Final recap from Dr. Adam Bagg's lecture at USCAP 2010 on molecular testing in lymphoma. The highpoint of the lecture for me was his comments on diffuse large B-cell lymphoma (DLBCL) since this is the most common lymphoma type we see in our practice. It so happens that there have been some recent papers that I've read that complement Dr. Bagg's lecture notes and I cite those below. I think if you consider his comments and have a look at these papers, you should be well-prepared for the next DLBCL you see.
DLCBL is the most common lymphoma in adults but it is a heterogeneous clinical, morphological, immunophenotypic, molecular, and genetic "category." Some differences in clinical behavior can be predicted by the recognition of certain between-defined morphological subtypes and/or clinical parameters. This presentation focused on current attempts to use molecular approaches to more consistently predict behavior and response to specific therapies.
The most commonly dysregulated genes in DLBCL are BCL6 (~40%), BCL2(~20%), BCL10 (~15-20%), and MYC (~5-10%).
The presence of a MYC translocation predicts a poor prognosis.
t(14;18)(q32;q21)--identical to that seen in FL--is present in 20% to 30% of DLBCL and is associated with germinal-center-B-cell (GCB) type. The detection of the translocation does not correlate with bcl2 protein expression. Some evidence suggests that detection of bcl2 protein expression in the absence of t(14;18) predicts a poorer prognosis.
The gene BCL6 in DLBCL is promiscuous with respect to translocation partners--with at least 20 different partners described. BCL6 protein expression associated with GCB type and a favorable prognosis. However, the effect of BCL6 translocation on prognosis depends on the translocation partner.
Gene array data using unsupervised analysis identifies three gene expression subgroups:
Germinal center B-cell-like (GCB): expresses genes characteristic of germinal centers. Detection of t(14;18) is the most specific abnormality in this group.
Activated B-cell-like (ABC): expresses genes characteristic of mitogenically-activated peripheral B-cells; associated with activation of the NFκB pathway.
Unclassified ("type 3"): neither sets of genes from GCB or ABC are highly expressed.
Although such expression profiling/gene chip array methods are not currently clinically available, this data can be translated into RT-PCR and IHC methods to further characterize DLBCL.
Measurement of only 6 genes by quantitative RT-PCR on routinely processed FFPE tissue appears sufficient to predict OS: expression of BCL6, LMD2, and FN1 (GCB genes) are associated with favorable prognosis, while expression of BCL2, CCND1, and SCYA2 (ABC genes) are associated with an adverse prognosis.
Using just three IHC markers--CD10, BCL6, MUM1--can segregate DLBCL into GCB and non-GCB types in about 80% of cases compared with gene expression profiling. Recently, a 5-marker panel adding GCET1 (Serpin A9) and FOXP1 improves concordance with gene expression profiling to about 93%.
There is currently no consensus regarding the optimal IHC panel to distinguish GCB from non=GCB DLBCLs.
Additional insights have been obtained using supervised array analyses.
Expression signatures associated with either curable or refractory disease identified protein kinase C beta (PKCβ) and c-AMP-specific phosphodiesterase (PDE4B) as associated with an adverse outcome and are potential therapeutic targets.
Numerous microRNA signatures have also been described and some of these appear prognostically relevant.
Some of the data appears contradictory and is confusing.
t(14;18) in DLBCL is considered a poor prognostic indicator but is associated with GCB-DLBCL which has a good prognosis.
There is no significant correlation with bcl2 protein expression and OS within the GCB group (which is typically associated with the t(14;18) translocation), while bcl2 protein expression does predict adverse OS in the ABC group (which has a higher frequency of 18q21 amplification--where BCL2 is located).
In addition, there are conflicting and confusing data regarding the use of IHC markers and/or surrogates for prognosis and predictive value in DLBCL. There are multiple possible reasons for this. Hence, there is a definite need for standardization and harmonization of these genetic and molecular markers before they can be used clinically to determine therapeutic strategies.
Gurbaxani S, Anastasi J, Hyjek E. Diffuse large B-cell lymphoma--more than a diffuse collection of large B cells. Arch Pathol Lab Med 2009;133:1121-1134.
Carbone A, Gloghini A, Aiello A, Testi A, Cabras A. B-cell lymphomas with features intermediate between distinct pathologic entities--from pathogenesis to pathology. Hum Pathol 2010;41:621-631.