Norman E. Sharpless
Researchers identified genomic alterations that defined four subtypes of diffuse large B-cell lymphoma that may predict which patients will respond to treatment and have longer OS, according to study findings published in The New England Journal of Medicine.
“These findings are the culmination of 2 decades of research at NCI and elsewhere, advancing our understanding of the effect of DNA mutations and gene expression on lymphoma biology and outcome,” Norman E. Sharpless, MD, director of the NCI, said in a press release. “This refined molecular classification will be instrumental in predicting prognosis and tailoring therapy for patients with DLBCL going forward.”
Gene expression profiling identified DLBCL subgroups, including activated B-cell like (ABC) — which has approximately a 40% survival rate — and germinal center B-cell like (GBC), with a 75% survival rate. However, many patients with the GBC subtype experience disease relapse after treatment.
Researchers used a multiplatform analysis of structural genomic abnormalities and gene expression in 574 fresh-frozen DLBCL biopsy samples. They hypothesized that this hybrid approach would yield a nosology of DLBCL based on shared genetic pathogenesis which would, in turn, improve understanding of treatment response and therapeutic targets.
Researchers conducted exome and transcriptome sequencing, array-based DNA copy-number analysis and targeted amplicon resequencing of 372 gene to identify recurrent aberrations. They then used an algorithm to define genetic subtypes based on the co-occurrence of genetic alterations.
“The first question we wanted to tackle was whether there were molecular features of the tumors that could help us explain why some people were well served by chemotherapy. The second, related question was if we could understand who was not responding well to treatment, could we understand the genetics of these tumors to suggest new potential therapies beyond chemotherapies?” Louis M. Staudt, MD, PhD, director of the NCI Center for Cancer Genomics and lead author on the study, said in a press release.
Multiplatform genomic analysis showed genes were altered at significantly different frequencies in ABC and GBC subtypes (P < .01).
Researchers enriched the cohort for unclassified DLBCL — because it’s genetic composition is unknown — to represent 20% of patients compared with 11.3% of a population-based cohort. In these cases, NOTCH2 mutations and BCL6 fusions co-occurred (P = 2.78 x 10-12), which appeared to be a distinguishing factor from other DLBCLs.
Unclassified cases also appeared enriched for mutations targeting SPEN, an inhibitor of NOTCH-dependent gene expression; 30.4% of cases had NOTCH2 or SPEN mutations.
Researchers also evaluated whether genetic aberrations correlated with the ABC-GBC prediction score — a metric with low values for most GCB-like cases and high values for most ABC-like cases — and found NOTCH2 mutations and BCL6 fusions were concentrated in the center of the distribution, whereas cases with CD79B and MYD88L265P mutations were enriched at the far ABC end of the spectrum and EZH2 mutations and BCL2 translocations significantly co-occurred on the GCB end.
Overall, researchers identified four genetic subtypes:
- MCD, based on co-occurrence of MYD88L265P and CD79B;
- BN2, based on BCL6 fusions and NOTCH2 mutations;
- N1, based on NOTCH1 mutations; and
- EZB, based on EZH2 mutations and BCL2 translocations.
The subtypes harbored different gene-expression signatures and responded to immunochemotherapy differently, which suggested they differed phenotypically.
Researchers observed favorable survival in BN2 and EZB cases and inferior outcomes in the MCD and N1 cases, which were dominated by ABC cases. EZB included mostly GCB cases, whereas BN2 represented all three gene-expression subgroups.
“This shows we’ve gone beyond where we were. Before, even with our most advanced molecular diagnosis, we would have said all ABC tumors are the ‘bad’ type and they need to be treated aggressively,” Staudt said. “Now we can implement this kind of classification and say that even if a patient has the ‘bad’ ABC type, they have the ‘good’ genetic type, BN2, so there’s a much better chance of chemotherapy curing the disease.”
Gene-expression predictor classifications of MCD, BN2, N1 and EZB cases allowed researchers to estimate that these four subtypes would comprise 46.6% of DLBCL cases.
“The goal is to find the right drug for the right person at the right time. And we feel this genetic understanding of diffuse lymphoma is a step forward in precision therapy,” Staudt said. – by Melinda Stevens
Disclosures: Staudt reports grants from NCI Strategic Partnering to Evaluate Cancer Signatures grant (5U01CA157581-05) during the conduct of the study, as well as a pending patent for Nanostring ABC/GCB predictor-U.S. Provisional Patent Application No. 15/035,101. Please see the study for a list of all other authors’ relevant financial disclosures.