Oncofusion

start / graph / concept-mrd-monitoring-ig-tcr-rearrangement

Explainerstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q34-mrd-monitoring

MRD by Ig/TCR rearrangement: the standard in ALL, and the assay infant KMT2A-rearranged disease strains hardest

also called Ig/TCR MRD · IG/TR MRD · ASO-PCR MRD · antigen receptor gene rearrangement MRD · allele-specific oligonucleotide PCR

The usual way to track childhood leukaemia is to read the unique antibody-gene rearrangement each B cell carries and use it as a barcode for the cancer clone. In babies with a broken KMT2A gene the barcode is a poor one: it is often missing, often several barcodes at once, and it can change while the leukaemia is being treated.

The same thing, for a scientist

Ig/TCR MRD is the default molecular MRD method in ALL: patient-specific allele-specific oligonucleotide real-time quantitative PCR against the junctional regions of clonally rearranged immunoglobulin and T-cell receptor genes, interpreted under the EuroMRD framework [E1], and the basis of risk stratification in virtually all frontline childhood ALL protocols [E2]. Infant KMT2A-rearranged ALL is the subtype where its assumptions fail hardest, and this was measured before it was explained. In 133 infants, Ig/TCR rearrangements were **less frequent and more oligoclonal** than in childhood precursor-B ALL, IGH rearrangements were present in 91% but IGK in only 21% and IGL in 13%, and MLL-AF4- and MLL-ENL-positive infants showed the most immature rearrangement patterns [E3]. Interfant-06 therefore made the genomic KMT2A breakpoint the preferred target, stating that earlier studies obtained more reliable MRD data with it than with the *occasionally absent, oligoclonal, or less stable* IG/TR rearrangements, and noting that false positivity of IG/TR PCR targets in regenerating marrow has been described [E4]. A fourth strain has now been characterised: in KMT2A::AFF1 the IGH locus shows proximally biased V(D)J with IGHV6-1 significantly elevated and more abundant de novo DJH recombination than other BCP-ALL subtypes, which the authors state has implications for MRD detection in clinical practice [E5]. That is treated on its own in `concept-proximal-vdj-bias-and-ig-mrd-in-kmt2a-aff1`.

Go deeper
What the assay assumes, and why this disease breaks each assumption

Ig/TCR MRD rests on four assumptions. In infant KMT2A-rearranged ALL each has a measured counter-example.

AssumptionWhat is measured in this disease
A clonal rearrangement existsIGH in 91% of infants, but IGK 21% and IGL 13%; cross-lineage TCRD 46%, TCRG 17%, TCRB 10%; less frequent overall than in childhood precursor-B ALL 3
The clone is monoclonalRearrangements in infant ALL were more oligoclonal than in childhood precursor-B ALL 3; in adult BCP-ALL, IGH clonal evolution at diagnosis was found in one third of 280 patients, and high prevalence and degree of clonal evolution driven by ongoing DH/VH-DJH recombination was associated with KMT2A rearrangement 6
The target is stable over timeInterfant-06 describes IG/TR rearrangements in this disease as occasionally absent, oligoclonal, or less stable 4
A positive signal means leukaemiaFalse positivity of IG/TR PCR targets in regenerating bone marrow has been described 4

The fourth is the one that produces the wrong clinical action rather than a missing result: a regenerating infant marrow after intensive induction is exactly the situation in which physiological rearrangements are abundant.

The cell of origin is the underlying reason

The rearrangement-pattern data point at one explanation. The immature Ig/TCR pattern in infant ALL correlated with young age at diagnosis, CD10-negativity and predominantly with the presence and type of MLL translocation, which the authors attribute to early prenatal transformation of an immature B-lineage progenitor with germline Ig/TCR genes combined with a short latency 3.

Independent genome-wide work reaches the same place from the repertoire side. Deep BCR sequencing of t(4;11)/MLL-AF4 infant samples (blasts >98%) found repertoires without significantly expanded VDJ-rearranged B-cell clones at diagnosis or relapse, against healthy peripheral blood; the comparator non-KMT2A BCP-ALL samples were all strongly clonal, with largest clones at roughly 3-40% of total BCR 7. If the founding cell is pre-VDJ, there may be no dominant rearrangement to use as a barcode - which is the most literal possible form of the target does not exist.

This also connects the MRD problem to this graph's cell-of-origin layer rather than leaving it as an assay curiosity. The MRD failure and the fetal-progenitor origin are the same fact seen from two instruments.

It is not useless, and the counter-case matters

Two results bound the pessimism. First, in Interfant-06 the overall agreement between KMT2A and IG/TR targets was high - 94% of 223 samples concordant, no significant discrepancy at end of induction (P = .60), only 7 of 90 samples differing by at least one log - so on most samples the two are interchangeable 4. Second, in KMT2A-germline infant ALL on the same protocol, where no KMT2A target exists and MRD is necessarily Ig/TR-based, end-of-induction MRD remained prognostic: 6-year DFS 61.4% for high MRD versus 87.9% for undetectable and 76.4% for intermediate 8.

The honest summary is that Ig/TCR MRD is a good assay applied to a subtype that violates its premises at a measurable but not catastrophic rate, and that the violations are concentrated in the direction of false negativity at the decision threshold: 19 Interfant-06 patients were positive above the cut-offs by KMT2A and negative by IG/TR, and 9 of them relapsed 4.

One technical partial rescue exists. Vdelta2-Jalpha rearrangements occur in 41% of childhood precursor-B ALL, show high sensitivity (10^-4 or better in most cases) and good stability (88% preserved at relapse), and have preferential Jalpha29 usage that is virtually absent in normal lymphoid cells 9. That is an unusually well-behaved Ig/TCR target - but 44% of them were themselves oligoclonal, and the study is childhood precursor-B ALL, not infants.

Facts and figures
assay classpatient-specific allele-specific oligonucleotide real-time quantitative PCR
analytegenomic DNA
targetjunctional region of clonally rearranged IGH/IGK/IGL/TRD/TRG/TRB
interfant06 roleadditional MRD-PCR target; the KMT2A genomic breakpoint was preferred
infant target availability PCT
{
  "IGH": 91,
  "IGK": 21,
  "IGL": 13,
  "TCRD": 46,
  "TCRG": 17,
  "TCRB": 10,
  "n_infants": 133,
  "source_pmid": "17268512"
}
failure modes
[
  "target absent (pre-VDJ cell of origin)",
  "oligoclonality at diagnosis",
  "ongoing DH/VH-DJH recombination and clonal evolution during therapy",
  "false positivity in regenerating marrow",
  "proximally biased IGHV usage narrowing primer design space"
]
discordance at cutoff interfant06
{
  "igtcr_negative_kmt2a_positive_n": 19,
  "relapses": 9,
  "source_pmid": "33405950"
}
better behaved target class
{
  "name": "Vdelta2-Jalpha",
  "frequency_childhood_bcp_all_pct": 41,
  "oligoclonal_pct": 44,
  "stable_at_relapse_pct": 88,
  "source_pmid": "14656882",
  "tested_in_infants": false
}
How this could bite us back
  • Ig/TCR MRD drives intensification in virtually every childhood ALL protocol, and in this subtype its errors at the decision threshold are asymmetric. In Interfant-06, 19 children were above the cut-off by the fusion target and below it by Ig/TR; 9 relapsed. Monitoring those children on Ig/TR alone would have withheld intensification from them.
  • False positivity in regenerating marrow points the other way and is the more insidious error, because it produces intensification of a child who does not need it, at an age where treatment-related mortality is a leading cause of death.
  • An oligoclonal diagnostic sample means the assay is built on one of several coexisting clones. Therapy selects between them, so a falling MRD value can record clonal replacement rather than response, and the surviving clone carries no assay.
  • The repertoire biology that breaks this assay is the same biology the graph cites for a pre-VDJ fetal cell of origin. If that is right, the assay is not merely difficult in this disease - it is measuring a feature the founding cell did not have, and improvements in Ig/TCR technology will not fix it.
  • Ig/TCR is the only molecular option in KMT2A-germline infant ALL and in any centre that cannot commission a breakpoint assay. Declaring it inferior without providing the alternative changes nothing clinically and risks eroding confidence in the only MRD result those children will get.
Still open
  1. How often does an Ig/TCR MRD target actually fail during therapy in infant KMT2A-rearranged ALL? Interfant-06 can only report concordance in samples where both targets existed, so the denominator that matters - patients in whom no usable IG/TR target could be established, or in whom it was lost - has never been published.
  2. Does the proximal IGHV bias change ASO primer design failure rates in practice? The bias is measured at diagnosis on RNA; whether it translates into failed or non-specific patient-specific assays is a question only an MRD laboratory's archive can answer.
  3. Would a Vdelta2-Jalpha29 target, which is virtually absent from normal lymphoid cells and stable at relapse in 88% of childhood precursor-B ALL, behave better in infants than an IGH target? Nobody has tested it in this subtype.
  4. Reconciling 91% IGH rearrangement detection with an absence of expanded VDJ clones on deep repertoire sequencing: are these the same cells, and does a clinically usable target exist in the patients whose repertoire looks polyclonal?
Every source behind this entry (9)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Guidelines for the interpretation of real-time quantitative PCR data for MRD analysis by Ig/TCR gene rearrangements were developed by the European Study Group on MRD detection in ALL, comprising 30 MRD-PCR laboratories worldwide, so that MRD data are interpreted identically between laboratories of the same protocol and comparably across protocols.

    PMID 17287850DOI
    What this evidence cannot show

    A consensus interpretive framework rather than a measurement of assay performance. It defines how to read a curve; it does not address whether a target exists in a given patient, which is the binding constraint in infant KMT2A-rearranged disease.

    Evidence that argues against this (1)
    • The guideline framework was written for real-time qPCR and needed adaptation for droplet digital PCR, where there are no standard curves and droplet generation and intra- and inter-assay variation must be controlled separately. Chen D et al. 2021
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    MRD monitoring is routine clinical practice in frontline treatment of virtually all childhood ALL and many adult ALL patients and is the strongest prognostic factor allowing risk-group assignment; flow cytometry and PCR analysis of rearranged immunoglobulin and T-cell receptor genes by allele-specific oligonucleotide PCR are the two established techniques, but classical flow cytometry does not reach a solid 10^-4 and classical ASO-PCR is time-consuming and labour intensive.

    PMID 25999452DOI
    What this evidence cannot show

    A review by the architects of the EuroFlow and EuroMRD programmes, so the framing of which technologies meet the criteria is not disinterested. It describes ALL as a whole and does not address infant KMT2A-rearranged disease specifically.

    Evidence that argues against this (1)
    • The claim that classical flow does not reach a solid 10^-4 is contested by the flow-versus-transcript comparison in 942 infant samples, where flow was only 2.0% uniquely positive against 10.5% for the molecular assay but identified rapid early response that the molecular assay did not. Popov A et al. 2023
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In 133 infants with ALL, immunoglobulin and T-cell receptor rearrangement analysis found IGH rearrangements in 91%, IGK in 21% and IGL in 13%, cross-lineage TCRD in 46%, TCRG in 17% and TCRB in 10%; compared with childhood precursor-B ALL, Ig/TCR rearrangements in infant ALL were less frequent and more oligoclonal; MLL-AF4 and MLL-ENL-positive infants showed immature rearrangements; and the immature pattern correlated with young age, CD10-negativity and predominantly with the presence and type of MLL translocation.

    PMID 17268512DOI
    What this evidence cannot show

    A cross-sectional diagnostic characterisation, not a longitudinal MRD study: it shows the targets are fewer and more oligoclonal at diagnosis and does not itself measure how often an MRD assay subsequently failed. Rearrangement detection used the methods of the mid-2000s, so some low-abundance rearrangements detectable by modern sequencing would have been scored absent.

    Evidence that argues against this (1)
    • Despite the immature and oligoclonal pattern, 91% of infants had a detectable IGH rearrangement, so the great majority do have at least one candidate Ig target; the problem is its quality and stability, not universal absence. Jansen MW et al. 2007
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In Interfant-06 the genomic KMT2A fusion breakpoint was the preferred MRD-PCR target because earlier studies obtained more reliable MRD data with it compared with the occasionally absent, oligoclonal, or less stable IG/TR rearrangements; false positivity of IG/TR PCR targets in regenerating bone marrow has been described; overall agreement between the two targets was nonetheless high (94% of 223 samples concordant, P = .60 at end of induction, 7 of 90 samples differing by at least one log); and when protocol cut-offs were applied 19 patients were positive by KMT2A and not IG/TR (9 relapses) against 12 the other way (1 relapse), P = .02.

    PMID 33405950DOI
    What this evidence cannot show

    Read in full text. The characterisation of IG/TR targets as occasionally absent, oligoclonal or less stable is the authors' summary of prior work, cited rather than measured in this paper. The 19-versus-12 comparison is post hoc with small numbers and wide confidence intervals. Concordance was computable only where both targets existed, so patients with no usable IG/TR target are excluded from the very analysis that would quantify how often that happens.

    Evidence that argues against this (1)
    • The primary agreement analysis in the same paper found no significant discrepancy between the two target types, and of the 7 samples differing by a log, 2 were higher by IG/TR - so the direction of failure is not uniform. Stutterheim J et al. 2021
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In 48 adult, 36 paediatric and 37 infant cases of BCP-ALL carrying t(4;11)/KMT2A::AFF1, whole-transcriptome sequencing showed strongly favoured usage of proximal IGHV genes in V(D)J recombination and more abundant de novo DJH recombination than in other BCP-ALL subtypes, and the authors state that this finding has implications for the detection of measurable residual disease in clinical practice because clonal evolution interferes with reliable detection of MRD based on IG rearrangements.

    PMID 38650597DOI
    What this evidence cannot show

    Read in full text. This is an RNA-sequencing repertoire analysis of diagnostic samples, not an MRD study: no remission samples were tested and no MRD assay failure was observed. The MRD implication is the authors' inference. Samples come from one commercial diagnostic laboratory plus a validation cohort, and the analysis is MiXCR read counts rather than DNA-based clone tracking.

    Evidence that argues against this (1)
    • Interfant-06 had already moved to the genomic breakpoint as the preferred target, on the stated ground that IG/TR rearrangements in this disease are occasionally absent, oligoclonal or less stable - three years before this mechanism was described. The finding explains a practice that was already in place rather than changing it. Stutterheim J et al. 2021
  6. E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In 280 adults with newly diagnosed BCP-ALL analysed by high-throughput sequencing using shared DNJ-stems to identify clonally related IGH sequences, IGH clonal evolution at diagnosis was identified in one third of patients, and in a subset of 167 patients with molecular subtype allocation a high prevalence and high degree of clonal evolution driven by ongoing DH/VH-DJH recombination was associated with the presence of KMT2A gene rearrangements, while VH replacements occurred more frequently in Ph-like and DUX4 BCP-ALL.

    PMID 37143651DOI
    What this evidence cannot show

    Adults, not infants, and the KMT2A-rearranged subgroup within 167 subtyped patients is small. Clonal evolution is inferred from sequence relationships at a single diagnostic timepoint rather than observed longitudinally, and the bioinformatic definition of a marker DNJ-stem is introduced in this paper rather than independently validated.

    Evidence that argues against this (1)
    • The same group's framework attributes much BCP-ALL clonal evolution to VH replacement, which was enriched in Ph-like and DUX4 rather than KMT2A-rearranged disease, so the mechanism operating in KMT2A-r (ongoing DH/VH-DJH recombination) is one of two and its MRD consequences may differ. Darzentas F et al. 2023
  7. E7moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Deep sequencing of B-cell receptor repertoires in t(4;11)/MLL-AF4-positive infant BCP-ALL peripheral blood samples with more than 98% blasts found no significantly expanded VDJ-rearranged B-cell clones at diagnosis or relapse compared with healthy peripheral blood, in contrast to five non-MLL BCP-ALL patients who were all significantly clonal with largest clones comprising roughly 3-40% of total BCR.

    PMID 30679323DOI
    What this evidence cannot show

    Read in full text. Four diagnosis-relapse pairs only, and the non-MLL comparator is five patients. Peripheral blood rather than marrow, and a PCR-based BCR sequencing method with molecular barcodes rather than a clinical DNA MRD assay - the absence of an expanded clone in this assay does not by itself prove that no ASO-PCR target could have been designed. The finding is presented as support for a pre-VDJ cell of origin, which is the paper's thesis.

    Evidence that argues against this (1)
    • A 133-infant clinical study found IGH rearrangements in 91% of infants with ALL, which is hard to reconcile with a complete absence of expanded clones and suggests the two assays are measuring different things - detectable clonal rearrangement versus relative clonal expansion within a repertoire. Jansen MW et al. 2007
  8. E8strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In 167 infants with KMT2A-germline ALL on Interfant-06, where MRD was measured by real-time quantitative PCR in 73 patients, those with high end-of-induction MRD at or above 5x10^-4 had a 6-year disease-free survival of 61.4% against 87.9% for patients with undetectable MRD and 76.4% for intermediate MRD (P = 0.02).

    PMID 34785111DOI
    What this evidence cannot show

    MRD was measured in only 73 of 167 patients and at varying timepoints, so this is a subset analysis. The paper does not separate which of those assays used Ig/TR targets, although by definition no KMT2A breakpoint target exists in KMT2A-germline disease. Event numbers are small (31 relapses in the whole cohort).

    Evidence that argues against this (1)
    • In the KMT2A-germline group age under 6 months was a favourable prognostic factor, the opposite direction to KMT2A-rearranged disease, so this cohort is not a clean control for assay performance in the rearranged subtype. Stutterheim J et al. 2022
  9. E9strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Clonal Vdelta2-Jalpha rearrangements were identified in 141 of 339 (41%) childhood and 8 of 22 (36%) adult precursor-B ALL; 44% of those in childhood precursor-B ALL were oligoclonal; Jalpha29 was used in 54% of rearrangements and Vdelta2-Jalpha29 rearrangements were virtually absent in normal lymphoid cells; and monoclonal Vdelta2-Jalpha rearrangements showed sensitivity of 10^-4 or less in most cases and were preserved at relapse in 88%.

    PMID 14656882DOI
    What this evidence cannot show

    Childhood and adult precursor-B ALL generally, not infants and not KMT2A-rearranged disease; the paper does not report a KMT2A-rearranged subgroup. The 88% stability figure is for this particular target class and does not transfer to IGH targets.

    Evidence that argues against this (1)
    • Cross-lineage TCRD rearrangements were found in only 46% of infants with ALL, and the infant pattern was more immature overall, so the subset of infants in whom a well-behaved Vdelta2-Jalpha target exists is smaller than the 41% figure from older children. Jansen MW et al. 2007
Connected to