Oncofusion

start / graph / target-prevention-exposure-avoidance-infant-leukemia

Targetmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q1b-genesis2

Prevention by exposure avoidance: verdict NOT FEASIBLE, and the transplacental-exposure literature is much weaker than its prominence suggests

also called maternal diet hypothesis · dietary topoisomerase II inhibitors · primary prevention of infant leukaemia

There is a well-known idea that things a mother eats or is exposed to during pregnancy cause infant leukaemia. The evidence for it is thin: the studies are small, the risk estimates are wildly imprecise, and the best-conducted one mostly did not find the effect. No modifiable exposure has been established, and this is not a realistic prevention route.

The same thing, for a scientist

The transplacental dietary topoisomerase II inhibitor hypothesis is biologically coherent and epidemiologically weak, and those two facts are routinely conflated. **Mechanistically it is well founded**: etoposide and related agents cause therapy-related AML with 11q23/KMT2A rearrangements, treatment-related breakpoints cluster differently within the breakpoint cluster region from de novo breakpoints [E3], and the covalent topoisomerase II cleavage complex is a documented route to KMT2A breakage (`mechanism-top2-cleavage-at-kmt2a`). **Epidemiologically it is not established.** The original signal came from 84 case-control pairs, with an odds ratio of 9.8 for infant AML but a 95% confidence interval of 1.1 to 84.8 [E1] — an interval spanning nearly two orders of magnitude, which is a signal of imprecision, not of effect size. The largest and best-designed follow-up, a Children's Oncology Group study of 240 incident infant leukaemias and 255 controls, reported 'little evidence of an association between the specific DNAt2 index and leukemia overall and by subtype', with the AML(MLL+) trend not reaching significance (P for trend = 0.10) and the most consistent finding being a *protective* association with maternal fruit and vegetable intake [E2]. **Verdict: not feasible as an intervention.** There is no modifiable exposure with an effect size and evidence base sufficient to justify advising pregnant women, and writing this up as though there were would be a disservice.

Go deeper
Why the evidence is weak, specifically1 deeper

Five structural problems, each of which would be enough on its own:

  1. Recall of diet during pregnancy, collected after a child has been diagnosed with leukaemia. Differential recall by case mothers is the textbook bias in this design and it is not correctable.
  2. Confidence intervals spanning two orders of magnitude. OR 9.8 (1.1-84.8) 1 is compatible with almost any effect including a trivial one.
  3. The index is a construct. 'DNA topoisomerase II inhibitor-containing foods' was a list of ten foods identified a priori 1; the assignment of foods to that list, and the dose actually reaching a fetal haematopoietic cell, are both unvalidated.
  4. Subgroup dependence. The association appears only in the infant AML stratum, and not in ALL, where most KMT2A-rearranged infant leukaemia actually sits 12.
  5. The best study is largely negative. The COG study had three times the sample size and found little evidence overall 2. In a literature this small, that should dominate.
What the COG study did find, and why it is not a prevention target either

Spector et al. reported significant or near-significant inverse linear trends between maternal fruit and vegetable intake and infant leukaemia overall, ALL(MLL+) and AML(MLL-) 2. Maternal fruit and vegetable intake is correlated with essentially every socioeconomic and health-behaviour variable there is, so this is the least interpretable kind of observational nutrition finding. It is recorded here because the same paper is routinely cited for the harmful DNAt2 association while its own headline protective finding is ignored, which is a sign of selective reading in the secondary literature.

The one exposure that IS established, and why it does not help

Therapeutic topoisomerase II inhibitors cause KMT2A-rearranged secondary leukaemia. This is not in doubt and is reflected in the breakpoint distribution: treatment-related cases cluster in a distinguishable part of the breakpoint cluster region relative to de novo cases 3. But the exposure is a cancer drug given for a reason, the population is patients already receiving chemotherapy, and the avoidance strategy (use less etoposide) is already standard oncology practice rather than a research programme. It does nothing for the de novo infant leukaemias this project is about, which occur in pregnancies with no such exposure.

Facts and figures
modalitypopulation-level exposure avoidance
verdictNOT FEASIBLE
reasonNo modifiable exposure is established for de novo infant KMT2A-rearranged leukaemia. The dietary topoisomerase II inhibitor hypothesis is biologically coherent but epidemiologically weak, and the largest study is essentially negative.
what would change the verdictA prospective pregnancy cohort with biomarker-based exposure assessment (not recall) linked to cord-blood fusion screening, with enough births to observe KMT2A-positive clones as the outcome rather than leukaemia itself.
How this could bite us back
  • Publicising a weak dietary association would impose guilt on mothers of affected infants for an exposure that has not been shown to matter. That is a real harm and it is the main reason this entry states the verdict plainly.
  • If avoidance advice were issued and were wrong, it would also consume the public-health attention that a genuine prevention route would later need.
Still open
  1. Could the exposure question be reframed with the fusion, not the leukaemia, as the outcome? Screening cord blood for KMT2A fusions in a prospective pregnancy cohort with measured (not recalled) exposures would raise the outcome rate by orders of magnitude and remove recall bias entirely. It depends on first establishing whether cord-blood KMT2A screening is even possible (`epi-fusion-positive-newborns-versus-leukemia-risk`).
  2. Is there a genetic modifier of susceptibility to topoisomerase-II-induced KMT2A breakage, such as an NQO1 or TDP2 variant, that would identify a small high-risk group in whom avoidance advice would actually be worth giving? Candidate-gene work in this area exists and was not assessed by this beat.
  3. Has anyone formally meta-analysed the maternal-exposure infant-leukaemia literature with modern bias assessment? The field cites the 1996 result far more often than the 2005 near-null, which suggests the synthesis has never been done properly.
Every source behind this entry (3)
  1. E1weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper

    In 84 matched case-control pairs, increasing maternal consumption of foods classified as containing DNA topoisomerase II inhibitors was associated with infant AML (OR 9.8, 95% CI 1.1-84.8 for medium consumption; OR 10.2, 95% CI 1.1-96.4 for high), with no association for ALL.

    PMID 8932918DOI
    What this evidence cannot show

    84 cases, 97 matched controls; retrospective maternal dietary recall after the child's diagnosis; confidence intervals spanning nearly two orders of magnitude; the association is confined to the AML stratum; the authors themselves describe it as preliminary and as a justification for collecting dietary history in future studies rather than as a finding.

    Evidence that argues against this (1)
    • The much larger Children's Oncology Group study found little evidence of an association overall or by subtype, with the AML(MLL+) trend not statistically significant. Spector LG et al. 2005
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    In 240 incident infant acute leukaemias and 255 controls, there was little evidence of an association between a maternal dietary topoisomerase II inhibitor index and leukaemia overall or by subtype; the AML(MLL+) trend was not significant (P for trend = 0.10) and maternal fruit and vegetable intake was inversely associated with risk.

    Maternal diet and infant leukemia: the DNA topoisomerase II inhibitor hypothesis: a report from the children's oncology group · Spector LG, Xie Y, Robison LL, Heerema NA, Hilden JM, Lange B, et al. · Cancer Epidemiology, Biomarkers & Prevention 2005
    PMID 15767345DOI
    What this evidence cannot show

    Still a retrospective case-control design with food-frequency questionnaire recall after diagnosis, and random-digit-dial controls, which selects for reachable and cooperative households. The inverse fruit/vegetable association is heavily confounded by socioeconomic status. This study's value is as a large, well-powered failure to replicate, not as positive evidence of anything.

    Evidence that argues against this (1)
    • An earlier study by the same cooperative group, with similar methodology, found the opposite in the AML stratum. Pooling three Children's Cancer Group childhood-leukaemia studies (303 infant cases, 468 matched controls originally interviewed; follow-up dietary data on 84 cases and 97 matched controls), a composite variable of 10 foods identified a priori as containing topoisomerase II inhibitors showed no significant trend overall or in ALL, but a statistically significant positive association within the AML stratum (odds ratio 9.8, P for trend 0.04). The 2005 null and this positive result come from the same consortium and the same questionnaire tradition, so the discrepancy cannot be attributed to method, and the entry should not present the null as settling the hypothesis. Ross JA et al. 1996
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    11q23 breakpoints in treatment-related AML are distributed differently within the KMT2A breakpoint cluster region from those in de novo acute leukaemia, and correlate with scaffold attachment regions and topoisomerase II consensus binding sites.

    PMID 8634439
    What this evidence cannot show

    Breakpoint mapping by Southern blot in a modest series; the de novo versus treatment-related distinction has been refined by later, larger breakpoint studies. Cited here for the established causal exposure (therapeutic topoisomerase II inhibitors), not for the dietary hypothesis.

    Evidence that argues against this (1)
    • The topoisomerase II consensus correlation does not survive direct break mapping - this is the exact 'sequence feature marks the neighbourhood, not the cut site' error. Mapping etoposide-induced double-strand breaks across the MLL breakpoint cluster region showed they colocalise with the RNA polymerase II binding site and the transcription initiation region of a newly identified gene-internal promoter that coincides with breakpoint hot spot II, and explicitly NOT with a nearby topoisomerase II consensus sequence. The authors conclude the genomic instability of MLL is linked to transcriptional processes. Southern-blot correlation of breakpoint distribution with scaffold attachment regions and topo II consensus sites therefore describes where such sites happen to lie, not where the enzyme cuts. Scharf S et al. 2007
Connected to