Directing Treatment by Molecular Subtype in Diffuse Large B-Cell Lymphoma: Ready for Primetime?

Rapidly evolving genetic and epigenetic “mutation-seeking” platforms, high-level cranial processors, and the ability to access patient tissue for research (which varies depending on both patients’ willingness and their country of residence) have resulted in multiple novel targets in diffuse large B-cell lymphoma (DLBCL). As a result, a new era of drug discovery is upon us, and many classes of agents are now jockeying for position to be used with competing upfront strategies or in combinations as part of nonchemotherapeutic strategies. In this issue of ONCOLOGY, Drs. Dunleavy and Wilson review the clinical significance of DLBCL subtyping and current treatment paradigms.[1]

DLBCL has been recognized as a microscopically heterogeneous disease for decades; this characterization has been confirmed at the molecular and epigenetic levels by gene expression profiling and other platforms. However, physicians cannot yet reliably obtain the results of a patient’s molecularly based disease characterization in time for the treatment consultation. Immunohistochemistry (IHC) stain-based algorithms that act as a surrogate for molecular-based subtype analysis have been widely used, but their reproducibility and utility continue to be debated.

Nevertheless, molecular classification of DLBCL, initially by gene expression profiling, has resulted in three distinct subtypes: germinal center B-cell–like (GCB), activated B-cell–like (ABC), and primary mediastinal B-cell lymphoma (PMBL).[2] As described by Drs. Dunleavy and Wilson, these discrete DLBCL subtypes appear to provide an explanation both for the lymphoma cell of origin and for the distinct prognosis of each type. Simplistically, the GCB phenotype evolves from a germinal center B cell, the ABC phenotype evolves from a post–germinal center B cell, and PMBL evolves from a resident thymic B cell. We now know that the GCB subtype is more than twice as common as the ABC subtype and confers a better overall outcome.[3] These conclusions are based on retrospective experiences with homogeneously treated patients. Prospective experience in newly diagnosed patients is lacking.

Another issue in the subtype discussions that was not commented upon in the review is how to interpret-neither overinterpret, nor ignore altogether-the multitude of IHC markers that are increasingly being reported (or being requested on tests) in DLBCL as surrogates for molecular subtyping. IHC stains that include CD30[4] and MYC[5] with more standard markers, such as CD10, BCL6, and MUM1 (Hans algorithm[6]), or an expansion set with GCET1 and FOXP1 (Choi algorithm[7]), often result in perplexing pathologic impressions without a change in the treatment paradigm from a decade ago. At this time, we still routinely use the Hans algorithm to distinguish between the GCB and ABC subtypes in DLBCL, and we often let the clinicopathologic situation dictate whether the patient has PMBL. We have yet to incorporate gene expression profiling or other analogous platforms into our clinical diagnostic workflow, although undoubtedly, technologic advances will eventually make the option of molecular subtype analysis a real possibility.

We await with great anticipation the results of the recently accrued Cancer and Leukemia Group B (CALGB) 50303 (ClinicalTrials.gov ID NCT00118209) phase III clinical trial co-led by Drs. Andrew Zelenetz and Wyndham Wilson. This trial randomly assigned patients with newly diagnosed DLBCL to treatment with either R-CHOP (rituximab + cyclophosphamide, doxorubicin, vincristine, and prednisone) or DA (dose-adjusted)–EPOCH-R (etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, and rituximab). Hopefully, this trial will not only help clarify the standard-of-care treatment in DLBCL, but will also expand upon current comparisons between molecular-based gene expression profiling and IHC subtype algorithms, and will confirm or refute prior cell of origin experiences. As has been suggested in the upfront setting, subtype analysis may also predict outcomes in relapsed and refractory DLBCL. In the subset analysis from the Collaborative Trial in Relapsed Aggressive Lymphoma (CORAL study) that utilized the Hans algorithm, those patients with the GCB phenotype who were treated with DHAP (dexamethasone, cytarabine, cisplatin) had an improved progression-free survival (PFS) compared with those who received ICE (ifosfamide, carboplatin, etoposide).[8] A GCB phenotype was also associated with a better outcome than an ABC phenotype overall. Although the patients in this trial were not randomized by subtype, the resulting data still suggest a divergence in sensitivity between the GCB and ABC subtypes. A similar situation may arise when the results of CALGB 50303 are reported, and if gene expression profiling or another platform is not ready for broad application with global acceptance, then the alternative surrogate IHC algorithms may remain relevant despite their lower sensitivity and specificity.

For the time being, we continue to use R-CHOP given every 21 days as our standard-of-care upfront regimen in DLBCL, regardless of molecular or IHC subtype, but we have increasingly felt the weight of the impressive results of the phase II CALGB study of DA-EPOCH-R (ClinicalTrials.gov ID NCT00032019).[9] Also, the regimen of dose-intensified doxorubicin, cyclophosphamide, vindesine, bleomycin, and prednisone (ACVBP) was shown in a randomized trial in Europe to have a significantly prolonged 3-year PFS compared with R-CHOP (87% vs 73%) in younger patients, although at the price of increased toxicity.[10] The regimen has not been studied in the United States because it includes the drug vindesine. The activity of bortezomib and ibrutinib in the ABC phenotype of DLBCL may drive further upfront strategies, especially if the results of CALGB 50303 favor DA-EPOCH-R.[11,12] An important point that is often glossed over is that if DA-EPOCH-R is to be used as curative therapy in any subtype of DLBCL-and if clinicians expect to see results similar to those in published experiences-the intent-to-dose-adjust paradigm might need to be followed. However, it is also possible that no dose modification, as with R-CHOP, might be equally efficacious.

We echo the sentiment of Drs. Dunleavy and Wilson that major strides have been made in DLBCL treatments. However, we worry that the molecular and epigenetic studies may only be available to a few, and we tremble at the possibility that the search for a better IHC algorithm may be abandoned. To date, both approaches have had significant implications and will continue to guide treatment paradigms in DLBCL. Not surprisingly, our heads hit the pillow later at night as the complexity of treating DLBCL increases, but we are well rested on account of the progress that is being made.

Financial Disclosure:Dr. Armitage serves as a consultant to GlaxoSmithKline, Seattle Genetics, Genentech, Roche, Spectrum Pharmaceuticals, and Ziopharm Oncology; he is also a member of the Board of Directors of Tesaro, Inc. Dr. Lunning has no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this commentary.

References:

1. Dunleavy K, Wilson WH. Appropriate management of molecular subtypes of diffuse large B-cell lymphoma. Oncology (Williston Park). 2014;28:326-34.

2. Rosenwald A, Wright G, Chan WC, et al. The use of molecular profiling to predict survival after chemotherapy for diffuse large-B-cell lymphoma. N Engl J Med. 2002;346:1937-47.

3. Lenz G, Wright G, Dave SS, et al. Stromal gene signatures in large-B-cell lymphomas. N Engl J Med. 2008;359:2313-23.

4. Hu S, Xu-Monette ZY, Balasubramanyam A, et al. CD30 expression defines a novel subgroup of diffuse large B-cell lymphoma with favorable prognosis and distinct gene expression signature: a report from the International DLBCL Rituximab-CHOP Consortium Program Study. Blood. 2013;121:2715-24.

5. Hu S, Xu-Monette ZY, Tzankov A, et al. MYC/BCL2 protein coexpression contributes to the inferior survival of activated B-cell subtype of diffuse large B-cell lymphoma and demonstrates high-risk gene expression signatures: a report from The International DLBCL Rituximab-CHOP Consortium Program. Blood. 2013;121:4021-31; quiz 250.

6. Hans CP, Weisenburger DD, Greiner TC, et al. Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray. Blood. 2004;103:275-82.

7. Choi WW, Weisenburger DD, Greiner TC, et al. A new immunostain algorithm classifies diffuse large B-cell lymphoma into molecular subtypes with high accuracy. Clin Cancer Res. 2009;15:5494-502.

8. Thieblemont C, Briere J, Mounier N, et al. The germinal center/activated B-cell subclassification has a prognostic impact for response to salvage therapy in relapsed/refractory diffuse large B-cell lymphoma: a bio-CORAL study. J Clin Oncol. 2011;29:4079-87.

9. Wilson WH, Jung SH, Porcu P, et al. A Cancer and Leukemia Group B multi-center study of DA-EPOCH-rituximab in untreated diffuse large B-cell lymphoma with analysis of outcome by molecular subtype. Haematologica. 2012;97:758-65.

10. Recher C, Coiffier B, Haioun C, et al. Intensified chemotherapy with ACVBP plus rituximab versus standard CHOP plus rituximab for the treatment of diffuse large B-cell lymphoma (LNH03-2B): an open-label randomised phase 3 trial. Lancet. 2011;378:1858-67.

11. Dunleavy K, Pittaluga S, Czuczman MS, et al. Differential efficacy of bortezomib plus chemotherapy within molecular subtypes of diffuse large B-cell lymphoma. Blood. 2009;113:6069-76.

12. Wilson WH, Gerecitano JF, Goy A, et al. The Bruton’s tyrosine kinase (BTK) inhibitor, ibrutinib (PCI-32765), has preferential activity in the ABC subtype of relapsed/refractory de novo diffuse large B-cell lymphoma (DLBCL): interim results of a multicenter, open-label, phase 2 study. Blood; ASH Annual Meeting Abstracts. 2012;120:abstr 686.