BCL6 Protein Expression Identifies Cases without BCL6 Gene Rearrangement in Non-Germinal Centre Diffuse Large B- Cell Lymphoma Cases

Research Article

J Dis Markers. 2023; 8(2): 1056.

BCL6 Protein Expression Identifies Cases without BCL6 Gene Rearrangement in Non-Germinal Centre Diffuse Large B- Cell Lymphoma Cases

Pumza Magangane¹*; Jenifer Vaughan²; Tracey Wiggill²; Yvonne Perner¹

1Department of Anatomical Pathology, University of the Witwatersrand/NHLS, South Africa

2Molecular Medicine and Haematology, University of the Witwatersrand/NHLS, South Africa

*Corresponding author: Pumza Magangane Lecturer, Department of Anatomical Pathology, Room 3D06 Medical School, Faculty of Health Science, South Africa. Tel: +27 11 489 8882 Email: pumza.magangane@wits.ac.za

Received: June 19, 2023 Accepted: July 22, 2023 Published: July 29, 2023

Abstract

BCL6 gene rearrangements are the most frequent cytogenetic abnormality in Diffuse Large B-Cell Lymphoma (DLBCL), occurring in up to 64% of cases, suggesting a poorer prognosis and response to therapy. The aim of this study was to calculate the accuracy and predictive power of BCL6 IHC in identifying BCL6 gene rearrangements. A search for DLBCL cases was performed on the laboratory information system. We retrospectively analysed 46 cases of DLBCL, and correlated BCL6 protein expression with gene rearrangement status. Of the 46 cases, 39 (84.78%) showed positive BCL6 protein expression by IHC. In comparison, only 10 (21.74 %) samples presented with gene rearrangements. There were 8 cases positive for BCL6 IHC with gene translocations and 5 cases were negative for both protein expression and gene rearrangements. BCL6 IHC had a sensitivity of 80% and a specificity of 14%. Furthermore, the Positive Predictive Value (PPV) and Negative Predictive Value (NPV) were 21% and 71%, respectively. BCL6 rearrangement was more common in cases with a Non-Germinal Centre (GC) Cell of Origin (COO) (45.5%) as compared to the GC group (14.3%). When stratifying the analysis according to the COO, the sensitivity and NPV of BCL6 IHC in the non-GC group were both 100%, while the same parameters were considerably poorer in the GC cases. BCL6 protein expression did not correlate with the presence of BCL6 gene rearrangements in HIV related DLBCL cases. However, lack of BCL6 protein expression may be used to identify cases without a BCL6 gene rearrangement, particularly in non-GC COO.

Keywords: NPV; PPV; Sensitivity; Specificity; Immunohistochemistry; Fluorescence in-situ hybridization

Introduction

Rearrangements of the BCL6 gene are the most frequent cytogenetic abnormality in Diffuse Large B-Cell Lymphoma (DLBCL), occurring in up to 64% of HIV unrelated cases [1-4] and 20% of HIV related cases [5]. The BCL6 gene is found on chromosome 3q27, and encodes a 96 kDa protein with C-terminal zinc-finger motifs and an N-terminal protein-protein interacting domain [1,6]. The BCL6 chromosomal translocation partners are many; a phenomenon termed promiscuous translocation, and may involve a variety of Immunoglobulin (IG) genes on differing loci [7].

BCL6 protein expression preferentially occurs in germinal centre B cells [8] and is thus included in the Hans algorithm, classifying DLBCL into Germinal Centre (GC) and non-GC phenotypes [9]. Protein expression of BCL6 is detected in up to 79% of HIV unrelated DLBCL cases [4,10] and 56% of HIV related DLBCL [11], with a strong nuclear staining pattern [12].

Gene rearrangements are detected using Fluorescence In-Situ Hybridization (FISH), which is technically intensive, time-consuming and expensive [13]. In contrast, Immunohistochemistry (IHC) is relatively cheaper, widely available and easier to interpret [13].

Diagnostic tests are used to establish the absence or presence of a condition. In contrast, screening tests are used to determine the risk of acquiring the condition. Screening tests are usually more accessible, less invasive and less expensive compared to diagnostic tests [14]. IHC has been previously shown to be a good screening tool that can detect some gene rearrangements with high accuracy as in the case of MYC [15-17]. The accuracy of a screening tool relative to a known standard test is informed by calculating sensitivity, specificity, Positive Predictive Value (PPV) and Negative Predictive Value (NPV) [14].

Previous studies on DLBCL found no association between BCL6 rearrangements and BCL 6 protein expression [11,12,18]. However, the values for accuracy and predictive power of BCL6 IHC were not calculated. The aim of this study was to determine the accuracy and predictive power of BCL6 IHC in identifying BCL6 gene rearrangements in DLBCL cases from South Africa.

Materials and Methods

Ethics Committee Approval and Patient Consent

This study was approved by the University of the Witwatersrand Human Research Ethics Committee (R14/49), according to the principles of the Declaration of Helsinki.

Cases Selection

A total of 46 cases of Diffuse large B-Cell Lymphoma (DLBCL) were retrieved from the TrakCare laboratory database of the Department of Anatomical Pathology for 2017. Cases with both FISH and IHC data on BCL6 were included in the study. The demographic data, IHC staining profile and FISH results were extracted from pathology reports and analysed.

Immunohistochemical Analysis and Scoring

IHC was performed as part of the diagnostic work-up. Briefly, DLBCL cases were sectioned at 4μm and baked at 56°C for at least an hour. The slides were pre-treated with heat induced epitope retrieval with high pH retrieval solution, using the Automated Link 48 (Dako, Glostrup, Denmark) instrument. Slides were blocked with hydrogen peroxide for 10 min prior to the application of primary antibody. The monoclonal antibody BCL6 (PG-B6P, Dako, Glostrup, Denmark) was used as part of the Hans Criteria [9]. The Hans algorithm classifies DLBCL cases into GC and non-GC subtypes, using the immunohistochemical staining patterns of CD10, BCL6 and MUM1. Visual detection was achieved using the Dako Envision horse radish peroxidase for 20 min.

IHC results were scored according to the Hans classification by using a 30% cut-off value [9]. Slides with ≥30% tumour staining were considered positive, while those with <30% tumour staining or with weak or focal staining were scored negative.

FISH Analysis and Scoring

BCL6 FISH analyses were requested during the diagnostic workup, if required. FISH analysis was performed on unstained 2μm tissue sections using a Vysis LSI BCL6 dual colour break-apart probe (Abbott Molecular Inc., Desplaines, IL, USA). Tumour areas were defined by haematoxylin & eosin stained tissue sections from representative formalin-fixed, paraffin-embedded tissue blocks. Images were captured with Cytovision 4.0 (Leica Biosystems Inc., Buffalo Grove, IL, USA) on an Olympus BX61 fluorescence microscope (Olympic Scientific Solutions, Waltham, MA, USA) as previously described [19].

Data Analysis

Sensitivity, specificity Positive Predictive Value (PPV) and Negative Predictive Value (NPV) were calculated based on [20], as follows:

Sensitivity= true pos/(true pos+ false neg) x 100

Specificity= true neg/ (true neg+ false pos) x100

PPV= true pos/ (true pos + false pos) x100

NPV= true neg/ (true neg+ false neg) x100

“True positive” was defined as cases with gene rearrangements and positive IHC expression

“True negative” was defined as cases with no gene rearrangements and negative IHC expression.

“False positive” was defined as cases with no gene rearrangements and positive IHC expression

“False negative” was defined as cases with gene rearrangements and negative IHC expression

Results

Our cohort consisted of 46 DLBCL cases diagnosed in 2017 (Table 1). The average age of the cohort was 40±8.996 years with 30.43 % (n=14) female and 69.57 % (n=32) male population, respectively. Three cases (6.52%) were diagnosed as double hit lymphomas with both MYC and BCL6 translocations. MYC was translocated in 10 (22.73 %) of the DLBCL cases. There were 35 (76.09%) cases with GC subtype and 11 (23.91%) with non-GC subtype phenotypes. Only 42 cases (91.30%) were tested for HIV infection, which was seropositive in 36 (85.71%) of the cases. The median CD4 count was 183±247cells/uL with more than half of the cases with CD4 counts less than 200cell/μL. The median viral load was 706.5±273 880 copies/mL, with majority of the cases with less than 1000 copies/mL (Table 1).