Acute myeloid leukemia: genetics and MRD at the right time

A 52-year-old man arrives at the emergency department with fatigue, purple spots on his legs and fever for a week. The blood count shows 68,000 leukocytes, hemoglobin of 7.4 g/dL and 21,000 platelets, with blasts in the blood. The bone marrow aspirate confirms acute myeloid leukemia. The team collects the marrow, admits the patient and schedules the start of induction chemotherapy for the next day. The genetic sequencing request goes out, but no one notes when the result should arrive. On the eighth day, during rounds, the resident asks whether the patient has an FLT3 mutation, and no one knows. The scene is fictional and was written for teaching.

This text uses the case as a clinical reasoning scenario for students and health professionals. The central question is why genetics and measurable residual disease, MRD, need to arrive at the right time in acute myeloid leukemia, and what is lost when they arrive late. The answer depends on three clues: what genetics changes in induction, what it changes after remission and when MRD should be measured. If you play the case of the day in NEXO, you will recognize the logic: a new clue changes the weight of the others.

Key points

• Acute myeloid leukemia is a group of diseases defined by genetics, and risk changes according to mutations and karyotype.
• The FLT3 mutation changes induction: in the RATIFY study, adding midostaurin to chemotherapy from the eighth day increased survival.
• The European LeukemiaNet genetic risk group guides the decision about transplantation in first remission.
• MRD measures the leukemic cells left after remission and helps predict relapse.
• NPM1 molecular MRD by PCR and MRD by flow cytometry have different sensitivities and complement each other.
• Time to transplant also matters and should be planned from diagnosis.

Why this is a reasoning scenario

Acute myeloid leukemia often arrives as an emergency, and the rush to start chemotherapy is real. The problem is that some important decisions depend on tests that take days, and the rush on one side cannot erase the deadline on the other. Organizing the timing of tests is part of treatment.

What the scene teaches

The scene shows a process error, not a knowledge error. The team knew that sequencing was necessary and requested the test, but did not agree on when the result needed to arrive. The reasoning requires asking, right on the first day: which result changes what I will do and on what day do I need it? With that question, the test request becomes part of the plan.

What is not at stake

The text does not discuss acute promyelocytic leukemia, which is an emergency with its own treatment, nor does it detail regimens for patients who cannot receive intensive chemotherapy. It also does not compare all FLT3 inhibitors or the new drugs for NPM1. The focus is the order and timing of genetic tests and MRD in the patient eligible for intensive chemotherapy. Each patient's management follows the service protocol and the team's decision.

The first clue: the genetics that changes induction

Some genetic alterations change treatment in the first week. The best known is the FLT3 mutation, present in about 30 percent of patients with acute myeloid leukemia, according to the Spanish PETHEMA registry (Romero Riquelme and colleagues, 2026).

The RATIFY study

RATIFY examined 3277 patients aged 18 to 59 with newly diagnosed acute myeloid leukemia for the FLT3 mutation, and randomized 717 with the mutation to receive standard chemotherapy with midostaurin or placebo. Midostaurin was given at a dose of 50 mg orally twice a day, from day 8 to day 21 of each induction and consolidation cycle, and then as maintenance. Overall survival was higher with midostaurin, with a hazard ratio for death of 0.78, and the median was 74.7 months versus 25.6 months with placebo (Stone and colleagues, 2017).

The FLT3 subtypes in the study

RATIFY separated patients by mutation type: a point mutation in the tyrosine kinase domain, called TKD, and an internal tandem duplication, called ITD, with a high or low proportion of the mutated allele. The benefit was analyzed in all subtypes, and allogeneic transplant was allowed. Serious adverse events had a similar frequency in both groups (Stone et al., 2017). For those who order the test, this means the report must state the mutation type, not just "positive" or "negative".

What this means for the clock

The study design sets a deadline: midostaurin starts on day eight. For that, the FLT3 result must be available beforehand. In the scene, the resident's question on day eight arrives exactly at the limit. If the result is not ready, the window the study tested is lost for that cycle.

The result in larger studies

In AMLSG 16-10, compared with historical controls from the same group, midostaurin was associated with better event-free survival, with a hazard ratio of 0.55, in both younger and older patients, and with better overall survival, with a hazard ratio of 0.57 (Döhner et al., 2022). The study also included allogeneic transplant and one year of maintenance with midostaurin, which shows that the decision about FLT3 extends well beyond the first week.

Older patients and real life

A phase 3b study with 301 patients, almost half of them over 60, showed that midostaurin with intensive chemotherapy was well tolerated with no upper age limit (Sierra et al., 2023). The German and Austrian group AMLSG 16-10, with 440 patients aged 18 to 70 with FLT3-ITD mutation, concluded that midostaurin should be standard in older patients as well (Döhner et al., 2022). In the PETHEMA registry, with 1658 adults with FLT3 mutation, composite complete remission was 81.4 percent with midostaurin versus 71.7 percent without it, and 30-day mortality was 2.1 versus 7.1 percent. Median overall survival was 47.2 months versus 19.3 months, and allogeneic transplant in first remission was more frequent in the midostaurin group, 48.9 versus 41.3 percent (Romero Riquelme et al., 2026).

How long genetics takes

Ordering the test is not enough. You need to know when the result arrives and what to do until then.

The deadline in practice

In the study that led to the approval of midostaurin, 3277 patients aged 18 to 59 with newly diagnosed acute myeloid leukemia were tested for FLT3 mutations, and the result was reported to investigators within 48 hours after the sample reached the laboratory (Stone et al., 2017). That deadline was part of the study and does not automatically apply to every service. The team needs to know, on the day of admission, which results arrive in time to change induction.

How to organize the wait

• Collect the marrow for karyotype, mutation panel and cytometry in the same procedure.
• Order the FLT3 test separately with an agreed deadline before day eight.
• Note in the medical record the expected date of each result.
• Store material for molecular MRD, if there is an NPM1 mutation.
• Review the results at the first-week meeting.

When to wait and when to start

Not every patient can wait. Very high leukocytosis, bleeding or severe infection may require immediate start. The decision to wait or not should be explicit and recorded, with the list of what will change when each result arrives. This is a direct application of clinical decision-making under pressure.

The second clue: the genetics that changes post-remission

After remission, the question becomes how to consolidate: more chemotherapy or allogeneic transplant. The genetic risk group of the European LeukemiaNet, ELN, is the main tool for this choice.

What changed in ELN 2022

In the 2022 version, mutations in genes related to myelodysplasia became an adverse-risk category. In a European cohort of 4978 patients treated with intensive chemotherapy, 34.1 percent had at least one of these mutations, with a lower complete remission rate, 65.7 versus 77.7 percent, and worse overall survival, with a hazard ratio of 1.45. The effect varied by gene: ASXL1, RUNX1, SF3B1 and U2AF1 behaved as adverse risk, while SRSF2 and STAG2 came closer to intermediate risk (Bill et al., 2026).

Does the classification change the outcome?

A German center analyzed 662 patients treated between 2010 and 2022. When the risk classification was updated in 2017 and post-remission treatment was intensified, above all with more transplants, 4-year overall survival rose from 40 to 60 percent. The improvement was clearest in the intermediate-risk group (Berdel et al., 2026). Four-year relapse-free survival also improved, from 35 to 54 percent. To rule out the bias of comparing different eras, the authors reclassified the more recent patients using the old criteria and confirmed the improvement in the intermediate group. The finding suggests that the classification is not just a label: applied to treatment, it changes the outcome.

Genetics and age

In an Irish series of 166 patients, with a median age of 70 years, age under 65 years, de novo disease, favorable or intermediate risk by ELN 2022 and absence of TP53 mutation were associated with better survival (Elhassadi et al., 2026). The same authors note that NPM1 mutations appear in about 30 percent of adults with the disease and FLT3 mutations in 25 to 30 percent, often together.

The third clue: MRD at the right moment

Complete remission is defined under the microscope, but the microscope does not see everything. MRD looks for the leukemic cells that remain in numbers too small to show up on the marrow aspirate.

Two methods that complement each other

MRD can be measured by flow cytometry, which recognizes the protein profile of the leukemic cells, or by PCR, which looks for a specific genetic alteration, such as the NPM1 mutation. A systematic review of 23 studies found sensitivity of 76.2 percent and specificity of 96.0 percent for flow cytometry when compared with PCR. The authors concluded that flow cytometry is fast and highly specific, but less sensitive than PCR, especially in acute myeloid leukemia, and that the two methods together stratify risk better (Alalwy et al., 2026).

MRD after the first cycle

In the British AML18 study, with 439 patients aged 60 or older without known adverse karyotype, patients who were not in remission with negative MRD after the first cycle could enter a second randomization to intensify treatment. Remission with negative MRD after the first cycle was 47 percent with daunorubicin, cytarabine and gemtuzumab, versus 29 percent with CPX-351 (Knapper et al., 2026). Three-year overall survival was 52 percent versus 35 percent, in favor of the gemtuzumab regimen. The study shows how MRD measured at a defined point can be used to decide the next step.

Molecular MRD for NPM1

In leukemia with the NPM1 mutation, PCR-based MRD tracks the amount of the mutated transcript. The European LeukemiaNet recommends monitoring MRD by quantitative PCR or digital PCR whenever possible, focusing on stable, clonal mutations, and the NPM1 mutation is the preferred target when present (Ferrari et al., 2026). At a university hospital in France, digital PCR detected low levels of the mutation in samples previously negative by quantitative PCR and allowed relapse to be predicted earlier in some patients (Ferrari et al., 2026). This allows time to discuss transplant or another approach before clinical relapse.

Why measure at fixed points

DRM can only be compared between patients and between studies when it is measured at defined time points, such as after the first and second cycle or before transplant. A DRM collected outside these points is hard to interpret, because there is no reference to say whether the value is good or bad. That is why the DRM schedule should be written into the plan together with the chemotherapy schedule, and the laboratory needs to know in advance which method will be used for each collection.

DRM and transplant

In an analysis of the MORPHO study, on maintenance with gilteritinib after transplant in patients with FLT3-ITD, the time between diagnosis and transplant and the use of a FLT3 inhibitor before transplant influenced DRM before the procedure. Patients transplanted within less than 120 days of diagnosis had higher DRM before transplant and were the ones who benefited most from maintenance (Levis and colleagues, 2026). The lesson is that DRM, timing and treatment interact, and the decision needs to look at all three together.

Table: what to order and when

Time point	Test	What changes	Main source
Diagnosis, first day	Karyotype, mutation panel, cytometry	Risk group and overall plan	Bill and colleagues
Before the eighth day	FLT3 test	Add midostaurin to induction	RATIFY
After the first cycle	DRM by cytometry or PCR	Intensify or follow the plan	AML18
During consolidation	Molecular NPM1 DRM, if there is a mutation	Predict relapse and plan transplant	Ferrari and colleagues
Before transplant	DRM and time since diagnosis	Maintenance after transplant	MORPHO

The conversation with the patient and the family

In acute myeloid leukemia, the patient receives a lot of information in a few days, almost always as an inpatient and frightened. The way the team explains the pending tests changes trust in the treatment and adherence to the next steps.

Explaining what is pending

Saying that "the genetic tests have not come back yet" without explaining what they change creates anxiety. A better explanation says what each test looks for, when it should come back and what can change in the plan. For example: "We are waiting for a test that shows whether the leukemia has an alteration called FLT3. If it does, we add a tablet medicine starting on the eighth day."

Preparing the family for transplant

When the genetic risk suggests transplant, the donor search may involve siblings and other relatives. Talking about this early, still during induction, allows time for compatibility tests and keeps the family from getting the news at the last minute. The conversation should make clear that the final decision depends on the response to treatment and on DRM.

Explaining DRM

Many patients think remission is a cure. Explaining that DRM looks for what the microscope does not see helps them understand why bone marrow tests continue after remission, and why a positive result can change the plan even when the blood count is normal. The conversation should also include uncertainty: DRM predicts risk, but it does not on its own determine what will happen.

Risks and reasoning errors

In acute myeloid leukemia, the most common errors are about timing and communication. They resemble the ones that appear in clinical deterioration on the ward: the sign was there, but no one had the task of looking.

Ordering without a deadline

The error in the scene is ordering the right test without agreeing when it should come back. A FLT3 result that arrives on the fifteenth day is correct, but useless for induction. A diagnostic pause on the first day, asking "which result changes what I do and when?", prevents this error.

Trusting only microscopic remission

Considering the patient cured because the marrow aspirate shows no blasts ignores DRM. Morphologic remission is the first step, not the final response.

Reading DRM out of context

A DRM result only makes sense when the method, the time of collection and the leukemia subtype are known. The systematic review showed that disagreement between cytometry and PCR does not have a single meaning (Alalwy and colleagues, 2026). Dealing with this ambiguity is an exercise in diagnostic uncertainty.

Leaving transplant for later

Referral to the transplant center and the donor search take time. If the genetic risk indicates transplant, the search should start early, with the team discussing the case together, as in interprofessional clinical reasoning.

How NEXO helps train this reasoning

Acute myeloid leukemia is a good example of a case in which the clues arrive on different days, and each one changes the plan. In NEXO, the case of the day works with progressive clues, and practicing this movement helps you notice when a pending result should be at the center of the decision. The method relies on learning based on clinical cases.

Three exercises with cases

• Rewrite the first day of the scene with a list of tests, the deadline for each one and what each result changes.
• Imagine that FLT3 comes back positive on the sixth day. Write what changes in the plan and who needs to know.
• Imagine that, after consolidation, the NPM1 MRD starts to rise again. List the first three actions.

You can also practice the order of hypotheses with the text on differential diagnosis with clues. Those who create cases can contribute through the Creators page.

Frequently asked questions

What is MRD?

It is measurable residual disease, that is, the leukemic cells left over after treatment in an amount too small to show up under the microscope. It can be measured by flow cytometry or by PCR.

Why does the FLT3 result need to come out fast?

Because in the RATIFY study midostaurin was given starting on the eighth day of induction (Stone et al., 2017). The result needs to be ready before that day.

Does ELN 2022 decide the transplant on its own?

No. It defines the genetic risk group, which is one of the bases of the decision. MRD, age, other diseases, the donor and the patient's preference also come into it.

Flow cytometry or PCR: which is better?

It depends on the subtype. PCR is more sensitive when there is a molecular target, such as the NPM1 mutation, and flow cytometry is fast and serves most patients. The two complement each other (Alalwy et al., 2026).

Can I start chemotherapy before the genetics?

Often yes, especially in severe patients. What matters is that the tests that change the first week, such as FLT3, have a defined deadline.

Final checklist

• I collected marrow for karyotype, mutation panel and flow cytometry at diagnosis.
• I ordered FLT3 with a deadline before the eighth day.
• I noted the expected date of each result.
• I classified the risk by ELN when the genetics arrived.
• I agreed on the timing and the method of MRD.
• I started the transplant evaluation early when the risk indicated it.
• I explained to the patient what each test changes.

Want to practice reasoning with progressive clues? Play the case of the day.

References

1. Alalwy A, Alshreef BS, Semoud I, et al. Diagnostic Accuracy and Prognostic Value of Flow Cytometry Versus PCR-Based Minimal Residual Disease Assessment in Leukemia: A Systematic Review and Meta Analysis. Cureus. 2026. PMID 42633347.
2. Berdel AF, Ronnacker J, Wenge DV, et al. Impact of European LeukemiaNet-guided postremission therapy on outcomes of patients with AML from 2010 to 2022. Blood Neoplasia. 2026. PMID 41509995.
3. Bill M, Eckardt JN, Döhner K, et al. Differential prognostic impact of myelodysplasia-related gene mutations in a European cohort of 4978 intensively treated AML patients. Leukemia. 2026. PMID 41145671.
4. Döhner H, Weber D, Krzykalla J, et al. Midostaurin plus intensive chemotherapy for younger and older patients with AML and FLT3 internal tandem duplications. Blood Adv. 2022. PMID 35486475.
5. Elhassadi E, O'Rourke E, Bannaga A, et al. Genomic and clinical determinants of outcome in Irish AML: a real-world analysis (2007-2025), single centre experience. Ann Hematol. 2026. PMID 42217025.
6. Ferrari C, Coulleray B, Kerneves P, et al. Optimization of Analytical Performances and Routine Laboratory Implementation of NPM1 Mutation Detection by Digital PCR in the Diagnosis and Monitoring of NPM1 -Mutated Acute Myeloid Leukemias. Int J Lab Hematol. 2026. PMID 42298870.
7. Knapper S, Dillon LW, Babu M, et al. CPX-351 vs daunorubicin, cytarabine, and gemtuzumab ozogamicin in older adults with non-adverse-risk AML: the NCRI AML18 trial. Blood. 2026. PMID 41237344.
8. Levis MJ, Hamadani M, Logan BR, et al. Regional practice differences significantly affect the benefit of post-HCT gilteritinib in FLT3-ITD AML. Blood Adv. 2026. PMID 42532509.
9. Romero Riquelme MA, Aspas Requena G, Lloret-Madrid P, et al. Real-World Outcomes of Midostaurin Plus Intensive Chemotherapy in FLT3-Mutated AML: The PETHRATIFY Study. Am J Hematol. 2026. PMID 42671173.
10. Sierra J, Montesinos P, Thomas X, et al. Midostaurin plus daunorubicin or idarubicin for young and older adults with FLT3-mutated AML: a phase 3b trial. Blood Adv. 2023. PMID 37581981.
11. Stone RM, Mandrekar SJ, Sanford BL, et al. Midostaurin plus Chemotherapy for Acute Myeloid Leukemia with a FLT3 Mutation. N Engl J Med. 2017. PMID 28644114.