Ready, Set, Go: Stratification of Diffuse Large B-Cell Lymphoma in Real-Time Based on Gene Expression Profiling

A Randomized Evaluation of Molecular Guided Therapy for Diffuse Large B-Cell Lymphoma With Bortezomib (REMoDL-B); ISRCTN 51837425.

NCT01324596

The UK National Cancer Research Institute Lymphoma Clinical Studies Group in collaboration with the Swiss Group for Clinical Cancer Research and the Haematological Malignancy Diagnostic Service (HMDS), St James’s University Hospital, Leeds, through the University of Southampton Clinical Trials Office, Southampton, UK

Patients newly diagnosed with diffuse large B-cell lymphoma (DLBL) will undergo full staging and commence treatment with R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, prednisolone). During the first three-week cycle, a gene expression profile will be generated from each patient’s original diagnostic tissue sample. To ensure uniformity, all samples will be processed and analyzed at HMDS. The molecular data will be used to assign cases to either germinal center B-cell (GCB) or activated B-cell (ABC) types; and beginning with the second treatment cycle, patients will be randomized either to continue conventional R-CHOP or to also receive bortezomib at 1.3 mg/m2 (R-CHOP-B) on days 1 and 8 of each subsequent 21-day cycle (i.e., cycles 2-6).

Initially, all patients will be randomized, but an iterative design will allow closure of randomization for GCB lymphomas if there is evidence of detriment in outcomes after 55 such patients treated with R-CHOP-B have been followed for at least six months. According to the protocol, patients in the GCB group will no longer receive R-CHOP-B if the interim analysis shows that the six-month progression-free survival (PFS) for that group is < 80 percent. A second analysis will be performed for futility in the GCB group after 73 of those patients have been randomized to receive R-CHOP-B and followed for one year. If the estimate of one-year PFS is < 85 percent, further exploration of the value of bortezomib in this group of patients is not warranted based upon data from retrospective series using molecular profiling. Because of these potential interim modifications and allowing for failure of mRNA extraction and unclassifiable gene expression profiles in some cases, the total number of patients taking part in the trial may vary between 567 and 940. The study aims to randomize a minimum of 260 ABC-type lymphomas to allow for detection of a statistically significant improvement in 30-month PFS of 10 percent in the R-CHOP-B arm.

Gene expression profiling of fresh frozen tissue samples has provided new insights into the biology of DLBL. Unsupervised hierarchical clustering identified two distinct subgroups of the disease: one with GCB-like and one with ABC-like patterns of gene expression. These molecular subgroups had distinct clinical outcomes after R-CHOP chemotherapy, with the ABC-like lymphomas having an inferior three-year PFS of 40 percent, compared with 75 percent PFS in the GCB group. Several oncogenic mechanisms distinguish the two subgroups. In particular, the constitutive activation of the nuclear factor-κB (NF-κB) signalling pathway appears central to cell survival in ABC-like lymphomas. The induction of the NF-κB pathway may suppress the apoptotic effect of cytotoxic chemotherapy, and this mechanism could contribute to the observed differences in outcome. This effect might be mitigated through the use of bortezomib, which, among other actions, can reduce proteosomal degradation of the natural NF-κB inhibitor, I-κB.

Attempts have been made to simplify molecular subclassification of DLBCL by using immunohistochemistry (IHC) methods, but to date, no such approach has proven sufficiently reproducible among laboratories or even among expert hematopathologists. Consequently, this trial uses the more technically complex array technology as the basis for stratification of the two subgroups.

This is the first study to use prospective gene expression profiling in lymphoma as a means of stratifying randomization of DLBCL subgroups between treatment with R-CHOP or R-CHOP-B. The goal of the trial is to determine if there is a subset of lymphomas in which bortezomib improves outcome. Additionally, the study affords an opportunity to analyze the clinical utility of several IHC algorithms in distinguishing subgroups of DLBCL, using the expression array data as the standard of comparison. The results of this aspect of the study will become increasingly important as more targeted therapies, requiring a reproducible means of identifying the appropriate subset of patients to treat, emerge.