Oncofusion

start / graph / ana-cd19-in-kmt2a-rearranged-patients-two-profiles-of-the-same-cohort-disagree

Analysisweak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.Q138-kmt2a-aff1-surfaceome

CD19 looks lower in KMT2A-rearranged patients in the RNA-seq arm of TARGET and identical in the microarray arm of the same study - so this graph should not record a CD19 deficit

also called CD19 in KMT2A-r, unresolved

There is a widespread belief that KMT2A-rearranged leukaemia carries less CD19 than other B-cell leukaemias, which matters because CD19 is what most immune therapy targets. Two measurements of the same patient group point in opposite directions, so this project records the question as open rather than picking the answer it expected.

The same thing, for a scientist

In TARGET Pediatric ALL Phase II, split on the study's `MLL_STATUS` attribute [E1]: | profile | MLL+ | MLL+ median | MLL- median | ratio | MLL+ above MLL- p90 | |---|---:|---:|---:|---:|---:| | RNA-seq RPKM | n=4 | 37.25 | 78.06 | **0.48** | 0 of 4 | | microarray | n=13 | 5,292.89 | 5,488.11 | **0.96** | 1 of 13 | The RNA-seq arm is suggestive: all four MLL-positive values (24.79, 31.44, 43.07, 78.09) sit at or below the MLL-negative median of 78.06. With n=4 that is a one-in-sixteen event under a null of no difference - **p = 0.0625, which does not reach 0.05**, and it is the entire basis of the apparent deficit. The microarray arm, with **three times the sample size from the same cohort**, gives a ratio of 0.96 and puts the MLL-positive median within 4% of the MLL-negative one. **The larger measurement shows no effect.** The honest conclusion is that this cohort cannot resolve it. A CD19 deficit in KMT2A-rearranged B-ALL may well be real - it is reported clinically and CD19-low or CD19-negative relapse after CD19-directed therapy is well documented in this subtype - but **these data do not establish it, and a deficit recorded from the RNA-seq arm alone would be a number picked because it agreed with expectation** [E2].

Go deeper
Why this is written up as a null rather than dropped

The RNA-seq result is the one that agrees with what this project already believes about CD19 in KMT2A-rearranged disease, and it is the one a beat looking for a story would report. It is also the smaller, weaker arm, and it fails its own significance test.

This project has recorded the general form of this failure before: a deflating caveat and an agreeable claim need the same standard of check, and the agreeable one usually gets waved through. Here the agreeable finding was checked against a second instrument in the same study and did not survive 1.

What survives is the cell-line observation, which is about something different and is not in conflict: in the three KMT2A::AFF1 lines CD19 is 229.6 (SEM), 400.9 (RS4;11) and 0.2 (MV4-11) nTPM. That is not a deficit; it is a complete absence in the monocytic line, and it is written up separately 3.

Facts and figures
scriptanalyses/kmt2a-aff1-surface-target-screen/run.py
statusrun
verdictUNRESOLVED. The larger arm of the same cohort shows no difference. No CD19 deficit is recorded by this project on these data.
result summaryTwo expression profiles of one cohort disagree about CD19 in KMT2A-rearranged B-ALL; the question is recorded as open.
microarray
ratio0.964ratio: 0.964mll_pos_n13.0mll_pos_n: 13mll_neg_n231.0mll_neg_n: 231mll_pos_median5,292.89mll_pos_median: 5292.89mll_neg_median5,488.108mll_neg_median: 5488.108
rna seq
{
  "mll_pos_n": 4,
  "mll_pos_values": [
    24.79,
    31.44,
    43.07,
    78.09
  ],
  "mll_pos_median": 37.253,
  "mll_neg_n": 168,
  "mll_neg_median": 78.055,
  "ratio": 0.477,
  "sign_test_p": 0.0625
}
microarray — the numbers behind the chart
{
  "mll_pos_n": 13,
  "mll_pos_median": 5292.89,
  "mll_neg_n": 231,
  "mll_neg_median": 5488.108,
  "ratio": 0.964
}
How this could bite us back
  • Reporting the 0.48 ratio alone would have created a graph entry that a later beat would have quoted as established, from n=4 at p=0.0625, in direct contradiction to the larger arm sitting in the same API response.
  • A null from an underpowered cohort is not evidence that CD19 levels are fine in this disease. If a clinician reads this entry as reassurance about CD19-directed therapy in KMT2A-rearranged infants, the entry has done harm; the claim is only that THIS cohort cannot answer it.
Still open
  1. Does a CD19 deficit at diagnosis exist in KMT2A-rearranged B-ALL when measured per blast by quantitative flow rather than in bulk expression? Antibody-binding-capacity measurements exist in the diagnostic literature and were not searched on this beat.
  2. Would adjusting for blast percentage reconcile the two profiles? The cohort carries day-8, day-15 and day-29 marrow blast fields that were not used here.
  3. The RNA-seq arm has 4 MLL-positive samples of the 48 the cohort labels MLL-positive. What happened to the other 44, and is the subset with RNA-seq a biased draw?
Every source behind this entry (3)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    In TARGET Pediatric ALL Phase II, CD19 median is 37.25 RPKM in 4 MLL-positive versus 78.06 in 168 MLL-negative samples in the RNA-seq profile (ratio 0.48), and 5,292.89 versus 5,488.11 in 13 versus 231 samples in the microarray profile of the same study (ratio 0.96).

    What this evidence cannot show

    n = 4 in the arm showing the difference. The two profiles are different platforms on overlapping patients, so they are independent measurements but not independent samples. All four MLL-positive RNA-seq values at or below the negative median gives p = 0.0625, above the conventional threshold.

    Evidence that argues against this (1)
    • Bulk tumour expression includes normal residual B cells, which are CD19-positive and are more abundant in samples with lower blast burden. A difference in blast percentage between arms would move CD19 without any per-blast difference existing. cBioPortal for Cancer Genomics 2026
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset

    This graph already holds entries on CD19-directed therapy and lineage switch in KMT2A-rearranged disease, including risk-lineage-switch-cd19-directed-therapy and target-cd19-car-t-infant-kmt2a-r.

    What this evidence cannot show

    A statement about this repository's contents on one date, verified by an id listing over all 2,570 entries rather than by recall.

    Evidence that argues against this (1)
    • Those entries concern antigen LOSS under therapeutic pressure, which is a different claim from a lower baseline level at diagnosis. Nothing here bears on them and they are not challenged by this null. cBioPortal for Cancer Genomics 2026
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    CD19 is 229.6 nTPM in SEM, 400.9 in RS4;11 and 0.2 in MV4-11, the three Cellosaurus-confirmed KMT2A::AFF1 cell lines.

    What this evidence cannot show

    MV4-11 is a monocytic AML line; its CD19 absence is a lineage fact, not a KMT2A::AFF1 fact, and it would be wrong to read it as a fusion-driven antigen deficit.

    Evidence that argues against this (1)
    • Two B-ALL lines and one AML line is not a sample from which any statement about typical KMT2A::AFF1 CD19 levels can be made. Human Protein Atlas 2026
Connected to