Abstract

Abstract 1777

Antigen stimulation through the B cell receptor (BcR) is critical in CLL ontogeny as evidenced by the distinct prognosis of cases with different mutational status of the clonotypic BcRs and the existence of subsets of patients with stereotyped BcRs. Of note, for several subsets analyzed thus far, stereotypy extends from shared primary BcR sequences to shared clinicobiological features and, even, outcome. Hence, it may be hypothesized that the clinical course of CLL perhaps reflects the outcome of ongoing BcR signaling in the context of other co-signals. Previous studies have documented high expression of activation-induced cytidine deaminase (AID) and in vivo class switch recombination (CSR) in a proportion of CLL cases, especially those with unmutated IGHV genes (U-CLL). Here we sought evidence of isotype switch events in 486 CLL cases, by far the largest series analyzed to-date, focusing on potential associations with the expression of particular BcRs. All 486 cases were selected due to expressing MD isotypes as determined by peripheral blood flow cytometry (FC) (298/486 caes; 61.3%) or for expressing tumor-specific Cμ and Cδ transcripts as determined by RT-PCR (188/486 cases; 38.7%). They were all subjected to RT-PCR for μ, δ, α and γ transcripts utilizing primers specific for the respective C genes along with leader primers corresponding to the IGHV subgroup of the tumor-specific VDJ rearrangement. All obtained RT-PCR products were sequenced in order to determine the correspondence with the clone-specific VDJ rearrangement. Using the 98% cut-off value for identity to the germline (GI), 340/468 cases (70%) expressed mutated IGHV genes (M-CLL), whereas the remaining (146/486 cases, 30%) cases concerned U-CLL. Stereotyped BcRs characteristic of major CLL subsets were expressed by 71 cases: subset #1, n=25; subset #2, n=23; subset #3, n=8; subset #6, n=7; subset #7, n=8. Except for subset #2 (IGHV3–21), all other subsets concerned U-CLL; subset #3, 6 and 7 BcRs utilize the IGHV1–69 gene, whereas BcRs of subset #1, the largest CLL subset overall, utilize IGHV genes of phylogenetic clan I (IGHV1/5/7). Overall, RT-PCR products for Cα and/or Cγ transcripts in addition to Cμ/Cδ were obtained in 78/486 cases (16%); in 55/78 cases, transcripts were tumor-specific VDJ sequences, for an overall frequency of 11.3% (55/486 cases). In particular, dominant cDNA bands of α isotypes were found in 27/486 cases (5.6%), dominant cDNA bands of γ isotypes were found in 21/486 cases (4.3%), while cDNA bands for both α and γ isotypes were found in 7/486 cases (1.4%). In keeping with the literature, the frequency of isotype switch events was significantly higher in U-CLL vs. M-CLL (33/146 vs. 22/340 cases, respectively; p<0.001). Of great interest, a significantly different frequency of such events was also observed in distinct stereotyped subsets. In particular, alternative tumor-derived Cα and/or Cγ transcripts were detected in 13/25 subset #1 cases vs. 0/23 subset #2 vs. 2/8 subset #3 vs. 0/7 subset #6 vs. 3/8 subset #8. In conclusion, we confirm and significantly extend previous reports that ongoing CSR occurs in a sizeable fraction of CLL, mostly U-CLL. Importantly, we provide for the first time evidence for subset-biased profiles of in vivo CSR, documenting a very high frequency of alternative tumor-derived transcripts in certain subsets, especially subset #1, reported to exhibit a particularly aggressive clinical behavior. Altogether, these findings link distinct CLL BcRs exhibiting particular biological features and behavior to specific modes of (micro)environmental interactions leading to clonal evolution and, perhaps, adverse outcome.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.