AML vs. MDS: Key Differences in Diagnosis and Treatment

Myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML) are both cancers of the blood-forming cells in bone marrow, but they differ in how aggressively abnormal cells crowd out healthy ones. The traditional dividing line has been a single number: the percentage of immature cells called blasts in the bone marrow. AML is defined by a high blast count, reflecting a severe block in blood cell maturation, while MDS involves lower blast counts alongside misshapen, poorly functioning blood cells. That distinction matters because it shapes everything from how urgently treatment begins to which therapies are on the table, and sometimes whether a cure is even the goal.

What Goes Wrong in the Bone Marrow

Your bone marrow is a factory that continuously produces red blood cells, white blood cells, and platelets from a pool of stem cells. In MDS, that factory still runs, but it runs badly. Stem cells mature into finished blood cells, yet those cells come out misshapen and reduced in number, a condition pathologists call dysplasia. Patients end up with low blood counts (cytopenias) even though the marrow itself may be packed with cells. In AML, the production line essentially jams: immature blast cells pile up instead of maturing into functional blood cells, flooding the marrow and often spilling into the bloodstream.¹

The bone marrow microenvironment also behaves differently in each disease. In MDS, the marrow tends to be inflamed, with an overactive immune system paradoxically killing off the very cells the body is trying to make. In AML, immune evasion takes center stage: leukemia cells suppress surrounding immune cells to protect themselves.² This shift from inflammation-driven cell death toward immune suppression is one reason MDS and AML feel like different diseases even though they share a biological family.

The Blast Percentage and How Diagnosis Works

For decades, a bone marrow blast count of 20 percent has served as the formal boundary between MDS and AML. If blasts make up at least 20 percent of the nucleated cells in the marrow, the diagnosis is AML. Below that threshold, and in the presence of dysplasia and cytopenias, the diagnosis is MDS.³ That cutoff replaced an older one of 30 percent, which had placed patients with 20 to 29 percent blasts in a category called “refractory anemia with excess blasts in transformation,” essentially a gray zone between MDS and AML.

Diagnosis is not just about counting blasts, though. A bone marrow biopsy allows pathologists to assess cellularity, fibrosis, and the physical arrangement of immature cells within the marrow tissue. In high-risk MDS, these architectural details help distinguish a disease that is inching toward leukemia from one that is relatively stable.⁴ Genetic testing adds another layer: certain chromosomal abnormalities and gene mutations now define specific subtypes of AML regardless of blast count, a recognition that biology sometimes matters more than a single number.

Shifting Classification Systems

If all of this sounds tidy, the reality at the diagnostic bench is less so. Two major classification systems were updated around 2022, and they introduced overlapping but not identical criteria. The International Consensus Classification (ICC) requires at least 10 percent blasts in the marrow or blood for an AML diagnosis (with certain genetic exceptions), while the 2022 World Health Organization system does not specify a blast cutoff for genetically defined AML subtypes.⁵ Both systems keep recurrent genetic abnormalities as the primary organizing principle, but the practical result is that the same patient can receive a different diagnosis depending on which system their hospital uses. For patients, this means it is worth asking your care team which classification they follow, because it can affect which clinical trials you qualify for.

The Genetic Landscape

One of the clearest differences between MDS and AML lies in the types of mutations their cells carry. MDS tends to harbor mutations in genes involved in the spliceosome (the machinery that processes genetic instructions) and chromatin remodeling (the way DNA is packaged). Genes like SF3B1, SRSF2, and U2AF1 are frequently mutated in MDS and also appear in AML that has evolved from prior MDS.⁶ AML that arises on its own, by contrast, is more likely to carry mutations in signaling pathways, particularly RAS and FLT3, which act like stuck accelerators driving uncontrolled cell growth.⁷

FLT3 mutations alone are found in roughly 30 percent of AML cases, and their presence generally signals a worse prognosis without targeted treatment.⁸ In AML that develops from MDS, the acquisition of FLT3 or RAS pathway mutations is often what tips the disease over the edge into outright leukemia, alongside new chromosomal abnormalities.⁹

How MDS Transforms Into AML

MDS is sometimes called a “pre-leukemic” condition because a meaningful fraction of patients eventually progress to AML. That transformation is not a simple straight line. Research using single-cell sequencing has shown that even at the MDS stage, stem cells harbor a surprisingly complex web of subclones, small genetically distinct populations coexisting in the marrow. Some of these subclones contribute to the low-grade disease, while others lurk quietly and later expand to drive AML.¹⁰ Crucially, the clone that dominates in AML is not always descended from the dominant MDS clone; sometimes a previously minor branch takes over.

This branching, nonlinear pattern of evolution explains why treating MDS early does not always prevent leukemia. The seeds of transformation can be planted in stem cells that are clinically invisible at the time of the initial MDS diagnosis. It also helps explain why AML arising from MDS (sometimes called secondary AML) is generally harder to treat than AML that appears without a prior blood disorder.

Precursor Conditions Before MDS

Before MDS itself, there are even earlier stages worth knowing about. Clonal hematopoiesis of indeterminate potential (CHIP) refers to the presence of certain mutations in blood-forming cells without any noticeable blood count abnormalities. Clonal cytopenia of undetermined significance (CCUS) is similar but includes low blood counts that don’t quite meet MDS criteria. Both conditions carry a real, if small, risk of eventually progressing to MDS or AML.¹¹

CHIP is increasingly recognized as more than just a hematologic curiosity. The same mutations that predispose to blood cancers also appear to increase the risk of cardiovascular disease and inflammatory conditions, making CHIP a systemic concern rather than a narrow oncology issue.¹² Specialized clinics are emerging to monitor people with CHIP, though the optimal surveillance strategy is still being worked out.

Risk Stratification in MDS

Not all MDS is the same. Some patients live for years with manageable anemia, while others progress to AML within months. The tool doctors use to sort this out is a prognostic scoring system. The Revised International Prognostic Scoring System (IPSS-R) has been the standard, relying on chromosomal abnormalities, marrow blast percentage, and the depth of low blood counts to place patients into one of five risk categories.¹³

A newer model, the Molecular IPSS (IPSS-M), incorporates mutations in 31 genes alongside the traditional clinical variables. It reclassified about 46 percent of patients compared to the older system, often shifting people into a higher or lower risk group than their IPSS-R score suggested. Mutations in TP53, FLT3, and MLL predicted the worst outcomes, while SF3B1 mutations were linked to more favorable prognosis, though that favorable signal weakened when SF3B1 appeared alongside certain co-mutations.¹⁴ In studies comparing the two systems in patients treated with hypomethylating agents, the IPSS-M outperformed the IPSS-R in predicting progression-free and leukemia-free survival.¹⁵

Risk stratification matters because it directly dictates treatment intensity. Lower-risk MDS may be managed with supportive care and drugs to boost blood counts, while higher-risk MDS is treated more aggressively, sometimes with chemotherapy or a stem cell transplant.

Treating Lower-Risk MDS

For patients with lower-risk MDS, the primary goal is improving blood counts and quality of life rather than trying to eliminate the disease entirely. The foundation is supportive care: red cell transfusions for symptomatic anemia, platelet transfusions when bleeding risk is high, and iron chelation therapy to manage iron overload from repeated transfusions. Recent attention has also turned to bone health, as MDS patients face elevated risks of osteoporosis.¹⁶

Beyond transfusions, a range of drugs can help. Erythropoiesis-stimulating agents (ESAs), which work like synthetic versions of the hormone that tells the body to make red blood cells, are often the first pharmacologic step. When ESAs fail, luspatercept, a protein that promotes late-stage red cell maturation, has shown benefits in reducing transfusion dependence.¹⁷ The treatment landscape also includes lenalidomide (particularly effective in a chromosomal subtype called del(5q)), hypomethylating agents like azacitidine and decitabine, and newer options such as imetelstat and roxadustat.¹⁸

Treating Higher-Risk MDS

Higher-risk MDS is treated with a greater sense of urgency because the disease is more likely to progress to AML and survival times are shorter without intervention. Hypomethylating agents (HMAs) like azacitidine have become the standard for patients who are not candidates for a stem cell transplant. In a nationwide Thai study comparing HMA-treated patients to those receiving only supportive care, median survival was 15 months with HMAs versus 6 months without, and HMA therapy was independently associated with improved survival in multivariable analysis.¹⁹

For eligible patients, allogeneic stem cell transplant remains the only treatment with a realistic chance of curing higher-risk MDS. Timing is critical: waiting too long risks disease progression, but transplanting too early exposes patients to transplant-related complications without clear benefit. Prospective trials have confirmed a survival advantage for transplant over non-transplant approaches in higher-risk MDS, and current research is focused on refining decisions around pre-transplant therapy and post-transplant maintenance.²⁰

AML Treatment for Younger and Fit Patients

AML treatment is typically more intensive than MDS treatment, reflecting the disease’s faster pace. For younger patients and older patients who are physically fit, the standard approach remains intensive induction chemotherapy, a regimen designed to wipe out leukemia cells quickly enough to allow normal blood production to recover. Induction therapy remains the standard of care for younger patients with favorable- or intermediate-risk disease, though not every patient is deemed eligible.²¹

After remission is achieved, consolidation therapy follows — additional chemotherapy cycles, and in many cases, a stem cell transplant for patients with higher-risk genetics. Transplant outcomes in AML that evolved from prior MDS or from treatment for another cancer (therapy-related AML) tend to be worse. A retrospective study of therapy-related AML and MDS patients reported two-year overall survival of 37 percent after transplant, with higher non-relapse mortality in patients who had received multiple prior treatment lines or who were transplanted more than six months after diagnosis.²²

AML Treatment for Older and Less Fit Patients

Most AML patients are older adults, and many cannot tolerate intensive chemotherapy. Over the past several years, the combination of azacitidine and venetoclax (a drug that blocks a protein called BCL-2 that leukemia cells rely on for survival) has become the standard of care for this population.²³ The pivotal VIALE-A trial, which enrolled patients aged 75 and older or those with conditions precluding intensive chemotherapy, showed a median overall survival of roughly 15 months with the combination versus about 10 months with azacitidine alone.²⁴

Even patients in their eighties and nineties can be treated with this combination, though dose reductions and extended cycle intervals are often needed. In a multicenter analysis of octogenarians and nonagenarians with newly diagnosed AML (without prior MDS), the complete remission rate was about 73 percent, and median overall survival was roughly 8 months, extending to about 13 months for those who achieved remission.²⁵ These are not curative numbers, but they represent a meaningful shift from an era when many elderly AML patients received little more than transfusion support.

Targeted Therapies in AML

The treatment landscape for AML has changed substantially over the past decade thanks to drugs aimed at specific molecular targets. FLT3 inhibitors are the most prominent example. First-generation inhibitors like midostaurin block multiple kinases, while second-generation drugs like gilteritinib and quizartinib are more specific to FLT3 and more potent.⁸ Drugs targeting IDH1 and IDH2 mutations represent another class that has earned regulatory approval. Additional targeted and cellular therapies are in development for other defined patient subgroups.²⁶

MDS does not yet have the same breadth of targeted options. Because MDS mutations tend to involve fundamental cell machinery like the spliceosome and epigenetic regulators rather than druggable signaling kinases, developing precision therapies has been harder. Some MDS patients do benefit from lenalidomide (targeting del(5q)) or luspatercept, but the kind of mutation-matched targeted therapy that has reshaped AML care is largely still aspirational for MDS.

Quality of Life and the Burden of Supportive Care

Because MDS often moves slowly and treatment centers on managing symptoms rather than pursuing cure, the day-to-day experience of MDS can be dominated by fatigue, transfusion appointments, and monitoring visits rather than the acute toxicity of chemotherapy. A pooled analysis of over 3,300 patients with newly diagnosed myeloid malignancies found that MDS patients reported meaningfully better scores than AML patients across several quality-of-life measures, including less nausea, less pain, less breathlessness, and better cognitive function at diagnosis.²⁷

That said, long-term transfusion dependence in MDS carries its own costs, both financial and physical. An older but illustrative cost analysis found that median annual transfusion costs per MDS patient were roughly $4,900 per year, driven largely by red cell transfusions. For those who progressed to AML, median transfusion costs jumped to over $13,000 per patient during the leukemia phase.²⁸ Iron overload from chronic transfusions can damage the liver and heart over time, adding another dimension to the care burden.

For AML patients, the quality-of-life trajectory depends heavily on treatment intensity. A study comparing patients who received intensive chemotherapy with those on non-intensive regimens found that the intensive group showed improvement in physical well-being over the first month, while the non-intensive group experienced a relative decline in physical function during that same period.²⁹ The explanation is not that chemotherapy feels good, but that achieving remission rapidly can resolve symptoms like severe fatigue and bleeding that were dragging quality of life down at diagnosis.

Pediatric AML and MDS

Both diseases can occur in children and young adults, though they are far less common in pediatric populations. When they do appear, the biology often differs from adult disease. Pediatric AML tends to feature a distinct set of genetic drivers, including KMT2A rearrangements and core-binding factor fusions, and generally relies less on the age-related mutations (like those in splicing factors and epigenetic regulators) that dominate adult MDS and secondary AML.³⁰ Pediatric MDS is rare enough that it is often studied in small case series, and treatment decisions are heavily individualized. Children with MDS are more likely to proceed to stem cell transplant relatively early, in part because they can tolerate the procedure better and have more years of life at stake.

Management of both diseases in older adults also requires a distinct approach. Because MDS and AML disproportionately affect people over 65, treatment plans need to account not just for the disease’s biology but for each patient’s overall fitness, other medical conditions, and personal goals.³¹ A healthy 70-year-old and a frail 70-year-old with the same bone marrow findings may reasonably receive entirely different treatments.