Intelligence

Analysis

AML Flow Cytometry MRD After Lower-Intensity Therapy: Prognostic Signal, Not a Treatment Rule

Haaksma and colleagues connect measurable residual disease after three cycles to subsequent AML relapse; a restricted remission cohort means this association must not become an unvalidated treatment rule.

Technical Review · DOI: 10.1002/ajh.70291. Source peer-reviewed correspondence with original retrospective analysis; CFCM editorial synthesis, not independent clinical validation.

Why the treatment context matters

An MRD threshold developed in intensive treatment is not automatically transferable to older patients receiving less intensive AML therapy. Antecedent myelodysplasia and clonal haematopoiesis can complicate interpretation of unusual phenotypes, making diagnostic context part of the assay question.

What the original 2026 analysis actually measured

In a retrospective analysis of HOVON135 and HOVON155, Haaksma and colleagues characterised diagnostic leukemia-associated immunophenotypes in 231 patients. In 29 paired diagnosis/relapse cases, diagnostic LAIPs remained identifiable, but 13 patients also showed new phenotypes at relapse. This combination argues for reviewing both stable and evolving populations rather than relying on an unchanged single gate.

A prognostic association with a narrow denominator

Only 65 patients met the remission and MRD assessment criteria; 56 were assessed after three cycles and 38 at nine months. An MRD-positive definition of at least 0.1% after cycle three was associated with higher cumulative incidence of relapse (subdistribution HR 2.65, 95% CI 1.34–5.24). The corresponding unadjusted overall-survival comparison was not statistically significant. These estimates do not license a treatment recommendation from a single assay result.

What should an expert laboratory verify?

Report the LAIP definition and parent denominator, available event count, specimen adequacy, exact sampling timepoint, and relevant therapy. Evaluate whether the tested regimen and population resemble those used in an intended clinical claim. The study is a peer-reviewed original analysis presented as correspondence, not a prospective validation of a treatment algorithm.

Original CFCM conceptual three-part reading guide: diagnostic LAIP context, cycle-three MRD measurement, prognostic association versus treatment decision; not reproduced patient data or a validated clinical algorithm.
Original CFCM conceptual diagram. Educational synthesis; not clinical validation.

Daily Learning Pill

A prognostic association can be real while the treatment action remains unvalidated. Report what the flow assay measured before proposing what a clinician should change.

Explore the related CFCM resources and troubleshooting guides. Full primary source: read the paper.