Significant advances in the understanding of the molecular complexity and alterations involved in the pathogenesis and progression of B-cell non-Hodgkin lymphomas (NHL) have been achieved with the emergence of high-throughput molecular technologies such as array comparative genome hybridization and next generation sequencing. Extensive research based on these novel technologies using material from surgical specimens (paraffin-embedded tissue and frozen specimens) and cell lines has unveiled a large number of chromosomal abnormalities, as well as somatic mutations of potential therapeutic importance.[1-4]
Recurrent point mutations involving the EZH2 and CD79B genes have been reported in follicular lymphoma, diffuse large B-cell lymphoma (DLBCL) and not in marginal zone lymphoma.[5-7] EZH2 (enhancer of zeste homolog 2) represents 1 of 3 subunits of polycomb-repressive complex 2 (PRC2) and is responsible for catalyzing the methylation of PRC2's target, lysine-27 of histone 3 (H3K27), an epigenetic marker associated with transcriptional silencing and involvement in cellular differentiation, morphogenesis, and organogenesis.[8-11] Mutations in the Tyr641 residue in the SET (Su[var]3-9, Enhancer-of-zeste, Trithorax) domain of EZH2 have shown to act in a dominant fashion increasing H3K27 tri-methylation, conferring a gain of function and involvement in carcinogenesis. For CD79B, point mutations causing the substitution of the first tyrosine of the ITAM (immunoreceptor tyrosine-based activation) motif (Y196) cause the B-cell receptor to form clusters in the cytoplasm, leading to the so-called “chronic active B-cell receptor signaling,” with subsequent constitutive downstream activation of the nuclear factor-κB pathway.[5, 13] These 2 mutations therefore have a major role in lymphomagenesis and stand out as putative candidates for molecular targeted therapies in the future.[12-14]
There is a recurrent concern over which is the most appropriate sample to be used for genomic analysis, based on the fact that changes in the molecular status may occur in malignant neoplasms between initial diagnosis and relapse, with the acquisition of novel abnormalities. In breast cancer, for example, there is mounting evidence that recurrent or metastatic tumors should be re-biopsied, due to differences seen in molecular profile and discrepancies in receptor status between primary and recurrent samples. Recently, differences in mutational status were found between paired ductal carcinomas in situ and the adjacent invasive component in breast cancer. Different chromosomal abnormalities and copy number alterations were also seen in follicular lymphomas and the transformed DLBCL samples and in primary and metastatic samples from lung and colon cancer.[18, 19]
Recent reports describe the use of cytology specimens as a source of molecular information, with reliable results using novel technologies and DNA extracted from archived cytological preparations and fresh samples.[20-22]
In this study, we sought to detect mutations in patients with B-cell NHL and compare mutational status over time using different cytological and histological preparations as the substrate for a customized multiplex mutation assay for the detection of point mutations involving EZH2 and CD79B using the MassARRAY spectrometry platform.
- Top of page
- MATERIALS AND METHODS
- FUNDING SOURCES
- CONFLICT OF INTEREST DISCLOSURE
We have demonstrated that the mutational status of B-cell NHL samples from the same patients collected at distinct times during the course of their disease varied in one-third of the positive cases. Successful and reliable results were obtained using minimal residual fresh unfixed material from FNAs stored in FTA cards as well as from archived smears and cytospin preparations. In addition, with the application of a customized multiplex mutation assay, we found mutations involving EZH2 and CD79B in cases of FL and DLBCL in a similar rate to that reported for surgical pathology specimens.
Comparison was possible for multiple samples from the same patient, with evaluation of the mutation status at different time points. Two previous reports on the analysis of EZH2 Y646 mutations in B-cell NHL have compared initial and recurrent samples from the same patients.[6, 27] In both studies, however, only histological specimens were used, and comparisons were available for only 2 specimens from the same case. In one of the studies, although there seemed to be a tendency for an increased frequency of EZH2 mutation in the cases with transformed samples, no association between EZH2 mutation and FL transformation was seen. These results are in agreement with our study, where EZH2 mutational status varied over the course of the disease within specimens from the same patient.
Discrepant mutational status of EZH2 over time seen in our series could be explained by genetic heterogeneity. The case with previous wild-type tumors and a subsequent mutated sample may have acquired genetic abnormalities over time, a common phenomenon in lymphomas.[17, 29, 30] For the 2 cases in which the mutation was only seen in the initial samples, discrepancies may be explained wherein samples may not be truly “recurrent,” but a de novo neoplasm unrelated to the primary samples and therefore exhibiting different mutational statuses. Other possibilities are tumor heterogeneity or parallel clonal evolution of the initial and recurrent tumors. In many instances, primary and sequential transformed biopsies of FL did not share all abnormalities, indicating that rather than evolving from the initial sample, the sequential biopsies could have often come from a common precursor, with parallel clonal evolution, and one of the “subclones” may have given rise to large-cell lymphomas. Our results support the idea of a shift from the analysis of the primary tumor alone, to a combination of primary tumor, circulating tumor cells, metastatic deposits, and cell-free DNA to capture the full extent of the genetic heterogeneity present. The successful use of all types of specimens, including limited resources such as archived smears and cytospin preparations, can further be explored to elucidate the discrepancies seen over time in a mutation-enriched cohort of patients.
Mutations at the tyrosine in the SET domain of EZH2 were seen in our patients at a similar rate to previous reports.[27, 31, 33, 34] For CD79B mutations, our study showed a lower rate, being observed in just 1 case. Variations of the mutations reported here for either EZH2 or CD79B were the same as previously described, and no new variants were seen. Of interest, all cases with 2 mutated samples showed the same type of nucleotide substitution. Our study and previous reports have demonstrated that tumor specimens are heterozygous for the Y646 EZH2 mutations.
The use of MassARRAY spectrometry combined with Sanger sequencing provided definitive results for all types of samples tested in our study, although some of the biospecimens were not initially collected for the purpose of performing molecular analyses. Only one case found to be positive for the EZH2 mutation using MassARRAY spectrometry was not confirmed by direct sequencing. A possible explanation for this discrepancy might be that a small percentage of tumor cells are present in this sample and may have fallen into the range of 5% to 10% of mutated tumor cells that could be detected by MassARRAY spectrometry but would not be detectable by direct sequencing.[35, 36] Alternatively, it might also represent a false positive case, a less probable situation because there was no notable presence of salt adducts (that may create noise and “false mutation” peaks) in the preliminary manual analysis of our samples, and all other positive cases were confirmed by Sanger sequencing, which minimizes the possibility of hairpins. However, because we established Sanger sequencing as our confirmatory method for the results from MassARRAY assays and there was no confirmation of the mutation by direct sequencing in 2 independent runs, that sample was excluded from the analysis of the mutated cases. Overall, considering all the assays that were successfully performed, our samples had a very high specificity, similar to other series using this technology.[38-41]
Our specifically designed panel has further validated this technique as a fast and reliable alternative for testing multiple samples in a clinical basis. MassARRAY spectrometry is a high-throughput platform that maximizes the amount of data extracted from mutation profiling. The advantages include the standardized assay conditions and the homogeneous reaction format, which simplifies and minimizes the reagents and the processing time. Sanger sequencing requires amplification of multiple fragments per sample, with optimization of multiple primers and reactions, which besides time and costs, uses a greater quantity of DNA. In addition, MassARRAY technology has been demonstrated to be more sensitive than Sanger sequencing, having detection limits that range from 2.5% to 10%, depending on the specific mutation tested.[35, 36] In the present study, mutations involving EZH2 and CD79B were chosen due to their potential role as therapeutic targets, but an increasing number of assays for various genes can be added to this platform. In the commercially available OncoCarta Panel (Sequenom, San Diego, Calif), for example, 238 mutations from 19 oncogenes can be simultaneously interrogated, and it has already been validated using limited samples, with satisfactory results.
In summary, mutational status may vary in samples from the same case along the course of the disease, which corroborates a concept currently gaining acceptance that patients should be treated according to their current molecular findings, not on the basis of results obtained from previous specimens. The successful use of minimal and residual material from cytological samples allied to an innovative high-throughput multiplex platform for finding EZH2 and CD79B mutations in patients with B-cell NHL exponentially increases the availability of specimens for clinical analysis and may lead to an increase in the number of patients eligible for clinical trials. Future larger studies using different molecular platforms to examine the pathogenesis of tumor progression may clarify the way in which mutations evolve over time and our understanding of the related molecular events.