Clinical Impact of Stereotyped Antigen Receptors in Chronic Lymphocytic Leukemia

In chronic lymphocytic leukemia (CLL), the molecular features of the clonotypic B-cell receptor immunoglobulins (BcR IG) are set from the birth of the clone and in contrast to genetic aberrations, remain stable overtime, rendering the BcR IG a reference biomarker that is usually not significantly affected by clonal evolution. Approximately 30% of CLL cases carry quasi-identical BcR IGs and can be assigned to distinct stereotyped subsets. While preliminary evidence alludes to BcR IG stereotypy being relevant from a clinical viewpoint, this aspect has never been explored systematically or in a cohort of adequate size to enable meaningful conclusions. In order to assess the clinical implications of BcR IG stereotypy, we evaluated clinicobiological data from 8593 CLL patients, particularly focusing on 14 major stereotyped subsets of cases with unmutated (U-CLL) or mutated IGHV genes (M-CLL). The largest subset was #2 (n=254, 3%), within which 156 and 98 cases were M-CLL and U-CLL, respectively, followed by U-CLL subsets #1 (n=173, 2%) and #7 (n=123, 1.4%). Amongst M-CLL, the largest subsets were #4 (n=94, 1.1%) and #148 (n=92, 1%). Stereotyped subsets, even of the same mutational category i.e. U-CLL or M-CLL, exhibited significant clinicobiological differences regarding: age at diagnosis (median age ranging from 53-68 years, p<0.001); gender distribution (male/female ratios ranging from 0.85-4.4, p<0.001); the frequency of advanced stage disease (Binet C) at diagnosis (ranging from 9-21%, p<0.001); CD38 positivity (median percentage of positive cells ranging from 2-73%, p<0.05). Moving to cytogenetic findings, we detected subset-biased spectra of FISH-detected aberrations which can be summarized as follows: (i) isolated del(13q) varied from 10% in subset #8 (U-CLL), to 29% in subsets #59 (U-CLL) and #16 (M-CLL), and >75% in M-CLL subsets #77 and #148 (p<0.001); (ii) trisomy 12 was absent in subsets #31 (U-CLL) and #77 (M-CLL), in 20% of subset #201 (M-CLL) and, 60% in subset #8 (U-CLL) (p<0.001); (iii) del(11q) ranged from 69% in subset #31 (U-CLL) to 12% in subset #8 (U-CLL) and 0% in M-CLL subsets #16 and #201 (p<0.001); (iv) del(17p): absent from all M-CLL subsets except #148 (7.5%); absent from U-CLL subsets #5 and #59, and present in 19% and 17% of stereotyped subset #8 and #6 cases, respectively (p<0.05). We considered subset #2 separately since it included both M-CLL and U-CLL and identified a distinctive clinicobiological profile characterized by a constellation of features, including high incidence of Binet stage C disease (21%), CD38 positivity (41%) and certain cytogenetic abnormalities i.e. del(13q) (56%) and del(11q) (23%) and, in contrast, low frequency of del(17p) (4.5%). Notably, no differences were found between U-CLL versus M-CLL subset #2 cases regarding the incidence of these parameters. Survival analysis for time-to-first-treatment (TTFT) revealed that each stereotyped subset exhibited a distinct clinical course, different from that of other subsets even with similar IGHV gene mutational status and/or expressing the same IGHV gene. Interestingly, subsets #1 and #2 demonstrated a similar TTFT to that of del(17p) positive cases, while within subset #2 the clinical outcome was independent of IGHV mutational status. By integrating BcR IG stereotypy into the Döhner cytogenetic prognostication model, we found that subsets #1, #2 and #4, collectively accounting for ~7% of our series, and representing both U-CLL and M-CLL, constituted distinct clinical entities, with the latter exhibiting a particularly indolent disease even when compared to isolated del(13q) cases or cases with no cytogenetic aberrations. Finally, in multivariate analysis for TTFT, assignment to subset #2 emerged as a new independent unfavorable factor (p=0.002); furthermore, when restricting the analysis to Binet A cases, subset #2 exhibited the second highest hazard ratio after U-CLL. In conclusion, the molecular classification of CLL based on BcR IG stereotypy refines prognostication beyond IG mutational status and improves the Döhner hierarchical model. Hence, we suggest that a compartmentalized approach focusing on and comparing different subsets sheds light on CLL clinical behavior and biology and that BcR IG stereotypy represents a companion molecular diagnostic for personalized medicine in CLL.