[1358] Comparative Analysis of Immunohistochemical Algorithms for Subtyping Nodal DLBCL According to Cell-of-Origin: Comparison with the Germinal Center B-Cell Marker HGAL

LIvia M Bacchi, Gabriela Gualco, Yaso Natkunam, Carlos E Bacchi. Universidade de São Paulo, São Paulo, SP, Brazil; Consultoria em Patologia, Botucatu, SP, Brazil; Stanford University, Stanford, CA

Background: Diffuse large B-cell lymphoma (DLBCL) can be molecularly separated into germinal center B-cell (GCB) and activated B-cell (ABC) subtypes. immunohistological algorithms have been proposed in order to facilitate the separation of these subtypes at the protein level in routine specimens. We analyzed 424 cases of nodal DLBCL applying three previously described algorithms (Hans, Choi and TALLY) and compared their results with a new combination including the germinal center B cell-associated marker, HGAL.
Design: TMA sections containing 424 cases of nodal DLBCL were stained using antibodies against CD10, BCL6, MUM1, FOXP1, GCET1, LMO2 and HGAL. All cases were classified as GCB or non-GCB by applying Hans (CD10, BCL6, MUM1), Choi (GCET1, MUM1, CD10, BCL6, FOXP1) and TALLY (CD10, GCET1/MUM1, FOXP1, with LMO2 as discriminator when score is equal) algorithms. Monoclonal anti-HGAL was used as an additional GC marker combined with the 3 algorithms above.
Results: Comparative results of different algorithms in 424 cases of nodal DLBCL are summarized in the table below.


In 49 GCET1-negative cases, HGAL expression was seen in 30 (61%); in 350 CD10-negative cases, HGAL expression was present in 67 (19%); in 185 BCL6-negative cases, HGAL expression was observed in 23 (12%); in 360 LMO2-negative cases, HGAL expression was seen in 71 (20%). Overall, HGAL was positive in 121 cases (29%), including 11 cases that were negative for all other GCB markers.
Conclusions: In 424 nodal DLBCL, using HGAL in combination with other markers (published algorithms), the frequency of GCB cases increased significantly. When HGAL was used isolated, 29% were classified as GCB and 71% as non-GCB. Our results showed a higher difference between the two cell-of-origin groups than observed in other series (42% vs. 58% by Hans; 57% vs. 43% by Choi and 45% vs. 55% by TALLY). These differences may be due to a larger case cohort and the inclusion of only nodal DLBCL in our series. This difference diminished when HGAL is used in combination. The sensitivity of HGAL in detecting GC-derived B-cell lymphomas may be a useful asset in an algorithm aimed at separating the suptypes of DLBCL.
Category: Hematopathology

Tuesday, March 20, 2012 1:00 PM

Poster Session IV # 182, Tuesday Afternoon


Close Window