Chromatin dynamics in CLL: a matter of balance Free

In this issue of Blood, Chapaprieta et al introduce a novel prognostic score in chronic lymphocytic leukemia (CLL) based on the balance of de novo chromatin changes associated with either progressive or indolent disease.¹ This balance score was validated across independent cohorts and surpassed IGHV gene mutational status as a prognostic marker.

Over the past 15 years, the application of genomic tools has revealed a highly heterogeneous genomic landscape in CLL, characterized by thousands of recurrently altered genes, often at low frequencies. To date, more than 200 genomic driver events have been identified in CLL, impacting critical cellular processes and signaling pathways, such as B-cell receptor/NF-κB signaling (NFKBIE, BIRC3), Notch signaling (NOTCH1, FBXW7), RNA metabolism (SF3B1, RPS15), and DNA repair (TP53, ATM).^2,3 Among these genomic events, alterations in over 50 genes have been associated with clinical outcome, frequently correlating with shorter time to first treatment (TTFT) and reduced overall survival.⁴ 

In addition to genomics, the role of epigenetics in CLL has been extensively studied.⁵ Most studies have focused on DNA methylation, revealing associations with cell of origin, evolution dynamics, as well as outcome prediction.^6-8 Specifically, by comparing the DNA methylation profiles of CLL cells with normal B cells, the disease has been classified into 3 subgroups or epitypes: clinically aggressive naivelike CLL (n-CLL), more indolent memory–like CLL (m-CLL), and intermediate CLL (i-CLL), which exhibits an intermediate prognosis.^7,8 Although less well characterized, various types of chromatin modifications also play a role in disrupted epigenetic regulation. A large-scale study using chromatin immunoprecipitation (ChIP) sequencing and ATAC sequencing identified extensive de novo acquisition of chromatin changes specific to CLL.⁹ Notably, patients with poor prognosis with unmutated IGHV genes (U-CLL) displayed a more open and active chromatin state than patients with indolent disease with mutated IGHV genes (M-CLL).

In the current study, Chapaprieta and colleagues conducted an in-depth analysis of the biological and clinical significance of chromatin activation. By reanalyzing previously generated ChIP-sequencing data for histone 3 lysine 27 acetylation (H3K27ac) from their earlier study,⁹ they identified de novo activation marks associated with either clinically aggressive or indolent disease courses. Separate scores were developed for progressive disease, S_PD, and indolent disease, S_ID, where the former was more prominent in patients with U-CLL and the latter in patients with M-CLL (see figure). Importantly, when a combined “balance” score, S_B, was calculated as the ratio of the S_PD to the S_ID score, it outperformed the individual scores, effectively stratifying patients into groups with low to high balance scores, which corresponded to longer or shorter TTFT (see figure). In multivariate analysis across independent cohorts, the S_B score emerged as a strong independent prognosticator of TTFT, remaining significant after adjustment for age, sex, and Binet stage. Notably, the balance score also outperformed IGHV gene mutational status as an independent prognostic factor.

Furthermore, the researchers examined the relationship between genetic drivers and the balance score. In M-CLL, a higher balance score was strongly associated with trisomy 12, del(11q), and SF3B1 mutations, and among patients with U-CLL, those with NOTCH1 and ATM aberrations exhibited higher scores. Interestingly, the researchers also analyzed patients within the i-CLL epitype, and patients carrying the IGLV3-21^R110 mutation had a higher balance score compared with those without the R110 mutation. This finding aligns with the understanding that patients with i-CLL harboring the R110 mutation have a worse prognosis, resembling the outcomes of patients with n-CLL.¹⁰ 

The researchers then explored the pathways associated with H3K27 activation marks, highlighting the activation of tumor necrosis factor-α/NF-κB and mammalian target of rapamycin signaling. Notably, based on the ID signature, they identified multiple H3K27 peaks linked to ATXN1 expression, a gene known to inhibit Notch signaling. Subsequent analysis of lymph node (LN) samples revealed a negative correlation between ATNX1 expression and the expression of several NOTCH1 target genes. Additionally, by examining sorted proliferative and resting fractions from peripheral blood (PB) samples as well as LN specimens, they found that the S_PD score was higher in LN compared with PB, and in proliferative vs resting fractions, and the S_ID remained relatively stable. Once again, the balance score showed a strong association with proliferation. Finally, when primary CLL cells were stimulated to proliferate using CpG oligonucleotides, the researchers observed a strong correlation between the balance score and Ki67⁺ cells.

Taken together, to our knowledge this study provides the first evidence of a dual role for CLL-specific histone activation. Regulatory programs linked to disease progression are effectively captured by the balance score, which emerges as a robust, independent predictor of TTFT in CLL. This finding was validated across independent cohorts and replicated using transcriptomic and proteomic data to construct similar scores. Additionally, a connection to the CLL tumor microenvironment was observed, with progressive activation marks predominantly observed in LN and proliferative fractions of PB. As the authors rightly note, the primary challenge of this study lies in translating these findings into routine practice, as technologies such as ChIP sequencing or RNA sequencing remain limited to research settings and are not yet integrated into clinical diagnostics. The authors propose the development of a gene expression-based assay, such as one leveraging NanoString technology, though such an assay would require extensive validation (including which genes to include and cutoffs to use) in both retrospective and prospective cohorts before clinical implementation. Furthermore, it will be crucial to assess the balance score’s impact on other clinical end points, such as progression-free survival and overall survival, as well as its relevance in the context of targeted therapies, such as BTK and BCL2 inhibitors. Finally, it would be valuable to determine whether the balance score could further enhance the stratification of CLL subgroups defined by, for example, genomic aberrations or IGHV gene mutational status. Nonetheless, this study underscores the value of multiomics analysis in CLL, providing critical insights into the complex interplay of epigenetic regulatory mechanisms driving gene expression and disease progression.

Conflict-of-interest disclosure: R.R. has received honoraria from AbbVie, AstraZeneca, Janssen, Illumina, Lilly, and Roche.