[1334] Classification of Diffuse Large B-Cell Lymphoma with a New Algorithm and Evaluation of GCET2 in the Classification System.

Zenggang Pan, Min Li, Omar Hameed, Wing C Chan, Zifen Gao. University of Alabama at Birmingham; Peking University Health Science Center, Beijing, China; University of Nebraska Med. Center, Omaha

Background: GCET1 has been incorporated in a new algorithm using 5 antibodies (GCET1, CD10, BCL6, MUM1 and FOXP1) to classify diffuse large B-cell lymphoma (DLBCL) into germinal center B-cell-like (GCB) and activated B-cell-like (ABC) subgroups (Clin Cancer Res. 2009: 5291-5303), which is significantly more accurate than the previous Hans' algorithm (Blood. 2004, 103: 275-282). However, a small proportion of DLBCLs remain difficult to be subclassified. We studied a series of DLBCLs with the new algorithm and evaluated the potential role of another GC B-cell marker, GCET2, in classification of DLBCL.
Design: Four tissue microarrays containing 92 DLBCL cases were stained using monoclonal antibodies against GCET1, GCET2, BCL6, CD10, MUM1 and FOXP1. These DLBCL cases were classified as GCB or ABC subtypes based on the new algorithm. The sensitivity and specificity of each GC B-cell markers were calculated.
Results: The 92 DLBCL cases were classified into 59 ABC-DLBCL (27 nodal and 32 extranodal) and 33 GCB-DLBCL (15 nodal and 18 extranodal). The ABC to GCB subtypes had a highest ratio in the extranodal, non-gastrointestinal locations (18:6).

Classification of DLBCL with the new algorithm

	ABC-DLBCL	GCB-DLBCL	Total
Nodal	27	15	42
Extranodal	32	18	50
GI tract	14	12	26
Non-GI tract	18	6	24
Total	59	33	92

Of the four GC B-cell markers, both GCET1 and CD10 were most accurate in classifying DLBCL into GCB subtype, and CD10 had a highest specificity (100%, 0/59) but lowest sensitivity (52%, 17/33). GCET2 and BCL6 had similar overall sensitivity and specificity, and GCET2 had a significantly higher specificity in classifying nodal DLBCLs than the extranodal cases (data not shown). Especially, in conjunction with other markers, GCET2 clearly subclassified a small proportion of DLBCLs that were initially difficult to classify by the new algorithm.

Sensitivity and Specificity of GC B-Cell Markers in Classification of DLBCL

	ABC-DLBCL		GCB-DLBCL
	Positive	Negative	Positive	Negative	Sensitivity	Specificity
GCET1	12	47	24	9	73% (24/33)	80% (47/59)
GCET2	21	29	22	8	73% (22/30)	58% (29/50)
BCL6	22	34	25	8	76% (25/33)	61% (34/56)
CD10	0	59	17	16	52% (17/33)	100% (59/59)

Conclusions: In our study, ABC-DLBCL is more common than GCB subtype in both nodal and extra-nodal locations compared with the Western series. GCET2 can be used for more accurate classification of DLBCL.
Category: Hematopathology

Wednesday, March 2, 2011 9:30 AM

Poster Session V # 178, Wednesday Morning