Clues to the Pathogenesis of Waldenström Macroglobulinemia and Other Monoclonal IgM Disorders Provided by the Analysis of Immunoglobulin Heavy Chain Gene Rearrangement and Clustering of B-Cell Receptors,

Abstract 3680

In B-cell malignancies, biased usage of immunoglobulin heavy chain variable (IGHV) genes and stereotyped clusters of immunoglobulin receptor suggest that a limited set of antigens or superantigens are involved in malignant transformation. In the present study we analyzed the IGH rearrangements of a large series of patients with Waldenström Macroglobulinemia (WM), IgM-monoclonal gammopathies of undetermined significance (IgM-MGUS) and IgM-related disorders (IgM-RD), with the aim to characterize IGH repertoire and to search for clusters of stereotyped receptors.

A total of 123 patients with IgM monoclonal gammopathies, including 59 WM, 55 IgM-MGUS and 9 IgM-RD underwent IGHV amplification and direct sequencing. The median age of patients was 63 years (range: 32–86). Diagnosis of WM, IgM-MGUS and IgM-RD was made according to the consensus criteria proposed at the 2nd International Workshop on WM. All IGHV-D-J rearrangements were analyzed using the IMGT databases and the IMGT/V-QUEST tool to identify HCDR3 aminoacid (AA) sequences. HCDR3 sequences were aligned with each other and compared to 28,400 HCDR3 sequences from public databases (EMBL, NCBI, IMGT/LIGM-DB) and from unpublished multi-laboratory databases. Alignments were performed using the multiple sequence alignment software ClustalX (2.0). Criteria for including sequences in a cluster were: >60% aminoacid identity, IGHV gene belonging to the same clan, identical HCDR3 length and same junction residues and/or same IGHD-J genes.

A productive monoclonal IGHV-D-J rearrangement was obtained in 99/123 patients, including 55 WM, 37 IgM-MGUS and 7 IgM-RD. IGHV genes were mutated in 94/99 patients (95%). The median somatic hypermutation rate was 93.3% (range 85.5%-97.9%). Four cases showed >98% identity with the germline sequences, whereas one single WM sample showed 100% identity with the germline sequence. When compared with the normal mature B-cell repertoire, a significant over-representation of the IGHV3 family and a significant under-representation of the IGHV1 and IGHV4 families were observed both in WM (87%, 7% and 2% respectively) and in IgM-MGUS (78%, 8% and 8% respectively). IGHV3-23 was the gene most commonly used (29%). Analysis of IGHV3-23 sequences for AA changes in the positions involved in superantigen recognition by IGHV3 subgroup genes showed that the majority of cases had two or more non-permissive AA changes that compromise the capacity to mediate superantigen recognition and binding. The frequency of novel N-glycosylation sites, introduced by the somatic hypermutation process, in this series of patients (12%) was not statistically different from that observed in normal memory B-cells, suggesting that WM, IgM-MGUS and IgM-RD cells do not need to interact with stromal GC environmental elements. The median HCDR3 length was 13 AA (range: 5–29) and was similar in WM, IgM-MGUS and IgM-RD. Intra-WM/IgM-MGUS/IgM-RD search for HCDR3 similarity showed no association that met minimal requirements to define stereotyped receptors. The comparison of WM/IgM-MGUS/IgM-RD sequences with non WM/IgM-MGUS/IgM-RD database showed that WM/IgM-MGUS/IgM-RD sequences are unrelated to known CLL or SMZL subsets. Only one IgM-MGUS occurring in a HCV-positive patient formed a novel subset with other 4 HCV-related lymphoproliferative disorders.

The findings of this analysis of WM, IgM-MGUS and IgM-RD IGH sequences indicate that: i) family-usage in the WM, IgM-MGUS and IgM-RD showed a skewed usage compared to normal B-cell repertoire; ii) WM, IgM-MGUS and IgM-RD-specific HCDR3 clusters do not occur to a frequency detectable with currently available databases; iii) WM, IgM-MGUS and IgM-RD sequences are not related to known CLL and SMZL HCDR3 clusters; iv) one IgM-MGUS showed similarities with sequences derived from pathological HCV-related lymphoproliferations, suggesting that, at least in a fraction of cases, HCV may have a pathogenetic role. For the large majority of IgM-related disorders, however, there is little evidence in favor of a B-cell receptor-driven pathogenesis.