Sequencing shows that large B-cell lymphoma is at least seven diseases, each defined by which faults occur together. The classification explains a great deal about how the disease behaves and currently changes almost nothing about how it is treated.
Two groups sequenced large series in 2018 and arrived at overlapping answers. Exome and transcriptome sequencing with copy-number analysis across 574 biopsies produced four prominent subtypes named for the lesions that co-occur: MCD for MYD88 L265P with CD79B mutation, BN2 for BCL6 fusions with NOTCH2 mutations, N1 for NOTCH1 mutations, and EZB for EZH2 mutations with BCL2 translocations; survival after immunochemotherapy was better in BN2 and EZB and worse in MCD and N1 (Schmitz 2018). Consensus clustering of 304 primary tumours found five subsets, including a low-risk activated B-cell group of extrafollicular or marginal-zone origin, two germinal-centre subsets with different outcomes, and a group defined by biallelic TP53 inactivation and CDKN2A loss that cuts across cell of origin entirely (Chapuy 2018).
LymphGen is what made these usable on one patient: an algorithm that returns the probability that a lymphoma belongs to one of seven genetic subtypes, and which showed that each subtype shares a pathogenesis with a particular indolent or extranodal lymphoma, so the groups are not statistical artefacts (Wright 2020).
The limits matter. It needs sequencing for mutations, not a stain; its authors designed it to accept any combination of mutation, copy number and BCL2 or BCL6 rearrangement data, with everything except the mutation data optional. More than a third of cases come back unclassified: LymphGen placed 329 of the 574 cases in its founding cohort, 63.1%, against the 46.6% that the four subtypes of 2018 had covered (Wright 2020). No regulator licenses anything on a LymphGen call. One randomised trial has assigned treatment by genetic subtype: the phase 2 GUIDANCE-01 trial randomised 128 newly diagnosed patients to R-CHOP alone or R-CHOP plus a targeted agent chosen by a simplified subtyping algorithm, and reported complete response in 88% against 66% and two-year progression-free survival of 88% against 63% (Zhang 2023). That is one phase 2 trial, not a change of standard. The other effect has been on how trial subgroups are read: in the PHOENIX trial, younger patients with MCD or N1 lymphoma given ibrutinib with R-CHOP had three-year event-free survival of 100%, against 42.9% and 50% with R-CHOP alone (Wilson 2021).
In plain words · MYD88 (Myeloid differentiation primary response protein MyD88) is a gene that drives cell growth when it is altered. The public catalogues list it as a drug target, an oncogene driver and a biomarker, and clinical evidence ties its variants to diagnosis, prognosis or drug response. Tied to Non-Hodgkin lymphoma, Leukaemia, Breast cancer and 4 more.
Showing the target this term concerns: MYD88.
The glossary entry explains the word; the readout page carries the scoring rule, the thresholds approvals use, the companion diagnostics and the tests.
Shares MYD88, CD79b, Cell of origin (GCB vs ABC), EZH2.
Shares EZH2 gain-of-function mutation (Tyr646, originally Tyr641), BCL6, The germinal centre reaction, Cell of origin (GCB vs ABC).
Shares B-cell receptor / BTK signalling (to NF-κB), Inflammation & NF-κB, Marginal zone lymphoma, Diffuse large B-cell lymphoma.
Shares B-cell receptor / BTK signalling (to NF-κB), Inflammation & NF-κB, Marginal zone lymphoma, Diffuse large B-cell lymphoma.
Shares EZH2 gain-of-function mutation (Tyr646, originally Tyr641), The germinal centre reaction, BCL-2, Diffuse large B-cell lymphoma.
Every source this page cites, grouped by who published it. Grouping is read from the web address, not assigned by hand.
A source that has moved or gone is reported by the weekly link check. /audit/