Abstract

Diffuse large B-cell lymphoma (DLBCL) is the most common type of non-Hodgkin lymphoma in the Western world, accounting for approximately 30,000 new cases per year in the United States. Modern immune chemotherapy cures approximately 60% of patients with DLBCL, indicating that approximately 40% of the patients do not benefit from current standard therapy. Poor-risk patients should be prospectively identified so they can be offered more effective therapy. So how can we identify patients who are unlikely to be cured with standard therapy? And once poor-risk patients are identified, how can we use this information to improve treatment outcome? For almost two decades, the most widely used prognostic factors model in DLBCL was the International Prognostic Index, which identified patients with good-, intermediate-, and poor-risk features. However, the International Prognostic Index failed to provide information that guided therapy. Accordingly, empirical approaches to test new strategies in patients with poor-risk features failed to make a significant impact. Subsequently, gene expression profiling (GEP) studies revealed that histologically uniform lymphoma subtypes are heterogeneous at molecular levels. In DLBCL, GEP studies identified two biologically distinctive groups on the basis of their cell of origin (COO): the germinal center B-cell (GCB) and the activated B-cell (ABC) origin subtypes. An additional small subgroup could not be precisely classified into these two entities. Patients with the ABC subtype had a lower chance of cure with modern therapy, and therefore, became the new biologically defined poor-risk subset with unmet medical needs. Because ABC-DLBCL is highly enriched with activated nuclear factor kappa B (NFB) signaling, this biologic model also provided information on which agents should be prioritized for testing in the clinical setting. Accordingly, agents that modulated NFB signaling were preferentially tested in patients with ABC subtype. To expedite the development of new agents, investigators and regulatory agencies became interested in designing and conducting clinical trials in patients with ABC-DLBCL, because this group has more urgent unmet medical needs and because it was believed that some agents demonstrated preferential clinical activity in this subset. Because standard GEP tests initially required the availability of fresh-frozen lymphoma tissues, surrogate immunohistochemistry (IHC)-based biomarker tests were developed, which could be adopted in a wide range of clinical hematopathology diagnostic laboratories. Concordance between GEP and IHC methods rarely exceeded 80%. Despite this limitation, IHC-based COO classification was frequently used to stratify patients in clinical trials that enrolled patients with DLBCL. Most recently, DLBCL COO was identified by using GEP methods applied to formalin-fixed paraffinembedded tissue sections and a more simplified digital geneexpression signature (Lymph2Cx) using a nanostring platform. The Lymph2Cx assay was shown to have a better correlation with standard GEP methods and excellent concordance of COO assignment between laboratories. In the accompanying article, Scott et al applied the Lymph2Cx assay to a large cohort of patients (n 344) treated with standard rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) therapy. The Lymph2Cx assay identified 32% to be of the ABC subtype and 56% to be of the GCB subtype. Similar to standard GEP methods, Lymph2Cx also identified a small unclassified group (11%). The 5-year progression-free survival (PFS) was 73% for the GCB group, 54% for the unclassified group, and 48% for the ABC group. In the same study, Scott et al examined the prognostic significance of MYC and BCL2 protein expression by using IHC methods. The authors confirmed prior observations of the prognostic significance of this assay. Patients with DLBCL who expressed high levels of MYC and BCL2 proteins had a 5-year PFS of 48% compared with 70% for those who had lower protein expression. The development of clinical assays that can be easily adopted into clinical practice to classify DLBCL on the basis of the COO is important in the field. These assays, including the Lymph2Cx, provide valuable reproducible tools to identify poor-risk patients with DLBCL and to facilitate the development of clinical trials capturing these patients. Despite these advances, one should not lose sight of the big picture. First, none of these prognostic models and diagnostic tools is perfect. For example, all of these models failed to identify a group of patients who are 100% cured with standard R-CHOP. In the study reported by Scott et al, even the favorable GCB subtype had a 5-year PFS of 73%, which means that 27% of patients received toxic therapy without benefit. Conversely, 54% of patients with the ABC subtype still had benefit from R-CHOP. Second, the COO classification is applicable JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 33 NUMBER 26 SEPTEMBER 1

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call