Step 5: molecular confirmation and naming the partner
also called RT-PCR · nested RT-PCR · multiplex RT-PCR · LDI-PCR · long-distance inverse PCR · anchored multiplex PCR · targeted RNA sequencing · fusion transcript
The last step reads the actual joined sequence. Doctors copy the leukaemia cell's RNA and use primers that sit in KMT2A and in the suspected partner gene: if a product appears, the two genes really are stitched together, and sequencing it shows exactly where. When the partner is unknown or rare, cleverer methods work outwards from KMT2A without needing to guess, and the exact join they find becomes the fingerprint used to hunt for leftover leukaemia for years afterwards.
The same thing, for a scientist
Molecular confirmation closes the diagnosis by reading the fusion itself. Standardised RT-PCR primer sets, including for t(4;11)/MLL-AF4, reach a target sensitivity of 10^-2 in single PCR and 10^-4 in nested format, and multiplex panels screen the common fusions simultaneously [E1]. For unknown or rare partners the method of record is long-distance inverse PCR on genomic DNA, which identifies any KMT2A rearrangement whose breakpoint lies within the breakpoint cluster region without prior knowledge of the partner; in the original 40-patient validation it found three novel fusions (ACACA, SELB, SMAP1) and one MLL-TIRAP fusion arising from an interstitial deletion [E2]. Anchored multiplex PCR with targeted next-generation sequencing achieves the same partner-agnostic property on the RNA side, validated at 100% sensitivity and specificity against reference assays across 319 samples [E4]. The partner and exon junction matter because t(4;11)/MLL-AFF1 is a formal high-risk stratifier [E5], because the KMT2A breakpoint intron determines which domains the fusion retains and shifts with age [E3], and because trial eligibility is written in terms of a confirmed rearrangement. The patient-specific genomic breakpoint sequence established here is the MRD tracking target - see `concept-mrd-monitoring-genomic-breakpoint` [E2][E3].
RT-PCR: reading the fusion in the RNA2 deeper
The logic is simple enough to state in one sentence: if KMT2A exon 9 is spliced to AFF1 exon 4 in this cell's messenger RNA, then a forward primer in KMT2A and a reverse primer in AFF1 will produce a product, and no product exists in a cell where those genes sit on different chromosomes.
The practical version: RNA is extracted from the diagnostic marrow, reverse-transcribed to cDNA, and amplified with a fusion-spanning primer pair. Sequencing the product gives the exact exon junction. The European BIOMED-1 concerted action standardised this across 14 laboratories in eight countries for nine well-defined fusions, t(4;11) with MLL-AF4 among them, designing external primers for single PCR, internal primers for nested PCR, and deliberately 'shifted' primers upstream or downstream of the external pair so that a positive result could be independently confirmed or excluded in a second reaction 1.
The sensitivity figures from that standardisation are the ones still quoted: a minimal target sensitivity of 10^-2 for virtually all single PCR analyses and 10^-4 for nested PCR 1. That is to say a single round finds the fusion when roughly one cell in a hundred carries it, and a nested round when one in ten thousand does. At diagnosis, where blasts dominate the marrow, either is ample; the 10^-4 figure matters later, for residual disease.
Turnaround is one to three working days for a targeted panel, which places molecular confirmation last in the diagnostic sequence but still comfortably inside the first week.
Why the exon junction, not just the gene pair, is the answer
'KMT2A::AFF1' is not one molecule. It is a family of fusions that differ in where inside KMT2A the break fell, and the difference is structural.
The recombinome shows an age-dependent breakpoint shift: breakpoints within KMT2A intron 11 are associated with acute lymphoblastic leukaemia and younger patients, while breakpoints within intron 9 predominate in AML and in older patients 3. The authors read this as evidence of different aetiologies in different age groups, and they explicitly note that molecular characterisation of the breakpoints allows the retained functional domains of KMT2A to be correlated with clinical outcome 3.
So the junction is doing three jobs at once. It says which KMT2A domains survive in the fusion protein and which are lost with the reciprocal product, which is the input to every mechanistic argument about what the fusion does. It carries an aetiological signal about when and in which cell the break happened. And, most concretely, it defines the sequence against which patient-specific primers and probes will be designed for quantitative MRD.
A report that says only 'KMT2A rearranged' or only 'KMT2A::AFF1 positive' has discarded all three.
Multiplex panels, and the failure mode built into them
Because the common fusions are few, laboratories screen them together. The BIOMED-1 panel covers nine: t(1;19)/E2A-PBX1, t(4;11)/MLL-AF4, t(8;21)/AML1-ETO, t(9;22)/BCR-ABL p190 and p210, t(12;21)/TEL-AML1, t(15;17)/PML-RARA, inv(16)/CBFB-MYH11 and the 1p32 microdeletion with SIL-TAL1 1. International guidance describes exactly this approach - routine RT-PCR for the prognostically decisive abnormalities - as a basic yet informative screen 5.
The failure mode is structural and worth stating plainly: a primer-based panel finds only what its primers encode. Seven partner genes plus partial tandem duplications account for more than 90% of KMT2A recombinations, but 37 rearrangements recur and 63 have been characterised only once 3. A KMT2A::MLLT10 or a KMT2A::ACACA case run through a panel built for AFF1, MLLT3 and MLLT1 returns negative, and the negative is indistinguishable from a genuinely KMT2A-germline result unless a break-apart FISH or a partner-agnostic method has been run alongside. That is the precise mechanism by which an infant can be stratified as standard-risk on a test that could not have found their lesion.
When you do not know the partner: LDI-PCR and anchored RNA sequencing
Two methods solve the same problem from opposite ends of the central dogma: both need to know only one side of the junction.
Long-distance inverse PCR (LDI-PCR), on genomic DNA. Historically, only the most common KMT2A translocations were pursued by RT-PCR after cytogenetic identification, while infrequent or unknown rearrangements were simply excluded from further analysis 2. LDI-PCR was built to end that. Genomic DNA is digested with restriction enzymes, the fragments are circularised by ligation, and PCR primers sited in KMT2A point outwards around the circle - so amplification runs from known KMT2A sequence into whatever unknown sequence is now joined to it. Sequencing the product gives the partner gene and the exact genomic breakpoint to the base pair. Applied to 40 patients known to carry KMT2A abnormalities, it characterised 36 known fusions (34 der(11), 2 reciprocal alleles), identified three novel partners - ACACA, SELB and SMAP1 - and resolved one patient with a genomic KMT2A::TIRAP fusion produced by an interstitial deletion that had removed 123 genes across the intervening 13 Mb 2. It detects any KMT2A rearrangement whose breakpoint lies inside the breakpoint cluster region.
This is the method behind the recombinome catalogues: 1590 samples in 2013, 2345 in 2017 and 3401 by 2023, from which 107 in-frame KMT2A fusions have been assembled 3.
Anchored multiplex PCR with targeted next-generation sequencing, on RNA. The RNA-side equivalent uses one gene-specific primer plus a universal adapter ligated to the unknown end, so rearrangements are detected without prior knowledge of the fusion partners while also reporting single-nucleotide variants, insertions, deletions and copy-number change from the same library. Validation of a gene-rearrangement panel on 319 formalin-fixed samples gave 100% sensitivity (95% CI 96.5-100%) and 100% specificity (95% CI 99.3-100%) against reference assays, and across 986 clinical samples the method identified previously unknown therapeutically relevant fusions 4. It is compatible with low nucleic-acid input, which matters when the specimen is an infant's marrow.
The two methods are complementary rather than redundant. LDI-PCR returns a genomic breakpoint - the actual DNA join, including the intronic sequence that a transcript-based method never sees - which is what a DNA-based MRD assay needs. Anchored RNA sequencing returns the expressed junction, which is what tells you whether the fusion is in frame and productive; the recombinome records 16 out-of-frame fusions among the catalogued rearrangements, so the distinction is not academic 3.
Why the exact partner and junction matter clinically
Risk stratification. t(4;11)(q21;q23)/MLL-AFF1 is named explicitly among the abnormalities with the most significant impact on the treatment and management of B-cell precursor ALL and is used for high-risk stratification 5. The name on the report routes the patient to a protocol arm.
Trial eligibility. Menin-inhibitor programmes and other mechanism-directed trials enrol on confirmed KMT2A rearrangement, and the confirmatory assay and the partner identity appear on the eligibility documentation. In a disease affecting a few hundred children a year worldwide, a trial slot is not a marginal consideration - and an infant whose rearrangement was never characterised beyond 'break-apart positive' can be harder to enrol than one with a sequenced junction. The partner-agnostic methods above exist partly for this reason.
Mechanistic interpretation. Which KMT2A domains the fusion retains, and which are lost, depends on the breakpoint intron, and the recombinome authors state directly that molecular characterisation of the breakpoints allows the functional domains of the gene to be correlated with clinical outcome 3.
The hand-off to MRD. This is where the diagnostic pathway ends and surveillance begins. Both of the partner-agnostic methods were designed with this in mind. The BIOMED-1 standardisation was explicitly framed as being for molecular classification at diagnosis and for MRD detection during follow-up 1. The LDI-PCR paper notes that the determined patient-specific fusion sequences are useful for minimal residual disease monitoring of KMT2A-associated acute leukaemias 2. The 2023 recombinome states the same conclusion in the strongest terms: genomic breakpoint sequences from these patients were used to monitor MRD, and the work therefore translates directly from bench to bedside 3.
The operational point for this node is narrow. The diagnostic marrow is the only opportunity to establish that patient-specific genomic breakpoint sequence. What is then done with it - assay design, sensitivity, time points, the relationship between MRD level and outcome - belongs to concept-mrd-monitoring-genomic-breakpoint and is not covered here.
| hands off to | concept-mrd-monitoring-genomic-breakpoint |
|---|---|
| note on scope | This node establishes the patient-specific genomic breakpoint sequence. MRD assay design, sensitivity, time points and the MRD-outcome relationship are owned by concept-mrd-monitoring-genomic-breakpoint and deliberately not covered here. |
assays
[
{
"name": "single RT-PCR, standardised primers",
"template": "RNA (cDNA)",
"needs_known_partner": true,
"sensitivity": "10^-2 (minimal target, BIOMED-1)",
"gives": "presence of the fusion transcript",
"turnaround_days": [
1,
3
]
},
{
"name": "nested RT-PCR",
"template": "RNA (cDNA)",
"needs_known_partner": true,
"sensitivity": "10^-4 (minimal target, BIOMED-1)",
"gives": "presence of the fusion transcript at low burden",
"turnaround_days": [
1,
3
]
},
{
"name": "multiplex RT-PCR panel",
"template": "RNA (cDNA)",
"needs_known_partner": true,
"gives": "simultaneous screen of the common fusions; BIOMED-1 panel covers nine including t(4;11)/MLL-AF4",
"blind_to": "any KMT2A partner not encoded in the primer set",
"turnaround_days": [
1,
3
]
},
{
"name": "long-distance inverse PCR (LDI-PCR)",
"template": "genomic DNA",
"needs_known_partner": false,
"gives": "partner gene identity plus the exact genomic breakpoint to the base pair; includes intronic sequence",
"constraint": "breakpoint must lie within the KMT2A breakpoint cluster region",
"turnaround": "specialist centres; days to weeks"
},
{
"name": "anchored multiplex PCR with targeted NGS (RNA)",
"template": "RNA",
"needs_known_partner": false,
"gives": "expressed junction, in-frame status, plus SNVs, indels and copy-number change from one library",
"validated_performance": "100% sensitivity / 100% specificity vs reference assays across 319 FFPE samples (solid tumour validation)",
"turnaround": "days to about two weeks depending on batching"
}
]why the junction matters
[ "determines which KMT2A functional domains are retained in the fusion and which are lost", "carries an aetiological signal: intron 11 breakpoints associate with ALL and younger patients, intron 9 with AML and older patients", "distinguishes in-frame from out-of-frame fusions (16 out-of-frame among the catalogued rearrangements)", "defines the patient-specific sequence used to design quantitative MRD primers and probes", "names the entity that appears on the risk-stratification sheet and on trial eligibility documentation" ]
recombinome scale
{
"patients_characterised": 3401,
"in_frame_fusions_catalogued": 107,
"out_of_frame": 16,
"recurrent_rearrangements": 37,
"seen_once_only": 63,
"top7_plus_ptd_share_pct": 90,
"source": "Meyer 2023, PMID 37019990"
}- A primer-based panel returns negative for any partner it does not encode, and that negative looks identical to a genuinely KMT2A-germline result. Because seven partners cover over 90% of cases, this failure is rare, invisible on audit, and lands on exactly the patients whose biology is least understood.
- The genomic breakpoint established at diagnosis becomes the MRD target, so any clonal evolution that alters or deletes that junction produces false-negative MRD. A patient can be reported MRD-negative on a sequence the relapsing clone no longer carries.
- Lineage switch does not change the fusion, so a breakpoint-based MRD assay can stay positive while an immunophenotype-based one goes blind - or a fusion-negative subclone can drive relapse while the tracked target stays clear. The two MRD modalities fail in different directions and neither is complete.
- RNA is fragile. An infant's marrow transported to a reference laboratory can degrade enough to lose a transcript-based result, and there is often no second marrow to take. A DNA-based genomic breakpoint assay is more robust but reports a junction that may never be expressed.
- An out-of-frame or partner-less rearrangement - 16 out-of-frame fusions and 18 patients with nothing fused to 5'-KMT2A in the 3401-patient catalogue - breaks the assumption that a positive molecular result means a functional fusion protein and therefore a menin-inhibitor-sensitive lesion.
- KMT2A partial tandem duplications are not reachable by any single platform: banding cannot see them, a break-apart probe reads them as normal, and copy-number-based detection is confounded by overlapping unbalanced rearrangements. A complete-looking molecular work-up can still miss them.
- Trial eligibility written as 'KMT2A-rearranged, confirmed by a defined assay' converts a laboratory capability gap into an access gap: a child in a centre without partner-agnostic sequencing can be ineligible for a drug their biology predicts would work.
- What fraction of infants historically classified as KMT2A-germline on trial - and therefore treated on a lower-risk arm - would be reclassified as rearranged if their diagnostic material were re-analysed with LDI-PCR or anchored RNA sequencing? The 63 partners seen only once suggest the answer is not zero.
- Does partner identity change the optimal exposure or response duration for menin inhibition? Trials enrol on 'KMT2A-rearranged' as a single category, while the recombinome shows 107 in-frame fusions with different retained domains on the 3' side.
- Should partner-agnostic sequencing be the first-line molecular test in infants rather than a reflex after a negative panel, given that the pre-test probability of a KMT2A rearrangement in an infant with CD10-negative pro-B ALL exceeds 70%?
- How often does a KMT2A breakpoint fall outside the breakpoint cluster region that LDI-PCR covers, and is there any assay currently deployed that would find those cases?
- For an out-of-frame KMT2A fusion, what is the actual disease biology - and how should such a case be stratified, given that risk assignment is keyed to the cytogenetic name rather than to whether a functional fusion protein is produced?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.review
The European BIOMED-1 concerted action standardised RT-PCR primer sets across 14 laboratories in eight countries for nine well-defined fusion transcripts including t(4;11) with MLL-AF4, using external primers for single PCR, internal primers for nested PCR and shifted primers for independent confirmation, and achieved a minimal target sensitivity of 10^-2 for virtually all single PCR analyses and 10^-4 for nested PCR, explicitly for molecular classification at diagnosis and for MRD detection during follow-up.
Standardized RT-PCR analysis of fusion gene transcripts from chromosome aberrations in acute leukemia for detection of minimal residual disease. Report of the BIOMED-1 Concerted Action: investigation of minimal residual disease in acute leukemia · van Dongen JJ, Macintyre EA, Gabert JA, Delabesse E, Rossi V, Saglio G, Gottardi E, Rambaldi A, Dotti G, Griesinger F, Parreira A, Gameiro P, Diaz MG, Malec M, Langerak AW, San Miguel JF, Biondi A · Leukemia 1999What this evidence cannot show
A 1999 methods standardisation, now more than 25 years old; primer chemistry, reverse transcriptases and quantitative real-time formats have all moved on, and the quoted sensitivities are 'minimal target' figures achieved on cell lines and positive control patient samples in dilution experiments, not prospective clinical performance. The panel covers nine fusions, so by construction it says nothing about rare KMT2A partners. Gene names are given in the pre-HUGO nomenclature (MLL-AF4 rather than KMT2A::AFF1).
Evidence that argues against this (1)
- Nine fusion transcripts is the design limit. The same beat records 135 distinct KMT2A rearrangements with 94 partners characterised at molecular level, so a standardised nine-transcript panel confirms the common fusions and returns negative for the long tail — and a negative RT-PCR is easily read as absence of rearrangement rather than as absence of the tested fusion. Leukemia 2001
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Long-distance inverse PCR on genomic DNA identifies any KMT2A rearrangement whose breakpoint lies within the breakpoint cluster region without prior knowledge of the partner gene; applied to 40 leukaemia patients with known KMT2A abnormalities it characterised 36 known fusions (34 der(11), 2 reciprocal alleles), identified three novel partners - ACACA, SELB and SMAP1 - and resolved one patient with a genomic KMT2A::TIRAP fusion arising from an interstitial deletion that removed 123 genes across a 13 Mb interval; the resulting patient-specific fusion sequences are useful for minimal residual disease monitoring.
Diagnostic tool for the identification of MLL rearrangements including unknown partner genes · Meyer C, Schneider B, Reichel M, Angermueller S, Strehl S, Schnittger S, Schoch C, Jansen MWJC, van Dongen JJ, Pieters R, Haas OA, Dingermann T, Klingebiel T, Marschalek R · Proceedings of the National Academy of Sciences of the USA 2005What this evidence cannot show
A 40-patient methods validation in patients already known to carry KMT2A abnormalities, so it demonstrates analytic reach rather than clinical sensitivity in an unselected cohort. The stated coverage is explicitly conditional - any rearrangement 'if located within the breakpoint cluster region' - so breakpoints outside that region are missed by design, and the method is labour-intensive and concentrated in a small number of specialist centres rather than available in routine diagnostic laboratories.
Evidence that argues against this (1)
- Long-distance inverse PCR requires the breakpoint to lie inside the breakpoint cluster region, and this project's own breakpoint entries record that breakpoint position SHIFTS with age — infants toward introns 9-10, older patients toward intron 11 — with rearrangements outside the classical BCR documented. The assay's coverage is therefore not uniform across the patients it would be used on. Blood Cells, Molecules and Diseases 2010
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Genomic breakpoint characterisation of 3401 acute leukaemia patients has assembled 107 in-frame KMT2A fusions plus 16 out-of-frame fusions, with the seven most frequent partner genes and partial tandem duplications accounting for more than 90% of recombinations, 37 rearrangements recurring and 63 seen only once; the earlier 2345-patient series showed an age-dependent breakpoint shift with KMT2A intron 11 breakpoints associated with ALL and younger patients and intron 9 breakpoints predominating in AML and older patients, allowing retained functional domains to be correlated with outcome; genomic breakpoint sequences from these patients were used to monitor minimal residual disease.
The KMT2A recombinome of acute leukemias in 2023. · Meyer C, Larghero P, Almeida Lopes B, et al., Marschalek R · Leukemia 2023What this evidence cannot show
A referral-based diagnostic series accumulated 2003-2022 at one centre, so partner frequencies reflect referral and the reach of LDI-PCR rather than a population distribution, and rearrangements with breakpoints outside the cluster region are under-represented by construction. The intron 9 versus intron 11 age shift is reported in the 2017 companion paper (PMID 28701730, DOI 10.1038/leu.2017.213), which shares patients with the 2023 update, so the two are not independent. The domain-outcome correlation is asserted as an enabled analysis, not demonstrated within these papers.
Evidence that argues against this (1)
- Not every KMT2A lesion is reachable by breakpoint-cluster-region sequencing at all. KMT2A partial tandem duplications are small intragenic insertions explicitly described as particularly challenging to detect by standard molecular and cytogenetic approaches; in 932 AML cases, confirming the 41 PTDs required cross-checking targeted NGS against MLPA and optical genome mapping, and unbalanced rearrangements overlapping KMT2A generated false-positive calls. No single genomic platform fully ascertains them. Seto A et al. 2024
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Anchored multiplex PCR for targeted next-generation sequencing detects gene rearrangements without prior knowledge of the fusion partners, alongside single-nucleotide variants, insertions, deletions and copy-number changes, is compatible with low nucleic-acid input, and on validation of a gene-rearrangement panel across 319 formalin-fixed paraffin-embedded samples showed 100% sensitivity (95% confidence limit 96.5-100%) and 100% specificity (95% confidence limit 99.3-100%) against reference assays, identifying previously unknown therapeutically important fusions across 986 clinical samples.
Anchored multiplex PCR for targeted next-generation sequencing · Zheng Z, Liebers M, Zhelyazkova B, Cao Y, Panditi D, Lynch KD, Chen J, Robinson HE, Shim HS, Chmielecki J, Pao W, Engelman JA, Iafrate AJ, Le LP · Nature Medicine 2014What this evidence cannot show
The validation cohort is solid tumours in FFPE tissue - lung, glioblastoma, cholangiocarcinoma, thyroid - not haematological malignancy and not KMT2A. The 100%/100% result is against reference assays whose own sensitivity bounds the comparison, and the panel only reports fusions whose gene-specific primers are included on the KMT2A side, so 'partner-agnostic' means agnostic to the OTHER side of the junction only. Transfer of this performance to infant marrow RNA is an extrapolation, though the underlying chemistry is the basis of the anchored RNA panels now used clinically in leukaemia.
Evidence that argues against this (1)
- The validation is in the wrong specimen and the wrong disease: 319 formalin-fixed paraffin-embedded samples, which is solid-tumour practice. Leukaemia is diagnosed on fresh marrow or blood with high-quality RNA, so the FFPE validation neither transfers nor is the limiting case — and the claim of detecting rearrangements "without prior knowledge of the partner" still depends on having an anchored primer in the right KMT2A exon. Leukemia 2001
- E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.review
t(4;11)(q21;q23)/MLL-AFF1 is among the abnormalities with the most significant impact on the treatment and management of childhood B-cell precursor ALL and is used for high-risk stratification, and these abnormalities can be routinely tested by reverse transcription polymerase chain reaction as a basic yet informative screen, with cytogenetics and particularly FISH as reliable alternatives depending on each protocol's technical approach.
Detection of prognostically relevant genetic abnormalities in childhood B-cell precursor acute lymphoblastic leukaemia: recommendations from the Biology and Diagnosis Committee of the International Berlin-Frankfurt-Munster study group · Harrison CJ, Haas O, Harbott J, Biondi A, Stanulla M, Trka J, Izraeli S · British Journal of Haematology 2010What this evidence cannot show
Consensus recommendations from 2010, predating routine RNA sequencing and anchored multiplex panels, and written for childhood B-cell precursor ALL generally rather than for infants specifically. It endorses RT-PCR as a screen without quantifying the miss rate for rare partners, which is the principal limitation relevant to this node. It also does not address trial-eligibility requirements, which are a later development.
Evidence that argues against this (1)
- This is the same review cited elsewhere in the beat for the same sentence, so it adds no independent support. RT-PCR as a basic screen is a recommendation about what to run first, not a measurement of sensitivity, and the review does not report what fraction of KMT2A rearrangements its recommended panel would miss. Blood Cells, Molecules and Diseases 2010
- ← related toCBFB::MYH11 (CBFbeta-SMMHC), inv(16)(p13;q22)
Assay-uniformity bound (Yang 2021, PMID 34771519): among 1629 CBFB break-apart FISH tests, 271 CBFB rearrangements included three with a NEW partner gene, one insertion, one rare variant and eight 3-prime CBFB deletions, so a single RT-PCR primer set mis-assigns a measurable fraction.
RNA-level confirmation alone cannot distinguish a rearrangement from a trans-spliced chimera
quantifies the false-positive burden of RNA-seq fusion calling in childhood ALL
the diagnostic assay a screen would have to build on, and its limits
the step that names the partner
- related to →MRD by genomic breakpoint: the only assay in this disease that is fusion-specific by construction
the diagnostic step that establishes the patient-specific MRD target
- ← related toMRD by fusion transcript: fast, standardised, partner-limited - and structurally blind to KMT2A*
the diagnostic use of the same chemistry
two independent molecular assays called the same fusion; reciprocal AFF1::KMT2A also detected
- ← supportsSkin first, blood later: a three-day-old referred for purple spots, with KMT2A::MLLT3 neonatal ALL
fusion transcript confirmed and GATA1 mutation excluded to rule out transient abnormal myelopoiesis
a case where the transcript was never confirmed, so the partner assignment rests on band position alone
RT-PCR confirmation of the partner, plus a genomic breakpoint PCR assay used for MRD
an insertion ins(5;11) rather than a reciprocal translocation, which a standard karyotype need not reveal
PCR named the partner where the karyotype could not, and the same assay then served as the quantitative MRD measure
sequence-level identity of the fusion across a complete change of immunophenotype is what distinguishes a switch from a second leukaemia
RNA fusion analysis confirmed the partner that the karyotype band could only suggest