Analysis
Clinical Flow
Collision marrow: distinct AML and Waldenström-related trajectories
A one-patient AML/WM-related case shows why flow and serologic evidence must remain compartment-specific when single-cell analyses are added.
One marrow can contain more than one hematologic malignancy. Cao, Lu and Huang examined an unusual patient with AML and a Waldenström-related lymphoplasmacytic population, integrating routine flow cytometry, serum markers, MYD88 testing and diagnostic single-cell RNA-seq. The value is the separation of different kinds of evidence, not a new diagnostic algorithm.

Two diagnoses before transcriptomic inference
The patient was 67 years old, with 44% blasts in peripheral blood, an IgM-kappa monoclonal serum band and marrow findings consistent with AML of myelomonocytic phenotype plus a restricted B/plasma-cell population. Flow cytometry and orthogonal MYD88 L265P supported the WM-related process. MYD88 is supportive in this setting, not diagnostic in isolation. The report notes incomplete quantification of lymphoplasmacytic infiltration and limited evaluable metaphases, with no comprehensive myeloid sequencing.
What the single-cell atlas contributes
Single-cell clustering separated AML-related myeloid, WM-related B/plasma, T/NK, monocyte/niche and erythroid/megakaryocytic regions. AML-associated areas carried immature/progenitor, granulocytic-primed and inflammatory/myelomonocytic transcriptional features. The other population comprised B-cell-like and plasma-cell-like states. Computational copy-number inference placed the dominant predicted CNV burden in AML-associated states but did not uniformly capture the mature B/plasma clone.
This difference is not a refutation of flow-defined clonality. Transcriptome-derived CNV algorithms may be insensitive to some clonal alterations or cell states. More importantly, these data alone cannot prove whether both disease compartments have a common ancestor or a specific causal relationship.
Responses are not interchangeable
After azacitidine plus venetoclax the reported marrow AML blast proportion fell from 35.9% to 0.2% at early follow-up; the abnormal B-cell fraction persisted at approximately 0.1%. Later the AML proportion rose to 4.4% while a small B-cell fraction remained. Serum IgM did not mirror the early AML change. Such numbers refer to different biological populations, denominators and time points, not a single overall remission scale.
AML-directed therapy and baseline-only single-cell profiling do not establish a WM-specific venetoclax-resistance mechanism. The computationally inferred ligand–receptor programs and KLRC1/NKG2A immune state are hypotheses without functional confirmation in this patient.
CFCM interpretation
Document each population's phenotype, gating strategy and sample fraction separately. Align flow-based measurements with the relevant serum and molecular assay, while preserving uncertainty about common clonality, resistance and the significance of small residual populations. A single-case observation cannot define universal treatment monitoring or a clinical assay limit.
Daily Learning Pill: A composite marrow needs compartment-specific denominators and orthogonal evidence; an overall blast percentage can hide a persistent B-cell clone.
Primary evidence: Cao X, Lu W and Huang H, Frontiers in Oncology 16:1844537 (11 September 2026), DOI 10.3389/fonc.2026.1844537. Original CFCM artwork; no paper figures reused.