14 is a ribosomal protein which is a component of the structural framework of the 40s ribosomal subunit. Interestingly, mutational inactivation of RPS19, another gene involved in ribosomal biogenesis, is implicated in the pathogenesis of Diamond-Blackfan anemia. Evidence from animal models shows that impaired ribosome biogenesis leads to p53 stabilization and activation that contributes to the hypoplastic anemia and increased apoptosis of erythroid precursors in del(5q) MDS.

Epigenetic changes in MDS, such as aberrancies in DNA methylation and/or histone acetylation, are also believed to contribute to the pathogenesis of MDS. Increased methylation of CpG dinucleotides concentrated in the promoter regions of some genes (also known as CpG islands) leads to silencing of the gene. In MDS, increased methylation of CpG islands has been reported with more advanced MDS displaying greater methylation than lower risk disease. DNA transcription can also be affected by how condensed the DNA is wrapped around histone cores. When histone cores are acetylated, the negatively charged acetyl groups neutralize the positive charge on the histone proteins, resulting in the release of the negatively charged DNA. As a consequence, the condensed chromatin relaxes and permits higher activity of gene transcription. A balance exists between histone acetyltransferases (HAT), which open gene transcription, and histone deacetylases (HDAC), which close gene transcription. Abnormalities in HAT and HDAC have been reported in patients with MDS.¹⁵^,¹⁶

The MDS microenvironment also contributes to the pathobiology of disease. The abnormal bone marrow microenvironment is believed to promote the survival and proliferation of neoplastic MDS cells. Perturbations of several cytokines in bone marrow and peripheral blood have been reported, including IL-6, IL-8, stem cell factor, erythropoietin, TGF-/β, and TNF-α<x¹⁷^,¹⁸^,¹⁹^,²⁰^,²¹ Interestingly, apoptosis is a prominent finding in early MDS compared with late MDS.¹⁰^,²² Moreover, apoptosis occurs in nonclonal cells, suggesting a collateral damage that impairs the bone marrow’s ability to support effective hematopoiesis.²³

Another unique aspect of the supportive MDS bone marrow microenvironment is heightened neovascularization. Patients with MDS exhibit increased blood vessel density in their bone marrow.²⁴^,²⁵^,²⁶^,²⁷ Furthermore, abnormal elevations in angiogenic cytokines such as VEGF, bFGF, and angiopoietins have been reported in MDS bone marrow.²⁵^,²⁸ Given that myelomonocytic precursors from MDS patients overexpress VEGF and VEGFR2, paracrine and autocrine loops are believed to further stimulate MDS cell proliferation while suppressing effective erythropoiesis.

There is mounting evidence that immune dysregulation plays an important role in MDS pathophysiology. First, the incidence of autoimmune disorders is increased in patients with MDS.²⁹ Additionally, MDS can share disease characteristics that overlap with aplastic anemia and large granular lymphocytic leukemia, which are known to involve autoreactive T lymphocytes. Clinically, immunosuppression with antithymocyte globulin, cyclosporine and prednisone are effective in select groups of patients with MDS. Important questions relating to the mechanisms of autoimmunity in MDS remain to be answered.

DIAGNOSIS

MDS patients typically present with symptoms or signs related to cytopenias (e.g., anemia, leukopenia, and/or thrombocytopenia) and are referred for hematology evaluation. The peripheral blood smear may show dysplastic and immature circulating peripheral blood cells (Fig. 45.1A, B). After excluding nutritional deficiencies, hemorrhage, infections, endocrine abnormalities, alcohol abuse, and heavy metal toxicity, bone marrow aspiration and biopsy are performed to assess for primary disorders in hematopoiesis. Bone marrow aspiration and biopsy are required to diagnose MDS. The bone marrow typically shows hypercellularity adjusted for the patient’s age. However, approximately 15% of patients present with a hypocellular bone marrow, indicating a potential pathogenetic overlap with aplastic anemia. A hallmark of MDS is the presence of dysplastic changes in bone marrow precursors (Fig. 45.1C, D). Dysplastic bone marrow cellular changes can include multinucleated or megaloblastoid erythropoiesis, decreased segmentation or granulation of neutrophils, oligonuclearity or nuclear dispersion in megakaryocytes, nuclear blebs, irregular nuclear contours, and nuclear-cytoplasmic dyssynchrony (such as an immature appearing nucleus associated with a mature appearing cytoplasm). Abnormalities in iron metabolism can lead to mitochondrial iron deposits around nuclei of erythroid precursors—so called ring sideroblasts (Fig. 45.1E). Dysplastic myelomonocytes are sometimes present in the bone marrow (Fig. 45.1F), and when these abnormal monocytes are present to a high degree in the peripheral blood (i.e., absolute monocyte count >1,000 per µ1), a diagnosis of chronic myelomonocytic leukemia (CMML) is made. In advanced MDS, excess myeloblasts are noted (Fig. 45.1G). There can be great diversity in cellularity, degree of dysplastic changes, and lineages affected among MDS patient bone marrow specimens. However, common to all diagnoses of MDS is at least 10% of bone marrow cells with dysplastic cytologic changes.

Initial morphologic and diagnostic classifications of MDS date to 1976, when the French-American-British classification was proposed. Since then, several revisions have been made to this classification along with the introduction of new systems, including the latest classification proposed by the World Health Organization (WHO) in 2008 (Table 45-1).³⁰ In this classification, cell morphology assessed by microscopy as well as cytogenetic information from the neoplastic MDS clone is used to categorize disease.

Despite these finite criteria, diagnosing MDS can be difficult, as morphologic changes can be subtle and nonspecific. Moreover, DNA abnormalities detected by commercially available molecular genetic techniques (i.e., Giemsa band staining
and fluorescent in situ hybridization) are uninformative in approximately half of de novo MDS and in 10% to 20% of therapy-related MDS. Thus, further molecular refinement of MDS classification is expected that will reflect the underlying disease molecular biology.

Figure 45.1 Micrographs of blood and bone marrow from patients with MDS s. A: Blood smear from a healthy individual. B: Circulating abnormal monocytes in a patient with MDS.C: Bone marrow aspirate from a healthy individual. D: Bone marrow from a patient with MDS showing dysplastic erythroid progenitors. E: Ringed sideroblasts in a patient with MDS. Arrows point to blue-stained iron deposits around nuclei of MDS cells. F: Bone marrow from a patient with CMML showing dysplastic monocytes. G: Bone marrow from a patient with MDS showing excess blasts amidst dysplastic cellular changes. H: Reticulin staining showing increased fibrosis in the bone marrow of a patient with MDS. (Ying Li, M.D.graciously provided these light micrographs.)

Table 45-1 WHO Classification System for MDS

Disease	Blood	Bone Marrow
Refractory cytopenia with unilineage dysplasia (RCUD):	Unicytopenia or Bicytopenia No or rare blasts (<1%)	Unilineage dysplasia ≤10% <5% blasts <15% of erythroid are RS
RA
Refractory neutropenia (RN)
Refractory thrombocytopenia (RT)
RARS	Anemia No blasts	≤15% of erythroid are RS Erythroid dysplasia only <5% blasts
Refractory cytopenia with multilineage dysplasia (RCMD)	Cytopenia(s) No or rare blasts (<1%) No Auer rods Absolute monocyte count <1,000	Dysplasia in ≤10% of cells in ≥2 lineages <5% blasts No Auer rods ±15% of erythroid are RS
RAEB-1	Cytopenia(s) <5% blasts No Auer rods Absolute monocytes count <1,000	Unilineage or multilineage dysplasia 5%-9% blasts No Auer rods
RAEB-2	Cytopenia(s)	Unilineage or multilineage dysplasia
	5%-9% blasts	10%-19% blasts
	Auer rods ±	Auer rods ±
	Absolute monocytes count <1,000
Myelodysplastic Syndromes Unclassifiable (MDS-U)	Cytopenias <1% blasts	Dysplasia <10% Cytogenetic abnl c/w MDS <5% blasts
MDS associated with del(5q)	Anemia Platelets: normal or increased No or rare blasts (<1%)	Megas: normal to increased <5% blasts Isolated del(5q)
		No Auer rods
Vardiman JW, Thiele J, Arber DA, et al. The 2008 revision of the World Health Organization (WHO) classification of myeloid neoplasms and acute leukemia: rationale and important changes. Blood. 2009;114(5):937-951.

PROGNOSIS

A major shortcoming of past and present classification systems for MDS is the wide variation in patient outcomes within diagnostic categories. For example, although it is understood that patients with refractory anemia with excess blasts (RAEB) have worse outcomes than patients with refractory anemia and ringed sideroblasts (RARS), there is still considerable heterogeneity of clinical outcomes among RARS patients. To accurately forecast outcomes for all MDS patients, prognostic scoring systems were developed. In 1997, evolution of prognostic tools and advancement of cytogenetic technology led to the landmark International Prognostic Scoring System (IPSS), which is the most widely used prognostic system for MDS patients (Table 45-2).³¹ This scoring system considers the number of peripheral blood cytopenias (with more lineages affected
indicating more advanced disease), enumeration of myeloblasts in the bone marrow (with more indicating advanced disease), and assessment of cytogenetic abnormalities. Based on these three disease characteristics at the time of diagnosis, an overall IPSS score can be assigned to the patient’s disease (Table 45-2). The IPSS score is important for determining the patient’s risk for reduced survival and progression to AML, with higher risk groups at increased risk of disease progression and death. The IPSS is a useful tool for determining who and when to administer disease-altering therapy.

Table 45-2 The IPSS for Patients with MDS

Score Value
	0	0.5	1	1.5	2
BM blasts	<5%	<5%-10%	–	11%-20%	21%-30%
Karyotype	Good	Intermediate	Poor
Cytopenias	0/1	2/3
Risk Assessment
Total Score	Risk Group	Time to AML (for 25%) (years)	Median Survival (years)
0	Low Risk	9.4	5.7
0.5-1.0	Intermediate-1 (Int-1)	3.3	3.5
1.5-2.0	Intermediate-2 (Int-2)	1.1	1.2
≤2.5	High Risk	0.2	0.4
Good cytogenetic abnormalities: normal, del(5q), del(20q)
Intermediate cytogenetic abnormalities: others
Poor cytogenetic abnormalities: complex (>3 abnormalities), chromosome 7 abnormalities
Greenberg P, Cox C, LeBeau MM, et al. International scoring system for evaluating prognosis in myelodysplastic syndromes. Blood. 1997;89(6):2079-2088.

Despite the utility of the IPSS, it has several important limitations. First, the IPSS was only validated in previously untreated patients with de novo MDS and nonproliferative CMML. In addition, the IPSS, which was created in 1997, included MDS patients with blasts >20% (currently assessed as AML by WHO 2008 criteria). Thus, the high-risk IPSS group may have worse outcomes by including patients with more advanced disease. Given evidence that need for transfusion is a strong predictor of poor outcome in MDS patients,³² another weakness of the IPSS is that it does not distinguish the severity of cytopenia (e.g., cytopenias so low that the patient requires transfusions vs. minimally decreased blood count with no need for transfusions). Finally, with further advancements in molecular genetic technology and better appreciation for MDS pathobiology, the limited number of cytogenetic abnormalities listed in the original IPSS (just five) are not sufficiently inclusive.

As a consequence, several next generation prognostic tools have been proposed and their validity are being tested. These newer prognostic systems build upon the WHO classification system and the 1997 IPSS by including additional disease characteristics such as need for transfusions, fibrosis in the bone marrow (Fig. 45.1H), CD34+ cell clustering, ferritin level, β-2- microglobulin level and new, recurrent DNA mutations.³³^,³⁴^,³⁵ For example, by combining WHO classification, IPSS score, and need for transfusions, the WHO-based prognostic scoring system was developed.³³ Moreover, this new tool has been validated for use throughout the MDS patient’s lifetime, allowing adjustment in disease expectation and not just limited to the time of diagnosis like the IPSS. Another new prognostic tool for MDS patients focuses on the IPSS lower-risk group (low/intermediate-1), which contains a heterogeneous group of individuals with different severities of cytopenias.³⁴ This new scoring system takes into account age, cytogenetic abnormalities, anemia, severity of thrombocytopenia and myeloblasts in the bone marrow, and stratifies lower-risk patients into those with favorable versus unfavorable prognosis. Use of this prognostic tool identifies lower-risk MDS patients who may benefit from therapies that bring about hematologic improvement (HI) alone or extend survival.

MANAGEMENT

At clinical presentation, precise disease classifications and prognostic tools are available to evaluate MDS patients. However, each individual brings his or her own unique condition. For example, sometimes bone marrow pathology may not fit cleanly in a WHO classification. Sometimes a crossover diagnosis between MDS and another disease, such as aplastic anemia or myelofibrosis, seems more appropriate than strictly limiting the diagnosis to one type. In recognition of the possibility of overlap
syndromes, the WHO classification systems have included the overlap subgroup, myelodysplastic syndrome/myeloproliferative neoplasm (MDS/MPN). This subgroup includes CMML, atypical chronic myeloid leukemia, juvenile myelomonocytic leukemia and introduced in 2008, the provisional entity, RARS with thrombocytosis. These myeloproliferative and dysplastic bone marrow conditions seem to have in common aberrancies in the as-Mitogen-Activated Protein Kinas (MAPK) signaling pathway.

Occasionally, a confounding disease like an autoimmune disease (e.g., systemic lupus erythematosis) or chronic infection (e.g., HIV) may present with cytopenias and a bone marrow consistent with WHO MDS classification. In this situation, the clinician must balance the probabilities of the correct diagnosis. More commonly, the MDS patient’s performance status and comorbid diseases place boundaries on the intensity of treatment. This is especially true in MDS where the median age at presentation is 76. Finally, interpreting treatment results from the published medical literature and applying it to the presenting MDS patient can be difficult. Nuances in clinical trial design necessitate that clinicians understand each trial’s patient selection, assignment of therapy, and criteria for response.

Indications for Treatment

In general, the indication to treat a patient with MDS is to ameliorate cytopenias, delay disease progression to leukemia, and extend survival. This is based on recognition that most MDS patients will succumb to infection or bleeding, and those that avoid these life-threatening complications with higher-risk disease will ultimately experience disease progression to AML. When to initiate treatment depends on risk category. For example, for patients with IPSS low or intermediate-1 (Int-1) risk MDS and no severe cytopenias, initiation of therapy may not be needed for several years. The treatment goal in this situation is hematologic improvements (HIs) and preventing or limiting complications associated with chronic cytopenias. In contrast, for patients with higher-risk MDS (Int-2/high IPSS risk) and transfusion dependence, initiation of therapy typically occurs as early as possible. The treatment goal in this situation is to alter the natural history of the disease and extend survival. This risk-adapted approach intends to recognize who will best benefit from therapy, while minimizing unnecessary exposure to treatment-related toxicities.

Principles of Treatment Selection

Because of the heterogeneity of MDS, choice of initial therapy is based on several factors, such as chromosomal or molecular abnormalities, disease risk category, patient comorbidities, and performance status. Patient age may also be considered but is often a surrogate for comorbidities and organ reserve.

Consensus Guidelines

Three expert working groups have convened to define MDS treatment guidelines for practicing clinicians: the National Cancer Center Network (NCCN) in the United States, the British Committee for Standards in Haematology (BCSH) in the United Kingdom, and the Italian Society of Hematology (SIE) in Italy. Each working group employs different methods to derive their recommendations.

The NCCN/US guidelines are developed by consensus of expert opinion and quality of data, base treatment decisions on IPSS risk category, and present choices as flow charts.³⁶ The BCSH/UK guidelines are based on consensus and systematic literature review, base treatment on disease risk category (IPSS or Sanz), and also present choices as flow charts.³⁷ The SIE/Italian guidelines are constructed from consensus of experts and systematic literature review, and are presented as graded responses to relevant questions in MDS management.³⁸

In general, the three working groups present similar recommendations for the diagnosis and management of MDS. In particular, the proposed initial evaluation and classification schemas are similar. For patients with lower-risk MDS, supportive care and growth factor support are recommended. For patients with intermediate-2 and high IPSS risk MDS, allogeneic hematopoietic cell transplantation (if donor available) and azanucleoside therapies are recommended. For patients with low and intermediate-1 IPSS risk MDS with a sole deletion of chromosome 5q, lenalidomide is recommended. The groups also recommend consideration of immunosuppressive therapy for patients <60 years, those with hypoplastic MDS, MDS with PNH, and MDS with HLA-DR15 phenotype. Also, common among the groups is the recommendation to consider iron chelation in low and intermediate-1 IPSS risk MDS patients.

Despite the mostly consistent recommendations among the groups, there are some differences. The BCSH/UK guidelines have not been updated since 2003, thus lack recommendations based on recent studies such as azanucleoside and immunomodulatory agents. Whereas the NCCN/US makes a preferred recommendation for 5-azacitidine (AZA, Vidaza) over 2′-deoxy-5-azacitidine (decitabine, DAC, Dacogen) based on survival data for AZA in patients with higher-risk disease,³⁹ the SIE/Italian group only recommends AZA and makes no recommendation for DAC. The groups also differ in first choice of iron chelating agent: the NCCN/US panel recommends subcutaneous deferoxamine (DFO), whereas the SIE/Italian group recommends oral deferasirox (Exjade) as the first-choice drug. In addition, the SIE/Italian group recommends using ferritin levels not as a decisive parameter to start chelation: rather as a marker of efficacy in the follow-up of chelation therapy.

Notably, where the groups cannot make a recommendation, largely because of lack of data, is whether a remission should be achieved in patients who intend to receive a transplant.

Response Criteria

Incorporating new therapeutic agents into the management of MDS should be based on convincing evidence of the intervention’s efficacy balanced by its safety. Moreover, when comparing therapeutic agents, similar response criteria are required for valid examination. Thus, precision in describing disease response is of crucial importance. Toward this end, an International Working Group (IWG) has created standardized
response criteria for MDS published originally in 2000 and revised in 2006.⁴⁰^,⁴¹ These criteria include responses relative to altering natural history of disease, cytogenetic responses, HIs, and quality of life measures (Table 45-3).

TREATMENT

Supportive Care

Nearly all patients with MDS require supportive care. For many MDS patients who are older in age and suffer from comorbidities, supportive care may be the only suitable management option. Supportive care usually refers to replacement of blood components (RBCs and platelets) in symptomatic patients, prevention and treatment of infection, and prevention and treatment of transfusional iron overload. The goal of supportive care is to improve quality of life and prolong survival. Transfusions and antimicrobials do not alter the natural history of MDS. However, these interventions do address life-threatening complications associated with MDS. Given evidence that transfusion requirement and iron overload are associated with reduced survival in MDS,³³^,⁴²^,⁴³ there is some evidence that iron chelation therapy may have a disease-modifying effect.⁴⁴^,⁴⁵^,⁴⁶

The major risks of blood product transfusions include febrile nonhemolytic transfusion reactions, hemolytic reactions, transfusion associated acute lung injury, bacterial infections, viral infections, volume overload, transfusion associated graft-versus-host disease (GVHD; in severely immunocompromised individuals), and iron overload. With repeated blood product replacements, patients can become alloimmunized against blood products and consequently become refractory to transfusions, in particular, platelet products. For those considered candidates for allogeneic hematopoietic cell transplantation, increased number of transfusions prior to transplant is a major risk factor for graft rejection and post-transplant survival. For these reasons, blood products for MDS patients should be leukoreduced and irradiated to reduce risks for immunologically mediated side effects and reactivated viral infections.

Table 45-3 MDS International Working Group 2006 Response Criteria

	Disease Remission		Hematologic Improvements (HI)
Category	Response (at least 4 wk)	Category	Response (at least 8 wk)
CR	BM ≤5% blasts, may have persistent dysplastic changes in marrow, normal CBC	Erythroid response (pre-rx Hgb <11)	Hgb increase by ≥1.5, RBC tx reduction by 4 units/8 wk
PR	Blasts decreased by ≥50%, but BM still has ≥5% blasts	Platelet response (pre-rx PLT <100 K)	If starting with >20 K, then increase by ≥30 K; If starting with <20 K, then >100% increase (and over 20 K)
Marrow CR	BM blasts ≤5% and decreased by ≥50%, if HI then noted individually
SD	No PR, but progression in >8 wk
Failure	Progression to a more advanced class, worsened counts, death because of MDS	Neutrophil response (pre-rx ANC <100)	At least 100% increase and absolute increase of 500
Relapse	Return to pre-rx blasts, grans or plts decrease by ≥ 50%, Hgb decrease by ≥ 1.5 or tx dependent	Progression or relapse after HI	At least 1 of the following:
Cytogenetic response	Complete: disappearance of abnormality		• At least 50% decrease from max response (ANC and PLTS)
	Partial: at least 50% reduction in abnormality		• Reduction of Hgb by 1.5
Disease progression	≥50% increase in blasts, decreased counts by ≥50%, tx dependent		• Transfusion dependence
Cheson BD, Greenberg PL, Bennett JM, et al. Clinical application and proposal for modification of the International Working Group (IWG) response criteria in myelodysplasia. Blood. 2006;108(2):419-425.

Iron overload is a major risk of excessive RBC transfusions in patients with MDS. Because there is no active mechanism to eliminate iron from the body, iron toxicity can occur after as few as 20 units of RBCs.⁴⁷ When transferrin binding capacity is exceeded, free iron then deposits in tissues such as the liver, pancreas, and heart and generates hydroxyl and oxygen-free radicals, which causes cellular toxicity and organ damage.⁴⁸^,⁴⁹^,⁵⁰^,⁵¹ Serum ferritin, which serves to store and transport iron in a nontoxic form, is elevated in situations of high-iron levels and is grossly proportional to total body iron load.⁵² In patients with lower-risk MDS, high-serum ferritin (>1,000 µg per ml) has been associated with an incremental reduction in survival owing to hepatic insufficiency, diabetes mellitus (pancreatic insufficiency), gonadal failure, cardiac failure, and progression to AML.³⁴^,⁵³^,⁵⁴ Accordingly, guidelines from the MDS Foundation’s Working Group on Transfusional Iron Overload define iron overload as serum ferritin levels consistently ≥ 1,000 µg per ml.⁵⁵ It should be noted that serum ferritin is also an acute phase reactant produced by the liver, and therefore, may not be highly specific for iron overload in situations of inflammation, hepatitis, and liver damage. Other methods to measure iron load, such as MRI, are under investigation.

As iron overload adversely impacts survival in MDS, strategies to chelate iron in transfusion-dependent patients are important. Retrospective studies indicate that iron chelation improves survival in transfusion-dependent, lower-risk MDS patients.⁴⁶ To eliminate the potential for selection bias in these retrospective studies, prospective randomized trials are now underway to better define the impact of iron chelation on survival. Generally, iron chelation is considered in patients who receive ≥20 units of RBCs annually and if serum ferritin levels are consistently elevated > 1,000 ng per ml, with the goal of maintaining serum ferritin levels between 500 and 1,000 ng per ml.⁵⁵^,⁵⁶ It should be noted that these thresholds are based on limited data and inferred from studies involving patients with thalassemia major. There is still no widespread consensus regarding lower thresholds, and they have yet to be validated in MDS.

There is limited data on the clinical benefit of iron chelation therapy in patients with MDS. In a retrospective collection of MDS cases receiving subcutaneous DFO (Desferal) for up to 5 years, iron chelation reduced transfusion need by half in nearly two-thirds of patients, and neutrophil and platelet counts improved in 78% and 64% of patients, respectively.⁴⁴ Serum ferritin decreased in 82% of patients. In addition, improvements in cardiac iron content were noted after DFO chelation.⁵⁷ In terms of survival, two retrospective studies provide evidence of improvement. In the first, investigators from Vancouver conducted a review of 178 MDS patients and identified 18 patients as having received DFO by subcutaneous pump and 18 case-matched controls.⁵⁸ For patients with lower-risk MDS (low/Int-1 IPSS risk), the chances of survival at 4 years was higher in those receiving iron chelation therapy (64% vs. 43%, P = .003). In a multivariate analysis, case characteristics associated with improved overall survival included iron chelation therapy (P = .02, HR 0.1 [0.01 to 1.0]), IPSS score (P = .008). In a second retrospective study, investigators from the Groupe Francophone des Myelodysplasies compared the clinical outcomes of three groups of MDS patients based on dose of iron chelation therapy: regularly scheduled infusions ≥ 3 times per week (standard chelation), intermittent bolus (low chelation), and those who did not receive iron chelation therapy.⁵⁹ In this series of 170 patients, of which 59% had low-risk MDS, median survival was longer in those receiving any kind of iron chelation even after adjusting for prognostic factors unequally distributed between groups (115 vs. 51 months, P = .0001). Interestingly, the investigation also detected improved survival in those receiving higher doses of iron chelation therapy (median survival 120 months in standard chelation group vs. 69 months in low chelation group, P < .001), suggesting a dosedependent effect of iron chelation on survival. Prospective trials are currently accruing.

Although DFO can be absorbed orally, its half-life is very short, and for this reason, it is administered via continuous intravenous or subcutaneous infusion. DFO binds iron and is excreted in the urine and bile. Dosing of DFO is typically 50 mg per kg delivered by subcutaneous infusion over 8 to 12 hours each night. Dosing is kept below 2.5 g to avoid side effects such as auditory loss. This regimen typically results in an iron loss of 20 to 50 mg per day, which monthly is roughly equivalent to the cumulative amount of 2 to 6 units of RBCs each month. An alternative administration for patients with cardiac arrhythmias and heart failure or those who do not tolerate subcutaneous injections is continuous 24-hour DFO via a central venous catheter. Patients should be monitored closely for toxicities involving the eyes, ears, and kidneys.

Recently once-daily oral deferasirox (Exjade) was approved by the FDA for the treatment of transfusional iron overload. Doses of 10 to 30 mg/kg/day resulted in significant reductions in serum ferritin levels.⁶⁰^,⁶¹^,⁶² The reductions in serum ferritin are dosedependent, irrespective of whether patients were chelation-naïve or previously chelated, and correlated with liver iron content as measured by MRI. In patients with evidence of hepatic toxicity due to iron overload, serum alanine aminotransferase levels decreased significantly.⁶² Unfortunately, approximately 50% of MDS patients had to discontinue oral deferasirox due to side effects such as abdominal pain, diarrhea, increased serum creatinine, nausea, vomiting, and rash. Moreover, the US FDA released after-market safety information for deferasirox including the added risks of renal insufficiency and failure, hepatic insufficiency and failure, and gastrointestinal hemorrhage in patients with marked thrombocytopenia. Thus, close monitoring of renal, hepatic and gastrointestinal systems is highly recommended, especially in patients ≥60 years. Dose adjustments may be necessary to avoid treatment-related toxicities.

Erythropoietic Stimulating Agents

Erythropoietic stimulating agents (ESAs) are usually considered the first step for management of anemia in lower-risk MDS (Low/Int-1 IPSS risk) by most guidelines. ESAs reduce transfusion requirement in 15% to 20% of unselected MDS patients who receive epoetin alpha (EPO) equivalent doses of 40,000 to 80,000 units weekly.⁶³ The addition of a myeloid
growth factor such as granulocyte-colony stimulating factor (G-CSF) improves the erythroid response rate to 40%.

The Nordic MDS Group published a meta-analysis of their clinical trials involving a total of 205 patients with MDS treated with EPO.⁶³ Thirty-three patients (16%) experienced an erythroid response to treatment. Patients with EARS had a significantly lower response rate than all other morphologic categories (7.5% vs. 21%, P = .01), as did patients who required transfusions compared with patients without the need for transfusion (10% vs. 44%, P < .001). The endogenous, serum erythropoietin level was significantly lower in responding patients and was inversely proportional to response rate.

Only gold members can continue reading. Log In or Register to continue