Oncofusion

start / graph / cell-type-monoblast

Cell typestrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q-cells-saturation

Monoblast — the immature monocyte that names FAB M5a, the dominant morphology of congenital KMT2A leukaemia

also called monoblast · monoblastic · acute monoblastic leukaemia · AML M5a

A monoblast is the youngest recognisable step on the way to becoming a monocyte: a large cell with a big pale nucleus that has committed to the monocyte path but cannot yet do a monocyte's job. When a leukaemia is made mostly of these cells it is called monoblastic, and that is the commonest appearance of leukaemia in newborn babies, most often with a KMT2A break.

The same thing, for a scientist

The monoblast (CL:0000040) is the earliest morphologically identifiable monocytic precursor, downstream of the GMP and upstream of the promonocyte and monocyte. It defines the FAB M5a category — acute monoblastic leukaemia, in which monoblasts constitute 80 percent or more of the monocytic component — as distinct from M5b, acute monocytic leukaemia, in which promonocytes and monocytes predominate [E1]. Its relevance here is that monoblastic morphology is the commonest presentation of neonatal leukaemia: in roughly two-thirds of patients neonatal leukaemia manifests as acute myeloid leukaemia, **frequently with monocytic/monoblastic characteristics**, and the most frequently observed molecular abnormality is **t(4;11)(q21.3;q23.3)/KMT2A-AFF1** [E2]. Two of the field's workhorse KMT2A cell lines, MOLM-13 and MOLM-14, are FAB M5a by their founding report [E3]. The category should nonetheless be handled with care: expert haematopathologists do not reliably separate monoblasts from promonocytes, and the current diagnostic practice is to combine them [E4].

Go deeper
What a monoblast is, and where it sits

Blood-cell maturation is a staircase, and each step has a name that comes from what the cell looks like down a microscope. On the monocyte staircase the steps are: granulocyte-monocyte progenitor (no distinctive appearance, identified by surface markers) → monoblast → promonocyte → monocyte → macrophage in tissue.

A monoblast is large, with abundant basophilic cytoplasm, a round nucleus with delicate chromatin and one or more prominent nucleoli. It stains positive for non-specific esterase, which is what distinguishes it cytochemically from a myeloblast heading for the neutrophil lineage. It has committed to the monocytic path but cannot yet phagocytose, present antigen, or perform the haem-iron recycling that its mature descendants do.

In the FAB scheme, a leukaemia is M5a (acute monoblastic) when monoblasts make up 80 percent or more of the monocytic cells, and M5b (acute monocytic) when the more mature promonocytes and monocytes predominate 1. Both are lumped together in most modern usage as 'monoblastic/monocytic', which is exactly how Cellosaurus annotates the KMT2A cell lines.

Why this is the face of congenital KMT2A leukaemia

Neonatal leukaemia — presenting within the first 28 days of life, and in most cases congenital — is rare, and in about two-thirds of patients the disease is acute myeloid leukaemia, frequently with monocytic/monoblastic characteristics. The most frequently observed cytogenetic abnormality across neonatal leukaemia is t(4;11)(q21.3;q23.3)/KMT2A-AFF1, followed by t(1;22)/RBM15-MKL1 and t(8;16)/KAT6A-CREBBP. Common clinical features include prominent hepatosplenomegaly and a high incidence of skin involvement, sometimes with no marrow disease at all, and some cases — particularly t(8;16) — undergo spontaneous remission 2.

That is a striking pairing for this graph. KMT2A-AFF1 is the archetypal B-lymphoblastic fusion of infant ALL, and it is also the commonest lesion in a neonatal population whose disease is usually monoblastic myeloid. The same fusion, at almost the same age, produces opposite lineages. Whatever determines that choice is the central question of mechanism-lineage-switch-kmt2a and hyp-menin-release-lineage-destination, and the neonatal series is the cleanest natural experiment available on it.

The category is not reliably measurable, and this constrains how it can be used

Enumerating blasts and promonocytes is essential to WHO classification of myelomonocytic neoplasms, and the field has tested how well it can be done. Eleven experienced haematopathologists blindly reviewed slides from 20 patients. For the sum of blasts plus promonocytes in marrow, agreement was good (Fleiss kappa 0.744; 0.949 in peripheral blood), and the AML-versus-CMML call agreed 74 percent of the time. But distinguishing abnormal from normal monocytes in peripheral blood did not achieve good concordance and showed strong evidence of systematic differences between observers (p < 0.0001). The study's own conclusion was to combine blasts and promonocytes into a single category 4.

A convolutional-neural-network study reached the same place from the other direction: models trained on 935 monocytic cells labelled by five morphologists could not separate monocyte, promonocyte and blast as three classes, but performed acceptably on two (monocyte versus promonocyte-plus-blast), reaching accuracy comparable to human reviewers 5.

The practical rule for this graph: 'monoblastic' is usable as a coarse lineage assignment and is not usable as a maturation-stage variable. Any analysis that treats M5a and M5b as separable strata is resting on a distinction that trained experts and trained networks both fail to make consistently.

Facts and figures
ontologyCL:0000040
locationBone marrow; in congenital and infant disease also skin (leukaemia cutis) and liver/spleen
immediate upstreamcell-type-gmp
immediate downstreamcell-type-monocyte
markers
[
  "non-specific esterase positive",
  "CD34 variable",
  "CD117 variable",
  "CD64+",
  "CD14- (acquired at the monocyte stage)",
  "myeloperoxidase negative or weak"
]
lineage potential
[
  "promonocyte",
  "monocyte",
  "macrophage"
]
fab categories
[
  {
    "code": "M5a",
    "name": "acute monoblastic leukaemia",
    "criterion": "monoblasts >= 80 percent of monocytic cells"
  },
  {
    "code": "M5b",
    "name": "acute monocytic leukaemia",
    "criterion": "promonocytes and monocytes predominate"
  },
  {
    "code": "M4",
    "name": "acute myelomonocytic leukaemia",
    "criterion": "mixed granulocytic and monocytic"
  }
]
kmt2a relevance
{
  "neonatal": "two-thirds of neonatal leukaemia is AML, frequently monocytic/monoblastic, and the commonest lesion across neonatal leukaemia is t(4;11)/KMT2A-AFF1",
  "cell_lines": "MOLM-13 and MOLM-14 are FAB M5a by their founding report",
  "measurement_caveat": "the M5a/M5b distinction is not reproducible between expert observers and should not be used as a stratification variable"
}
atlas refs
[
  {
    "label": "Neonatal leukaemia clinical and molecular review",
    "pmid": "29806701",
    "title": "Neonatal leukaemia.",
    "authors": "Roberts I, Fordham NJ, Rao A, et al.",
    "venue": "Br J Haematol",
    "year": 2018,
    "title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
  },
  {
    "label": "Interobserver concordance for blasts plus promonocytes",
    "pmid": "32297416",
    "title": "Concordance among hematopathologists in classifying blasts plus promonocytes: A bone marrow pathology group study.",
    "authors": "Foucar K, Hsi ED, Wang SA, et al.",
    "venue": "Int J Lab Hematol",
    "year": 2020,
    "title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
  }
]
How this could bite us back
  • Monoblastic congenital leukaemia can present in skin alone without marrow involvement, so a marrow-based response assessment can be clean while disease persists in a compartment nobody sampled.
  • Because expert observers cannot reliably separate monoblasts from promonocytes, a trial that stratifies or measures response by monocytic maturation stage will carry measurement error large enough to obscure a real effect.
  • If monoblastic identity carries the monocytic metabolic configuration, a venetoclax-containing backbone will be weakest in the youngest patients, in whom monoblastic disease is commonest.
Still open
  1. Why does KMT2A-AFF1 produce monoblastic myeloid disease in the first weeks of life and B-lymphoblastic disease months later? The neonatal series is the cleanest natural experiment on lineage determination available in this disease and has never been used as one.
  2. Do the t(8;16)/KAT6A-CREBBP neonatal cases that remit spontaneously share anything with KMT2A-AFF1 neonatal cases that do not? Spontaneous remission of a monoblastic congenital leukaemia is an unexplained phenomenon with direct therapeutic interest.
  3. Since the M5a/M5b distinction is unreliable, is there a quantitative surface or transcriptional maturation index that would let monocytic maturation stage be used as a variable in KMT2A-rearranged AML at all?
  4. Do skin-limited congenital KMT2A-rearranged presentations with no marrow disease represent a distinct biology or an earlier stage of the same one?
Every source behind this entry (5)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    The French-American-British cooperative group's revised criteria defined the monocytic AML subtypes, separating M5a (acute monoblastic, monoblasts predominating) from M5b (acute monocytic, more mature monocytic cells predominating) and M4 (acute myelomonocytic), on morphological and cytochemical grounds.

    Proposed revised criteria for the classification of acute myeloid leukemia. A report of the French-American-British Cooperative Group · Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DA, Gralnick HR, Sultan C · Annals of Internal Medicine 1985
    PMID 3862359DOI
    What this evidence cannot show

    1985 consensus criteria with no PubMed abstract; the exact percentage thresholds quoted in modern textbooks derive from this and its 1976 predecessor. Read: PubMed record only, no abstract exists.

    Evidence that argues against this (1)
    • The WHO 5th edition organises AML primarily by defining genetic abnormality, making 'AML with KMT2A rearrangement' its own entity and reserving differentiation-based categories for cases lacking a defining genetic lesion. FAB M5a is therefore a description of cell morphology, not a disease classification, and a KMT2A-rearranged case is classified by its fusion regardless of how monoblastic it looks. Read: abstract only. Khoury JD et al. 2022
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.review

    Neonatal leukaemia, defined as occurring within the first 28 days of life and in most cases congenital, manifests in two-thirds of patients as acute myeloid leukaemia, frequently with monocytic/monoblastic characteristics; the most frequently observed cytogenetic/molecular abnormality is t(4;11)(q21.3;q23.3)/KMT2A-AFF1, followed by t(1;22)/RBM15-MKL1 and t(8;16)/KAT6A-CREBBP, with prominent hepatosplenomegaly, frequent skin involvement sometimes without marrow disease, and spontaneous remission in some cases, particularly t(8;16).

    Neonatal leukaemia · Roberts I, Fordham NJ, Rao A, Bain BJ · British Journal of Haematology 2018
    PMID 29806701DOI
    What this evidence cannot show

    A narrative review rather than a primary cohort, so the two-thirds figure is a synthesis across small series and case reports and carries substantial ascertainment bias — neonatal leukaemia is rare and skin-limited or spontaneously remitting cases are under-ascertained. Down-syndrome transient abnormal myelopoiesis is explicitly excluded. Read: abstract only.

    Evidence that argues against this (1)
    • The lineage distribution reverses within months of the neonatal window. Across infant (under one year) KMT2A-rearranged leukaemia the dominant presentation is B-lymphoblastic, and multi-omics clustering of 84 infants placed cases across stages of B-lymphopoietic and early hemato-endothelial development, with the KMT2A-AFF1 group forming its own HOXA-type cluster. The monoblastic association is specific to the first weeks of life, not to KMT2A-AFF1, so it should not be generalised to infant ALL. Read: abstract only. Isobe T et al. 2022
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    MOLM-13 and MOLM-14, two of the most-used KMT2A-rearranged AML cell lines, were established from the peripheral blood of one patient at relapse of acute monocytic leukaemia FAB M5a, express monocyte-specific esterase and MLL-AF9 fusion mRNA, and can be induced to macrophage-like morphology by interferon-gamma with or without TNF-alpha.

    PMID 9305600DOI
    What this evidence cannot show

    Both lines came from AML that had evolved from myelodysplastic syndrome and carry trisomy 8 present already in the MDS phase, so their genetic background is not that of de novo paediatric KMT2A-rearranged disease. Read: abstract only.

    Evidence that argues against this (1)
    • Cellosaurus does not carry the monoblastic assignment for MOLM-14, listing only NCIt C9154 'Adult acute myeloid leukemia' and ORDO Orphanet_519 with no monocytic term, while giving MOLM-13 the monocytic terms. The M5a assignment for MOLM-14 therefore exists only in the founding paper, and any automated lineage annotation built from the database will disagree. Bairoch A; SIB Swiss Institute of Bioinformatics 2026
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Among 11 experienced haematopathologists blindly reviewing blood and marrow smears from 20 patients, agreement on the combined blast-plus-promonocyte count was good (Fleiss kappa 0.744 in marrow, 0.949 in blood) and the AML-versus-CMML classification agreed at 74 percent, but distinguishing abnormal from normal monocytes in peripheral blood did not achieve good concordance and showed strong evidence of differences between observers (p < 0.0001); the authors concluded that blasts and promonocytes should be combined into a single category.

    Concordance among hematopathologists in classifying blasts plus promonocytes: A bone marrow pathology group study · Foucar K, Hsi ED, Wang SA, Rogers HJ, Hasserjian RP, Bagg A, et al., Bueso-Ramos CE · International Journal of Laboratory Hematology 2020
    PMID 32297416DOI
    What this evidence cannot show

    Twenty patients, 12 CMML and 8 AML, with limited non-specific esterase staining available; the cases are predominantly CMML rather than KMT2A-rearranged AML, and agreement on the combined category was in fact good, so the result limits the granularity of the monocytic call rather than invalidating it. Read: abstract only.

    Evidence that argues against this (1)
    • The same study is also evidence that the coarse monocytic call is reliable: kappa of 0.744 in marrow and 0.949 in blood for blasts plus promonocytes, and 74 percent concordance on AML versus CMML, are good numbers. Treating 'monoblastic' as unusable would overread this result; what it shows is that the M5a/M5b split is unreliable, not that the monocytic assignment is. Read: abstract only (same paper). Foucar K et al. 2020
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Convolutional neural networks trained on 935 digital images of monocytic cells labelled by consensus of five experienced morphologists could not reliably predict three monocytic subtypes (monocyte, promonocyte, blast) but performed acceptably on two classes (monocyte versus promonocytes plus blasts), reaching accuracy comparable to human reviewers (0.78 +/- 0.10 versus 0.86 +/- 0.05).

    Classification of Monocytes, Promonocytes and Monoblasts Using Deep Neural Network Models: An Area of Unmet Need in Diagnostic Hematopathology · Osman M, Akkus Z, Jevremovic D, Nguyen PL, Roh D, Al-Kali A, et al., Salama ME · Journal of Clinical Medicine 2021
    PMID 34073699DOI
    What this evidence cannot show

    935 cells from a single institution, ground truth defined by morphologist consensus rather than by an independent standard, so the network's ceiling is the morphologists' own agreement. A failure of a CNN on 935 images is weak evidence that the distinction is intrinsically unmakeable rather than under-sampled. Read: abstract only.

    Evidence that argues against this (1)
    • Morphology is not the only available discriminator, so the difficulty demonstrated here does not mean the monoblast stage is unmeasurable. Flow-cytometric immunophenotype (CD34, CD117, CD14, CD64, CD11b) and single-cell transcriptomics both separate monocytic maturation stages on quantitative grounds, and a single-cell taxonomy of blood resolved four monocyte subtypes that morphology does not distinguish at all. The limitation is of light microscopy, not of the biology. Read: abstract only. Villani AC et al. 2017
Connected to