Restricted MHC Phenotypes in Lymphoproliferative Disorders with Non-Random Immunoglobulin Variable Gene Usage: Evidence for Antigen-Specific Interactions between Tumor and Microenvironment.

Abstract

Several lines of evidence implicate antigenic stimulation in the pathogenesis of chronic B-lymphoproliferative disorders. This is illustrated by B-cell chronic lymphocytic leukaemia (B-CLL) in which there is non-random use of immunoglobulin (Ig) variable heavy chain (Vh) genes and a gene expression profile and immunophenotype in keeping with antigen exposure and activation. Sequence analysis of IgVh genes reveals that within B-CLL there is a subset of patients whose tumors utilize Vh3–21 and show such homology between antigen binding sites that the Ig must recognize the same antigen. In splenic marginal zone lymphoma (SMZL) overuse of the Vh1–02 gene is well documented and there is similarly some restriction of antigen binding sites. Engagement of the B cell receptor by a relatively restricted range of antigens might thus select cells that subsequently form the malignant clone. T lymphocytes may also be involved in this process since IgVh genes in about 50% of B-CLL patients show evidence of somatic hypermutation and CD40L expressing T lymphocytes surround proliferating tumor cells in pseudofollicles in the bone marrow and lymph node. Although both Vh3–21 B-CLL and Vh1–02 SMZL have a relatively low Vh gene mutational rate, it is thought that somatic hypermutation has taken place indicating possible interaction with T cells. To investigate whether such T cell/tumor interactions might be antigen specific we looked for evidence of MHC restriction in homologous cases of Vh3–21 B-CLL and Vh1–02 SMZL. Medium resolution HLA typing was performed by PCR-SSO and the phenotypes compared with normal controls and other disease groups by Fishers exact test with Bonferroni correction (shown following the uncorrected values). Groups included: Vh3–21 B-CLL with/without homologous CDR3 (n=33/7), B-CLL cases with random Vh gene use (without Vh1–02 or Vh3–21, n=22), SMZL with (n=18) and without (n=18) Vh1–02, B-CLL with Vh1–02 (n=17) and 1667 normal controls drawn from the same population. Of 18 patients with Vh1–02 SMZL, 11 (61%) had an HLA-DR15 phenotype compared to only 441/1667 (26%) controls (p=0.0022, 0.029) and 2/18 (11%) of non Vh1–02 SMZL (p=0.0045, 0.058). Of the 7 Vh1–02 SMZL that did not express HLA-DR15, 5 expressed HLA-DR4. Non Vh1–02 SMZL was associated with HLA-DR4 (p=0.0023, 0.023) but not HLA-DR15. A number of other associations between IgVh homologous subgroups and MHC phenotype were noted including Vh3–21 B-CLL and HLA-B15 and non Vh3–21/Vh1–02 B-CLL and HLA-B58 however once corrected for the number of variables, statistical significance was lost. In general the Vh genes of both normal and neoplastic human marginal zone B cells show evidence of somatic hypermutation although whether this occurs through a T-dependent or independent pathway is currently unknown. The finding of MHC restriction in Vh subsets of SMZL however suggest that this disorder derives from a marginal zone B cell that has responded to a T-dependent antigen. Whether similar mechanisms apply in B-CLL is unclear from the current data and further study of a larger number of cases is therefore warranted.