.Chronic lymphocytic leukemia (CLL) is the most common adult leukemia in the Western world. It is an incurable disease, yet very heterogeneous in terms of clinical course and outcome, most probably reflecting underlying biological heterogeneity. Over the last decades, it became obvious that CLL is subdivided in categories with distinct biological characteristics, regarding both the genomic and epigenetic profile (however without certain defects being detected in a large proportion of patients) , as well as structural and signaling features of the clonotypic B-cell receptor immunoglobulin (BcR IG)247.
Immunogenetic analysis of the clonotypic BcR IG has proved to be instrumental, not only for understanding the pathogenesis of CLL but also for defining clinical subgroups with strikingly different behavior and outcome. That said, the full potential of the IG molecule was not immediately realized, rather initial studies62-64 set in action a train of events culminating with the current notion that CLL is an antigen-driven disease74,76,100. Nowadays, the successful development of drugs targeting the BcR signaling pathway is revolutionizing CLL therapeutics214,216.
Besides BcR signaling, CLL clones appear to actively interact with their microenvironment, receiving prosurvival and proliferation signals while at the same time efficiently escaping immune surveillance217,219. Immunomodulatory drugs acting at the level of the immune synapse between clonal B cells and their cognate T cells are currently being tested in clinical trials with promising results218,231.
The immunogenetic approach of CLL pathogenesis needs to be able to also explain the existence of CLL-like monoclonal B-cell lymphocytosis (MBL). The term describes the presence of a clonal B cell expansion with CLL immunophenotype in the peripheral blood, at a size that does not reach the consensus cut-off for CLL diagnosis (i.e. <5000 clonal B cells/μL). The incidence of CLL-like MBL in the general population is ~10% and rises even further with age, or when more sophisticated flow cytometry protocols are applied. It is generally perceived as a premalignant condition of CLL, as it progresses to CLL requiring treatment at a 1-2%/year rate154. However, if MBL cells are CLL progenitors, why don't all MBL progress to CLL?
We investigated CLL ontogenesis from an immunogenetic point of view, at three separate directions: (i) BcR IG immunoprofiling in MBL and comparison to CLL (especially early-stage disease), (ii) BcR IG immunoprofiling of IG-switched CLL in order to examine its ontogenetic relation to the common IgM/IgD variant, and, (iii) T-cell receptor (TR) immunoprofiling in order to gain insight into the possible role of antigen in the selection and activation of cognate T cells.
We performed a detailed immunogenetic profiling of 333 CLL-like MBL cases (60 LC and 273 HC). Low-count (LC) and high-count (HC) MBL had distinct IG gene repertoires, whereas the HC-MBL IG gene repertoire exhibited clear similarities to early-stage CLL (CLL Rai stage 0, CLL-0). Furthermore, we sought for the expression of stereotyped BcR IG through a cluster analysis of the MBL sequences together with all CLL sequences from our cohort and non-CLL IG sequences retrieved from the IMGT/LIGM-DB sequence database. Overall, only 5.5% LC-MBL rearrangements could be clustered with other sequences. In contrast, HC-MBL included a significantly higher frequency of ‘CLL-specific’ BcR stereotypes, with 23.3% of cases clustering together with either MBL or CLL cases. This frequency was similar to that observed in CLL-0 (20.2%). Collectively, the frequency of BcR IG stereotypy seemed to increase in parallel with the absolute count of CLL-like cells, starting with 5.5% in LC-MBL, raising to 23.3-20.2% in in HC MBL/CLL-0 and peaking at 30.4% in the entire CLL cohort. Overall, these findings suggest that LC-MBL is not a true premalignant condition, but rather the result of chronic antigenic stimulation or immune senescence. In contrast, HC-MBL is just a step behind early-stage CLL, requiring either additional genetic hits or simply time to cross the numerical cut-off that discriminates it from CLL. Therefore, the identification of molecular biomarkers that may predict progression of HC MBL/CLL-0 into CLL requiring treatment is strongly warranted.
Next, we performed a comprehensive immunogenetic comparison of 169 IgG-switched CLL patients (G-CLL) versus 1087 IgM/IgD CLL patients (classic MD-variant, MD-CLL). The IG gene repertoire was significantly different among G-CLL and MD-CLL, with overrepresentation of the IGHV4-34 and IGHV4-39 genes and underrepresentaion of the IGHV1-69, IGHV3-21, IGHV1-2 and IGHV3-48 genes, respectively. G-CLL included significantly more cases with mutated IGHV genes compared to MD-CLL, and significantly more MD-CLL cases carried BcR IGs with no SHM compared to G-CLL. The extreme skewing of the G-CLL repertoire merely reflected the fact that almost one-third of all cases concerned three CLL subsets: mutated stereotyped subsets #4 and #16 utilizing the IGHV4-34 gene, and truly unmutated subset #8, utilizing the IGHV4-39 gene (18.3%, 4%, and 7.3% of all G-CLL, respectively).These subsets, especially subsets #4 and #8, are polar opposites in terms of prognosis, with subset #4 cases exhibiting particularly indolent disease whereas subset #8 cases experience an aggressive disease course often complicated by Richter’s syndrome. In clear contrast to these 3 subsets utilizing the IGHV4-34 and IGHV4-39 genes, all major CLL subsets utilizing the IGHV1-69, IGHV3-21, IGHV1-2 and IGHV3-48 genes were exclusively represented in MD-CLL. Even when restricting the comparison solely to cases with mutated BcR IGs, G-CLL still exhibits an overall distinct immunogenetic profile from MD-CLL, thus prompting speculations about distinct cell of origin and/or distinct immune trigerring and raising questions regarding the timing of CSR in regards to malignant transformation21. That said, the paradigmatic case of truly unmutated, G-switched subset #8 suggests that CSR and SHM may occur independently in CLL.
Finally, considering the critical role of antigenic stimulation in CLL pathogenesis, we sought to gain insight into the immune pathways shaping the T-cell repertoire in CLL. Our preliminary approach entailed classic subcloning techniques, followed by Sanger sequencing. We analysed 58 CLL patients, selected upon their clonotypic BcR IG molecular characteristics so as to represent major stereotyped CLL subsets (namely subsets #1, #2, and #4), where the CLL clone is most evidently selected by antigen. Our study revealed skewed TR repertoire and oligoclonality. Moreover, we identified common T-cell clonotypes among different patients, alluding to selection by a shared antigenic elements. However, low-throughput methods such as classic subcloning, despite widely used, are inherently limited in describing other than monoclonal immune repertoires, thus precluding definite conclusions.
Therefore, we aimed to advance our analytical depth by employing high-throughput, next-generation sequencing techniques, in order to obtain a comprehensive profile of the T-cell repertoire in CLL. Considering PCR-based NGS limitations, we followed a strict experimental approach including multiple types of controls, and we developed in-house, purpose-built bioinformatics tools for the management, curation and interpretation of the produced (big) data. In total, 32 CLL patients were analyzed. Again, we focused on patients assigned to major stereotyped subsets (namely subsets #1, #2, and #4),but we also included patients expressing heterogeneous clonotypic BcR IGs (mutated and unmutated), in order to seek for T-cell immunogenetic signatures that may be ubiquitous in CLL. The study confirmed TR repertoire skewing, with oligoclonality that came in sharp contrast to the polyclonal profile of age-matched healthy controls, alluding to antigenic selection. Longitudinal analysis showed that T-cell clones persist and expand over time, suggesting a persistent antigenic drive. Moreover, the identification of shared clonotypes among different patients, most specially patients assigned to the same stereotyped subset, indicates common antigenic stimulation, perhaps in a CLL subset-specific context.
Figure 27. A common scenario for CLL ontogenesis.
How could all these pieces of information contribute in a plausible scenario for CLL ontogenesis? There is no doubt that many aspects of disease pathogenesis remain elusive or debatable. However, we could postulate that, in a premalignant phase, B cells with or without genetic predisposition for unrestrained clonal expansion are subjected to persistent stimulation by exogenous or endogenous antigens and are driven to oligo/polyclonal expansion. In this context, they may undergo SHM/CSR depending on their origin (B2, B1, marginal zone B cells?) and/or the nature of antigenic trigerring, and while they proliferate, they may acquire genetic lesions that translate to aberrant (i.e. CLL) phenotype. Thereby they enter the LC-MBL state, which remains stable over time unless further genetic/epigenetic hits and/or the functional properties of the clonotypic BcR IG (e.g. polyreactivity or capability of autonomous cell signaling through homotypic BcR interactions) lead to HC-MBL, which, given time and appropriate microenvironmental signals will progress to CLL requiring treatment (Figure 27).