Six relapses and a change of lineage: KMT2A::AFDN T-lymphoblastic lymphoma that came back as monoblastic leukaemia, and a seventh remission that held
also called t(6;11) T-LBL to AMoL lineage switch case · MLL-MLLT4 T-lymphoblastic lymphoma case
A 27-year-old man was diagnosed with a T-cell lymphoma growing as a mass in his chest; it came back five times, in his brain and spinal fluid and in his marrow. At the sixth relapse the disease had changed type entirely, from a T-cell lymphoma into a myeloid leukaemia - but the same genetic fault was in both, proving it was the same disease wearing a different face, and after one course of an antibody drug and a cord blood transplant he had been in remission for more than three years.
The same thing, for a scientist
A 27-year-old man developed T-cell lymphoblastic lymphoma with a mediastinal mass, carrying t(6;11)(q27;q23). He relapsed repeatedly in the central nervous system and marrow. At the fifth relapse the marrow contained 42.8% blasts expressing CD4, CD5, CD7, CD10, CD33, CD34, HLA-DR and cytoplasmic CD3. While in complete remission on nelarabine, a sixth relapse occurred two months later with 6.8% marrow blasts that were now monoblastic: myeloperoxidase, CD4, CD33, CD56, CD64 and HLA-DR positive, and negative for cytoplasmic CD3, CD5, CD7, CD10 and CD34. Marrow cells at the fifth (lymphoid) and sixth (myeloid) relapses carried the same t(6;11)(q27;q23) and the same MLL::MLLT4 (KMT2A::AFDN) fusion transcript, and the fusion sequence in the initial mediastinal cells matched that in peripheral blood at the sixth relapse. He received one course of gemtuzumab ozogamicin followed by cord blood transplantation and remained in a seventh remission for more than three years. [E1]
How it began, and what was found
With a mass in the chest. A 27-year-old man presented with T-cell lymphoblastic lymphoma and a mediastinal mass, carrying t(6;11)(q27;q23) 1.
That starting point already distinguishes this entry from every other in the arm. Nothing here is an infant with pallor and a big spleen. A mediastinal mass in a young adult is the presentation of T-lineage disease, and the KMT2A partner - AFDN at 6q27 - is one the graph carries as a canonical fusion. This is what fusion-kmt2a-afdn looks like when the breakpoint lands on the T-cell side.
The breakpoint position predicts the lineage. Work on t(6;11) separates two groups: breakpoints in the KMT2A major breakpoint cluster region, introns 9-10, associated mainly with AML M1/M4/M5, and breakpoints in the minor cluster region, introns 21-23, associated with T-ALL 2. This patient presented with T-lineage disease and relapsed with monoblastic disease, which is the pairing that work was trying to explain. The report does not state which cluster region this patient's breakpoint fell in, so the connection is suggested and not established here.
What was done and what happened
He relapsed repeatedly, in the central nervous system and in the marrow 1.
Fifth relapse, in the marrow: 42.8% blasts, expressing CD4, CD5, CD7, CD10, CD33, CD34, HLA-DR and cytoplasmic CD3. Cytoplasmic CD3 is the lineage-defining T-cell marker.
He achieved complete remission on nelarabine. Two months later, the sixth relapse: 6.8% marrow blasts with monoblastic morphology, positive for myeloperoxidase, CD4, CD33, CD56, CD64 and HLA-DR, and negative for cytoplasmic CD3, CD5, CD7, CD10 and CD34.
By the ordinary rules of lineage assignment, that is a different disease. Cytoplasmic CD3 positive to cytoplasmic CD3 negative, myeloperoxidase negative to myeloperoxidase positive - a T-lymphoblastic lymphoma had become an acute monoblastic leukaemia in two months.
It was not a different disease. Marrow cells at both the fifth and sixth relapses carried t(6;11)(q27;q23) and the same MLL::MLLT4 fusion transcript. More than that: the fusion sequence documented in the initial mediastinal cells was the same as that in peripheral blood at the sixth relapse 1. One clone, carried from a chest mass at presentation through five relapses to a monoblastic leukaemia, with its sequence unchanged.
Then it stopped. One course of gemtuzumab ozogamicin - a CD33-directed conjugate, given because the switched clone was now CD33-positive - followed by cord blood transplantation. He continued in a seventh remission for more than three years at the time of report.
Why this case is in the graph
Two reasons, and they pull in opposite directions.
First, it is the arm's clearest sequence-level proof of clonal identity across a lineage switch. The lineage-switch entry elsewhere in this arm rested on partial NG2 positivity. This one rests on the same fusion sequence in the mediastinal cells at presentation and in the blood at the sixth relapse - the two ends of the disease, separated by years and by a complete change of immunophenotype. Whatever a KMT2A fusion does to a cell, it does not require the cell to stay in its lane.
Second, it is the only durable survivor in the arm, and it was rescued after the switch, not before. More than three years in a seventh remission, after one course of an antibody-drug conjugate chosen on the basis of the antigen the switched clone had acquired. The switch, which is normally recorded as a disaster, was here also the thing that made the disease targetable: CD33 appeared and was used.
Three honest qualifications. This is an adult, so the infant biology may differ. The switch occurred during chemotherapy and nelarabine rather than under CD19-directed pressure, so it is not the same phenomenon as immunotherapy-driven escape - it says the plasticity is intrinsic, not induced. And one patient in a seventh remission is one patient.
| age at dx | 27 years |
|---|---|
| sex as reported | male |
| fusion | fusion-kmt2a-afdn |
| karyotype | t(6;11)(q27;q23) with an MLL::MLLT4 (KMT2A::AFDN) fusion transcript; breakpoint intron not stated |
| clonality evidence | same t(6;11)(q27;q23) and same MLL::MLLT4 fusion transcript at the fifth (lymphoid) and sixth (myeloid) relapses; the fusion sequence in the initial mediastinal cells matched that in peripheral blood at the sixth relapse |
| cns status | several relapses in the central nervous system |
| course | six relapses; fifth relapse with 42.8% marrow blasts of T-lineage phenotype; complete remission on nelarabine; sixth relapse two months later with 6.8% monoblastic blasts; one course of gemtuzumab ozogamicin, chosen because the switched clone expressed CD33, followed by cord blood transplantation |
| outcome | in a seventh remission for more than three years at the time of report; the longest documented survival in this arm; no later follow-up available |
| source type | case report |
| consent note | Published peer-reviewed case report on an adult patient. This entry records only what the publication discloses and omits the treating institutions from the body text. |
| privacy review | No detail added, inferred or reconstructed beyond the publication. |
presentation
[ "T-cell lymphoblastic lymphoma with a mediastinal mass", "subsequent relapses in the central nervous system and bone marrow" ]
immunophenotype fifth relapse
[ "CD4", "CD5", "CD7", "CD10", "CD33", "CD34", "HLA-DR", "cytoplasmic CD3" ]
immunophenotype sixth relapse
[ "myeloperoxidase", "CD4", "CD33", "CD56", "CD64", "HLA-DR" ]
immunophenotype sixth relapse negative
[ "cytoplasmic CD3", "CD5", "CD7", "CD10", "CD34" ]
- Lineage plasticity in KMT2A-rearranged disease does not require an immunotherapy to provoke it. Any strategy that assumes a stable immunophenotype - diagnostic classification, MRD panels, antigen-directed therapy - is building on something the clone can change under ordinary chemotherapy.
- A clone whose fusion sequence is unchanged from a mediastinal mass at diagnosis to a monoblastic marrow leukaemia six relapses later is a clone that survived every intervening therapy without needing to mutate. Genomic resistance screening would have found nothing to report at any point.
- If the same fusion can drive T-lineage and monocytic disease in the same patient, then a menin inhibitor's activity must be lineage-independent to be useful here, and every trial that enrols by disease label rather than by lesion is partly mis-stratified.
- Does the KMT2A breakpoint position predict which lineage a switch moves towards? The major-versus-minor cluster distinction predicts the lineage at diagnosis in t(6;11); nobody has asked whether it also predicts the direction of a later switch, and archived switched cases could answer it.
- This switch happened under chemotherapy and nelarabine, with no antigen-directed drug involved. What fraction of lineage switches in KMT2A-rearranged disease are spontaneous rather than immunotherapy-driven? Distinguishing the two changes whether CD19-directed therapy should be avoided or merely monitored.
- The switched clone acquired CD33 and was treated with a CD33-directed conjugate. Should a lineage switch trigger immediate re-phenotyping specifically to look for a newly targetable antigen, rather than being recorded only as a failure?
- Seven remissions in one patient implies the clone was repeatedly reducible but never eliminated until cord blood transplant. What did the transplant do that six prior treatments did not - graft-versus-leukaemia, or simply a deeper reduction?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
A 27-year-old man with T-cell lymphoblastic lymphoma and a mediastinal mass carrying t(6;11)(q27;q23) relapsed repeatedly in the central nervous system and marrow; at the fifth relapse the marrow contained 42.8% blasts positive for CD4, CD5, CD7, CD10, CD33, CD34, HLA-DR and cytoplasmic CD3, and two months later, while in complete remission on nelarabine, a sixth relapse contained 6.8% monoblastic blasts positive for myeloperoxidase, CD4, CD33, CD56, CD64 and HLA-DR and negative for cytoplasmic CD3, CD5, CD7, CD10 and CD34; marrow cells at both relapses carried t(6;11)(q27;q23) and the same MLL::MLLT4 fusion transcript, and the fusion sequence in the initial mediastinal cells matched that in peripheral blood at the sixth relapse; after one course of gemtuzumab ozogamicin followed by cord blood transplantation he continued in a seventh remission for more than three years.
Lineage switch with t(6;11)(q27;q23) from T-cell lymphoblastic lymphoma to acute monoblastic leukemia at relapse · Higuchi Y, Tokunaga K, Watanabe Y, et al. · Cancer Genetics 2016What this evidence cannot show
Abstract read, not full text; the abstract of this paper is unusually complete and carries the whole case, including both immunophenotypes, the clonality evidence and the outcome, but the presenting symptoms, blood counts at diagnosis, the KMT2A breakpoint intron and the details of the five earlier relapses are not available to this entry. A single case: it demonstrates that this happened and establishes no frequency, and a patient in a seventh remission after six relapses is published precisely because that is extraordinary. This is a 27-year-old adult with T-lineage disease, so nothing transfers directly to infant B-lineage leukaemia. The durability is reported as 'more than 3 years' at the time of writing in 2016; no later follow-up is available. The attribution of the sustained remission to gemtuzumab ozogamicin cannot be separated from the cord blood transplant that followed it.
Evidence that argues against this (2)
- Against reading the lineage switch here as the same phenomenon as immunotherapy-driven escape: this switch occurred under conventional chemotherapy and nelarabine, not under antigen-directed pressure, which argues that lineage plasticity in KMT2A-rearranged disease is intrinsic rather than created by CD19-directed drugs - and therefore that avoiding those drugs would not avoid the switch. source 2018
- Against generalising the favourable outcome: a lineage switch to AML during induction for early T-cell precursor ALL with the same t(6;11)(q27;q23)/KMT2A-AFDN lesion has been reported separately in a paediatric patient, and that report's title alone establishes that the same switch occurs in children, where the course is not documented here as favourable. source 2021
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Patients with t(6;11)(q27;q23) divide into two groups by breakpoint position: those with breakpoints in the KMT2A major breakpoint cluster region (introns 9-10) are associated mainly with AML M1/M4/M5, and those with breakpoints in the minor cluster region (introns 21-23) are associated with T-ALL; in a cell-culture model the exMLL-AF6 and AF6-shMLL combination induced both a myeloid-specific and a T-cell-specific gene signature, which the authors propose may explain the T-ALL phenotype in patients with minor-cluster breakpoints.
The role of reciprocal fusions in MLL-r acute leukemia: studying the chromosomal translocation t(6;11) · Kundu A, Kowarz E, Marschalek R · Oncogene 2021What this evidence cannot show
Abstract read, not full text. This is a transfected cell-culture model with doxycycline-inducible expression over 48 hours and 3'-end RNA sequencing, not patient material, so the link from breakpoint position to clinical lineage is an inference the authors draw rather than a result they measure in patients. It is cited here only to frame the breakpoint-lineage association; the case report itself does not state which cluster region its patient's breakpoint fell in, so the two cannot be joined.
Evidence that argues against this (1)
- The two-group division for t(6;11) is drawn from series small enough that the phenotype associations rest on a handful of patients per cell, and breakpoint-to-phenotype mappings of this kind have not generally replicated. For a single patient story the association provides context and cannot explain this patient's course. Statistics in Medicine 2006
- related to →KMT2A::AFDN (MLL-AF6), t(6;11)(q27;q23)
t(6;11)(q27;q23) with the same MLL::MLLT4 fusion sequence in the mediastinal mass at presentation and in blood at the sixth relapse
counter-example: a complete T-lymphoid to monoblastic switch occurring under chemotherapy and nelarabine with no antigen-directed drug involved, so lineage plasticity here is intrinsic rather than induced by CD19-directed therapy
cytoplasmic CD3 positive and MPO negative became cytoplasmic CD3 negative and MPO positive in one clone in two months
sequence-level identity of the fusion across a complete change of immunophenotype is what distinguishes a switch from a second leukaemia
t(6;11)(q27;q23) carried unchanged through six relapses
several of the six relapses occurred in the central nervous system
cord blood transplantation after a CD33-directed conjugate delivered the first durable remission, at the seventh attempt
two lineage switches in KMT2A-rearranged disease with opposite provocations and opposite outcomes: spontaneous under chemotherapy with long survival, versus drug-induced under blinatumomab with death
the same KMT2A::AFDN fusion presenting as T-lymphoblastic lymphoma with a mediastinal mass in one patient and as monocytic AML in another
identical fusion transcript across a T-lymphoid to monoblastic switch