Retrospective analysis of the Hans algorithm and BCL2 expression in HIV positive and negative patients with diffuse large B-cell lymphoma

Abstract

Introduction: Diffuse large B-cell lymphoma (DLCBL) is the most commonly occurring non-Hodgkin lymphoma. Several DLBCL variants are described, with varying morphological, genetic and molecular characteristics, as well as clinical features. With the use of a wide range of diagnostic tools,these tumours have been usefully classified and prognosticated. The cell of origin (COO) theory and its immunohistochemical interpretation by the Hans algorithm is currently regarded as the most useful immunohistochemical tool for classification and prognostication. It recapitulates whole genome sequencing by approximately 80%, separating DLBCL into a germinal centre (GCB) and an activated B cell (ABC) phenotypes. Tumours of germinal centre phenotype have a significantly better prognosis than those of activated B cell phenotype. The B-cell lymphoma 2 (BCL2) family of genes encode for a group of proteins that regulate cellular apoptosis. Twenty-five proteins are included in this group; one of these is known as BCL2. The role of BCL2 upregulation as a predictor of disease prognosis in DLBCL is controversial. Studies yield inconclusive and sometimes conflicting results, with some authors demonstrating adverse outcomes in BCL2 expressing neoplasms of the ABC DLBCL subgroup.Other studies have shown BCL2 protein to be a marker of poor prognosis in germinal centre phenotype DLBCL but not in activated status phenotype DLBCL. In HIV positive individuals, DLBCL is recognised to be a more aggressive disease than in HIVnegative individuals. V Methods and Aims: The interrelationship of these various factors forms the basis of this study, which is to establish the immunohistochemical classification of DLBCL according to the Hans algorithm, whether thetumours express BCL2 immunophenotypically and whether these two prognostic factors have anyrelationship to the HIV status. Results: The majority of patients included in the study were HIV positive (63/72, 87,5%) and the minority were negative (9/72, 12,5%). Consensus microscopy of the cases determined the majority to be of the centroblastic morphological subtype (62/72, 86,1%) with others classified as anaplastic (10/72, 13.9%). No immunoblastic or “other” morphological subtypes were identified. Of the cases evaluated, the majority were of GCB by the Hans immunohistochemical panel (49/72, 68%), and the minority were ABC subtype (23/72 32%). Of the HIV positive cases, 37.9% (25/66) were BCL2 negative and 62.1% (41/66) were BCL2 positive. Of the HIV negative cases, 0.00 % (0/9) were BCL2 negative and 100% (9/9) were BCL2 positive. A statistically significant difference in BCL2 positivity rates in HIV positive and HIV negative patients was seen (100% versus 37.9%) (two-tailed Fisher’s Exact test, p-value = 0.02). Conclusion: Our findings demonstrate a higher prevalence of DLBCL in HIV infected patients compared to uninfected patients. We were not able to find a reliable conclusion as to the relationship between COO by Hans algorithm and HIV status. We showed a higher correlation of BCL2 positivity with ABC COO subtype, over GCB subtype. We were able to demonstrate a significant association between BCL2 positivity and HIV uninfected DLBCL patients, compared to a markedly lower association in HIV infected patients.