NEW YORK—The importance of cell of origin in choosing a treatment for diffuse large B-cell lymphoma (DLBCL) is a topic “that has been kicking around for 16 years,” according to a speaker at the NCCN 11th Annual Congress: Hematologic Malignancies.
Cell of origin was first described in the year 2000 as a distinguishing factor in large-cell lymphoma, said the speaker, Andrew Zelenetz, MD, PhD, of Memorial Sloan Kettering Cancer Center in New York, New York.
The cell of origin in DLBCL—whether it’s germinal center B cell (GCB), activated B cell (ABC), or unclassified—contributes to biological and clinical heterogeneity of the disease.
“And more importantly, activated B-cell diffuse large B-cell lymphoma and germinal center diffuse large B-cell lymphoma are simply different diseases,” Dr Zelenetz said.
He then elaborated on the importance of cell of origin in treating DLBCL.
Dr Zelenetz noted that ABC and GCB lymphomas have different molecular pathways. ABC lymphomas are very dependent on the NF-kB pathway and have more active signaling through the B-cell receptor.
The GCB lymphomas tend to have more tonic regulation, and the PI3 kinase/mTOR pathway is more critical. GCB lymphomas have more genomic instability.
“So cell of origin determination identifies tumors with distinct biology, may provide prognostic information, and may be predictive for treatment selection,” Dr Zelenetz said. “Unfortunately, cell of origin is not the whole story.”
Gene mutations occur in large-cell lymphoma “just like every other cancer,” Dr Zelenetz said. And the vast majority occur in both lymphoma subtypes, he added, further complicating our understanding of the biology of these tumors.
Some of these mutations predict for sensitivity to treatment, while others predict for resistance. For example, CARD11 predicts for resistance to ibrutinib, while CD79b predicts for sensitivity.
Determining the cell of origin
Gene-expression profiling on fresh tissue is considered the gold standard, but “it is clearly not a clinical tool,” Dr Zelenetz said. It requires the Wright classifier, a statistical method based on Bayes’ rule, to make patient-level assignments to 1 of the 3 subgroups.
Immunohistochemistry is widely available, but reproducibility may be difficult. Many assays exist, such as the Hans, Choi, and Muris assays, but, in many studies, there may be a lack of correlation with gene-expression profiling.
In the last few years, gene-expression profiling of formalin‐fixed paraffin‐embedded (FFPE) tissue has emerged as a reliable method. The assay is reproducible between laboratories, and it’s reproducible between different sets of reagents.
“[T]here is tremendous correlation between the Lymph2Cx assay and the gold standard,” Dr Zelenetz added.
“So here we have a robust assay,” he said, which allows investigators to explore whether the cell of origin is prognostic in large-cell lymphoma.
In a data set of 339 patients with de novo DLBCL treated with R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone), the Lymph2Cx assay showing cell of origin was predictive of overall and progression-free survival.
However, the same exact assay applied to the RICOVER-60 data from the German High-Grade Non-Hodgkin Lymphoma Study group was not predictive, Dr Zelenetz reported, based on a personal communication from one of the investigators. There was a slight trend in favor of GCB tumors, but it was not statistically significant.
And the REMoDL-B study, using gene-expression profiling of FFPE tissue with the DASL assay, also didn’t show any difference in outcome between ABC or GCB tumors.
So gene-expression profiling of FFPE tissue does not universally show a prognostic difference, Dr Zelenetz said.
Influence of chemotherapy by cell of origin
CALGB 59910 showed that GCB tumors had a superior event-free, progression-free, and overall survival with dose-adjusted EPOCH-R (etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, and rituximab) compared to ABC tumors.
“However, this is a phase 2, hypothesis-generating experiment,” Dr Zelenetz pointed out, and the results of the confirmatory study comparing dose-adjusted EPOCH-R and R-CHOP21 (CALGB 50303) will be presented later this year at the ASH Annual Meeting.
Sequential, non-cross-resistant chemotherapy
Data from the Memorial Sloan Kettering study (MSKCC 01-142/08-146; NCT00712582) of sequential therapy with R-CHOP followed by ICE (ifosfamide, carboplatin, and etoposide) demonstrate “excellent” progression-free and overall survival, Dr Zelenetz said.
“When we analyzed the outcome by cell of origin, there was a suggestion that the patients with the non-germinal center tumors were actually doing better than the germinal center tumors,” he added.
He pointed out one of the limitations of the study is that the cell of origin was determined by the Hans model. Nevertheless, the study raised another testable hypothesis: sequential therapy might overcome the adverse impact of the non-germinal center tumors.
Cell of origin analysis of the prospective, randomized study (LNH 03-2B) comparing R-ACVBP (rituximab, doxorubicin, cyclophosphamide, vindesine, bleomycin, and prednisone) to R-CHOP showed that whether patients with GCB tumors received CHOP or ACVBP didn’t make “a whit of difference,” Dr Zelenetz said, in terms of progression-free and overall survival.
However, patients with ABC tumors demonstrated an “enormous difference in favor of R-ACVBP,” he said.
“Again, evidence that you can overcome the adverse effect of the ABC tumors with chemotherapy.”
Dr Zelenetz pointed out that R-ACVBP and R-CHOP followed by ICE are “actually remarkably similar regimens.” Both are sequential, both include consolidation, and both incorporate high-dose ifosfamide and etoposide.
“[S]o they actually reinforce each other,” he said, “demonstrating a similar result.”
Lenalidomide in the relapsed/refractory setting has modest activity in DLBCL, with most of the benefit accruing to patients with non-germinal center tumors.
Two clinical studies evaluated the impact of adding lenalidomide to standard chemotherapy.
In an Italian series using lenalidomide (L) plus R-CHOP21 in elderly untreated patients, the combination produced outstanding progression-free and event-free survival, but with no significant differences between the subtypes.
A US study of RL-CHOP versus R-CHOP included 87 matched historical controls treated with R-CHOP and 64 patients treated with RL-CHOP. Patients with non-germinal center tumors treated with RL-CHOP fared much better than historical controls treated with R-CHOP.
However, among germinal center tumors, “there was not a hint of any difference,” Dr Zelenetz noted.
Two studies—E1412, using an unselected population, and the international ROBUST study, selecting for patients with ABC tumors—are underway to confirm that the benefit with lenalidomide is in patients with activated B-cell tumors.
Ibrutinib, a Bruton’s tyrosine kinase inhibitor, also has modest activity as a single agent in an unselected patient population with relapsed/refractory DLBCL. And most of the patients who demonstrated benefit had activated B-cell tumors.
Upon further analysis, investigators found that response was enhanced by the CD79b mutation, but it was not necessary for a response. And patients with CARD11 had no response.
MYD88 mutations seemed to cause resistance to ibrutinib, unless the mutation was associated with the CD79b mutation, and then patients had a “great” response, Dr Zelenetz explained.
In the upfront setting, a phase 1b study of R-CHOP plus ibrutinib demonstrated the safety of the combination, which had an overall survival rate of 100% and a complete response rate of 91%.
The prospective, randomized, phase 3 PHOENIX trial (NCT01855750) evaluating the combination in newly diagnosed non-germinal center DLBCL has completed accrual, but analysis is still pending.
“The prognostic significance of cell of origin is still controversial,” Dr Zelenetz wrapped up, “although I actually believe there is a prognostic difference in unselected registry patients.”
Sequential chemotherapy with ifosfamide and etoposide consolidation does very well in activated B-cell tumors, both in phase 2 and phase 3 studies.
“Importantly, small molecules seem to have differential effects totally predictable based on the biology of the difference between activated B-cell and germinal center tumors,” Dr Zelenetz said.
“But the big wild card here is somatic mutations further complicate things and will have to be incorporated into our understanding in the selection of patients.”