Richter transformation is rare complication of CLL/SLL with a dismal prognosis and is distinct from the progression of CLL/SLL . DLBCL,NOS is the most frequent type of RT, and the minority of RT is represented by other aggressive lymphomas. In the present study, for the first time we describe HGBL-11q as the result of CLL/SLL transformation. HGBL-11q, earlier defined as Burkitt-like lymphoma with 11q aberration, is relatively new entity and the knowledge regarding this disease is still developing [1, 2]. First studies presented, that this lymphoma, which morphologically and phenotypically resembles Burkitt lymphoma (BL), has unique chromosome 11q aberrations (11q gain/loss) instead of MYC rearrangement [16, 17]. Subsequent data described, that some rare cases with 11q gain/loss have simultaneously rearrangement or amplification of MYC and are diagnosed as BL or HGBL,NOS, or even double-hit lymphoma with BCL2 rearrangement [18,19,20,21]. All these reports showed, that HGBL-11q is primary disease, however, as we present in our study, it can also be the result of clonally unrelated CLL/SLL transformation.
In recent years, there has been a deeper understanding of RT evolution from CLL. Generally, the main problem with neoplasia evolution investigations is that they are based on the genetic analysis of single time-point samples. In this context, RT evolution studies have the advantage due to availability of paired CLL-RT samples. According to current data, in vast majority of RT cases the transformation occurs through a linear model . In this type of evolution the predominant clone acquires novel genetic lesions leading to more aggressive disease . For comparison, CLL progression is based on both linear and branching clonal evolution, where in branching model precursor clone diverge into separate lineages . Our result confirmed, that clonal evolution of RT has predominantly features of linear model. In majority of cases, the abnormalities identified in the CLL/SLL phase were maintained in the RT phase, and RT was characterized by acquisition of new, secondary aberrations. The average numbers of karyotype and CN aberrations in CLL/SLL-phase samples were significantly lower than in RT samples, reaching in CLL/SLL phase 1 ~ 2 alterations, while in RT samples they were as high as 11 ~ 14. These results are consistent with literature data, in which RT typically arises from the predominant CLL clone by acquiring of ~ 20 genetic aberrations/case .
RT is characterized by high complexity of genetic alterations and by frequent lesions of TP53, CDKN2A and MYC [8, 22, 25]. Accordingly, almost all RT patients in our study presented the high degree of genome heterogeneity with ~ 13 karyotype aberrations or ~ 14 CNAs per case and the above aberrations were the most frequent alterations. They were observed in near half of FISH study group and in 30%-40% of microarray study group. This is consistent with published data, in which TP53 alterations (deletions and mutations) occur in 40% to 80% of RT, deletion of CDKN2A occurs in ~ 30% of cases and MYC structural alterations are observed ~ 30% of RT [22, 26, 27]. Disruption of TP53, one of the most prominent tumour suppressor genes, explains the chemorefractory phenotype of RT; on the other hand, cell-cycle deregulation by CDKN2A deletion and MYC deregulation may result in the progressive biology of RT . Important role of the MYC activation in RT pathogenesis is confirmed not only by high frequency of MYC alterations in RT, but also by two mutually exclusive ways of MYC stimulation in RT: translocations/gains of MYC or NOTCH1 mutations . In our study, CDKN2A deletion and MYC alterations were generally acquired at the time of transformation, what is consistent with current knowledge [8, 27]. On the other hand, deletion of TP53 was observed both before and during transformation, similar to other studies [26, 29].
Data regarding the role of IGH rearrangement in biology of RT are scarce, nevertheless IGH appears to be target of recurrent lesion in RT. Published data showed, that IGH translocation leading to activation of oncogenes, is one of the most frequent aberration at clonal evolution in CLL, has intermediate-adverse prognostic impact in CLL and a distinct mutational profile from other classic cytogenetic CLL subgroups [30, 31]. In our study and previously published report, the frequency of IGH rearrangement in RT was remarkably high and present before and after transformation . We observed that this aberration led to MYC::IGH fusion in some cases, and to fusions with unrecognized partners in the remaining cases.
Deletion of 13q14 is described as infrequent, but recurrent lesion in RT [22, 32]. This aberration is known to be an early event in CLL and it is not surprising to find this alteration maintained from CLL phase. According to current data, this alteration has never been acquired at RT . In contrast to these data, in our report we demonstrate, that incidence of 13q14 deletion in RT was quite high and this alteration was observed not only at CLL/SLL phase, but it was also acquired during transformation also.
There are limited published literature regarding deletions of 14q in RT. Based on these data, 14q deletions are mapped between 14q23.2-q32.33 and TRAF3 is considered as the putative target of these lesions . Similarly, we observed, that 14q deletions were heterogenous in size, they covered centromeric or telomeric regions of IGH and, in some cases, TRAF3. It might be interesting to investigate this lesion in other CLL-RT populations, considering, that IGH deletions were the only alterations in our RT patients, which have never been acquired at transformation, but maintained from CLL/SLL phase. This feature of IGH deletion suggests their primary character. Wlodarska et al. described, that 5’IGH deletions reflect physiological event, however, Quintero-Rivera et al. considered these lesions as important early events in the pathogenesis of CLL [33, 34]. Quintero-Rivera F et al. based their thesis on high frequency of 5’IGH deletions in CLL patients, presence of these deletions in BM derived cells not yet exposed to antigen, and lack of these deletions in normal cases.
In contrary to above alterations, gain of 5q35.2 in RT has not been describe to date. This region encompass THOC3, which belongs to genes coding RNA-binding proteins (RBP). These genes play an important role in post-transcriptional regulation and the activity of RBP-RNA networks has been shown to be closely related to tumour development .
To date, the molecular profile of RT-DLBCL,NOS has been considered not to overlap with the genetics of de novo DLBCL . However, comprehensive genetic analysis of primary DLBCLs by Chapuy et al. allowed to identify five robust DLBCL subsets, including subset “cluster 2” with genetic profile resembling profile of RT in our study . This subset was characterised by frequent CN aberrations and harboured losses of TP53, CDKN2A, 13q14, 14q32.31 together with gains of 8q24.22, 5q,11q. Such a genetic signature contrasted with genetic profiles of remaining subsets, which were characterised mainly by mutational drivers.
Single nucleotide polymorphism arrays are powerful method, which can complement cytogenetics and CGH with unique ability to find a hidden chromosomal lesions, as CN-LOHs are. According to current knowledge, regions affected by acquired CN-LOHs include suppressor genes important for cancer genesis and evolution. The frequency of CN-LOHs varies in different types of hematological malignancies, reaching ~ 80% in acute lymphoblastic leukemia or acute myeloid leukemia . In CLL patients, CN-LOH frequency is of 6–7%, which is lower than other malignancies . However, little is known about CN-LOHs in RT, and reported regions are different from CN-LOHs detected in our patients . Interestingly, the CLL/SLL-paired RT cases in our report demonstrated, that transformation was not accompanied by acquisition of LOHs, contrary to CNAs and cytogenetic alterations.