Myelodysplastic Syndromes

Meet the patient

A 76-year-old man is referred for a haemoglobin of 84 g/L found on routine bloods. His MCV is 108 fL, his platelets are 90×10⁹/L, and the film shows hypogranular neutrophils and a few pseudo-Pelger-Huët cells. His B12, folate and thyroid are normal, he does not drink heavily, and his copper level is normal.[1]

He is a classic MDS presentation — an older adult with a macrocytic cytopenia and dysplastic film, with the nutritional and toxic mimics excluded. The marrow biopsy will give the clonal plasma-cell... no, the clonal haematopoietic picture, the blast percentage, the cytogenetics and the molecular profile that drive the entire management fork: is this lower-risk disease to support, or higher-risk disease to treat?[1]

The paradox — a cellular marrow with empty blood

Myelodysplastic syndromes are clonal haematopoietic stem-cell neoplasms in which ineffective haematopoiesis produces one or more persistent peripheral cytopenias — anaemia most often, but also neutropenia and thrombocytopenia — with a variable propensity to transform into acute myeloid leukaemia. They are the commonest myeloid malignancy of older adults.[1]

The central lesion is ineffective haematopoiesis: the clone proliferates but its offspring die by apoptosis before reaching the blood. The result is the paradox that defines MDS — a hyper- or normocellular marrow producing peripheral cytopenia, the opposite of the empty marrow of aplastic anaemia. A parallel process of blast accumulation and clonal evolution accounts for progression to AML in about a quarter to a third of patients. The 20 percent marrow or peripheral blast threshold separates MDS from AML (the WHO lowered the old FAB 30 percent to 20 percent in 2001).[1][7]

The diagnostic triad — DYC

The diagnosis rests on a triad that an examiner will always probe:[1]

1. Dysplasia of at least 10 percent in one or more myeloid lineages (erythroid, granulocytic, megakaryocytic), or an MDS-defining cytogenetic abnormality such as isolated del(5q).
2. A persistent unexplained cytopenia (at least four to six months, not attributable to a secondary cause).
3. Exclusion of secondary and reactive causes — vitamin B12 and folate deficiency, copper deficiency, alcohol, viral infection (HIV, parvovirus B19), and drug effect.[1]

[1]

WHO 2022 classification — genetics now runs the show

MDS classification has moved from morphology alone (the 1982 FAB system) to the WHO 5th edition (2022) and the parallel International Consensus Classification (ICC, 2022), both reorganising the disease around genetics. The thresholds a student must reproduce are the 10 percent dysplasia cut-off, the 5 percent blast line separating low-blast from high-blast MDS, and the 20 percent blast line that defines AML.[7]

[7]

The older FAB classification is still encountered: refractory anaemia (RA), RA with ringed sideroblasts (RARS), RA with excess blasts (RAEB), RAEB in transformation (now reclassified as AML), refractory cytopenia with multilineage dysplasia (RCMD), and the 5q- syndrome. Chronic myelomonocytic leukaemia (CMML) is no longer MDS — it is an MDS/MPN overlap neoplasm defined by persistent peripheral monocytosis of at least 1.0×10⁹/L.[1]

The 20 percent line — where MDS becomes AML

WHO 2022 and ICC 2022 agree on most categories and both retain the 20 percent blast threshold for AML when no defining AML genetics are present. They recognise MDS-biTP53 and the SF3B1 and del(5q) genetic entities. The ICC, however, sets the AML threshold at 10 percent blasts for cases with a defining AML genetic abnormality (such as t(8;21) or an NPM1 mutation), and merges some low-blast categories. In practice, treat the IPSS-R and IPSS-M risk scores — not the classification name — as the primary guide to therapy.[7]

Who gets it — and the CHIP spectrum

MDS has an overall incidence of about 4 per 100,000 per year, rising to over 30 per 100,000 in those over 70. The median age is about 76 with a slight male predominance. It is substantially more common than AML in the elderly.[1]

Recognised acquired and environmental risk factors:[1]

• Increasing age (the dominant factor — clonal haematopoiesis becomes almost universal beyond 70), male sex, tobacco smoking, benzene and solvent exposure, and ionising radiation.
• Prior cytotoxic therapy — the single most important modifiable risk factor.[1]

Therapy-related MDS (t-MDS) arises after chemo- or radiotherapy and carries two subtypes: an alkylating-agent or radiation-related form (latency 5 to 7 years, chromosomes 5 and 7 abnormalities, complex karyotype, TP53 mutation, poor prognosis) and a topoisomerase-II-inhibitor-related form (latency 2 to 3 years, balanced translocations of 11q23/KMT2A or RUNX1, often presenting as AML).[1]

The antecedent spectrum explains most "de novo" MDS. Clonal haematopoiesis of indeterminate potential (CHIP) — a somatic mutation (DNMT3A, TET2, ASXL1, TP53, JAK2) at a variant allele fraction of at least 2 percent in someone with a normal count — is detectable in over 10 percent of people over 70. Add an unexplained cytopenia without enough dysplasia to call MDS and it is clonal cytopenia of undetermined significance (CCUS); without a clone it is idiopathic cytopenia of undetermined significance (ICUS).[1]

Inherited bone-marrow-failure and germline-predisposition syndromes (Fanconi anaemia, Shwachman-Diamond, dyskeratosis congenita and other telomere-biology disorders, and germline GATA2, RUNX1, ETV6, DDX41 and CEBPA) evolve to MDS — a young patient with MDS should trigger germline testing, which changes donor selection and conditioning.[1]

The mechanism — ineffective haematopoiesis

An acquired somatic mutation in a multipotent haematopoietic stem or progenitor cell confers a clonal advantage, and the disease then evolves through multiple genetic hits:[1]

• Early founding mutations in DNA-methylation and chromatin genes — TET2, DNMT3A, ASXL1 — shared with CHIP and establish the clone.
• Spliceosome mutations (SF3B1, SRSF2, U2AF1, ZRSR2) commonly follow; SF3B1 is the hallmark of ringed-sideroblast disease.
• Late, progression-associated mutations — TP53 (multi-hit), RUNX1, EZH2, NRAS or KRAS, FLT3, IDH1 or IDH2 — drive blast accumulation and evolution to AML.[1]

The dominant consequence is ineffective haematopoiesis: maturing progenitors undergo excessive intramedullary apoptosis (up to three times normal), so the marrow is frequently hyper- or normocellular while the blood is cytopenic. A pro-inflammatory, dysfunctional marrow microenvironment selectively favours the clone.[1]

Recurrent cytogenetic lesions and their prognostic signal:[2]

• Favourable or good: isolated del(5q), del(20q), -Y, a normal karyotype (about half of all MDS).
• Intermediate: del(7q), trisomy 8 (+8), i(17q).
• Adverse or poor: monosomy 7 (-7), del(5q) with other abnormalities, inv(3) or t(3;3), a complex karyotype (at least 3 unrelated abnormalities); multi-hit TP53 commonly coexists with a complex karyotype.[2]

Recurrent gene mutations: SF3B1 (ringed sideroblasts — favourable); TP53 multi-hit (complex karyotype, therapy-related — very adverse); ASXL1, RUNX1, EZH2, NRAS or KRAS (adverse); TET2, DNMT3A (common, largely intermediate). These — captured by the molecular IPSS-M — refine prognosis well beyond the IPSS-R.[1][5]

How it walks in — usually an incidental count

The commonest presentation is an incidental finding of cytopenia, macrocytosis or dysplastic morphology on a routine full blood count in an older adult — frequently asymptomatic. When symptoms occur they reflect the affected lineage:[1]

• Anaemia (present in over 80 percent at diagnosis) — fatigue, pallor, exertional dyspnoea, worsening of angina or cardiac failure.
• Neutropenia — recurrent bacterial infection (chest, skin, perianal), fever, and occasionally neutropenic sepsis as the first clue.
• Thrombocytopenia — easy bruising, petechiae, purpura, mucosal bleeding, menorrhagia.[1]

Bicytopenia or pancytopenia is common; an isolated cytopenia (usually anaemia) is the other pattern. Splenomegaly and lymphadenopathy are uncommon in MDS — their presence should prompt consideration of an MDS/MPN overlap syndrome (CMML), primary myelofibrosis, CML, or a separate lymphoid process.[1]

Atypical and paraneoplastic presentations that examiners test deliberately: acute febrile neutrophilic dermatosis (Sweet syndrome) and cutaneous vasculitis may accompany or precede MDS, as may autoimmune phenomena. In the very elderly, MDS may present as isolated fatigue, falls, delirium, or an exacerbation of cardiac failure from unrecognised anaemia.[1]

The differential — dysplasia alone is not MDS

Dysplasia alone is not MDS — reactive and secondary causes must be excluded before diagnosing a clonal disorder. This single principle generates more exam questions than any other in the topic.[1]

[1]

The single most important teaching point: always send B12, folate and a serum copper level, take an alcohol and drug history, and exclude HIV before labelling dysplasia as MDS.[1]

The bedside round

The focused assessment documents the consequences of the cytopenia, screens for alternative diagnoses, and gathers the information that drives intensity of therapy (especially transplant eligibility).[1]

Run it in order: general (pallor, bruising, petechiae and purpura, infection; document performance status and comorbidity burden — both gate transplant decisions); abdomen (hepatosplenomegaly suggests an overlap syndrome, CMML, myelofibrosis or infiltration; lymphadenopathy suggests lymphoma); cardiovascular (the consequences of anaemia — tachycardia, a flow murmur, cardiac failure); and a focused drug, occupational (benzene, solvents, radiation) and prior-cancer-treatment exposure history, plus a family history of haematological malignancy (germline predisposition).[1]

There are no pathognomonic bedside signs of MDS — the diagnosis is made in the laboratory.[1]

Investigations — the marrow and the iron stain

First-line bloods: full blood count and film (macrocytosis — MCV often 100 to 120 fL, hypogranular or hyposegmented pseudo-Pelger-Huët neutrophils, giant hypogranular platelets, inappropriately low reticulocytes, a peripheral blast count), LDH (raised in higher-risk disease), and a screen to exclude secondary causes (B12, folate, copper, ferritin, HIV, an alcohol history).[1]

Bone-marrow aspirate and trephine biopsy is the diagnostic cornerstone and every patient needs one:[1]

• Cellularity — hyper-, normo- or hypocellular (hypoplastic MDS mimics aplastic anaemia).
• Dysplasia of at least 10 percent — erythroid (nuclear budding, multinucleation, ringed sideroblasts), granulocytic (hypogranular, hyposegmented Pelgeroid forms), megakaryocytic (micromegakaryocytes).
• Blast percentage — counted precisely (the 5, 10 and 20 percent lines all matter).
• Prussian-blue (Perls) iron stain — for ringed sideroblasts (erythroid precursors with at least 5 siderotic granules encircling at least one-third of the nucleus); 15 percent or more with an SF3B1 mutation defines MDS-SF3B1.
• Reticulin fibrosis — grades MDS-f.[1]

Cytogenetics: conventional G-banding karyotype plus FISH for the common lesions (-7 or del(7q), del(5q), del(20q), trisomy 8, complex karyotype). Molecular next-generation sequencing for TP53, SF3B1, ASXL1, RUNX1, TET2, EZH2, NRAS, FLT3, IDH1 or IDH2 refines prognosis and guides therapy — lenalidomide for del(5q), venetoclax consideration, TP53-directed trials.[1]

Erythropoietin level — a serum EPO under about 500 mU/mL with a low transfusion burden (under 2 units per month) predicts a good response to erythropoiesis-stimulating agents.[1]

Risk stratification — IPSS-R and the molecular IPSS-M

Risk stratification drives every management decision. The IPSS-R (Greenberg, Blood 2012) combines four clinical variables into five risk categories; the IPSS-M (Bernard, NEJM Evidence 2022) layers gene-mutation data on top to reclassify about 20 percent of patients — usually upwards.[2][5]

The IPSS-R combines marrow blast percentage, cytogenetic risk group, haemoglobin and platelet count:[2]

[1]

The blast percentage, cytogenetic group, haemoglobin (cut-points at 9 and 10.5 g/dL) and platelet count (cut-points at 50 and 100×10⁹/L) combine into a score distributing patients into five categories:[2]

[2]

The molecular IPSS-M incorporates the IPSS-R clinical variables plus 31 gene mutations — weighted most heavily for multi-hit TP53 (adverse) and SF3B1 (favourable), with FLT3 also adverse:[5]

[5]

In practice, the IPSS-M reclassifies about one patient in five, most often from lower to higher risk — which is why molecular testing is now standard in fit patients. Iron-overload monitoring: serial serum ferritin, and consider chelation when ferritin exceeds about 1000 micrograms per litre (or after 20 to 25 red-cell units).[1]

The first threats — anaemia, bleeding, neutropenic sepsis

[1]

Treat the acute cytopenic consequences first, then start risk-adapted disease therapy. The immediate threats are severe symptomatic anaemia, major bleeding, and neutropenic sepsis.[1]

• Severe or symptomatic anaemia — red-cell transfusion. Use leucodepleted products, irradiated if the patient is a transplant candidate, and CMV-negative if CMV-negative and a transplant candidate. Transfuse one unit at a time in older patients and aim for the lowest haemoglobin that relieves symptoms (often 8 to 9 g/dL, higher in cardiac disease).
• Thrombocytopenia — platelet transfusion for active bleeding, before procedures, or as prophylaxis when platelets are under 10×10⁹/L (or under 20 with fever or bleeding risk). Thrombopoietin-receptor agonists carry a caution for blast progression and are not routine.
• Febrile neutropenia — a medical emergency: blood cultures and empirical broad-spectrum antipseudomonal beta-lactam (piperacillin-tazobactam 4.5 g IV every 8 hours or ceftazidime) within one hour.
• Iron overload from chronic transfusion — start chelation when ferritin exceeds about 1000 micrograms per litre.[1]

[1]

Lower-risk MDS — supportive and growth-factor therapy

Therapy is risk-adapted using the IPSS-R (and increasingly IPSS-M) and calibrated to age, fitness and goals. Lower-risk disease (IPSS-R very low, low, and many intermediate) is managed with supportive care and growth factors. The aim is to relieve cytopenia, improve quality of life and delay transformation — not to eradicate the clone.[1][3]

[1] [1] [4] [6]

Higher-risk MDS — HMA, venetoclax, transplant

The aim in higher-risk disease is to modify the disease course and pursue the only curative option — allogeneic transplant in fit patients. The standard disease-modifying agent is a hypomethylating agent (HMA) — azacitidine first-line — increasingly combined with the BCL-2 inhibitor venetoclax.[3][8]

[1] [3] [8] [1]

Allogeneic haematopoietic stem-cell transplant is the only potentially curative therapy. It exploits a graft-versus-leukaemia effect — the donor immune system recognises and destroys residual malignant cells. Selection balances the high transplant-related mortality (10 to 20 percent at one year) and graft-versus-host disease against the poor natural history of untreated higher-risk disease. TP53-mutated and therapy-related MDS carry the worst post-transplant outcomes, but transplant still offers the best chance of long-term survival in selected patients, best paired with a clinical trial.[1]

The subtypes that bite

[1]

When it goes wrong — complications and the traps that cost marks

Disease complications — progressive cytopenia (infection, bleeding, transfusion dependence), transfusional iron overload with end-organ damage (cirrhosis, cardiac failure, diabetes, hypogonadism, hypothyroidism), and transformation to AML in about 25 to 30 percent overall (higher in high-blast and TP53-mutated disease). Infection is the leading cause of death in MDS.[1]

Treatment complications — examiners test these by drug:[1]

[1]

The classic diagnostic pitfalls: mistaking B12, folate or copper deficiency (or alcohol) for MDS; overcalling reactive dysplasia (post-infection, growth-factor effect); failing to send cytogenetics and molecular testing on the marrow; not excluding HIV and other viruses; and diagnosing MDS from dysplasia alone without a persistent cytopenia or a clone. Treatment pitfalls: over-treating lower-risk disease like leukaemia; failing to chelate iron in the chronically transfused; delaying transplant referral for fit higher-risk patients; and using thrombopoietin-receptor agonists without monitoring blast progression.[1]

Prognosis and disposition

Prognosis is highly heterogeneous and driven by the risk score. The IPSS-R median survivals range from about 8.8 years (very low) down to about 0.8 years (very high); the IPSS-M refines this further, from about 8.8 years (very low) down to about 0.9 years (very high).[2][5]

Adverse prognostic factors: a higher marrow blast percentage, adverse or complex cytogenetics (especially -7, inv(3) and multi-hit TP53), a low haemoglobin, a low platelet count, a high LDH, older age and poor performance status. Favourable factors include isolated del(5q), an SF3B1 mutation, a normal karyotype, and a low blast count with limited transfusion need.[2]

Allogeneic stem-cell transplant is the only modality with curative potential, selected by age, comorbidity (HCT-CI), performance status and donor availability; the graft-versus-leukaemia effect is central to its efficacy. Follow-up is by serial full blood counts and transfusion requirement, periodic marrow with blast and cytogenetic reassessment, ferritin and iron monitoring, and surveillance for infection and treatment toxicity.[3]

Special populations

• Younger patients — prioritise a donor search, germline (inherited-predisposition) testing (GATA2, RUNX1, ETV6, DDX41), and early allogeneic transplant referral; a younger patient with apparent "de novo" MDS always warrants a germline work-up before selecting a donor.[1]
• Elderly and frail — supportive care, transfusion and iron chelation, with an HMA and reduced-intensity transplant reserved for fit higher-risk individuals selected by physiological rather than chronological age (a fit 72-year-old may be transplanted, a frail 60-year-old may not).[1]
• Therapy-related MDS — poor-risk biology; discuss goals of care, an HMA and clinical trials; transplant selected case-by-case.
• Inherited bone-marrow-failure syndromes (for example Fanconi anaemia) — a modified, fludarabine-based reduced-intensity protocol avoiding radiation, and donor or relative genetic screening to avoid using an affected relative.[1]
• Resource-limited settings — transfusion, affordable HMA where available, and supportive antimicrobial prophylaxis; transplant access and novel-agent cost are major barriers, so access — not biology — often drives treatment choice.[1]
• Pregnancy (very rare) — supportive care with transfusion; defer disease-modifying therapy; lenalidomide is absolutely contraindicated (teratogenic).[1]

The evidence and regional differences

The WHO 5th edition (2022) and the International Consensus Classification (ICC, 2022) both reorganise MDS around genetics, recognise MDS-biTP53 and the SF3B1 and del(5q) entities, and retain the 20 percent blast threshold for AML (the ICC defining AML by genetics at the 10 percent blast count for defining recurrent genetic abnormalities).[7]

The Revised IPSS (2012, Greenberg) remains the standard bedside risk tool, while the molecular IPSS-M (2022, Bernard) incorporates mutations (multi-hit TP53, SF3B1, FLT3 and others) to refine risk and reclassify about one patient in five.[2][5]

[1] [1]

The mantra, and the mnemonics

[1]

The mantra: Dysplasia plus a cytopenia is MDS only when the mimics are excluded; 20 percent blasts is AML; lower-risk is supported, higher-risk gets an HMA and the only cure is a transplant — and copper deficiency is the great mimic.[1][3]

Ward-round test — three stems, thirty seconds each

Stem 1 — the man from the top of the topic (answer)

A 76-year-old with a haemoglobin of 84 g/L, MCV 108 fL, platelets 90×10⁹/L, hypogranular neutrophils and pseudo-Pelger-Huët cells, normal B12, folate, copper and TSH, no alcohol. What is the diagnosis, the next test, and how do you risk-stratify? Model: This is myelodysplastic syndrome — a macrocytic cytopenia with dysplastic morphology, the nutritional and toxic mimics excluded. The next test is a bone-marrow aspirate and trephine with a Prussian-blue iron stain, a karyotype plus FISH, and a molecular (NGS) panel. Confirm the diagnostic triad (dysplasia at least 10 percent plus a persistent cytopenia plus exclusion of secondary causes) and check the blast percentage — 20 percent or more is AML, not MDS. Risk-stratify with the IPSS-R (blasts, cytogenetics, haemoglobin, platelets) and, in a fit patient, the IPSS-M (which adds TP53, SF3B1, FLT3 and other mutations). Lower-risk disease is supported with an ESA; higher-risk disease gets an HMA and, if fit, an allogeneic transplant referral.[1][2]

Stem 2 — the cytopenia after gastric bypass (answer)

A 60-year-old woman two years after a Roux-en-Y gastric bypass has a normocytic anaemia and neutropenia with vacuolated erythroid and myeloid precursors and a numb, tingling feet. The registrar books a bone-marrow biopsy for suspected MDS. What is the likely diagnosis, and what single test confirms it? Model: This is copper deficiency — the classic MDS mimic, common after gastric or bariatric surgery (or zinc excess), producing anaemia plus neutropenia with vacuolated precursors and ringed sideroblasts and a sensory (dorsal-column) neuropathy. The single confirmatory test is a serum copper level (and caeruloplasmin). Copper deficiency corrects with replacement — copper 2 to 4 mg orally daily — and the cytopenia and dysplasia resolve. Diagnosing it as MDS would expose the patient to a marrow transplant work-up for a reversible nutritional deficiency. Always send B12, folate and copper before labelling dysplasia as MDS.[1]

Stem 3 — the fit 68-year-old with higher-risk MDS (answer)

A 68-year-old fit man (ECOG 1) has MDS-IB2 with 12 percent marrow blasts, a normal karyotype, and a serum beta-2 microglobulin of 5 mg/L. What is the management plan, and what is the goal? Model: This is higher-risk MDS. The goal is disease modification and the pursuit of the only curative modality — allogeneic stem-cell transplant. Start an HMA (azacitidine 75 mg/m² for 7 days every 28 days), plus or minus venetoclax, for cytoreduction while a donor search proceeds (matched sibling first, then matched unrelated, then haploidentical), and send molecular testing (TP53 especially, which would worsen the prognosis). Proceed to allogeneic transplant with reduced-intensity conditioning in this fit patient, exploiting the graft-versus-leukaemia effect. If transplant-ineligible, continue the HMA for as long as it works. Throughout, give transfusion and antimicrobial support, and chelate iron once ferritin exceeds about 1000 microg/L.[1][3][8]

References

1. [1]Cazzola M Myelodysplastic Syndromes N Engl J Med, 2020.PMID 32997910
2. [2]Greenberg PL, Tuechler H, Schanz J, et al. Revised international prognostic scoring system for myelodysplastic syndromes Blood, 2012.PMID 22740453
3. [3]Fenaux P, Mufti GJ, Hellstrom-Lindberg E, et al. Efficacy of azacitidine compared with that of conventional care regimens in the treatment of higher-risk myelodysplastic syndromes: a randomised, open-label, phase III study Lancet Oncol, 2009.PMID 19230772
4. [4]Fenaux P, Platzbecker U, Mufti GJ, et al. Luspatercept in Patients with Lower-Risk Myelodysplastic Syndromes N Engl J Med, 2020.PMID 31914241
5. [5]Bernard E, Tuechler H, Greenberg PL, et al. Molecular International Prognostic Scoring System for Myelodysplastic Syndromes NEJM Evid, 2022.PMID 38319256
6. [6]List A, Dewald G, Bennett J, et al. Lenalidomide in the myelodysplastic syndrome with chromosome 5q deletion N Engl J Med, 2006.PMID 17021321
7. [7]Khoury JD, Solary E, Abla O, et al. The 5th edition of the World Health Organization Classification of Haematolymphoid Tumours: Myeloid and Histiocytic or Dendritic Neoplasms Leukemia, 2022.PMID 35732831
8. [8]DiNardo CD, Jonas BA, Pullarkat V, et al. Azacitidine and Venetoclax in Previously Untreated Acute Myeloid Leukemia N Engl J Med, 2020.PMID 32786187