Alizadeh at al. (2000) first described
two distinct subtypes of DLBCL based on their cell of origin group using
expression microarrays (figure ***) (20). These are germinal centre b-cell like
(GCB) and activated b-cell like (ABC) DLBCL and these have been recognised in
the 2016 revision of the World Health Organization classification
of lymphoid neoplasms(17). These subtypes have differing gene
expression profiles, different overall survival following standard treatment
and different oncogenic pathways involved(21).
Mutations in common with both subtypes include B2M,
which is involved in immune surveillance, dysregulation of BCL6, possibly
through MEF2B or FBX011 mutations, and changes to histone modification genes
CREBBP and EP300(17,22).
GCB DLBCL originates from germinal
centre B cells. It responds favourably to rituximab and anthracycline
chemotherapy, and has a better outcome than ABC DLBCL. Mutations include GNA13,
t(14;18) which leads to deregulation of BCL2, this occurs in 45% of GCB DLBCL,
PTEN deletion and miR17-92 amplification which lead to deregulation of the
p13K/mTOR pathway, which occurs in 15% of GCB patients, and ING1 deletion, MDM2
gain or amplification and p53 mutation which lead to genomic instability(23,24).
ABC DLBCL is derived from B cells that
are in the process of differentiating into plasma cells. Mutations that arrest
differentiation are PRDM1 mutation or deletion and BCL6 translocation(21,25). 30% of patients with ABC DLBCL have
TNFAIP3 mutations or deletions and MYD88 mutations(17,22). It is also associated with
deregulation of BCL2 and constitutive action of the NF?B pathway due to CARD11
and CD79A/B mutations(21,23). ABC DLBCL is associated with a poorer
outcome, and most patients succumb to the disease(21).
Gene expression profiling using
microarrays can provide additional information over just the subtype of GCB or
ABC DLBCL. Lenz et al. (2009) described a survivor predictor model for patients
treated with rituximab and anthracycline chemotherapy(21,26). This analyses different gene expression
signatures – germinal-centre B cell, which parallels the distinction between GCB
and ABC DLBCL and is associated with a favourable outcome. Stromal-1 and
stromal-2 signatures reflect non-malignant cells. Stromal-1 is associated with
a favourable outcome and shows extracellular matrix deposition and histiocytic
infiltration. Stromal-2 is associated with an unfavourable outcome and reflects
tumour blood vessel density. These profiles are used to give patients scores
and split them into quartiles of 3-year overall survival (of 89%, 82%, 74% and
48%) and 3-year progression free survival (of 84%, 69%, 61% and 33%)(26).
Next gen sequencing (NGS)
First generation, or Sanger, sequencing
was first developed by Sanger et al. in 1977 and has been the most widely used
sequencing method for the past 40 years(27,28). In recent years there has been a
drift away from Sanger sequencing as there is a need for newer technologies to
sequence large numbers of human genomes.
There are varying methods of NGS,
described by Shendure et al. (2008), but they have principles in common(29). These methods of NGS use a massively
reduced reaction volume and an extended number of sequencing reactions. Arrays
of several hundred thousand sequencing templates on plates can be analysed in
parallel, compared to a maximum of 96 templates in Sanger sequencing(30). A large advantage to NGS is that
information about rare transcripts can be gathered without prior knowledge of
the genes involved(28).
So far 12 human genome sequences using
NGS have been published, marking the beginning of personalised genomics. NGS
can produce an enormous amount of data cheaply in comparison to older methods and
the cost should continue to drop, making it more viable for clinical practise(28).
Applications in Diffuse Large B Cell Lymphoma
NGS allows high throughput DNA
sequencing and has identified single nucleotide variants (SNVs) in DLBCL which
are more common in particular subtypes(17,31). Some SNVs occur in actionable targets
or correlate with antitumor response and could be used to guide future targeted
therapies. Using NGS to look at mutations that affect disease progression could
also help prognosis and staging.
Dubois et al. created a ‘lymphopanel’
to look at the mutations in 34 genes important in lymphomagenesis using NGS(31). They grouped these genes into eight
pathways – immunity, NOTCH, apoptosis/cell cycle, NF?B, epigenetic regulation,
MAP kinase, JAK-STAT, and BCR. They found that ABC DLBCL was dominated by NF?B
pathway mutations and GCB was dominated by epigenetic regulation pathway
mutations, which fits with current literature(21,23). This could be useful to further
define the genetic profiles of the subtypes of DLBCL, but also in clinical
practise to stratify patients by their actionable mutations and treatment
options. Present mutations that will make certain targeted therapies less
effective are also important. For example, doubly mutated MYD88 and CD79B
common in ABC is more responsive to single therapy with the BTK inhibitor
Ibrutinib, but CARD11 and TNFAIP3 mutations make Ibrutinib and Sotrastaurin (a
protein kinase C inhibitor) less effective(31–33).