Peripheral blood AML MRD: promising concordance, different adequacy rules

What the study establishes — and what it does not

A prospective single-center study in Blood Advances evaluated peripheral-blood measurable residual disease (MRD) in acute myeloid leukemia using a 12-color flow-cytometry assay developed for marrow, with blood-specific preanalytics and interpretation. A 53-patient paired training set informed the criteria; the prospective cohort included 118 patients across treatment stages. Blood specimens underwent prelysis and were processed within 24 hours. Among 118 blood samples, 26 were MRD-positive, 86 negative and six inadequate. Agreement was assessed in 109 evaluable paired blood–marrow comparisons, not all 118 patients: 107 of 109 pairs agreed (98%). Eight pairs were excluded from that comparison because of inadequate results, and an extramedullary case was considered separately. These figures describe this assay and case mix; they do not establish universal interchangeability with marrow, transferable detection limits or independent clinical validation. Treatment heterogeneity, follow-up and clinical-risk confounding warrant further assessment. Primary reference: Blood Advances, DOI 10.1182/bloodadvances.2025018607.

CFCM interpretation: define the specimen’s intended use

First distinguish an interim blood-based surveillance signal from a proposed replacement for scheduled marrow testing. The acceptable uncertainty is different. A negative peripheral-blood result cannot be used as proof that the marrow contains no residual AML, particularly when the phenotype of interest is rare or poorly represented in circulation. Document the clinical question before choosing acquisition targets, quality flags and follow-up rules.

Build an adequacy framework before declaring a negative

Preanalytical acceptance should record EDTA collection, time to processing, transport conditions, viable-cell recovery and any prelysis effect. Analytical acceptance should be population-aware: total acquired nucleated events and the presence of evaluable reference populations are related but distinct measures. The laboratory should define, using its own data, when a blood result is negative, limited or inadequate. A marrow-specific CD34-positive precursor minimum should not be transplanted mechanically into blood, where such cells may normally be scarce.

Investigate discordance instead of averaging it away

For paired specimens, retain the native blood and marrow results and their acquisition metadata. Stratify disagreements by treatment phase, target phenotype, leukocyte recovery, collection interval, gating decision and possible therapy-associated immunophenotypic change. Review both leukemia-associated immunophenotypes and differences from normal background without assuming that either approach alone captures every evolved subclone. Document whether the discordance may arise from specimen biology, sampling, analytical sensitivity or interpretation; do not infer one mechanism merely because two numbers differ.

A practical validation plan for laboratories

Prespecify the intended patient population and treatment windows. Include paired positive, negative and low-level samples, specimen-quality challenges and clinically relevant phenotypic variants. Establish within-laboratory detection limits, precision near the low-count boundary, interoperator agreement and reproducibility after realistic transport and processing delays. Report confidence intervals for agreement and sensitivity, not only point estimates. Assess how often specimens fail adequacy requirements, and analyze those failures rather than omitting them from headline concordance. For clinical claims, prospective follow-up and independent cohorts must test whether an additional blood result improves decisions beyond the existing marrow pathway.

Reporting and escalation

A responsible report should specify specimen type, number and nature of evaluable events, MRD phenotype, assay-specific limit and quality qualification. Use an explicit category for insufficient material, rather than converting an uninterpretable sample into a negative. Design a clinician-facing escalation route for a low-level or unexpected positive blood result; this is a proposed workflow to validate, not an established treatment rule. Keep the raw classification and analytical reservations accessible for longitudinal review.

Daily Learning Pill — biological adequacy

Analytical sensitivity is constrained not only by instrument event counts but also by where the target cells reside. Normal immature myeloid precursors may be plentiful in marrow and scarce in peripheral blood. A large blood acquisition therefore does not guarantee equivalent sampling of the relevant biological compartment. Always interpret “not detected” together with specimen representativeness and the assay’s validated detection capability.

Evidence and reuse boundaries

This CFCM methodological synthesis is original editorial analysis based on the article's Introduction, Material and methods, Results and Discussion. The supplementary files were not separately inspected. No figure or prose from the CC BY-NC-ND 4.0 publication is reproduced; the accompanying conceptual diagram is CFCM original artwork. Source: DOI 10.1182/bloodadvances.2025018607.