What every physician needs to know:
The myelodysplastic syndromes (MDS) include a heterogeneous group of clonal bone marrow failure syndromes characterized by cytopenias, clonally restricted hematopoiesis (associated with an abnormal G-banded metaphase karyotype in about 50% of cases), genomic instability, and a risk of progression to acute myeloid leukemia (AML).
More than 85% of MDS cases are idiopathic or “de novo”, while 10-15% are secondary to a recognized prior exposure to a DNA damaging agent such as alkylating agent / topoisomerase inhibitor chemotherapy, or ionizing radiation.
Although MDS were once known as “preleukemia”, only about 25-30% of patients ever progress to AML. About 50% die of complications of cytopenias – most commonly infection, followed by hemorrhage – and about 25% die of other conditions, succumbing with MDS rather than from MDS.
While MDS are most commonly found in older patients (median age at diagnosis, 71 years), MDS can be diagnosed at any age, including childhood. Pediatric MDS cases often present with pancytopenia and may be a consequence of a germline DNA repair defect.
Allogeneic hematopoietic stem cell transplantation is the only curative therapy for MDS, but currently <10% of patients undergo transplant due to advanced age, comorbid conditions, or lack of a suitable donor. All other treatments for MDS are only palliative in nature, although some therapies can extend survival and improve quality of life.
Are you sure your patient has a myelodysplastic syndrome? What should you expect to find?
Almost all patients with MDS (>95%) present with anemia; 30-50% also have neutropenia or thrombocytopenia at diagnosis. The prevalence of cytopenias increases with higher risk disease and with time from initial diagnosis.
Anemia associated with MDS is typically macrocytic or normocytic; microcytosis is rare and suggests concomitant thalassemia (acquired or congenital) or iron deficiency.
It’s also rare to have isolated thrombocytopenia without anemia in MDS, so other causes (such as immune thrombocytopenia or a drug effect) should be ruled out if a patient presents with isolated thrombocytopenia and no anemia or neutropenia.
The minimal diagnostic criteria for MDS proposed by an international working group include the presence of at least one meaningful peripheral blood cytopenia (Hb <11 g/dL, ANC <1500/uL, or platelet count <100 x 109/L) of at least 6 months duration, plus at least one “decisive criterion”. The 4 decisive criteria are:
>10% dysplastic cells in one or more cell lineages (isolated erythroid dysplasia is the least specific finding and should be viewed with suspicion),
an abnormal karyotype typical for MDS (the World Health Organization does not consider trisomy 8, loss of the Y chromosome, or isolated del(20q) as specific enough to define MDS), or
other evidence of restricted clonality of hematopoiesis (e.g., by FISH or another test).
Patients with cytopenias who do not meet these decisive criteria, but for whom another diagnosis is not apparent, may be termed “Idiopathic Cytopenias of Undetermined Significance (ICUS)”. ICUS cases may resolve spontaneously, may be found to have a different diagnosis over time, or may more clearly become overt MDS or AML. Patients with ICUS need to be monitored carefully.
The bone marrow in MDS is usually hypercellular for age, but about 10% of cases are accompanied by a hypocellular marrow, and such cases may be difficult to distinguish from aplastic anemia.
Beware of other conditions that can mimic myelodysplastic syndromes:
MDS “mimics” include:
Vitamin B12 (cyanocobalamin) or folate deficiency
Copper deficiency (beware of this particularly in patients with co-existing neuropathy, or a history of celiac disease, bariatric surgery, or chronic use of zinc supplements)
Medication effect (especially methotrexate, azathioprine, or chemotherapeutics – all can cause a macrocytic anemia)
Alcohol abuse (ethanol can cause megakaryocytes to look dysplastic)
Autoimmune disorders (e.g., immune thrombocytopenia (ITP), Felty syndrome)
Congenital syndromes (e.g., congenital or cyclic neutropenia, Fanconi anemia, germline mutations in RUNX1, which can cause Familial Platelet Disorder with Predisposition to MDS/AML, aka FPD-AML)
In addition, MDS can sometimes be difficult to distinguish from other hematological disorders, including aplastic anemia, large granular lymphocyte disorders, myeloproliferative neoplasms (MPN) (MDS/MPN overlap syndromes do occur), paroxysmal nocturnal hemoglobinuria (PNH) (PNH clones can appear in MDS, but are usually small – <10% of cells), or acute myeloid leukemia.
Which individuals are most at risk for developing a myelodysplastic syndrome:
MDS are primarily diseases of older patients; the median age at diagnosis in the United States is about 71 years. In China, Korea, and Southeast Asia, the median age at diagnosis is younger – between 45 and 55 years. The reason for this is not known.
Exposure to ionizing radiation, alkylating agents (e.g., melphalan, chlorambucil), and topoisomerase inhibitors (e.g. etoposide, topotecan, anthracyclines) are known risk factors for developing MDS.
There is a moderate male predominance (<1.5:1), probably related to occupational exposures. Work in agriculture and the petroleum industries and cigarette smoking may increase risk slightly.
The 5q- syndrome, an uncommon type of MDS with a relatively indolent natural history, has a female predominance in most reports.
Patients diagnosed with MDS under age 50 who do not have a history of exposure to a DNA damaging agent might have a germline disorder predisposing to MDS, such as Fanconi anemia (which is diagnosed in adulthood in up to 25% of cases).
MDS may arise in patients with aplastic anemia or PNH.
What laboratory studies should you order to help make the diagnosis, and how should you interpret the results?
After performing a history and physical and excluding other causes of low blood counts, the diagnosis of MDS is usually made with 3 tests:
Complete blood count (CBC) with peripheral smear (which should show cytopenias, and may show dysplastic morphologic changes such as neutrophil hypogranularity or hypolobation, giant or hypogranular platelets, or oval macrocytic erythrocytes),
Bone marrow aspirate and trephine core biopsy (which should show dysplasia in >10% of cells, and may show an increase in myeloid blasts, monocyte precursors, or the presence of pathologic ring sideroblasts), and
G-banded karyotyping (abnormal in 50% of cases, including >80% of the patients with secondary MDS).
Other tests such as flow cytometry, MDS fluorescent in situ hybridization (FISH) panels, and molecular genetic tests are supplementary. In the future, molecular tests will become increasingly important. MDS FISH panels should only be used if the patient refuses a marrow exam or if karyotyping fails.
Blast proportion assessment should be based on morphology, not flow, since flow results depend on gating and may be affected by sample preparation (e.g. ammonium chloride lysis of mature red cells may also destroy some dysplastic erythroid progenitors).
How are myelodysplastic syndromes classified?
The current classification of MDS is the 4th edition of the World Health Organization (WHO) Classification of Tumours of Haematopoietic and Lymphoid Tissues, published in 2008. The 2008 WHO MDS classification is a minor modification of the 3rd edition WHO classification formally published in 2001, which in turn was built on the 1976 leukemia classification and 1982 MDS classification developed by the French-American-British (FAB) Co-operative Group.
Important classification factors in the current WHO MDS classification includethe specific cell lineages in which dysplasia is present (e.g., isolatedanemia is typical for refractory anemia (RA) or RARS), the marrow blastproportion (5-10% blasts is RAEB-1, 11-19% blasts is RAEB-2), the presence of ring sideroblasts (>=15% are required to qualify asRARS), and, to a limited extent, the presence of cytogenetic abnormalities, such as an interstitial deletion of the long arm of chromosome 5 (required for diagnosis of del(5q) MDS).
If an MDS-associated karyotypic abnormality is present but the patient does not have obviously dysplastic morphology, MDS may still be diagnosed.These patients are considered to have unclassifiable MDS (MDS-U).
By definition, MDS are associated with <20% marrow blasts. Patients with >= 20% blasts have AML.
Treatment-related MDS are grouped together with treatment-related AML by the WHO in the category t-MDS/t-AML, because the outcome in such patients is poor regardless of the blast count.
What imaging studies (if any) will be helpful in making or excluding the diagnosis of myelodysplastic syndromes?
Imaging studies are not part of the routine diagnostic workup of MDS. Conditions that can resemble MDS may have radiographic findings – these should be pursued only if clinical features provide an indication for imaging.
If you decide the patient has a myelodysplastic syndrome, what therapies should you initiate immediately?
MDS are typically subacute diseases and the majority of cases do not require emergent management. Exceptions are cases in which abnormalities in the peripheral blood can lead to life-threatening conditions. These include:
Severe bleeding with thrombocytopenia or hypofunctioning platelets – Platelet function testing can help identify patients with abnormal platelet function; however, emergent management should include platelet transfusion, cessation of anti-platelet agents, and correction of coagulopathies if present.
Severe thrombocytopenia – patients with platelet counts less than 10,000 per microliter would receive prophylactic platelet transfusions in order to prevent catastrophic bleeding.
Fever and neutropenia – All infections in patients with MDS should be treated aggressively with antimicrobial agents since neutrophil dysfunction can be present despite a normal neutrophil count. However, febrile patients (temperature > 100.4 degrees F) with an absolute neutrophil count less than 500 per microliter should be treated with empiric broad spectrum antibiotics. Treatment with filgrastim or other myeloid growth factors may help increase the neutrophil count after several days, but may not overcome abnormalities in neutrophil function.
Severe or symptomatic anemia – Patients with MDS do not suddenly become anemic, and because of compensatory responses during the slow evolution of their condition, can usually tolerate a significant reduction in their hematocrit. However, those patients with a hemoglobin level less than 7-8 g/dL should be considered candidates for transfusion if symptomatic, as should patients with higher hemoglobin levels who have comorbid conditions such as coronary artery disease, chronic obstructive pulmonary disease, or a recent history of gastrointestinal bleeding. Erythropoiesis stimulating agents (ESAs) are not useful emergently since they can take weeks to increase the hemoglobin level and MDS patients can be refractory to treatment.
More definitive therapies?
How should myelodysplastic syndromes be treated definitively?
Goals of MDS therapy depend in part on the stage of disease, and include symptom control, reduction of transfusion needs, delay of disease progression, improvement of health-related quality of life, and extension of survival.
With the exception of allogeneic hematopoietic stem cell transplantation there are no curative therapeutic options for MDS. Advanced age, the presence of comorbidities, and a lack of a suitable donor limit the availability of allogeneic transplantation, but use of reduced-intensity conditioning approaches and alternative stem cell sources (e.g., umbilical cord blood) are expanding the roster of potentially eligible patients. Therefore, patients with MDS who are potentially candidates for transplantation should be evaluated early in the disease course by a physician with expertise in stem cell transplantation.
As of 2011, 3 medications now have specific FDA approval for MDS-related indications: azacitidine (Vidaza, approved May 2004), decitabine (Dacogen, approved May 2006), and lenalidomide (Revlimid, approved December 2005).
In addition, the iron chelators (deferasirox [Exjade], deferoxamine [Desferal]) are approved for patients with transfusion-related iron overload, which may complicate MDS. A third iron chelator, deferiprone [Ferriprox], was approved by the FDA in 2011 for a narrow thalassemia-associated iron overload indication and is not approved for MDS.
Treatments approved for other non-MDS indications that are commonly or occasionally used off-label in MDS include hematopoietic growth factors (epoetin [Procrit], darbepoetin [Aranesp], filgrastim [Neupogen], sargramostim [Leukine], pegfilgrastim [Neulasta], romiplostim [Nplate], eltrombopag [Promacta]), immunosuppressive and immunomodulatory therapies (equine antithymocyte globulin (ATG) [Atgam], cyclosporine A, tacrolimus, thalidomide [Thalomid], alemtuzumab [Campath]), androgens, antifibrinolytics (e.g. epsilon aminocaproic acid [Amicar], tranexamic acid [Lysteda, Cyklokapron]), and cytotoxic agents (e.g., cytarabine, AML-like induction chemotherapy).
What is a typical treatment approach to mylelodysplastic syndromes?
All patients with MDS should receive supportive care, including transfusions as indicated and antibiotics for infections.
Iron chelation therapy can be considered for patients who have received 25-30 units of blood and have an elevated ferritin >1500 mcg/L on at least 2 occasions several weeks apart. A standard dose with deferasirox is 20 mg/kg/day orally, titrating to response. Creatinine must be monitored closely.
For lower-risk patients:
If the patient is asymptomatic, observation alone may be recommended.
If the patient has anemia only:
If the patient has deletion 5q, lenalidomide at a starting dose of 10mg/day orally once daily is appropriate. A starting dose of 5 mg/day is associated with lower cytogenetic response rates and no difference in the rate of cytopenias (MDS-004 study, unpublished). However, if 10mg/day is not tolerated, dose reduction may be required.
If the serum erythropoietin (EPO) level is <500 U/L, epoetin alfa 40-60000 IU weekly or darbepoetin alfa 500 mcg once every 3 weeks can be considered. If there is no response or the patient has RARS, addition of a myeloid growth factor may be helpful. The Centers for Medicare and Medicaid Services (CMS) will not pay for epoetin or darbepoetin if the Hb is >=10 g/dL.
If the serum EPO level is >=500 U/L, an ESA such as epoetin or darbepoetin is unlikely to be effective. Immunosuppressive therapy with ATG and cyclosporine or tacrolimus can be considered, especially if the patient is young (<55 years) and has HLA-DR15, a PNH clone, trisomy8, or a normal karyotype.
If the patient has neutropenia, thrombocytopenia, or both:
A hypomethylating agent, immunosuppressive therapy, or clinical trial enrollment are appropriate. Optimal therapy is undefined in this group.
For higher-risk patients:
If the patient is a transplant candidate, allogeneic stem cell transplant is the treatment of choice. Azacitidine or decitabine may be used as a bridge to transplant.
If the patient declines transplant or is not a transplant candidate, a hypomethylating agent (i.e., azacitidine or decitabine) is the most appropriate initial therapy.
Azacitidineis preferred because a survival advantage has been demonstrated compared to “conventional care” (which in the AZA-001 study included supportive care, low-dose cytarabine, or 3&7 AML induction). The standard dose is 75 mg/m2/day x 7 days, IV or SC, once every 4 weeks. At least 6 cycles should be administered before treatment failure is declared, unless the patient does not tolerate the drug or there is obvious progression.
Decitabine may work more rapidly than azacitidine, and may be associated with higher response rates, but at the cost of a higher rate of febrile neutropenia. A survival advantage has been demonstrated with decitabine in AML in the elderly, but not in MDS, possibly due to suboptimal study design. The most commonly used decitabine schedule is 20 mg/m2/day IV for 5 days once every 4-6 weeks.
Patients failing the above therapies should be referred for clinical trials, as optimal treatment is not defined.
The most appropriate approach to maintenance therapy is unclear, but patients who achieve a response to azacitidine or decitabine should be continued on this treatment indefinitely, as cessation of therapy usually results in relapse within 6 months.
What other therapies are helpful for reducing complications?
ESAs (epoetin, darbepoetin) may reduce red blood cell (RBC) transfusion needs. Their long-term safety in MDS is unknown. ESAs in MDS are not subject to the ESA APPRISE REMS program, as per guidance from the FDA issued in early 2011.
Myeloid growth factors (filgrastim, pegfilgrastim, sargramostim) may increase the absolute neutrophil count reduce recurrent infections in patients with neutropenia or granulocyte dysfunction, but have no known survival benefit.
Thrombopoietinagonists / thrombopoiesis stimulating agents (romiplostim, eltrombopag) can increase the platelet count and reduce platelet transfusion needsand bleeding events, but in some cases stimulate blast growth as well, and should only be used with extreme caution.
In the case of mucosal bleeding due to thrombocytopenia and platelet dysfunction, antifibrinolytic agents (e.g. aminocaproic acid) may behelpful. Thrombosis may complicate this treatment.
Iron chelation therapy with deferasirox or deferoxamine may help reduce complications associated with transfusional hemosiderosis. Iron chelation is an extremely controversial area in MDS. Iron chelators can reduce ferritin, but have not yet been shown to decrease morbidity or mortality. In two Phase II studies less than half of enrolled patients were still taking the study drug (deferasirox) after 12 months.
There are anecdotal reports of hematopoietic response to vitamin D3 therapy.
Splenectomymay be useful in patients with concomitant ITP and MDS, who typically have thrombocytopenia out of proportion to other cytopenias.
What should you tell the patient and the family about prognosis?
The prognosis of patients with MDS is highly variable. Some patients see little change in their disease over the course of years while other progress to acute myeloid leukemia or die from complications of cytopenias after only a few months.
a. Prognostic scoring systems
Prognostic scoring systems help predict an individual patient’s risk of dying of their disease or progressing to acute myeloid leukemia.
i. Adverse prognostic factors – features common to most scoring systems include:
Older age (>60 years);
An increased proportion of blast cells in the bone marrow;
Specific chromosomal abnormalities (particularly abnormalities of chromosome 7 or complex karyotypes with more than 2 abnormalities);
Cytopenias in more than one cell line instead of just a single cytopenia;
Dependence on transfusion support;
Comorbidities and an impaired functional status are also associated with a shorter survival after diagnosis;
Acquired point mutations of certain genes in MDS cells have been shown to predict shorter survival, even after considering known prognostic features. Testing for these genetic abnormalities may become commercially available in the next few years, but point mutations are not considered by any of these prognostic scoring systems in routine clinical use today.
b. Risk-Adapted Therapy
Patients and families should know that the choice of therapy is often tied to the prognosis and the factors that help predict it. For example, younger patients with a recent diagnosis of MDS and normal karyotype are more likely to benefit from immunosuppressive therapy. Patients with anemia and deletions of the long-arm of chromosome 5, a favorable prognostic marker, are more likely to respond to lenalidomide. Treatment with the hypomethylating drugs azacitidine and decitabine are typically reserved for patients predicted to have a high risk of death or progression to AML since this group has been shown to have the greatest proportional benefit from these agents.
There are several prognostic scoring systems (risk models) currently in use for risk stratification of patients with MDS. These include:
the 1997 International Prognostic Scoring System (IPSS), which includes blast proportion, cytogenetics, and number of cytopenias;
the 2005 World Health Organization-based Prognostic Scoring System (WPSS), which includes WHO subtype, cytogenetics, and transfusion dependence (the WPSS was revised in 2009 to include marrow fibrosis as a high risk factor, and again in 2011 to change transfusion requirement for hemoglobin level);
the 2008 MD Anderson Cancer Center MDS risk model, which is a 5-tier, 17-point system with more than 8 parameters;
the 2011 revised International Prognostic Scoring System (IPSS-R), which is a minor modification of the 1997 system, rebalancing cytogenetics and blast proportion and including more karyotypes.
Each of these risk models has merits and limitations. The 1997 IPSS is the system most familiar to clinicians and the one most widely used for determining clinical trial eligibility, but applies only to “de novo”, previously untreated disease, and overemphasizes the important of excess blasts at the expense of cytogenetics.
"What if" scenarios.
Patients with congenital sideroblastic anemias are sometimes erroneously labeled as MDS. For instance, a 35 year old woman with isolated anemia, a normal karyotype, and ring sideroblasts is more likely to have a germline ALAS2 mutation or another congenital syndrome than RARS.
Flow cytometry should not be used as a substitute for a 500-cell-aspirate-differential to assess blast proportion.
FISH panels are not a substitute for conventional karyotyping.
Patients with higher-risk MDS are sometimes left untreated because of outdated notions of lack of therapeutic options beyond ESAs. Because hypomethylating agents can improve quality of life and extend survival, this is no longer appropriate unless the patient is too frail or too sick to be a treatment candidate.
Patients with MDS and neutropenia are sometimes given a standard vial of pegfilgrastim (Neulasta), 6 mg, but this dose is too high for many patients can lead to leukemoid reactions or spontaneous splenic rupture. If pegfilgrastim is to be used, a lower starting dose should be chosen.
Clinicians sometimes terminate treatment with lenalidomide or a hypomethylating agent without an adequate therapeutic trial. Response is delayed in some cases. If at all possible, an 8-12 week trial of lenalidomide, 4 cycles of decitabine, or 6 cycles of azacitidine should be administered before declaring the specific therapy a failure.
There are two retrospective reviews that suggest patients who get 7 days of azacitidine per cycle, regardless of whether there is a break for the weekend, have higher complete response (CR) rates than those who get less than 7 days per cycle.
Decitabine rarely works after azacitidine failure, and azacitidine rarely works after decitabine failure. Patients for whom a hypomethylating agent has failed have a poor prognosis (median survival <6 months) and should be considered for clinical trial enrollment.
Myelodysplastic syndromes are a collection of related diseases of the bone marrow. While they are fairly heterogeneous in their presentation, they share many pathophysiologic features.
a. Shared features:
MDS are clonal disorders – they arise from an abnormal hematopoietic precursor (a self-renewing or disease initiating cell) that has clonally expanded and evolved over time.
Ineffective hematopoiesis – dysplasia and increased apoptosis of differentiating cells leads to both quantitative deficiencies (too few) and qualitative defects (impaired function) of mature cells in the peripheral blood.
Risk of progression to acute myeloid leukemia – clonal evolution of disease cells can lead to the expansion of highly proliferative, undifferentiated myeloblasts characteristic of AML.
b. Subclasses with distinguishing clinical characteristics:
Clinical features can distinguish types of MDS with shared pathophysiologic features –
Refractory anemia with ring sideroblasts (RARS)
Ring sideroblasts – defined as having >15% of nucleated erythroid cells with iron laden mitchondria surrounding the nucleus on Prussian blue reaction (“iron stain”).
SF3B1 mutations – found to have high rate of mutations (>60%) in the splicing factor gene SF3B1.
Refractory anemia with ring sideroblasts and thrombocytosis (RARS-T) – RARS with myeloproliferative features similar to essential thrombocythemia.
Ring sideroblasts – defined as having >15% of nucleated erythroid cells with iron laden mitchondria surrounding the nucleus on an iron stain.
Thrombocytosis – platelet count of greater than or equal to 450,000 per microliter (450 x 109 per liter).
JAK2 and MPL mutations – over 50% of RARS-T cases have a V617F JAK2 mutation and about 5% have an activating mutation in the thrombopoietin receptor gene MPL – a mutation pattern identical that seen in patients with essential thrombocythemia.
SF3B1 mutations – found to have high rate of mutations (>60%) in the splicing factor gene SF3B1.
5q-deletion – an interstitial or terminal deletion of the long arm of chromosome 5, usually involving band q31 (Del[5q]) and often isolated, without other karyotypic abnormalities.
Female predominance – the female to male ratio in 5q- syndrome is nearly 2:1, compared with a ratio of 0.7:1 for other types of MDS.
Relative thrombocytosis – preserved or elevated platelet count compared with other types of MDS.
Marked impairment in erythropoiesis – significant anemia characterized by increased apoptosis of erythroid cells and activation of p53.
Prognostically favorable – lower risk of progression to AML compared with other MDS types.
Predicts response to treatment – high rate of response to the immunomodulatory class of drugs such as lenalidomide (Revlimid).
Unlike more uniform myeloid disorders, such as chronic myelogenous leukemia and polycythemia vera, that have identical molecular abnormalities in almost every case, patients with MDS have a wide range of molecular abnormalities that contribute to the development and progression of their disease. While these abnormalities are not present in every patient, many are recurrent, indicating that several discrete molecular mechanisms contribute to the pathogenesis of MDS.
c. Chromosomal abnormalities
Chromosomal abnormalities are present in 50% of cases assessed by standard karyotyping –
Chromosome 5q deletion – present in 8% of cases as an isolated abnormality, 2.5% with one other lesion. Seen in half of cases with complex cytogenetics.
Prognostically favorable as an isolated abnormality – still favorable if found with a single other abnormality.
Haploinsufficiency – loss of a single copy of one or more genes on 5q is believed to be the pathogenic event caused by del(5q). Recurrent point mutations in 5q-genes have not been identified to date.
Commonly Deleted Regions – areas recurrently deleted in cases of 5q-loss.
5q32-33 region – associated with 5q-minus syndrome – contains >40 genes.
RPS14 – encodes ribosomal protein – haploinsufficiency produces severe defect in erythropoiesis, p53 activation in erythrocytes, and increased apoptosis.
miR-145 & miR-146 – loss of these microRNAs can cause increased platelet counts and may lead to a selective growth advantage for diseased cells.
5q31 region – associated with complex karyotypes and therapy related MDS.
HSPA9 – encodes a 70 kD mitochondrial heat shock protein. Loss of function leads to impaired erythropoiesis.
CTNNA1 – encodes alpha-catenin – degree of DNA methylation of this gene is associated with decreased expression and more advanced disease.
EGR1 – loss of one copy of this gene associated with increased stem cell self-renewal.
CDC25Cand PPP2CA – these genes encode phosphatases potentially inhibited by lenalidomide, possibly explaining the favorable response of del(5q) patients to this drug.
distal 5q region – not within commonly deleted regions, but often lost as part of larger 5q-deletions.
APC – encodes a negative regulator beta-catenin signalling, a pathway implicated in self-renewal of disease initiating cells in myeloid malignancies.
NPM1 – encodes nucleophosmin – frequently mutated in AML leading to aberrant localization of the protein product. Hemizygous loss of NPM1 in mice produces an MDS-like phenotype and a predisposition to various hematologic malignancies.
Chromosome 7q deletion and Chromosome 7 loss – Present in 10% of cases, mostly with other abnormalities. Seen in 50% of therapy related MDS, related to alkylator or radiation exposure.
Haploinsufficiency – loss of a single copy of one or more genes on 7q is believed to be the pathogenic event caused by del(7q).
Commonly Deleted Regions – areas recurrently deleted in cases of 7q-loss.
7q22 region – no candidate gene identified in this region. Knockout of syntenic region in the mouse had no phenotype.
7q35-36 region – Contains the EZH2 gene which is recurrently mutated in MDS, although not when del(7q) is present. Therefore, deletion of other genes in this region may be pathogenic.
Trisomy 8 – Only recurrent chromosomal amplification in MDS. Present in 8% of cases as a sole abnormality.
Intermediate prognostic risk
Pathogenic Mechanism – Unclear, but associated with upregulation of anti-apoptotic genes. May have increased response to immunosuppressive therapy.
Chromsome 20q deletion – Rare but recurrent abnormality present in <5% of MDS cases as a sole abnormality.
Prognostically favorable as an isolated abnormality
Commonly Deleted Region – includes a few candidate genes (MYBL2, L3MBTL1) but none proven conclusively.
Chromsome Y loss – seen in <5% of cases, and believed to be a benign age-related event in most if not all cases, since also found in healthy age-matched controls with normal CBC.
Prognostically favorable as an isolated abnormality
Chromosome 3 rearrangements – Rare but recurrent abnormality present in <5% of MDS cases as a sole abnormality; associated with topoisomerase inhibitor therapy.
Pathogenic Mechanism – Disrupts MECOM gene locus on chromosome 3q26, increasing expression of EVI1 – a poor prognostic marker associated with impaired differentiation and increased self-renewal. The MECOM locus is a frequent oncogenic retroviral insertion site.
Complex karyotype – defined as 3 or more karyotypic abnormalities of any type. Occurs in 13% of cases.
Pathogenic Mechanism – Genomic instability – mutations of TP53 seen in 50% of cases with complex karyotype, 90% if an abnormality of chromosome 17p is present.
Translocations – much rarer in MDS than in other myeloid malignancies.
Variable Prognostic Significance – depends on the abnormality, but usually considered intermediate risk.
Other Significance – certain translocations can be used as presumptive evidence of an MDS diagnosis even if other criteria such as clear dysplasia are not met.
d. Epigenetic dysregulation
Covalent alterations of chromatin (DNA and its associated histones) play a large role in determining the pattern of gene expression in cells. MDS cells show changes in these epigenetic marks compared to normal cells. In particular, increases in DNA methylation in the promoter regions of tumor suppressor genes has been shown to decrease the expression of these genes. Greater levels of this type of hypermethylation have been associated with more advanced disease. This forms the rationale for the use of azacitidine and decitabine in MDS which are known to inhibit the DNA methyltransferases responsible for DNA hypermethylation, although it remains unclear whether azacitidine or decitabine are truly working via epigenetic mechanisms. In addition, many of the recurrently mutated genes identified in MDS function as regulators of epigenetic state.
e. Somatic point mutations
Nearly 70% of patients with MDS have identifiable acquired mutations in one or more genes present in their abnormal hematopoietic cells. No single gene is mutated in the majority of cases nor are any gene mutations unique to MDS. All of the genes known to be mutated in MDS can also be found in other myeloid malignancies, albeit often in a different proportion of cases. Patients with similar clinical findings can have different patterns of mutations, although some gene mutations are associated with particular disease phenotypes. The mutated genes seen in MDS can be roughly categorized by function.
Tyrosine kinase signalling genes – receptor tyrosine kinases (MPL, FLT3, CSF1R) or downstream signalling genes (JAK2, CBL, CBLB, PTPN11, NRAS, KRAS, BRAF) are mutated in the minority of non-CMML cases of MDS (<10%). NRAS mutations are the most frequent of these (3-5%) and are associated with poor prognostic features such as elevated blast count and thrombocytopenia. Mutations in these genes are more common in chronic myelomonotic leukemia (MML) and MDS/MPN overlap disorders.
Transcription factors and co-regulators –
RUNX1 – Mutations of this canonical hematopoietic transcription factor occur in 7% of MDS and are enriched in cases of therapy-related disease. Mutations are clustered in the DNA-binding domain, some of which have been shown to produce a protein product with dominant negative activity against wild-type RUNX1. Mouse models of RUNX1 mutations produce MDS-like or AML-like phenotypes depending on the nature of the mutations examined. Germline mutations of RUNX1 cause a familial platelet disorder with a propensity for acute myeloid leukemia. In MDS patients, RUNX1 mutations are associated with thrombocytopenia as well as elevated blast proportion and are independent predictors of decreased survival.
ETV6 – This ets-like transcription factor is essential for normal hematopoiesis and is frequently a translocation partner in various hematopoietic malignancies including rare cases of MDS. Point mutations of ETV6 have been described in 3% of MDS and are associated with a poor prognosis.
ASXL1 – This polycomb gene encodes an transcriptional co-activator linked to the retinoic acid receptor and is involved in the regulation of Hox genes. It is mutated in 14% of MDS and nearly 40% of CMML cases. Mutations result almost exclusively in heterozygous frameshifts or premature stop codons and are associated with decreased survival independently of prognosis predicted by the IPSS.
WT1 – Confoundingly, increased expression of WT1 is associated with advanced MDS and transformation to AML, yet mutations of this zinc-finger-containing transcription factor occur in AML and MDS at low frequency. The prognostic significance of these mutations in MDS is not known.
PHF6 – Recurrent and highly disruptive mutations of this gene containing PHD-type zinc finger DNA binding domains have recently been described. These mutations have also been observed in T-cell ALL and AML although their pathogenic mechanism and prognostic significance in MDS are not known.
Epigenetic modifiers – Several MDS-related genes are involved in epigenetic regulation of gene expression, including genes that may have other functions such as RUNX1, IDH1, IDH2 and ASXL1. Mutations in these genes are not always mutually exclusive of each other suggesting that they each have some independent pathogenic function.
DNMT3A – Mutations of this DNA methyl transferase occur in 10% of MDS and are believed to impair its ability to methylate cytosine residues in DNA. In AML, mutations of DNMT3A are associated with a poor prognosis, but in MDS this is not clear nor is it known if they predict response to treatment with hypomethylating agents.
EZH2 – Mutations of this histone methyltransferase occur in 6% of MDS and impair its ability to methylate lysine-27 on the tail of histone-3 (H3K27). This histone mark is associated with chromatin compaction and gene silencing. EZH2 mutations are not associated with known prognostic features and are therefore often seen in patients predicted to have lower risk disease. However, EZH2 mutations are strongly associated with a poor prognosis.
TET2 – Mutations of TET2 have been found in nearly all myeloid malignancies and are the most frequent genetic abnormality identified in MDS. They occur in over 20% of MDS cases and are enriched in CMML (> 40%). The TET2 protein is an α-ketoglutarate dependent enzyme that catalyzes the conversion of 5-methyl-cytosine to 5-hydroxymethyl cytosine, thus altering this epigenetic mark. It is associated with prolonged survival in CMML but is considered prognostically neutral in MDS. A recent study suggests that TET2 mtuations may be associated with a slightly higher rate of response to hypomethylating agents, although no difference in overall survival was seen.
ATRX – Germline mutations of this X chromosome gene are associated with mental retardation and alpha-thalassemia in boys. Rare patients with MDS can acquire mutations of ATRX leading to a profound alpha-thalassemia with notable hemoglobin H inclusions in red cells. ATRX has been shown to associate with EZH2 and the suppression of alpha-globin epxression is believed to occur through epigenetic silencing of the alpha-globin locus. How ATRX mutations contribute to the pathogenesis of MDS is not known.
Cell cycle and apoptosis – Mutations of TP53 occur in 7% of MDS and are associated with several poor prognostic features including complex karyoptype (50% of complex karyotype cases have point mutations of TP53), elevated blast proportion, and thrombocytopenia. However, TP53 mutations are strongly associated with a worse prognosis than would be predicted by these clinical measures alone and has been linked to treatment resistance.
NPM1 – Distal frameshift mutations of NPM1 are among the most frequent molecular abnormalities in AML but are seen only rarely in MDS (<3%). These mutations disrupt the C-terminal nuclear localization sequence and mislocalize the protein to the cytoplasm.
IDH1 and IDH2 – These genes encode NADPH-dependent enzymes that convert isocitrate to α-ketoglutarate in the cytoplasm (IDH1) and mitochondria (IDH2). They have been found to carry recurrent mutations that result in an enzymatic change of function. Mutated IDH proteins convert α-ketoglutarate to 2-hydroxyglutarate, an oncometabolite that is believed to alter the function of α-ketoglutarate dependent proteins including TET2. In AML, where these mutations are more common that they are in MDS (10% vs. 3%), IDH mutations are largely exclusive of TET2 mutations, suggesting that they have a shared pathogenic mechanism.
SF3B1 – Mutations of this member of the U2 spliceosome complex were described in 2011 during whole exome sequencing studies of CLL and RARS. They are present in 10% of MDS cases overall and are highly enriched in the RARS subtypes (>60%). Mutations appear to be exclusively missense and heterozygous. They occur at a small number of hotspots suggesting that they cause a gain of function in the protein product. Their pathogenic mechanism and independent prognostic significance are not known. At least 8 other splicing factors are known to be mutated in MDS and SRSF2 may have IPSS-independent adverse prognostic importance.
f. Changes in the bone marrow microenvironment
The non-hematopoietic elements of the bone marrow are not normal in patients with MDS. These changes are believed to be due to the presence of abnormal MDS cells, but might be intrinsic abnormalities of the stroma in some cases. Abnormal stromal function may explain why some patients with MDS can have cytopenias even though their disease cells only occupy a fraction of the bone marrow alongside normal hematopoietic cells. Some of the microenvironmental changes seen in MDS include:
Abnormal cytokine profiles – increases in tumor-necrosis factor alpha and vascular endothelial growth factor have been observed in patients with MDS.
Adaptive immune response -Patients with MDS can respond to immunosuppressive therapy indicating that an immune response can be partially responsible for features of the disease. Oligoclonal T-cell populations, presumed to be autoreactive, have been identified in these patients before treatment. T-cell populations then become more polyclonal (normal) after therapy.
Intrinsic stromal abnormalities – Studies in mice have shown that stromal abnormalities can give rise to disordered and inefficient hematopoiesis, even thought the blood forming cells are normal. In one example, the genetically modified stroma led to the development of clonal malignant myeloid cells. Other studies have shown genetic abnormalities in the stromal cells of rare patients with MDS. The environmental risk factors for acquiring MDS, such as radiation and chemotherapy, are likely to cause damage to the stroma as well as to the hematopoietic cells in the bone marrow. The extent to which toxic stromal injuries contribute to MDS is not known. Finally, a subset of patients with MDS will develop bone marrow fibrosis which can severely limit hematopoiesis. This is a late event and therefore unlikely to contribute to disease pathogenesis, but certainly adds to the morbidity associated with the progression of MDS.
What other clinical manifestations may help me to diagnose a myelodysplastic syndrome?
Patients with MDS may present with peculiar paraneoplastic manifestations, such as vasculitis or Behcet syndrome. These often improve with treatment.
Older patients with macrocytic anemia who have normal levels of serum B12 and red blood cell (RBC) folate should be referred for hematological evaluation, as MDS is very common in this group.
What other additional laboratory studies may be ordered?
Initial diagnostic evaluation should include:
CBC with leukocyte differential and peripheral smear;
bone marrow aspirate, biopsy, and cytogenetics;
serum B12 levels (supplemented by methylmalonic acid if borderline);
RBC folate level;
serum ferritin level;
serum lactate dehydrogenase (LDH);
serum chemistry group, including creatinine and liver tests;
HIV serology, if risk factors for infection are present;
serum copper level, if risk factors for copper deficiency are present;
human leukocyte antigen (HLA) typing, if the patient is a candidate for allogeneic transplant or for immunosuppressive therapy;
MDS FISH panel, if the patient refuses a bone marrow biopsy or karyotyping fails;
prothrombin time (PT) and activated partial thromboplastin time (aPTT), if the patient has a history of bleeding;
type and crossmatch, if the patient is known to be anemic and will require transfusion.
What’s the evidence?
Maciejewski, JM, Steensma, DP, Gregory, SA, McCrae, KR. “Marrow failure syndromes”. American Society of Hematology – Self Assessment Program. 2010. [Anannotated bibliography of MDS references can be found in Chapter 15, pages 443-474. The 5th edition of ASH-SAP will be published in 2013.]
Steensma, DP, Bennett, JM. “The myelodysplastic syndromes: diagnosis and treatment”. Mayo Clin Proc.. vol. 81. 2006. pp. 104-30. [A detailed bibliography of clinically relevant MDS papers (>300 references) can be found here.]
Guidelines for MDS. [Includes evidence-based recommendations and a comprehensive list of references relevant to diagnosis and treatment.]
Swerdlow, SH, Campo, E, Harris, NL. World Health Organization (WHO) Classification of Tumours of Haematopoietic and Lymphoid Tissues. 2008. [Current classification of MDS and related disorders.]
Valent, P, Horny, HP, Bennett, JM. ” Definitions and standards in the diagnosis and treatment of the myelodysplastic syndromes: Consensus statements and report from a working conference”. Leuk Res. vol. 31. 2007. pp. 727-36. [Minimal diagnostic criteria for MDS.]
Copyright © 2017, 2013 Decision Support in Medicine, LLC. All rights reserved.
No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.
- Myelodysplastic syndromes
- What every physician needs to know:
- Are you sure your patient has a myelodysplastic syndrome? What should you expect to find?
- Beware of other conditions that can mimic myelodysplastic syndromes:
- Which individuals are most at risk for developing a myelodysplastic syndrome:
- What laboratory studies should you order to help make the diagnosis, and how should you interpret the results?
- What imaging studies (if any) will be helpful in making or excluding the diagnosis of myelodysplastic syndromes?
- If you decide the patient has a myelodysplastic syndrome, what therapies should you initiate immediately?
- More definitive therapies?
- What should you tell the patient and the family about prognosis?
- "What if" scenarios.
- What other clinical manifestations may help me to diagnose a myelodysplastic syndrome?
- What other additional laboratory studies may be ordered?
- What’s the evidence?