Study of VH3-21 in a Large Cohort of Chronic Lymphocytic Leukemia Patients Reveals Evidence for Antigen Selection and Confirms Its Predictive Value for Early Disease Progression.

The use frequency of the immunoglobulin (Ig) heavy chain variable region gene (VH) 3–21 in chronic lymphocytic leukemia (CLL) varies among studies on various cohorts of European patients, ranging from 0.9%–10%. Such variation could be due to geographic/population differences and/or sample-size limitations. We examined a large cohort (N=2,190) of CLL patients evaluated in the United States by the CLL Research Consortium (CRC) and found 56 (2.6%) used IgVH3-21. Thirty-five of the 56 cases (63%) expressed Ig light chains, whereas only 821 (38%) of the 2,134 cases that used IgVH other than IgVH3-21 used light chains, a difference that was highly significant (P < 0.001). Cases that used IgVH3-21 and light chains had significantly fewer amino acid residues in Ig heavy chain third complementarity determining region (CDR3) (m = 11.5 ± 5.3, S.D.) than did VH3-21 cases with light chains (m = 18.4 ± 4.8) (P<0.001). Twenty-eight of the 56 cases (50%) used unmutated IgVH3-21, defined as having >98% homology to germline VH3-21. Twenty (43%) or 18 (38%) of the 47 cases examined by flow cytometry expressed ZAP-70 or high-level CD38, respectively. Although there was frequent concordant expression of ZAP-70 and/or CD38 with unmutated IgVH3-21, such associations were not absolute, as had been noted for CLL cases that did not use IgVH3-21. Thirty-two percent (18/56) of the cases had a previously described common amino-acid motif (ARDANGMDV) in the otherwise highly variable Ig heavy-chain CDR3. Seventeen (94%) of such cases used light chains typically encoded by V3-21/J3. In addition, we identified a novel amino-acid consensus motif (DPSFYSSSWTLFDY) in the Ig heavy chain CDR3 for 3 of the 56 cases (5.4%). We examined the time from diagnosis to initiation of therapy as per established NCI-Working Group guidelines in 40 patients for whom complete clinical data were available. With a median follow-up of 4.2 years from the date of diagnosis, 25 of the 40 patients had received therapy at the time of this analysis. The median time to treatment (TTT) for all 40 patients was 3.5 years, which was significantly shorter than the median TTT of 6.6 years noted for a previously-described CRC cohort of 307 patients that were not selected for use of IgVH3-21 (NEJM2004; 351: 893–901) (P<0.001). The median TTT of 19 patients that used unmutated IgVH3-21 in this subset (3.0 years) appeared shorter than that noted for the 21 patients that had mutated IgVH3-21 (5.4 years), but this difference did not reach statistical significance. We conclude that a small proportion of patients studied in the United States by the CRC use IgVH3-21, which encodes Ig heavy chains that frequently have canonical motifs in the CDR3 and that typically are paired with certain Ig light chains, providing strong evidence for Ig selection by antigen(s). Finally, patients with IgVH3-21-expressing CLL have a higher risk for early disease progression than do patients with CLL not selected for use of IgVH3-21.