To explore the prognostic impact and interdependence of the cell-of-origin (COO) classification, dual expression (DE) of MYC and BCL2 proteins, and MYC, BCL2, and BCL6 translocations in two prospectively randomized clinical trials of patients with diffuse large B-cell lymphoma (DLBCL).

Overall, 452 formalin-fixed paraffin-embedded samples from two prospective, randomized DLBCL trials (RICOVER-60, prospective, randomized study for patients > 60 years, all IPI groups; and R-MegaCHOEP, prospective, randomized study for patients ≤ 60 years with age-adjusted IPI 2,3) of the German High-Grade Non-Hodgkin Lymphoma Study Group were analyzed with the Lymph2Cx assay for COO classification, with immunohistochemistry for MYC and BCL2, and with fluorescent in situ hybridization for MYC, BCL2, and BCL6 rearrangements.

COO classification was successful in 414 of 452 samples. No significant differences with respect to COO (activated B-cell [ABC]–like DLBCL v germinal center B-cell [GCB]–like DLBCL) were observed in event-free survival, progression-free survival, and overall survival in patients treated with rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) in the RICOVER-60 trial. Also, no differences with respect to COO were observed in multivariable analyses adjusted for International Prognostic Index factors in event-free survival (hazard ratio [HR] of ABC-like disease v GCB-like disease, 1.0; 95% CI, 0.6 to 1.6; P = .93), progression-free survival (HR, 1.1; 95% CI, 0.6 to 1.8; P = .82), and overall survival (HR, 1.0; 95% CI, 0.6 to 1.8; P = .96). Similar results were observed in the R-MegaCHOEP trial. In patients treated with R-CHOP, DE status was associated with significantly inferior survival compared with nonDE within the GCB, but not within the ABC subgroup. DE status was associated with significantly inferior outcome compared with patients with ABC-like DLBCL without DE (5-year PFS rate, 39% [95% CI,19% to 59%] v 68% [95% CI, 52% to 85%]; P = .03) and compared with patients with GCB-like DLBCL without DE. When data from patients with nonDE were analyzed separately, the outcome of patients in the ABC subgroup was inferior to that of patients in the GCB subgroup (5-year PFS rate, 68% [95% CI, 52% to 85%] v 85% [95% CI, 74% to 96%]; P = .04).

COO profiling in two prospective randomized DLBCL trials failed to identify prognostic subgroups, whereas dual expression of MYC and BCL2 was predictive of poor survival. Evaluation of prognostic or predictive biomarkers in the management of DLBCL, such as the COO, within prospective clinical trials will be important in the future.

Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous disease category that contains germinal center B-cell (GCB) –like , activated B-cell (ABC) –like, and unclassified types.1,2 Stratification of DLBCL into GCB-like and ABC-like types describes diseases with different pathogenetic mechanisms. Retrospective analyses have shown an inferior prognosis in ABC-like DLBCL3-6; hence, this type may mandate novel treatment strategies.7-9 Recently, quantification of gene expression (GE) in formalin-fixed paraffin-embedded (FFPE) specimens has become possible,10 and excellent agreement with gene expression profiling (GEP) of fresh-frozen tissue has been demonstrated.11,12 The Lymphoma/Leukemia Molecular Profiling Project has constructed a parsimonious digital GE (NanoString) –based test (Lymph2Cx assay) for cell of origin (COO) assignment in FFPE tissue. This 20-gene assay recapitulates COO assignment by using the original model13 on matched frozen tissue, and it has yielded > 95% concordance of assignments between laboratories.12 An application of this test to a population-based series of patients with DLBCL treated with rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) confirmed the prognostic power of the test. The same study confirmed dual expression (DE) of MYC and BCL2 proteins as a significant risk factor in DLBCL.14

We have previously shown that biomarkers, such as immunohistochemistry (IHC)-based COO classification, do not necessarily retain prognostic power when tested within prospective clinical trials15,16 and that risk indicators may not be robust when tested in different patient populations, such as elderly (61 to 80 years) and younger (≤ 60 years) high-risk patients.17 The aims of this study were to apply the Lymph2Cx COO assay in two clinical trials and to correlate the results to DE status and clinical variables.

Study Cohorts

All patients had been enrolled in two prospective, randomized, multicenter clinical trials for patients with CD20+ aggressive B-cell lymphoma. In the RICOVER-60 trial, patients older than age 60 years were assigned to six or eight cycles of cyclophosphamide, doxorubicin, vincristine, and prednisone every 2 weeks (CHOP-14) with (ie, R-CHOP-14) or without eight applications of rituximab (375 mg/m2). In this trial, six R-CHOP-14 cycles significantly improved event-free survival (EFS), progression-free survival (PFS), and overall survival (OS) compared with CHOP-14, and there was no benefit of response-adapted chemotherapy beyond six cycles.18 In the R-MegaCHOEP trial, young high-risk patients (age 18 to 60 years) were randomly assigned to eight cycles of cyclophosphamide, doxorubicin, vincristine, etoposide and prednisone repeated every two weeks (CHOEP-14) or to sequential high-dose therapy with four cycles of MegaCHOEP-21 (high-dosed CHOEP repeated every three weeks) supported by repeated infusions of autologous stem cells. Both treatment arms included six infusions of rituximab (375 mg/m2). There was no survival advantage of R-MegaCHOEP compared with conventional R-CHOEP therapy, and R-MegaCHOEP was associated with more side effects.19 Diagnostic samples were reviewed by expert hematopathologists according to the 2008 WHO classification.2 Trials were conducted in accordance with the Helsinki declaration. The protocol had been approved by the ethics review committee of participating centers.

GEP

Digital GEP was performed on 200 ng of RNA extracted from FFPE tissue samples with the Qiagen AllPrep DNA/RNA FFPE Kit (Qiagen, Hilden, Germany) and using the NanoString nCounter technology for research use (NanoString Technologies, Seattle, WA). COO assignment through GE (Lymph2Cx assay) was performed as described by Scott et al,12 but the high-sensitivity setting was used instead of the standard-sensitivity setting for processing on the nCounter PrepStation. Thirty RNA samples from the cohort described by Scott et al12 were run on the new lot of the NanoString code set (Appendix Fig A1, online only) for calibration. The calculated calibration constant of 277 points was subtracted from the linear predictor score from the new code set lot.

IHC and Fluorescent In Situ Hybridization Analyses

Tissue microarray (TMA) assembly of tumors, antibodies used, IHC staining and scoring parameters, and fluorescent in situ hybridization (FISH) analysis for BCL2, BCL6, and MYC were performed as described.15-17 MYC and BCL2 stains were considered positive when more than 40% and 50% tumor cells, respectively, expressed the protein.1 Assignment of tumors to GCB and non-GCB subtypes according to Hans et al21 was done as described. Within the study cohort of the R-MegaCHOEP trial, other algorithms had been investigated.21-25

Statistical Analysis

EFS (events included progression, start of salvage treatment, additional [unplanned] treatment, relapse, or death as a result of any cause), PFS (disease progression, relapse, or death as a result of any cause; data from patients with complete response [CR] or unconfirmed complete response [CRu] and additional treatment were censored), and OS (defined by death as a result of any cause) were measured from the time of random assignment. EFS, PFS, and OS were estimated according to Kaplan-Meier curves, and log-rank tests were performed. Proportional hazard models for the Lymph2Cx result (ABC v GCB) were adjusted for International Prognostic Index (IPI) or age-adjusted IPI (aaIPI) factors, respectively.26 Hazard ratios (HRs) with 95% CIs and P values are presented. For correlation of Lymph2Cx results (ABC v GCB) with qualitative data (morphology, IHC, and FISH data) and for differences in patient characteristics, χ2 and, if necessary, Fisher’s exact tests were used. The two-sided significance level was P ≤ .05. Statistical analyses were performed with IBM SPSS Statistics 23 (IBM, Armonk, NY).

Patient Characteristics and COO Classification

Overall, 949 and 202 patients with a diagnosis of DLBCL were enrolled in the RICOVER-60 and the R-MegaCHOEP trials, respectively.18,19 Median observation times for EFS, PFS, and OS were 82, 82, and 82 months, respectively, in RICOVER-60 and were 41, 41, and 43 months, respectively, in R-MegaCHOEP. Clinical characteristics of the patient subgroups investigated in this study matched those of the trial cohorts with the exception of extranodal involvement in RICOVER-60 (51% and 57% in the analysis and trial cohorts, respectively; P = .007; Appendix Table A1, online only). No substantial differences in outcomes (EFS, PFS, OS) were observed between analyzed and nonanalyzed patients in either study. COO classification was obtained for 326 tumors in RICOVER-60 and for 88 tumors in R-MegaCHOEP. In RICOVER-60, 137 (42%) of 326 tumors were classified as ABC-like DLBCL, 142 (44%) were classified as GCB-like DLBCL, and 47 (14%) were unclassified. Proportions of ABC-like, GCB-like, and unclassified DLBCL types were distributed equally among patients treated with CHOP (n = 158; ABC, 42%; GCB, 42%; unclassified, 15%) and R-CHOP (n = 168; ABC, 42%; GCB, 45%; unclassified, 14%). In R-MegaCHOEP, 24 (27%) of 88 tumors were ABC, 53 (60%) were GCB, and 11 (13%) were unclassified. Distribution of COO subgroups was similar in the R-CHOEP (n = 45; ABC, 24%; GCB, 62%; unclassified, 13%) and R-MegaCHOEP (n = 43; ABC, 30%; GCB, 58%; unclassified, 12%) treatment arms. Significantly more GCB-like DLBCL (53 [69%] of 77 tumors) were identified in the R-MegaCHOEP study than in the RICOVER-60 study (142 [51%] of 279 tumors; P = .005). Significantly more GCB-like tumors were also observed in the R-MegaCHOEP study compared with RICOVER-60 patients with two or more factors of the aaIPI (according to the population of the R-MegaCHOEP trial; 33 [36%] of 91 patients; P < .001).

In the entire RICOVER-60 cohort, ABC-like DLBCL were associated with increased lactate dehydrogenase (P = .02), Eastern Cooperative Oncology Group performance status greater than 1 (P = .01), stages III to IV (P = .002), high IPI scores (P = .02), and B symptoms (P = .003; Table 1). A similar trend was seen for advanced disease stage, IPI scores, and lactate dehydrogenase in the R-CHOP treatment arms (Appendix Table A2, online only). Within R-MegaCHOEP, the ABC subtype correlated with an Eastern Cooperative Oncology Group performance status greater than 1 (P = .03), high aaIPI scores (P = .02), and B symptoms (not significant probably because of the small sample size; Table 2).

Table

Table 1. COO Subtype Correlations With Clinical, Morphologic, IHC, and FISH Data in the RICOVER-60 Trial

Table

Table 2. COO Subtype Correlations With Clinical, Morphologic, IHC, and FISH Data in the R-MegaCHOEP Trial

Correlation of COO With Cytomorphology, IHC, and Genetics
Cytomorphology.

In RICOVER-60, more immunoblastic lymphomas were assigned an ABC subtype (14 [14%] of 98) than a GCB subtype (two [2%] of 93; P = .002; Table 1). This correlation was not observed in R-MegaCHOEP (Table 2).

IHC

In RICOVER-60, the GCB subtype was associated with a greater proportion of tumors that expressed CD10 (64% v 10%; P < .001) and BCL6 (94% v 78%; P = .001), whereas BCL2 (78% v 50%) and IRF4/MUM1 (97% v 65%) were enriched in ABC-like DLBCL (P < .001; Table 1). Similar distinct differences were seen, although they were not uniformly significant because of the smaller sample size, in R-MegaCHOEP for CD10 (58% v 13%; P < .001), BCL6 (64% v 45%; P = .17), and LMO2 (87% v 32%; P < .001) protein expression enriched in GCB-like DLBCL and BCL2 (96% v 66%; P = .007), IRF4/MUM1 (91% v 37%; P < .001), and FOXP1 (55% v 6%; P < .001) expression in ABC-like DLBCL (Table 2). The number of DLBCL that expressed MYC was higher in ABC-like DLBCL, especially within the R-MegaCHOEP trial (50% v 26%; P = .06). In both study cohorts, a higher proportion of tumors with DE were in the ABC group (RICOVER-60: ABC, 33 [33%] of 99 tumors; GCB, 16 [17%] of 93 tumors; P = .01; R-MegaCHOEP: ABC, 10 [46%] of 22 tumors; GCB, six [17%] of 35 tumors; P = .02; Tables 1 and 2).

When unclassified types were excluded, concordances between the Hans classifier and the Lymph2Cx assay were 82% (172 of 210) and 73% (44 of 60) in RICOVER-60 and R-MegaCHOEP, respectively (Appendix Tables A3 and A4, online only). In R-MegaCHOEP, other classifiers also assigned GCB-like and non–GCB-like DLBCL with 59% to 84% agreement (Appendix Table A4).

FISH

In RICOVER-60, BCL2 translocations were exclusively observed in GCB-like DLBCL (P < .001). MYC breaks also clustered in GCB-like DLBCL (14% v 3% in ABC; P = .005). BCL6 breaks were equally distributed between ABC and GCB types (33% and 29%, respectively). All nine MYC/BCL2 double-hit (DH) lymphomas were GCB-like DLBCL (9% of GCB; P = .001). Three MYC/BCL6 DH lymphomas were equally distributed in the GCB, ABC, and unclassified groups (Table 1). In R-MegaCHOEP, BCL2 translocations were enriched in GCB-like DLBCL (11[26%] of 42) versus in ABC-like DLBCL (one [4%] of 23; P = .04). BCL6 translocations were more common in ABC-like DLBCL (48%) than in GCB-like DLBCL (17%; P = .007). All four MYC/BCL2 DH lymphomas and one triple-hit lymphoma (MYC/BCL2/BCL6) were classified as GCB-like DLBCL. One MYC/BCL6 DH lymphoma was classified as ABC-like, and one other as GCB-like DLBCL (Table 2).

Survival Analyses

In the entire RICOVER-60 cohort and in patients treated with CHOP-only, ABC-like DLBCL had an inferior clinical outcome compared with GCB-like DLBCL for EFS (RICOVER-60: HR, 1.4; 95% CI, 1.0 to 1.9; P = .04; CHOP-14 within RICOVER-60: HR, 1.6; 95% CI, 1.0 to 2.4; P = .05), PFS (RICOVER-60: HR, 1.5; 95% CI, 1.1 to 2.1; P = .02; CHOP-14 within RICOVER-60: HR, 1.6; 95% CI, 1.0 to 2.5; P = .05), and OS (RICOVER-60: HR, 1.5; 95% CI, 1.0 to 2.1; P = .04; CHOP-14 within RICOVER-60: HR, 1.7; 95% CI, 1.0 to 2.8; P = .05). This was not observed in patients treated with R-CHOP; there were no significant differences between ABC and GCB types in EFS (HR, 1.2; 95% CI, 0.8 to 2.0; P = .36), PFS (HR, 1.4; 95% CI, 0.8 to 2.3; P = .21), or OS (HR, 1.3; 95% CI, 0.8 to 2.2; P = .36; Fig 1A-1C), and results were confirmed in multivariable analyses adjusted for IPI factors with HRs close to 1 for the comparison of ABC with GCB (Table 3). We observed the same results for all patients in RICOVER-60 after adjustment for IPI factors with regard to EFS (HR, 1.1; 95% CI, 0.8 to 1.6; P = .48), PFS (HR, 1.2; 95% CI, 0.8 to 1.7; P = .34), and OS (HR, 1.2; 95% CI, 0.8 to 1.7; P = .38) for the comparison of ABC and GCB and after adjustment for IPI factors, bulky disease, sex, and age older than 70 years (data not shown). No survival differences were observed between ABC-like DLBCL and GCB-like DLBCL in the two arms of the R-MegaCHOEP trial (Figs 1D-1F), and these results were confirmed in multivariable analysis adjusted for the aaIPI score for EFS (HR, 1.2; 95% CI, 0.5 to 2.6; P = .70), PFS (HR, 1.0; 95% CI, 0.4 to 2.3; P = .93), and OS (HR, 1.1; 95% CI, 0.4 to 3.3; P = .82) for ABC versus GCB.

Table

Table 3. Multivariable Analyses of COO and Prognostic Factors

In R-CHOP–treated patients in RICOVER-60, MYC expression identified a subgroup with inferior survival among GCB-like DLBCL (P = .001, P = .001, and P = .002 for EFS, PFS, and OS, respectively), whereas there was only a trend toward inferior EFS in patients with ABC-like DLBCL (P = .08). BCL2 expression indicated a trend toward inferior outcome within GCB-like DLBCL, but not within ABC-like DLBCL (Appendix Fig A2, online only). Patients with DE status had significantly inferior survival in R-CHOP–treated GCB-like DLBCL compared with nonexpressers (P < .001 for each outcome of EFS, PFS, and OS) and with BCL2 expressers only (P < .001, P = .003, and P = .004 for EFS, PFS, and OS, respectively); these results did not occur in the ABC group (Fig 2). When patients with MYC/BCL2 DH tumors were removed from the analysis, a nonsignificant trend for inferior survival in DE compared with nonDE was noted among R-CHOP–treated patients with GCB-like DLBCL. Among all (GCB and ABC) patients treated with R-CHOP who had no MYC/BCL2 DH or triple-hit status, significantly inferior OS was observed for patients with DE (P = .03). Within the DE group, COO stratification failed to segregate prognostic subgroups. When patients with DE status were removed from the analysis, patients with ABC-like DLBCL showed a trend toward inferior outcome compared with patients with GCB-like DLBCL; 5-year rates were 65% (95% CI, 48% to 82%) versus 80% (95% CI, 68% to 92%; P = .06); 68% (95% CI, 52% to 85%) versus 85% (95% CI, 74% to 96%; P = .04); and 75% (95% CI, 60% to 90%) versus 88% (95% CI, 77% to 98%; P = .12) for EFS, PFS, and OS, respectively. In a comparative analysis, MYC/BCL2 dual expressers had significantly shorter survival compared with patients with ABC-like DLBCL without DE; 5-year rates were 34% (95% CI, 15% to 52%) versus 65% (95% CI, 48% to 82%; P = .02); 39% (95% CI, 19% to 59%) versus 68% (95% CI, 52% to 85%; P = .03); and 42% (95% CI, 22% to 62%) versus 75% (95% CI, 60% to 90%; P = .008) for EFS, PFS, and OS, respectively and also compared with patients with GCB-like DLBCL without DE (Figs 3A-3C). These results were confirmed in a multivariable model adjusted for IPI factors for the comparison of MYC/BCL2 DE with patients without MYC/BCL2 in the ABC subgroup. There was no significant difference in survival between ABC-like and GCB-like DLBCL in groups without MYC/BCL2 DE for EFS, PFS and OS (Table 3). In keeping with these findings, an IHC score of greater than 1 or a IHC-FISH score of greater than 1 remained of prognostic value also in multivariable analysis adjusted for COO results and IPI factors (Table 3).15 Because of the small numbers of patients, a similar subgroup analysis was not possible for the R-MegaCHOEP trial. Presence of MYC breaks predicted significantly inferior survival in R-CHOP–treated patients with GCB-like DLBCL for EFS, PFS and OS, but this effect could not be investigated within patients with ABC-like DLBCL because of the low sample size (data not shown).

The Lymphoma/Leukemia Molecular Profiling Project has developed a diagnostic FFPE COO assay (Lymph2Cx) that yielded highly concordant results between laboratories and with different reagent lots.12,14 Application of this test to a large series of patients with DLBCL treated with R-CHOP resulted in stratification of patients into groups with significantly different outcomes independent of the IPI.14 This study, which tested the assay in two large prospective, randomized clinical trials, failed to identify prognostic subgroups among R-CHOP–treated or R-MegaCHOEP–treated patients, which thus challenges the broad applicability of COO testing for the identification of high-risk patients. Retrospective GEP analyses of large patient cohorts treated with R-CHOP have assigned inferior survival to patients with ABC-like DLBCL.3,4,27,28 More recent, yet preliminary, results from prospective studies, however, also indicate a survival difference between patients with the GCB type and the ABC type of DLBCL.29,30 Notwithstanding the differences in the results reported here, this study confirms that the Lymph2Cx assay robustly identifies biologic tumor subgroups according to the presumed COO. Specifically, BCL2 and MYC breaks distinctly clustered in the GCB subtype, as had been reported before14,31,32, and there was a correlation of COO assignment with the results of immunohistochemical algorithms in both trials. In addition, COO testing disclosed prognostic subgroups in CHOP-only–treated patients in RICOVER-60. This is not surprising, because CHOP-only–treated patients have a significantly worse prognosis; hence, outcome differences in subgroups are more likely to be detected. The failure of this study to detect prognostic differences in R-CHOP–treated patients, however, is more difficult to explain. The results of the calibration study reported here between the two institutions (BC Cancer Agency and Würzburg/Stuttgart) makes technical differences unlikely to be reason (Appendix Fig A1). Although the cohorts were from two large, randomized clinical trials, the evaluable sample size was relatively small, especially in the R-MegaCHOEP cohort. Nevertheless, within the RICOVER-60 trial, there was, at a significance level of 5%, a two-sided 80% power to detect an outcome difference for ABC versus GCB at 3 years of 17% or a hazard ratio of 1.7, which is in an order of published IPI factors.

Therefore, the question of different selection criteria of patients enrolled in the respective studies arises. The cohort of Scott et al14 was drawn from a population registry of 1,194 R-CHOP–treated patients selected on the basis of the availability of a matched fresh-frozen biopsy, whereas the study populations of RICOVER-60 and R-MegaCHOEP were selected on the availability of FFPE tissue for analysis. It is not entirely clear whether patients in such trials fully represent the overall patient population. Data from the literature that have shown survival differences between GCB-like DLBCL and ABC-like DLBCL have been obtained mostly from retrospectively analyzed patient cohorts,3,4,27,28 but this study represents retrospective data accrual from prospective clinical trials. The RICOVER-60 study treated approximately 30% of patients with DLBCL in Germany, and there was a high adherence to the protocol (relative dose intensity > 96%), which possibly equalized differences in clinical and/or biologic features in subgroups. It is debatable whether the more-aggressive therapy (R-CHOP-14 v R-CHOP-21) used in the RICOVER-60 study might be a major reason for the results of this study, but this point cannot be addressed with the current data. At least two studies, however, have failed to show survival differences between R-CHOP-14 and R-CHOP-21.33,34 Although there is no doubt that the molecular characterization of patients with DLBCL into ABC and GCB subtypes has important biologic implications and may help to find better targeted therapies, our results challenge the conclusion that ABC-type DLBCL invariably carries a significantly worse prognosis than the GCB type, regardless of patient age, other important clinical characteristics like the IPI, and treatment details. The principal value of COO testing, however, is not hampered by the identification or nonidentification of prognostic impact. The updated WHO classification1 has reinforced the concept that ABC-type and GCB-type DLBCLs constitute different diseases with particular biologic, genetic, and clinical features, and different therapeutic options may be mandated by their identification.8,9,35 Biologic and predictive information rendered by COO, therefore, will be increasingly important.

This study confirmed the negative prognostic impact of MYC/BCL2 DE status in DLBCL.27,31,36 In keeping with data from Scott et al,14 patients with DE had inferior outcomes in GCB-like DLBCL, but this effect was not seen in ABC-like DLBCL. Our results are not in agreement with Hu et al,27 who reported that the outcome differences between ABC-like DLBCL and GCB-like DLBCL vanish upon exclusion of DE from analysis. In contrast, in our study, a prognostic difference between patients with ABC-like DLBCL and GCB-like DLBCL was detectable only after exclusion of DE lymphomas (Fig 3). This argues for a more complex interplay between MYC/BCL2 status and COO, as seen in the results by Scott et al,14 and argues that both assays provide complementary prognostic information.

In summary, in this first, to our knowledge, FFPE study on GE-based COO profiling in prospective, randomized clinical trials, molecular ABC and GCB subclassifications of DLBCL alone failed to identify prognostic subgroups, whereas MYC/BCL2 DE status was highly prognostic of poor survival. These data also exemplify that prognostication in DLBCL is a complex task. Given that tailored strategies have entered modern therapy management in DLBCL,8,9,35-43 predictive biomarkers such as the COO will become of pivotal importance in the future.

© 2017 by American Society of Clinical Oncology

Supported by the Robert-Bosch-Stiftung, Stuttgart, Germany.

Conception and design: Annette M. Staiger, Marita Ziepert, Andreas Rosenwald, German Ott

Financial support: Andreas Rosenwald, German Ott

Provision of study materials or patients: Thomas F.E. Barth, Heinz-Wolfram Bernd, Alfred C. Feller, Wolfram Klapper, Michael Hummel, Harald Stein, Martin-Leo Hansmann, Sylvia Hartmann, Peter Möller, Sergio Cogliatti, Lorenz Trümper, Michael Pfreundschuh, Andreas Rosenwald, German Ott

Collection and assembly of data: Annette M. Staiger, Marita Ziepert, Thomas F.E. Barth, Heinz-Wolfram Bernd, Alfred C. Feller, Wolfram Klapper, Monika Szczepanowski, Michael Hummel, Harald Stein, Dido Lenze, Martin-Leo Hansmann, Sylvia Hartmann, Peter Möller, Sergio Cogliatti, Lorenz Trümper, Norbert Schmitz, Michael Pfreundschuh, Andreas Rosenwald, German Ott

Data analysis and interpretation: Annette M. Staiger, Marita Ziepert, Heike Horn, David W. Scott, Georg Lenz, Lorenz Trümper, Markus Löffler, Norbert Schmitz, Andreas Rosenwald, German Ott

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

Clinical Impact of the Cell-of-Origin Classification and the MYC/BCL2 Dual Expresser Status in Diffuse Large B-Cell Lymphoma Treated Within Prospective Clinical Trials of the German High-Grade Non-Hodgkin's Lymphoma Study Group

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jco/site/ifc.

Annette M. Staiger

No relationship to disclose

Marita Ziepert

No relationship to disclose

Heike Horn

No relationship to disclose

David W. Scott

Honoraria: Roche

Consulting or Advisory Role: Celgene, Janssen

Patents, Royalties, Other Intellectual Property: Named inventor on a pending patent describing gene expression profiling in prognostication in classical Hodgkin lymphoma (Inst); As a member of the Lymphoma/Leukemia Molecular Profiling Project, I am potentially a named inventor on a pending patent on the use of gene expression profiling to assign cell-of-origin in diffuse large B-cell lymphoma (Inst); I am a named inventor on a pending patent on the use of gene expression profiling to determine the proliferation signature in mantle cell lymphoma (Inst).

Thomas F.E. Barth

No relationship to disclose

Heinz-Wolfram Bernd

No relationship to disclose

Alfred C. Feller

No relationship to disclose

Wolfram Klapper

Consulting or Advisory Role: Takeda (Inst), Roche (Inst), Janssen (Inst)

Research Funding: Amgen (Inst), Novartis (Inst), Bayer (Inst), Celgene (Inst), Takeda (Inst)

Monika Szczepanowski

Employment: Roche (I)

Michael Hummel

No relationship to disclose

Harald Stein

Employment: HistoGene Company HCC GmbH

Dido Lenze

Employment: Bayer Schering Pharma (I)

Stock or Other Ownership: Bayer Schering Pharma (I)

Travel, Accommodations, Expenses: Pfizer, AstraZeneca

Martin-Leo Hansmann

No relationship to disclose

Sylvia Hartmann

No relationship to disclose

Peter Möller

No relationship to disclose

Sergio Cogliatti

No relationship to disclose

Georg Lenz

Honoraria: Janssen-Cilag, Celgene, Gilead, Bayer, Roche

Consulting or Advisory Role: NanoString, Roche, Janssen, Celgene, Gilead, Bayer

Research Funding: Janssen-Cilag, Bayer, Celgene

Travel, Accommodations, Expenses: Janssen-Cilag, Celgene

Lorenz Trümper

Consulting or Advisory Role: Hexal (Inst), Celgene (Inst), Seattle Genetics (Inst)

Research Funding: Roche (Inst), Amgen (Inst)

Markus Löffler

No relationship to disclose

Norbert Schmitz

Stock or Other Ownership: Celgene

Honoraria: Roche, Celgene, CTI Life Sciences, Riemser, Janssen Oncology

Consulting or Advisory Role: Roche, Janssen, CTI Life Sciences, Riemser, Janssen Oncology

Travel, Accommodations, Expenses: Roche, Janssen, Celgene, CTI Life Sciences, Riemser, Janssen Oncology

Michael Pfreundschuh

Honoraria: Roche

Consulting or Advisory Role: Roche, Celgene, Amgen, Spectrum

Research Funding: Roche, Amgen, Spectrum

Andreas Rosenwald

Consulting or Advisory Role: Roche, Celgene, Morphosys

Patents, Royalties, Other Intellectual Property: NanoString Technologies, Celgene

Travel, Accommodations, Expenses: Roche, Celgene

German Ott

Patents, Royalties, Other Intellectual Property: NanoString Technologies

Table

Table A1. Demographic and Clinical Characteristics of DLBCL Study Cohorts

Table

Table A2. Clinical Characteristics of COO Subtype Within DLBCL Patients Treated With R-CHOP in the RICOVER-60 Trial

Table

Table A3. Correlation of COO Subtypes and IHC/FISH Scores Within Patients of the RICOVER-60 Trial

Table

Table A4. Comparison of Immunohistochemical and Lymph2Cx-Based COO Subtyping in the R-MegaCHOEP Trial

ACKNOWLEDGMENT

We thank Theodora Nedeva (Würzburg), Petra Hitschke, Daniela Pumm, and Katja Bräutigam (Stuttgart) for excellent technical assistance.

1. Swerdlow SH, Campo E, Pileri SA, et al: The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood 127:2375-2390, 2016 Crossref, MedlineGoogle Scholar
2. Swerdlow SH, Campo E, Harris NL, et al. WHO classification of Tumours of Haematopoetic and Lymphoid Tissues. Lyon, France, IARC Press, 2008 Google Scholar
3. Barrans SL, Crouch S, Care MA, et al: Whole genome expression profiling based on paraffin embedded tissue can be used to classify diffuse large B-cell lymphoma and predict clinical outcome. Br J Haematol 159:441-453, 2012 Crossref, MedlineGoogle Scholar
4. Gutiérrez-García 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 117:4836-4843, 2011 Crossref, MedlineGoogle Scholar
5. Lenz G, Wright GW, Emre NC, et al: Molecular subtypes of diffuse large B-cell lymphoma arise by distinct genetic pathways. Proc Natl Acad Sci USA 105:13520-13525, 2008 Crossref, MedlineGoogle Scholar
6. 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 346:1937-1947, 2002 Crossref, MedlineGoogle Scholar
7. Dunleavy K, Roschewski M, Wilson WH: Precision treatment of distinct molecular subtypes of diffuse large B-cell lymphoma: Ascribing treatment based on the molecular phenotype. Clin Cancer Res 20:5182-5193, 2014 Crossref, MedlineGoogle Scholar
8. 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 117:5058-5066, 2011 Crossref, MedlineGoogle Scholar
9. Roschewski M, Staudt LM, Wilson WH: Diffuse large B-cell lymphoma-treatment approaches in the molecular era. Nat Rev Clin Oncol 11:12-23, 2014 Crossref, MedlineGoogle Scholar
10. Geiss GK, Bumgarner RE, Birditt B, et al: Direct multiplexed measurement of gene expression with color-coded probe pairs. Nat Biotechnol 26:317-325, 2008 Crossref, MedlineGoogle Scholar
11. Masqué-Soler N, Szczepanowski M, Kohler CW, et al: Molecular classification of mature aggressive B-cell lymphoma using digital multiplexed gene expression on formalin-fixed paraffin-embedded biopsy specimens. Blood 122:1985-1986, 2013 Crossref, MedlineGoogle Scholar
12. 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 123:1214-1217, 2014 Crossref, MedlineGoogle Scholar
13. Wright G, Tan B, Rosenwald A, et al: A gene expression-based method to diagnose clinically distinct subgroups of diffuse large B-cell lymphoma. Proc Natl Acad Sci USA 100:9991-9996, 2003 Crossref, MedlineGoogle Scholar
14. Scott DW, Mottok A, Ennishi D, et al: Prognostic significance of diffuse large b-cell lymphoma cell of origin determined by digital gene expression in formalin-fixed paraffin-embedded tissue biopsies. J Clin Oncol 33:2848-2856, 2015 LinkGoogle Scholar
15. Horn H, Ziepert M, Becher C, et al: MYC status in concert with BCL2 and BCL6 expression predicts outcome in diffuse large B-cell lymphoma. Blood 121:2253-2263, 2013 Crossref, MedlineGoogle Scholar
16. Ott G, Ziepert M, Klapper W, et al: Immunoblastic morphology but not the immunohistochemical GCB/nonGCB classifier predicts outcome in diffuse large B-cell lymphoma in the RICOVER-60 trial of the DSHNHL. Blood 116:4916-4925, 2010 Crossref, MedlineGoogle Scholar
17. Horn H, Ziepert M, Wartenberg M, et al: Different biological risk factors in young poor-prognosis and elderly patients with diffuse large B-cell lymphoma. Leukemia 29:1564-1570, 2015 Crossref, MedlineGoogle Scholar
18. Pfreundschuh M, Schubert J, Ziepert M, et al: Six versus eight cycles of bi-weekly CHOP-14 with or without rituximab in elderly patients with aggressive CD20+ B-cell lymphomas: A randomised controlled trial (RICOVER-60). Lancet Oncol 9:105-116, 2008 Crossref, MedlineGoogle Scholar
19. Schmitz N, Nickelsen M, Ziepert M, et al: Conventional chemotherapy (CHOEP-14) with rituximab or high-dose chemotherapy (MegaCHOEP) with rituximab for young, high-risk patients with aggressive B-cell lymphoma: An open-label, randomised, phase 3 trial (DSHNHL 2002-1). Lancet Oncol 13:1250-1259, 2012 Crossref, MedlineGoogle Scholar
20. Scott DW, Chan FC, Hong F, et al: Gene expression-based model using formalin-fixed paraffin-embedded biopsies predicts overall survival in advanced-stage classical Hodgkin lymphoma. J Clin Oncol 31:692-700, 2013 LinkGoogle Scholar
21. 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 103:275-282, 2004 Crossref, MedlineGoogle Scholar
22. 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 15:5494-5502, 2009 Crossref, MedlineGoogle Scholar
23. 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 29:200-207, 2011 LinkGoogle Scholar
24. Muris JJ, Ylstra B, Cillessen SA, et al: Profiling of apoptosis genes allows for clinical stratification of primary nodal diffuse large B-cell lymphomas. Br J Haematol 136:38-47, 2007 Crossref, MedlineGoogle Scholar
25. 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 22:1094-1101, 2009 Crossref, MedlineGoogle Scholar
26. Ziepert M, Hasenclever D, Kuhnt E, et al: Standard international prognostic index remains a valid predictor of outcome for patients with aggressive CD20+ B-cell lymphoma in the rituximab era. J Clin Oncol 28:2373-2380, 2010 LinkGoogle Scholar
27. 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 121:4021-4031, quiz 4250, 2013 Crossref, MedlineGoogle Scholar
28. Lenz G, Wright G, Dave SS, et al: Stromal gene signatures in large-B-cell lymphomas. N Engl J Med 359:2313-2323, 2008 Crossref, MedlineGoogle Scholar
29. Gleeson M, Jack A, Cunningham D, et al: The activated B-cell subtype of diffuse large B-cell lymphoma as determined by whole genome expression profiling on paraffin-embedded tissue is independently associated with reduced overall and progression free survival in the rituximab era: Results from the UK NCRI R-CHOP 14 v 21 phase III trial. Blood 128;2016 (abstr 1746) Google Scholar
30. Vitolo U, Trneny M, Belada D, et al: Obinutuzumab or rituximab plus CHOP in patients with previously untreated diffuse large B-cell lymphoma: Final results from an open-label, randomized phase 3 study (GOYA). Blood 128; 2016 (abstr 470) CrossrefGoogle Scholar
31. Johnson NA, Slack GW, Savage KJ, et al: Concurrent expression of MYC and BCL2 in diffuse large B-cell lymphoma treated with rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone. J Clin Oncol 30:3452-3459, 2012 LinkGoogle Scholar
32. Visco C, Tzankov A, Xu-Monette ZY, et al: Patients with diffuse large B-cell lymphoma of germinal center origin with BCL2 translocations have poor outcome, irrespective of MYC status: A report from an International DLBCL rituximab-CHOP Consortium Program study. Haematologica 98:255-263, 2013 Crossref, MedlineGoogle Scholar
33. Cunningham D, Hawkes EA, Jack A, et al: Rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisolone in patients with newly diagnosed diffuse large B-cell non-Hodgkin lymphoma: A phase 3 comparison of dose intensification with 14-day versus 21-day cycles. Lancet 381:1817-1826, 2013 Crossref, MedlineGoogle Scholar
34. Delarue R, Tilly H, Mounier N, et al: Dose-dense rituximab-CHOP compared with standard rituximab-CHOP in elderly patients with diffuse large B-cell lymphoma (the LNH03-6B study): A randomised phase 3 trial. Lancet Oncol 14:525-533, 2013 Crossref, MedlineGoogle Scholar
35. Wilson WH, O’Connor OA, Czuczman MS, et al: Navitoclax, a targeted high-affinity inhibitor of BCL-2, in lymphoid malignancies: A phase 1 dose-escalation study of safety, pharmacokinetics, pharmacodynamics, and antitumour activity. Lancet Oncol 11:1149-1159, 2010 Crossref, MedlineGoogle Scholar
36. Green TM, Young KH, Visco C, et al: Immunohistochemical double-hit score is a strong predictor of outcome in patients with diffuse large B-cell lymphoma treated with rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone. J Clin Oncol 30:3460-3467, 2012 LinkGoogle Scholar
37. Cerchietti LC, Hatzi K, Caldas-Lopes E, et al: BCL6 repression of EP300 in human diffuse large B-cell lymphoma cells provides a basis for rational combinatorial therapy. J Clin Invest 120:4569-4582, 2010 Crossref, MedlineGoogle Scholar
38. Delmore JE, Issa GC, Lemieux ME, et al: BET bromodomain inhibition as a therapeutic strategy to target c-Myc. Cell 146:904-917, 2011 Crossref, MedlineGoogle Scholar
39. Dunleavy K, Pittaluga S, Czuczman MS, et al: Differential efficacy of bortezomib plus chemotherapy within molecular subtypes of diffuse large B-cell lymphoma. Blood 113:6069-6076, 2009 Crossref, MedlineGoogle Scholar
40. Dunleavy K, Little RF, Pittaluga S, et al: The role of tumor histogenesis, FDG-PET, and short-course EPOCH with dose-dense rituximab (SC-EPOCH-RR) in HIV-associated diffuse large B-cell lymphoma. Blood 115:3017-3024, 2010 Crossref, MedlineGoogle Scholar
41. Pfeifer M, Grau M, Lenze D, et al: PTEN loss defines a PI3K/AKT pathway-dependent germinal center subtype of diffuse large B-cell lymphoma. Proc Natl Acad Sci USA 110:12420-12425, 2013 Crossref, MedlineGoogle Scholar
42. Wilson WH, Young RM, Schmitz R, et al: Targeting B cell receptor signaling with ibrutinib in diffuse large B-cell lymphoma. Nat Med 21:922-926, 2015 Crossref, MedlineGoogle Scholar
43. Younes A, Romaguera J, Fanale M, et al: Phase I study of a novel oral Janus kinase 2 inhibitor, SB1518, in patients with relapsed lymphoma: Evidence of clinical and biologic activity in multiple lymphoma subtypes. J Clin Oncol 30:4161-4167, 2012 LinkGoogle Scholar

COMPANION ARTICLES

No companion articles

ARTICLE CITATION

DOI: 10.1200/JCO.2016.70.3660 Journal of Clinical Oncology 35, no. 22 (August 01, 2017) 2515-2526.

Published online May 19, 2017.

PMID: 28525305

ASCO Career Center