Molecular subtypes of diffuse large B cell lymphoma are associated with distinct pathogenic mechanisms and outcomes

Chapuy, Shipp and colleagues carried out a comprehensive genetic analysis of 304 primary diffuse large B-cell lymphomas, capturing recurrent mutations, somatic copy-number alterations, structural variants and low-frequency changes together rather than separately, and then clustered the combined signatures.

Five robust subsets came out. One was a previously unrecognised group of low-risk activated B-cell tumours of extrafollicular or marginal zone origin. Two were distinct germinal-centre groups with different outcomes and different targetable alterations. One was independent of the activated and germinal-centre division altogether, defined by biallelic inactivation of TP53, loss of CDKN2A and the genomic instability that follows. The coordinate genetic signatures predicted outcome independently of the clinical International Prognostic Index.

The paper was published three weeks after Schmitz's from the National Cancer Institute, which used a different algorithm on a different cohort and produced four subtypes rather than five. That the two classifications overlap but do not coincide is the central problem the field then had to solve, and Wright's probabilistic tool in 2020 was the attempt to solve it.