CXCR4 function in KMT2A-rearranged blasts is an output of the fusion's own HOXA/MEIS1 program, so a menin inhibitor should switch off the niche axis without a CXCR4 antagonist

[
  "Menin inhibition in a KMT2A::AFF1 line reduces CXCL12-directed migration more than it reduces viability at the same exposure.",
  "SYTL1 moves with HOXA9/MEIS1 under menin inhibition in deposited RNA-seq, whether or not CXCR4 transcript does.",
  "The effect is absent or much smaller in a non-KMT2A-rearranged B-ALL line with low HOXA9/MEIS1.",
  "If the effect is real and functional, a menin inhibitor produces a transient early rise in circulating blasts in a PDX before marrow burden falls."
]

[
  "ana-cxcr4-under-menin-inhibition-geo-reanalysis"
]

[]

[
  "F1 - TRANSCRIPTIONAL ARM. Cell type: human KMT2A::AFF1 B-precursor ALL lines SEM and RS4;11. System: 7-day revumenib at a concentration verified in the same wells to reduce HOXA9 and MEIS1 transcript by at least 70% versus DMSO by RT-qPCR. Kill criterion: surface CXCR4 falls by LESS THAN 25% in geometric mean fluorescence intensity (anti-CXCR4 clone 12G5, unpermeabilised, isotype-subtracted) AND SYTL1 transcript falls by less than 25%, in 3 independent experiments. If the fusion's output program collapses by 70% and neither the receptor nor its trafficking partner moves by a quarter, CXCR4 is not an output of that program in the index fusion.",
  "F2 - FUNCTIONAL ARM, scored independently of F1 because the SYTL1 route can move function without moving transcript. Cell type: SEM and RS4;11. System: 5-micron transwell migration toward 100 ng/mL recombinant CXCL12 for 4 hours, after the same 7-day revumenib exposure as F1. Kill criterion: migration index is NOT reduced by at least 30% versus DMSO, in 3 independent experiments. F1 and F2 both failing kills the hypothesis outright; F1 failing while F2 succeeds relocates the mechanism to trafficking and the hypothesis survives in amended form.",
  "F3 - PUBLIC-DATA ARM, runnable this week with no bench work. Data: deposited human menin-inhibitor transcriptomes in KMT2A-rearranged AML, GSE281506 or an equivalent series carrying a menin-inhibitor-alone arm. Kill criterion: in the menin-inhibitor arm versus vehicle, HOXA9 and MEIS1 both fall with log2 fold change below -1 at adjusted p < 0.05 (confirming the perturbation worked) WHILE CXCR4 and SYTL1 both show absolute log2 fold change below 0.3 at adjusted p > 0.05. That is a scoreable negative in a system where the drug demonstrably did its job.",
  "F4 - IN VIVO MOBILISATION PREDICTION. Cell type: a KMT2A::AFF1 patient-derived xenograft. System: NSG mice at established leukaemic burden, revumenib at a published efficacious dose, peripheral blood sampled at 0, 24, 48 and 72 hours. Kill criterion: circulating human CD19+/CD45+ blasts per microlitre show LESS THAN a 2-fold rise over vehicle at every timepoint within 72 hours, before any fall in marrow burden. No early egress means menin inhibition is not detaching the blast, whatever the transcript does.",
  "F5 - THE SPECIFICITY FALSIFIER. Cell type: a non-KMT2A-rearranged B-ALL line with low HOXA9/MEIS1 (for example REH, ETV6::RUNX1). System: identical revumenib exposure and identical CXCR4 flow and transwell readouts to F1 and F2. Kill criterion: CXCR4 surface level or CXCL12-directed migration falls by at least as much as in SEM and RS4;11. An equal effect in a line whose HOXA/MEIS1 program the drug does not run through means the readout is an off-target or general-toxicity effect and not an output of the fusion program."
]