DLBCL Cell of Origin: What Role Should It Play in Care Today?

September 18, 2018

In this article, we review the methods of determining cell of origin (COO); use of COO in clinical practice; clinical trials in DLBCL according to COO; and future directions of tailoring treatment, including alternate categorization of genetic subtypes or clusters in DLBCL.

Diffuse large B-cell lymphoma (DLBCL) is curable in about two-thirds of patients. Research has focused on determining which patients have less favorable prognoses so that they can be considered for novel targeted-treatment strategies. In 2000, gene expression profiling was used to define two principal DLBCL molecular subtypes, germinal center B-cell–like (GCB) and activated B-cell–like (ABC). Patients with GCB DLBCL have more favorable outcomes than those with ABC DLBCL when treated with standard immunochemotherapy. Alternate strategies to characterize molecular subtype include approximation with immunohistochemistry algorithms, and more recently the NanoString gene expression platform. Numerous studies have investigated novel agents in DLBCL with respect to GCB and ABC (or non-GCB) subtypes, but R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) remains the standard of care for most patients. Here we review the methods of determining cell of origin (COO); use of COO in clinical practice; clinical trials in DLBCL according to COO; and future directions of tailoring treatment, including alternate categorization of genetic subtypes or clusters in DLBCL.

Methods of Determining Cell of Origin (COO)

From a panel of genes involved in lymphocyte development and activation, Alizadeh et al created a unique DNA microarray (termed the Lymphochip) and determined that a subset of diffuse large B-cell lymphoma (DLBCL) cases expressed genes associated with the germinal center B-cell reaction, while another group expressed genes involved in lymphocyte activation in the postgerminal center state.[1] The former group, called germinal center B-cell–like (GCB) lymphomas, expresses genes including those encoding for CD10 and BCL6. The latter, the activated B-cell–like (ABC) group, notably expresses IRF4 and BCL2. The investigators found a difference in overall survival (OS) between GCB and ABC DLBCL patients. A larger follow-up study of DLBCL patients treated with standard anthracycline-based therapy reported a 60% 5-year OS rate for GCB patients vs 35% in ABC patients.[2] ABC patients were more likely to be older than age 60 and to have an Eastern Cooperative Oncology Group (ECOG) performance status score > 1. Subsequent studies have led to an increased understanding of the distinct pathways that lead to lymphomagenesis in GCB vs ABC DLBCLs that may represent therapeutic targets.[3]

Because gene expression profiling is not widely accessible, researchers approximated molecular subtypes using immunohistochemical (IHC) patterns.[4-8] The Hans algorithm, which uses CD10, BCL6, and MUM1 status to distinguish between GCB and non-GCB DLBCL, remains the most widely used of these algorithms (Figure). The nomenclature includes GCB and non-GCB rather than ABC. Each algorithm has confirmed that GCB DLBCL has better outcomes than non-GCB.[9] However, immunochemistry is imperfect in assessing molecular subtype, and more precise strategies have been pursued.[10]

The NanoString is a newer, alternate method of determining COO using formalin-fixed paraffin-embedded tissue. Lymph2Cx is one assay that employs NanoString technology; it uses a 20-gene panel to assess COO and may be more accurate than the Hans algorithm.[11,12] This method is currently being studied in the clinical trial setting. The ROBUST trial, an international phase III study of R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) plus lenalidomide or placebo, has employed the Lymph2Cx. The investigators recently reported that 2,110 samples were assessed; 788 were ABC, 1,010 were non-ABC, and the remaining 312 (15%) were unable to be processed based on technical problems such as inadequate tissue amount.[13] They concluded that Lymph2Cx is a feasible method of determining COO in this setting in which quick turnaround time is desired. Though it is infrequently used in clinical practice at this time, this type of technology may become more widespread, particularly if clinical implications grow.

Use of COO in Clinical Practice

The majority of pathologists who report COO do so based on IHC patterns. While multiple studies have demonstrated that ABC (or non-GCB) has a less favorable prognosis than GCB DLBCL,[1,2] the data are insufficient to treat the molecular subtypes differently in the absence of cytogenetic abnormalities. Six cycles of R-CHOP generally remain the standard of care for most patients with advanced-stage DLBCL. Randomized clinical trials incorporating targeted treatments in these frontline regimens are described in detail below and in Table 1.

COO is somewhat helpful in determining which DLBCL patients are appropriate to assess for double-hit (DHL) or triple-hit lymphoma (THL). These entities, which are referred to as high-grade B-cell lymphoma (HGBL) with MYC and BCL2 and/or BCL6 rearrangements in the World Health Organization 2016 classification, are characterized by such chromosomal rearrangements.[14] Double protein expressor lymphomas (DELs) have increased protein expression of MYC and BCL2 in the absence of cytogenetic abnormalities. While both entities have poorer response to standard immunochemotherapy compared with DLBCL not otherwise specified, HGBL-DHL/THLs are particularly aggressive and are typically treated with more intensive chemotherapy regimens, such as dose-adjusted (DA) EPOCH-R (etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, and rituximab).[15,16]

DHL/THLs are more commonly of a GCB subtype; in contrast, DELs are more likely of ABC subtype (or non-GCB).[17] A recent investigation of 1,222 DLBCL patients assessed fluorescence in situ hybridization (FISH) tests for MYC, BCL2, and BCL6, as well as COO using Lymph2Cx and IHC.[18] HGBL-DHL/THLs comprised 7.9% of DLBCL cases and were almost entirely GCB (13.3% GCB vs 1.7% ABC). The study determined that a method of screening DLBCL patients in which FISH for MYC, BCL2, and BCL6 is done solely on those cases determined to be GCB by either Lymph2Cx or IHC (using the Hans criteria)-and that also demonstrate double protein expression of MYC and BCL2-would limit testing to 11% to 14% of DLBCLs; however, about one-third of HGBL-DHL/THL cases would be missed. As a result, there is no consensus on screening strategy for deeper assessment of DHL/THLs within specific COO subtypes.

Clinical Trials With DLBCL According to COO

A number of clinical trials have assessed novel agents in newly diagnosed and relapsed/refractory DLBCL patients based on molecular subtype. Early studies suggested potential benefit of lenalidomide, bortezomib, and ibrutinib more selectively in ABC/non-GCB compared with GCB DLBCL.

Lenalidomide was initially studied in 40 heavily pretreated relapsed/refractory DLBCL patients. Investigators assessed responses based on IHC using the Hans criteria and found that the overall response rate (ORR) of the 23 GCB patients was 8.7%, compared with 23.5% in non-GCB, with a difference in progression-free survival (PFS) of 1.7 vs 6.2 months in favor of non-GCB patients.[19] A phase II study of R-CHOP plus lenalidomide in newly diagnosed DLBCL patients between 60 and 80 years of age found similar results in non-GCB and GCB subtypes. The 2-year OS rate of 92% and 2-year PFS rate of 80% were notable, since non-GCB patients have historically had inferior outcomes compared with GCB patients.[20] The phase III REMARC study randomized patients with newly diagnosed DLBCL (aged 60 to 80 years) to maintenance lenalidomide or placebo for up to 2 years following treatment with R-CHOP.[21] Interestingly, there was a more favorable PFS in GCB than in non-GCB patients treated with lenalidomide maintenance when IHC was used to determine COO, but not when NanoString was used. The ECOG 1412 study enrolled newly diagnosed stage II bulky and stages III/IV DLBCL patients to R-CHOP with lenalidomide vs placebo. Results have not yet been reported for this study or for the ROBUST study, which is investigating the same combination.

Bortezomib was first assessed in 49 patients with relapsed/refractory DLBCL.[22] Gene expression profiling and IHC were performed in about half of patients; results were concordant in all but one. Twenty-three patients were treated with single-agent bortezomib (part A); responding patients received DA-EPOCH-R and an additional group received the combination of bortezomib and DA-EPOCH-R (part B). Complete response (CR) rates were strikingly higher in ABC patients (41.5% vs 6.5%), and median OS was 10.8 months for ABC vs 3.4 months for GCB patients (P = .0024). Multiple trials have subsequently studied bortezomib plus chemotherapy in newly diagnosed DLBCL patients. A phase I/II study of 14 GCB and 18 non-GCB patients found similar PFS and OS in both groups treated with bortezomib plus R-CHOP.[23] Three randomized studies (PYRAMID, LYM2034, and REMoDL-B) assessed chemotherapy vs chemotherapy plus bortezomib in ABC (non-GCB) patients.[24-26] The REMoDL-B study also included GCB patients. PYRAMID and LYM2034 reported no differences in CR and PFS rates. While final results from the REMoDL-B study are pending, awaiting 30-month follow-up, patients with ABC and GCB DLBCL had similar PFS rates.

Ibrutinib was also investigated as a single agent in relapsed/refractory DLBCL.[27] Response rates were significantly higher in ABC (14/38) vs GCB (1/20) patients, with median OS of 10.32 vs 3.35 months in the two groups. A phase IB study of R-CHOP plus ibrutinib included 11 patients with COO data; 5 of 7 GCB patients had a CR vs 4 of 4 non-GCB patients. The majority were able to complete 6 planned cycles of R-CHOP.[28] Another study (PHOENIX) randomized patients with newly diagnosed non-GCB DLBCL to receive either R-CHOP or R-CHOP plus ibrutinib. Results have not yet been reported.

While many of the randomized trials in patients with newly diagnosed DLBCL are either ongoing or pending final results, those reported have shown improved results compared with prior studies of standard immunochemotherapy. One explanation is that the “screening steps” required to enroll in a prospective trial of a targeted agent in a DLBCL subtype result in a more favorable patient population than those included in the retrospective studies in which COO differences were established. This unintended selection bias could blunt the impact of the addition of a targeted agent in a DLBCL setting. Finally, a number of combination studies with these novel agents (including ibrutinib and lenalidomide) are ongoing in relapsed/refractory DLBCL patients (eg, ClinicalTrials.gov identifier: NCT02077166). At the current time, there is insufficient evidence to incorporate these drugs into therapy for newly diagnosed DLBCL outside of a clinical trial setting.

Future Directions for Tailoring Treatment in DLBCL

Over the past 10 years, DHL/THL patients have emerged as the DLBCL cases with the worst outcomes. DEL cases have an intermediate prognosis compared with DHL/THL and standard DLBCL. Strategies are being undertaken to try to improve outcomes in these patients with more aggressive disease. To date, however, there have been no clinical trials specifically enrolling DHL/THL and DEL patients. Based on retrospective data, the current standard of care for newly diagnosed DHL/THL patients is DA-EPOCH-R. A multicenter study of 311 DHL/THL patients found that the 61 patients who received DA-EPOCH-R had better CR and partial response rates than those who received other regimens.[15] PFS was improved in patients who received intensive regimens compared with R-CHOP, but outcomes remained poor, with median OS of 21.9 months. A second retrospective study at MD Anderson included 129 patients, the majority of whom had DHL/THL. Those treated with DA-EPOCH-R had a CR rate of 68% and a 67% 2-year event-free survival (EFS) rate. The CR rate was significantly higher than in patients who received R-CHOP (40%), and the 2-year EFS rate was higher than that of patients who received a Hyper-CVAD–based regimen (32%).[16] An investigator-initiated phase I trial of the regimen of DA-EPOCH-R plus the BCL2 inhibitor venetoclax (ClinicalTrials.gov identifier: NCT03036904) is currently enrolling patients with newly diagnosed DLBCL. A subsequent trial of chemotherapy plus venetoclax in DHL/THL and DEL is being planned through the Alliance for Clinical Trials in Oncology.

A more comprehensive understanding of DLBCL molecular aberrations was revealed in two recently published studies (Table 2). One of these identified four genetic subtypes of DLBCL.[29] The researchers investigated 574 DLBCL samples using exosome and transcriptome sequencing, deep amplicon resequencing, and DNA copy-number analysis. Of particular significance is that they were able to better characterize the 20% of cases that were unclassified (ie, neither GCB nor ABC) in their dataset. Little was previously known about this group of patients. The genetic subtypes were characterized by shared mutations and were termed MCD (MYD88 L265P and CD79B mutations), BN2 (BCL6 fusions and NOTCH2 mutations), N1 (NOTCH1 mutations), and EZB (EZH2 mutations and BCL2 translocations). MCD and N1 were primarily ABC DLBCLs; EZB comprised mostly GCB DLBCLs; and BN2 included ABC, GCB, and unclassified cases. BN2 cases frequently had abnormalities affecting the nuclear factor kappa B (NF-κB) pathway. A clinical assessment of 240 newly diagnosed DLBCL patients showed 5-year OS rates for the MCD, N1, BN2, and EZB subtypes of 26%, 36%, 65%, and 68%, respectively. Interestingly, ABC patients in the BN2 group had better PFS and OS than ABC patients in the MCD and N1 groups. The group with the worst outcomes, MCD, had a trend toward extranodal involvement and shared mutations with primary central nervous system lymphomas.

The second study used consensus clustering of genetic alterations and defined five subsets within 304 primary DLBCLs.[30] Cluster 1 consisted primarily of ABC DLBCLs with BCL6 structural variants (SVs) and NOTCH2 signaling. Cluster 2 included both ABC and GCB DLBCLs with biallelic inactivation of TP53 and 17p loss. Cluster 3 included mainly GCB DLBCLs with BCL2 mutations positioned beside IgH enhancer and contained mutations in chromatin modifiers KMT2D, CREBBP, and EZH2. Cluster 4 also consisted of GCB DLBCLs, but these cases had mutations involved in immune evasion, BCR/PI3K signaling, and the NF-κB and RAS/JAK/STAT pathways. Cluster 5 consisted primarily of ABC DLBCLs and included cases demonstrating 18q gains and BCL2 and MALT1 expression. Cluster 0 included a small number of cases without a pattern of cohesive genetic alterations. An assessment of patient outcomes found that clusters 0, 1, and 4 had better outcomes than clusters 3 and 5, while outcomes in cluster 2 were intermediate, with ongoing progression over time. Of interest is that mutations involving MYC, BCL2, and BCL6 were found in all clusters; therefore, DHL/THL appears to be much more complex than previously described.

The heterogeneity of DLBCL has been a key limitation in the improvement of outcomes for this disease. We anticipate that a more thorough understanding of DLBCL molecular abnormalities will enable researchers to develop accurate targeting strategies. For example, MCD cases may be particularly targetable by Bruton tyrosine kinase inhibitors. Other shared mutations may help to elucidate which agents are more likely to be effective in the different genetic subtypes/clusters. Next-generation sequencing techniques may allow investigators to perform gene mutation testing along with testing for MYC, BCL2, and BCL6 status to assess these subtypes/clusters at diagnosis. We envision a clinical trial structure that would first determine a DLBCL patient’s genetic features and then ensure that the patient received an appropriate targeted therapy based on the result. The identification of novel genetic groups is an important breakthrough that expands our understanding of DLBCL beyond COO and may lead to therapies that are more effective than R-CHOP.

Financial Disclosure:Dr. Rutherford has served as a consultant for Janssen Scientific Affairs. Dr. Leonard has served as a consultant for Celgene and Genentech/Roche.


1. Alizadeh AA, Eisen MB, Davis RE, et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature. 2000;403:503-11.

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. 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.

5. 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.

6. Muris JJ, Meijer CJ, Vos W, et al. Immunohistochemical profiling based on Bcl-2, CD10 and MUM1 expression improves risk stratification in patients with primary nodal diffuse large B cell lymphoma. J Pathol. 2006;208:714-23.

7. Nyman H, Jerkeman M, Karjalainen-Lindsberg ML, et al. Prognostic impact of activated B-cell focused classification in diffuse large B-cell lymphoma patients treated with R-CHOP. Mod Pathol. 2009;22:1094-101.

8. Natkunam Y, Farinha P, Hsi ED, et al. LMO2 protein expression predicts survival in patients with diffuse large B-cell lymphoma treated with anthracycline-based chemotherapy with and without rituximab. J Clin Oncol. 2008;26:447-54.

9. Meyer PN, Fu K, Greiner TC, et al. Immunohistochemical methods for predicting cell of origin and survival in patients with diffuse large B-cell lymphoma treated with rituximab. J Clin Oncol. 2011;29:200-7.

10. Gutierrez-Garcia G, Cardesa-Salzmann T, Climent F, et al. Gene-expression profiling and not immunophenotypic algorithms predicts prognosis in patients with diffuse large B-cell lymphoma treated with immunochemotherapy. Blood. 2011;117:4836-43.

11. Scott DW, Wright GW, Williams PM, et al. Determining cell-of-origin subtypes of diffuse large B-cell lymphoma using gene expression in formalin-fixed paraffin-embedded tissue. Blood. 2014;123:1214-7.

12. Yoon N, Ahn S, Yong Yoo H, et al. Cell-of-origin of diffuse large B-cell lymphomas determined by the Lymph2Cx assay: better prognostic indicator than Hans algorithm. Oncotarget. 2017;8:22014-22.

13. Nowakowski GS, Chiappella A, Witzig TE, et al. Results of real-time cell-of-origin subtype identification by gene expression profiling in patients with ABC-type diffuse large B-cell lymphoma in the phase III trial of lenalidomide plus R-CHOP vs placebo plus R-CHOP (ROBUST). J Clin Oncol. 2018;36(suppl):abstr 7548.

14. Swerdlow SH, Campo E, Pileri SA, et al. The 2016 revision of the World Health Organization classification of lymphoid neoplasms. 2016;127:2375-90.

15. Petrich AM, Gandhi M, Jovanovic B, et al. Impact of induction regimen and stem cell transplantation on outcomes in double-hit lymphoma: a multicenter retrospective analysis. Blood. 2014;124:2354-61.

16. Oki Y, Noorani M, Lin P, et al. Double hit lymphoma: the MD Anderson Cancer Center clinical experience. Br J Haematol. 2014;166:891-901.

17. Aukema SM, Siebert R, Schuuring E, et al. Double-hit B-cell lymphomas. Blood. 2011;117:2319-31.

18. Scott DW, King RL, Staiger AM, et al. High-grade B-cell lymphoma with MYC and BCL2 and/or BCL6 rearrangements with diffuse large B-cell lymphoma morphology. Blood. 2018;131:2060-4.

19. Hernandez-Ilizaliturri FJ, Deeb G, Zinzani PL, et al. Higher response to lenalidomide in relapsed/refractory diffuse large B-cell lymphoma in nongerminal center B-cell-like than in germinal center B-cell-like phenotype. Cancer. 2011;117:5058-66.

20. Vitolo U, Chiappella A, Franceschetti S, et al. Lenalidomide plus R-CHOP21 in elderly patients with untreated diffuse large B-cell lymphoma: results of the REAL07 open-label, multicentre, phase 2 trial. Lancet Oncol. 2014;15:730-7.

21. Thieblemont C, Tilly H, Gomes da Silva M, et al. Lenalidomide maintenance compared with placebo in responding elderly patients with diffuse large B-cell lymphoma treated with first-line rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone. J Clin Oncol. 2017;35:2473-81.

22. 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.

23. Ruan J, Martin P, Furman RR, et al. Bortezomib plus CHOP-rituximab for previously untreated diffuse large B-cell lymphoma and mantle cell lymphoma. J Clin Oncol. 2011;29:690-7.

24. Leonard JP, Kolibaba KS, Reeves JA, et al. Randomized phase II study of R-CHOP with or without bortezomib in previously untreated patients with non-germinal center B-cell-like diffuse large B-cell lymphoma. J Clin Oncol. 2017;35:3538-46.

25. Offner F, Samoilova O, Osmanov E, et al. Frontline rituximab, cyclophosphamide, doxorubicin, and prednisone with bortezomib (VR-CAP) or vincristine (R-CHOP) for non-GCB DLBCL. Blood. 2015;126:1893-901.

26. Davies AJ, Caddy J, Maishman T, et al. A prospective randomised trial of targeted therapy for diffuse large B-cell lymphoma (DLBCL) based upon real-time gene expression profiling: the Remodl-B study of the UK NCRI and SAKK lymphoma groups (ISRCTN51837425). Blood. 2015;126:abstr 812.

27. Wilson WH, Young RM, Schmitz R, et al. Targeting B cell receptor signaling with ibrutinib in diffuse large B cell lymphoma. Nat Med. 2015;21:922-6.

28. Younes A, Thieblemont C, Morschhauser F, et al. Combination of ibrutinib with rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) for treatment-naive patients with CD20-positive B-cell non-Hodgkin lymphoma: a non-randomised, phase 1b study. Lancet Oncol. 2014;15:1019-26.

29. Schmitz R, Wright GW, Huang DW, et al. Genetics and pathogenesis of diffuse large B-cell lymphoma. N Engl J Med. 2018;378:1396-407.

30. Chapuy B, Stewart C, Dunford AJ, et al. Molecular subtypes of diffuse large B cell lymphoma are associated with distinct pathogenic mechanisms and outcomes. Nat Med. 2018;24:679-90.