Oncofusion

start / graph / line-per-485-kmt2a-aff1-infant-relapse

Explainerstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q21-index-fusion-models

PER-485 (CVCL_W984) — the relapse half of a matched infant t(4;11) pair, and it has lost CD19

also called PER-485A · CVCL_W984

PER-485 was grown from the same baby as PER-494, but after her leukaemia came back. The relapse cells have thrown away the B-cell marker CD19 that most modern immunotherapies aim at.

The same thing, for a scientist

PER-485 (CVCL_W984) is the relapse counterpart of PER-494 (CVCL_A8JU), both from patient P272, female, diagnosed at 336 days and relapsed in bone marrow 3 months later; Cellosaurus records the same-individual relationship explicitly (OI CVCL_A8JU). The line was derived from peripheral blood and carries der(11)t(4;11)(q21;q23) within a complex karyotype: 47,X,-X,add(1)(p36.1),-4,+5,del(5)(q12q33),+6,add(7)(p13),add(8)(q24.3),der(11)t(4;11)(q21;q23),+mar. Its immunophenotype is the striking part: **CD19 <1%**, CD10 4%, HLA-DR 6%, iCD79a <1%, against CD4 100%, CD15 100%, CD33 99%, CD11b 65%, CD11c 60%, CD65 86% — a monocytoid phenotype. The matched primary relapse sample was still CD19 69% and CD33 66%, so the line is further along the myeloid axis than the patient's relapse was. Doubling time 48-53 h.

Go deeper
A matched diagnosis-relapse pair in an infant with the index fusion

This is a rare resource: two continuous lines, same child, same fusion, diagnosis and relapse, both infant-derived 12. The diagnostic line PER-494 is CD19 100%/CD10 100%; the relapse line PER-485 is CD19 <1%/CD15 100%. That pair brackets the lineage-switch phenomenon this graph already holds entries on, in human material from one patient, without any CD19-directed therapeutic pressure — the child was treated on CCG 1883 and CCG 0922 in 1992, decades before blinatumomab or CD19 CAR-T.

That is the most important single fact about this line. Lineage switch in KMT2A::AFF1 can happen without CD19-directed immunotherapy, which bears directly on whether the switch seen after blinatumomab is caused by the drug or merely revealed by it.

But the switch may be partly a culture phenomenon

The patient's relapse marrow was CD19 69%. The line derived from it is CD19 <1% 1. So culture selected a CD19-negative subpopulation, or drove further differentiation, or both. The honest reading is that this line demonstrates the capacity of an infant KMT2A::AFF1 clone to become CD19-negative and myeloid, and does not measure how much of that happened in the child. The same paper's PER-703 line shows the identical pattern from a t(1;11) patient (CD19 95% in the patient, <1% in the line, CD14 100%), so it is a reproducible property of these derivations rather than a one-off.

Facts and figures
accessionCVCL_W984
resolved byCellosaurus accession lookup 2026-09-19; primary name read back as PER-485
age at derivation1 year (relapse; same patient diagnosed at 336 days)
infant derivedtrue
sexFemale
sitePeripheral blood
lineagemyelomonocytic; CD19-negative
fusionKMT2A::AFF1, der(11)t(4;11)(q21;q23)
karyotype47,X,-X,add(1)(p36.1),-4,+5,del(5)(q12q33),+6,add(7)(p13),add(8)(q24.3),der(11)t(4;11)(q21;q23),+mar
same individual asCVCL_A8JU (PER-494, diagnosis)
in depmapfalse
crispr screen publishedfalse
synonyms
[
  "PER-485A"
]
immunophenotype
{
  "CD19": "<1%",
  "CD10": "4%",
  "HLA-DR": "6%",
  "iCD79a": "<1%",
  "CD4": "100%",
  "CD15": "100%",
  "CD33": "99%",
  "CD11b": "65%",
  "CD65": "86%"
}
doubling time h
[
  48,
  53
]
How this could bite us back
  • CD19-directed therapy has no target in this line, so every CD19 CAR-T and blinatumomab result in KMT2A::AFF1 has an escape phenotype that already exists in cell culture.
  • The pair shows a lineage switch happening in a patient treated in 1992, before any CD19-directed agent existed. If the switch is intrinsic, combination strategies designed to prevent drug-induced switching are aimed at the wrong cause.
  • Do not count PER-485 and PER-494 as two infants. They are one.
Still open
  1. Is the PER-494/PER-485 pair clonally continuous at the fusion junction? If the two lines share one breakpoint sequence, they are one clone that switched lineage; if not, the relapse arose from a different subclone and the pair means something else entirely.
  2. Does PER-485, being CD19-negative, still depend on menin? A CD19-negative KMT2A::AFF1 cell that keeps the menin dependency would say the dependency is fusion-driven rather than lineage-driven, which is a cleaner test than any comparison across patients.
Every source behind this entry (2)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    PER-485 was established at relapse from patient P272 (female, 336 days at diagnosis, relapsed in bone marrow at 3 months, died of disease at 5 months); the line carries der(11)t(4;11)(q21;q23) in a complex karyotype and is CD19 <1%, CD10 4%, HLA-DR 6%, iCD79a <1%, CD4 100%, CD15 100%, CD33 99%, CD11b 65%, whereas the matched primary relapse sample was CD19 69% and CD33 66%.

    Preclinical Evaluation of Carfilzomib for Infant KMT2A-Rearranged Acute Lymphoblastic Leukemia · Cheung LC, de Kraa R, Oommen J, et al. · Frontiers in Oncology 2021
    PMID 33937032DOI
    What this evidence cannot show

    Full text read via PMC (PMC8082024), Tables 1-3. Line karyotyped in 2020, 28 years after the 1992 primary sample. The immunophenotype panels for patient and line were not run contemporaneously and several antibodies are blank for the primary sample, so the patient-versus-line comparison is incomplete for some markers.

    Evidence that argues against this (2)
    • Loss of CD19 in a cultured line is a well-known artefact class and an equally well-known real phenomenon, and this design cannot separate them. The same paper's PER-703 line, from a t(1;11) infant, shows the same shift (patient CD19 95% to line CD19 <1%, CD14 100%), which argues the derivation process itself selects for the myelomonocytic state. Cheung LC et al. 2021
    • A retroviral MLL-Af4 model shows that MLL-Af4 cells can assume a myeloid state under environmental pressure while retaining lymphoid-lineage potential, and the authors tie this to patients whose t(4;11) leukaemia evaded CD19-directed therapy by myeloid-lineage switch. That reading makes the switch a property of the fusion plus environment rather than of the drug, which cuts against attributing clinical lineage switch to blinatumomab specifically — and equally against attributing PER-485's phenotype to culture alone. Lin S et al. 2016
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    Cellosaurus accession CVCL_W984 records PER-485 as a female human cancer cell line, age at sampling 1 year, derived in situ from peripheral blood, disease 'Infant acute lymphoblastic leukemia with MLL rearrangement' (NCIt C122617), with microarray and RNA-seq transcriptomes, doubling times of 53 h and 48 h, and an originating-from-same-individual cross-reference to CVCL_A8JU (PER-494).

    Cellosaurus cell line PER-485 (CVCL_W984) · Bairoch A, et al. (SIB Swiss Institute of Bioinformatics) · Cellosaurus 2025
    What this evidence cannot show

    Fetched by accession 2026-09-19. Cellosaurus records the age as 1 year, which is the relapse age, not the diagnostic age of 336 days; the two records in this pair therefore carry different ages for one patient and neither is wrong. No gene-fusion annotation is present.

    Evidence that argues against this (1)
    • Because Cellosaurus stores no fusion annotation for PER-485, and stores its age as 1Y rather than as an infant diagnosis, this line is invisible both to a fusion-keyed query and to a strict age<1y filter. Any inventory of infant KMT2A::AFF1 material built from Cellosaurus fields alone will miss it, which is one mechanism by which this project came to believe no such material exists. Beesley AH et al. 2006
Connected to