Immunogenotypic characteristics of diffuse large B-cell lymphoma in Malaysia

Abstract

Diffuse large B-cell lymphoma (DLBCL) is the most common aggressive type of non Hodgkin lymphoma with variable clinical outcomes. The immunogenotypic features of this heterogenous disease in Malaysia were not well characterized. In total 141 local series of DLBCL cases were retrospectively studied inclusive of cell-of-origin, CD5-positive DLBCL, co-expression of MYC and BCL2 proteins by immunohistochemistry; gene status of several genes of interest by FISH using split signal probes; determination of subtypes using Lymph2Cx RNA assay (NanoString®) as well as to compare with routine immunohistochemistry for validity purpose. Of these cases investigated using tissue microarray sections constructed by FFPE samples from UKM Medical Centre, we classified our patients into two subtypes: 32.7% (37/113) GCB and non-GCB 67.3% (76/113) by Hans algorithm and the results were compared with those classified by Choi algorithm, yielding strong agreement with κ = 0.828 (95% CI), P<0.01. We found that in between two subtypes, age (P=0.03), races (P=0.04), R-IPI (P=0.04) and B symptoms (P=0.03), the expression of CD10 (P<0.01), BCL6 (P<0.01) and MUM1 (P=0.02) showed statistically significance. Survival analysis indicated significance difference in between GCB and non-GCB subtypes (HR, 2.29; 95% CI, 1.40 to 3.75; P=0.01) with 4-year overall survival was 48.5% for GCB patients whereas 10.5% for non-GCB patients and the presence of B symptoms (HR, 1.62; 95% CI, 1.01 to 2.58; P=0.04). Ann Arbor staging and R-IPI did not yield significant results in our analyses. We observed 12% DLBCL cases were CD 5 positive and 81.8% of them died of the disease (P=0.076). Analysis on dual expression of MYC/BCL2 revealed that there is no significant difference in DE and non-DE groups (HR, 0.965; CI 95%, 0.49 to 1.89; P=0.916). FISH study reported there were 9.22% (13/141) rearranged cases observed in our population at which BCL6 indicated highest frequency of rearrangements (76.9%), followed by MYC (15.4%) and BCL2 (7.7%), no BCL10 and MALT-1 rearrangement found regardless of using TMAs or whole tissue samples. More cases of MYC protein overexpression observed compared to MYC translocation. There was no incidence of double hits detected in our patient group. Survival analysis showed that NanoString (P=0.017) was considered as better predictor compared to Hans algorithm (P=0.212). The degree of correlation is good with predictive values at 82% for GCB and 81% for non-GCB. In conclusion, relatively lower frequency of GCB tumours and low gene rearrangement rates found in Malaysian population. A national study is therefore warranted to know better the immunogenotypic characteristics of DLBCL in Malaysia and their implications on the survival.