Clinicobiological characteristics and treatment efficacy of novel agents in chronic lymphocytic leukemia with IGLV3-21R110

(IG) heavy chain (IGH) (<98% germline homology), IGHV ( 98% germline homology), of CLL grouped in subsets recurrent amino acid (aa) sequence similarities in IGH


TO THE EDITOR:
The molecular composition of the leukemia-specific B cell receptor (BCR) is of key importance in chronic lymphocytic leukemia (CLL) [1]. Firstly, using the burden of somatic hypermutation (SHM) present in the immunoglobulin (IG) heavy chain (IGH) variable region, a distinction can be made between patients with mutated IGHV (<98% germline homology), that generally have indolent disease, and those with unmutated IGHV (≥98% germline homology), with generally more aggressive disease. Secondly, up to 41% of CLL patients can be grouped in subsets based on recurrent amino acid (aa) sequence similarities in their IGH rearrangement [2]. Patients within these IGH stereotyped subsets often demonstrate consistent clinicobiological profiles [3,4]. However, individually, these IGH stereotyped subsets are rare: the largest subset, #2, represents 2.5% of all CLL cases [2].
Recently, a novel immunogenetically defined CLL subset was described, based on recurrent sequence similarities within the IG light chain (IGL) [5,6]. Patients within this subset, referred to as IGLV3-21 R110 , are characterized by rearrangement of a specific IG lambda variable gene, IGLV3-21, with a distinctive somatic hypermutation (SHM) present in the linker region between the IGLJ and IGLC genes (G110R). This mutation, which is detectable in CLL patients years prior to diagnosis, allows for antigenindependent, light chain based auto-aggregation of BCRs on the cell surface [7,8]. This auto-aggregation is additionally dependent on germline residues exclusively present in alleles IGLV3-21*01 or IGLV3-21*04, establishing these alleles as heritable risk factors for the development of IGLV3-21 R110 CLL [5,9]. Importantly, this newly defined IGL stereotyped subset accounts for approximately 20% of all CLL patients and is associated with shorter time-to-first treatment (TTFT) and inferior overall survival, irrespective of IGHV mutational status [5]. Although two previous studies have examined the molecular characteristics of IGLV3-21 R110 CLL, the combined cytogenetic, immunogenetic and mutational landscape of IGLV3-21 R110 CLL remains incompletely characterized [5,6]. In addition, the clinical impact of this IGL genotype in the context of therapy is still largely unknown.
In order to address these knowledge gaps, we characterized the light chain genotype, clinicobiological features, and response to therapy of patients enrolled in the HOVON-139/GIVE trial and the Dutch sub-cohort of the HOVON-141/VIsion trial. The HOVON-139/ GIVE trial is a phase-II trial, evaluating the efficacy of first-line MRDguided duration of treatment with obinutuzumab and venetoclax in unfit CLL patients [10]. The HOVON-141/VIsion trial is a phase-II trial in relapsed or refractory (R/R) CLL patients, evaluating the efficacy of MRD-guided ibrutinib and venetoclax combination treatment [11,12]. To determine the IG light chain rearrangement, we performed PCR amplification and Sanger sequencing on pretreatment cDNA. In the absence of cDNA, the IGLV3-21 R110 light chain genotype was determined by PCR amplification and Sanger sequencing on gDNA, using novel, custom-designed primer sequences that target the IGLV3-21 R110 leader region and the 5' end of the intron sequence between the IGLJ and IGLC gene.
For 65/70 patients from the HOVON-139/GIVE trial and 129/133 patients from the Dutch cohort of the HOVON-141/VIsion trial, samples were available for IGL sequencing. The IGLV3-21 R110 genotype was present in 16/65 patients (25%) in the first-line cohort, and at a similar frequency, in 32/129 patients (25%) in the R/R cohort. For an overview of clinical characteristics, stratified by IG light chain genotype, see Supplementary Table 1.
Taken together, in our study, we have characterized the clinicobiological features of the largest cohort of IGLV3-21 R110 patients (n = 48) reported thus far. The prevalence of the IGLV3-21 R110 genotype in our cohorts was 25%, which is higher compared to the prevalence of 7-17% identified in previously studied populations [5,6]. This discrepancy could be explained by enrichment of IGLV3-21 R110 in clinical trial populations, as IGLV3-21 R110 CLL is associated with a shorter TTFT, or due to geographical differences in the frequency of the IGLV3-21*01 and IGLV3-21*04 alleles [5].
We demonstrate that CLL with the IGLV3-21 R110 is typified by a distinct molecular profile. In these patients, lesions targeting SF3B1 and ATM are enriched. This is in line with the data reported by Nadeu et al., who characterized the genetic landscape of a cohort of 28 IGLV3-21 R110 patients using whole-genome sequencing [6]. Moreover, genomic array analysis revealed enrichment of del13q14 in IGLV3-21 R110 patients, but mutual exclusivity with trisomy 12 and del17p13 in our cohort. These patterns suggest that CLL with IGLV3-21 R110 genotype may rely on distinct intracellular signaling pathways, including aberrant splicing and dysfunctional DNA damage repair mediated through loss of ATM, but not TP53. This hypothesis warrants additional functional validation. In addition, IGLV3-21 R110 patients were characterized by distinctive IGH rearrangements, using phylogenetically related IGHV genes with markedly shorter HCDR3s and a borderline IGHV SHM imprint. These findings may reflect additional (patho) physiological selection pressure involving not only the IG light chain, but also the IG heavy chain. Lastly, despite its distinct clinicobiological profile, there was no evidence for a predictive impact of the IGLV3-21 R110 genotype on the efficacy of the novel therapies employed in the HOVON-139/GIVE and HOVON-141/ VIsion trials. This is supported by the stable prevalence of the IGLV3-21 R110 genotype in our first-line and R/R cohort. Our results suggest that the evaluated regimens of novel targeted therapies may mitigate the adverse risk profile of IGLV3-21 R110 CLL. These observations require further validation in a larger series, as the number of IGLV3-21 R110 patients that we evaluated was limited, especially in the HOVON-139/GIVE trial. Furthermore, to determine whether these patients should preferentially receive novel therapies, characterization of the predictive impact of IGLV3-21 R110 in the setting of chemoimmunotherapy is warranted.

METHODS
For an extended version of the methods, please refer to the Supplementary Information.