Follicular lymphoma (FL) is one of the most common B-cell non-Hodgkin's lymphomas. The initiating genetic event found in B90% of FL is the t(14;18), causing constitutive expression of the antiapoptotic BCL-2 protein. The exact secondary alterations leading to full FL development are still poorly defined. In this review, we address (i) the genetic pathways associated with tumorigenesis and progression of FL, (ii) the role of micro-environmental factors with emphasis on B-cell receptor ligands and (iii) lymphoma models in mice and what they teach us about lymphomagenesis in man.

Primary cutaneous B-cell lymphomas are B-cell non-Hodgkin’s lymphomas that arise in the skin. The major subtypes discerned are follicle center cell lymphomas, immunocytomas (marginal zone B-cell lymphomas), and large B-cell lymphomas of the leg. In this study, we analyzed the variable heavy chain (VH) genes of 7 of these lymphomas, ie, 4 follicle center cell lymphomas (diffuse large-cell lymphomas) and 3 immunocytomas. We show that all these lymphomas carry heavily mutated VH genes, with no obvious bias in VH gene usage. The low ratios of replacement versus silent mutations observed in the framework regions of 5 of the 7 lymphomas suggest that the structure of the B-cell antigen receptor was preserved, as in normal B cells that are selected for antibody expression. Moreover, evidence for ongoing mutation was obtained in 3 immunocytomas and in one lymphoma of large-cell type. In addition, in 1 immunocytoma, both IgG- and IgA-expressing clones were found, indicative of isotype switching. Our data provide insight into the biology of primary cutaneous B-cell lymphomas and may be of significance for their classification.

The expansion of follicular lymphomas (FLs) resembles, both morphologically and functionally, normal germinal center B-cell growth. The tumor cells proliferate in networks of follicular dendritic cells and are believed to be capable of somatic hypermutation and isotype switching. To investigate the relation between somatic mutation and heavy chain isotype expression, we analyzed the variable heavy (VH) chain genes of 30 FL samples of different isotypes. The VH genes of the FLs were heavily mutated (29.3 mutations on average). In addition, isotype-switched lymphomas contained more somatic mutations than immunoglobulin M–positive lymphomas (33.8 mutations per VH gene versus 23.0, respectively). In all but one of the FLs, the ratios of replacement versus silent mutations in the framework regions were low, independent of the absolute number of somatic mutations and the level of intraclonal variation. Analysis of relapse samples of 4 FLs showed no obvious increase in somatic mutation load in most FLs and a decrease in intraclonal variation in time. In 3 of 4 cases, we obtained evidence for selection of certain subclones, rather than clonal evolution. Our findings question if intraclonal variation is always a reflection of ongoing somatic hypermutation. This may have implications for the concept of antigen-driven lymphomagenesis.