Gene expression profiling has yielded a new molecular taxonomy of DLBCL in which 3 subtypes are distinguished; germinal center B-cell (GCB) subtype, derived from a germinal center B-cell; activated B-cell (ABC) subtype, derived from a post-germinal center B-cell; and primary mediastinal B-cell (PMBL) subtype, likely derived from a “thymic” B-cell. To define the effect of pathobiology on treatment outcome, immunophenotype has been used as an “approximate surrogate” biomarker for molecular taxonomy. Recent studies suggest that rituximab (R) benefit is primarily in BCL-6− (

Winter et al
Blood
107
:
4207
,
2006
) and BCL-2+ (
Mounier et al
Blood
101
:
4279
,
2003
) DLBCL, both more frequent in post-germinal center subtypes. Conversely, R-CHOP appears equivalent to CHOP in BCL-6+ DLBCL, which comprise up to 75% of all cases. We analyzed the outcome of DA-EPOCH-R using the immunophenotypic biomarkers BCL-6, BCL-2, MIB-1, CD10, MUM-1, and GCB vs ABC subtypes by the method of (Hans et al Blood 103:275, 2003). Patients had untreated de novo DLBCL excluding PMBL, stage II-IV, HIV-, and adequate organ function. No patients received radiation. Characteristics of 72 patients include median (range) age 50 (19–85); stage III/IV 69%; and High Intermediate/High International Prognostic Index (3–5) (IPI) 40%. Of 71 evaluable patients, response is CR/CRu 94%. At the median potential follow-up of 43 months, PFS is 82% and OS is 79% (Fig 1). PFS by Low/Low Intermediate (0–2) and High Intermediate/High (3–5) IPI is 93% and 64%, respectively, at 43 months. PFS by BCL-6 (Fig 1), and BCL-2 and GCB vs. ABC (Fig 2) is shown. PFS of DLBCL with MIB-1 < and > 80% was 92% and 81%, respectively. There was no significant difference in survival outcome by CD10 or MUM-1 expression (data not shown). Distribution by immunophenotype is shown below for patients with available tissue.

Biomarker Distribution

BiomarkerNumberPercent
BCL-6 Neg 14 24 
BCL-6 Pos 44 76 
BCL-2 Neg 24 41 
BCL-2 Pos 35 59 
GCB 34 67 
ABC 17 33 
MIB-1 < 80% 13 33 
MIB-1 > 80% 39 67 
CD10 Neg 36 63 
CD10 Pos 21 37 
Mum-1 Neg 30 64 
Mum-1 Pos 17 36 
BiomarkerNumberPercent
BCL-6 Neg 14 24 
BCL-6 Pos 44 76 
BCL-2 Neg 24 41 
BCL-2 Pos 35 59 
GCB 34 67 
ABC 17 33 
MIB-1 < 80% 13 33 
MIB-1 > 80% 39 67 
CD10 Neg 36 63 
CD10 Pos 21 37 
Mum-1 Neg 30 64 
Mum-1 Pos 17 36 
View Large

In conclusion, DA-EPOCH-R appears equally effective among all biomarker subgroups. Of note, DA-EPOCH-R is highly effective in BCL-6+ DLBCL, in which R does not appear to be very useful, suggesting the DA-EPOCH regimen itself may be highly effective in this group. The CALGB has initiated a Phase III randomized study of R-CHOP vr. DA-EPOCH-R to determine if DA-EPOCH-R represents a treatment advance. Gene expression profiling will be performed to assess the effect of the new molecular taxonomy and tumor biology on outcome.

Figure
View largeDownload slide
Figure
View largeDownload slide

Figure

Topics:
biological markers, diffuse large b-cell lymphoma, epoch protocol, ki-67 antigen, neprilysin, gene expression profiling, immunophenotyping, r-chop, cyclophosphamide, doxorubicin, prednisone, and vincristine, follow-up

Disclosures: Not all agents in the DA-EPOCH-R regimen are approved for the upfront treatment of DLBCL.

Author notes

*

Corresponding author

2006, The American Society of Hematology
2006
Sign in via your Institution