Novel case of cytoplasmic CD3 positive, surface CD4 positive high-grade B-cell lymphoma with MYC rearrangement

Abstract

High-grade B-cell lymphoma, not otherwise specified (HGBCL-NOS), is an aggressive mature B-cell neoplasm whose diagnosis remains challenging due to significant morphologic overlap with diffuse large B-cell lymphoma and Burkitt lymphoma, and the absence of defining molecular alterations. Aberrant antigen expression can further complicate diagnoses and may result in lineage ambiguity. We report a unique case of HGBCL-NOS with MYC rearrangement demonstrating aberrant cytoplasmic CD3 and surface CD4 expression, leading to an initial misdiagnosis as a T-cell lymphoma at an outside hospital. A 50-year-old man with a previous diagnosis of peripheral T-cell lymphoma presented to our institution with central nervous system symptoms and a persistent bladder mass. Evaluation by bladder biopsy and cerebrospinal fluid flow cytometry revealed a high-grade B-cell neoplasm with diffuse CD10 and CD79a expression, patchy CD20 positivity, strong PAX5 expression, lambda light-chain restriction, Ki-67 approaching 100%, and flow cytometry confirmed expression of cytoplasmic CD3 and dim surface CD4 expression in a clonal B-cell population. Fluorescence in situ hybridization demonstrated a t(8;14)(q24;q32) MYC::IGH rearrangement without BCL2 or BCL6 rearrangements. This extremely rare phenotype has not previously been described. This case expands the spectrum of aberrant immunophenotypes reported in both B-cell lymphomas and HGBCL-NOS and highlights the importance of comprehensive lineage assessment in lymphomas with unusual antigen expression.

Introduction

High-grade B-cell lymphomas (HGBCL) are a relatively new diagnostic category of aggressive, mature B-cell neoplasms with rapid evolution in their diagnostic criteria. B-cell lymphomas (BCL) that have morphologic, immunophenotypic, or genetic features intermediate between Burkitt lymphoma (BL) and diffuse large B-cell lymphoma (DLBCL) have long been recognized. The 2008 World Health Organization (WHO) criteria defined these entities as “B-cell lymphoma, unclassifiable, with features intermediate between DLBCL and BL” [1]. The revised 2016 WHO classification stratified these tumors by molecular characteristics into HGBCL with MYC and BCL2 and/or BCL6 rearrangements, so-called double-hit (DHL) or triple-hit (THL) lymphomas, and HGBCL-NOS [2]; the 2022 WHO criteria further added a new entity of HGBCL with 11q aberration (HGBCL-11q) and integrated all DLBCL and HGBCL with MYC and BCL2 rearrangements into one category as DLBCL/HGBCL-MYC/BCL2 [3]. This progressive genetic stratification has left HGBCL-NOS as a diagnosis of exclusion whose criteria depend on morphologic characteristics that do not fit into other well-defined categories with significant interobserver variability [4, 5]. Despite this variability, HGBCL-NOS remains a useful, though imperfect, contemporary diagnostic category for intermediate aggressive BCLs without definitional molecular signatures and it is not wholly without its own immunophenotypic and clinical characteristics [5].

In addition to morphologic ambiguity, atypical immunophenotypic findings may further complicate the diagnosis of HGBCL-NOS. As a rare entity, atypical antigen expression is not well described in HGBCL-NOS but is a known occurrence in BL and is well described in DLBCL [6, 7]. Atypical antigen expression can be limited to a loss or downregulation of B-cell markers, or in rare cases can result in aberrant antigen expression from non-B-cell lineages, potentially causing significant diagnostic uncertainty and an increased risk of misdiagnosis [8–10]. Here we present a HGBCL-NOS with MYC rearrangement (HGBCL-MYC-R) with cytoplasmic CD3 (cCD3) expression and co-expression of surface CD4 that resulted in an initial misdiagnosis. To our knowledge, this immunophenotypic description has not been previously described in literature.

Clinical history

A 50-year-old man presented to a hospital in South Africa with abdominal pain, hematuria, and an LDH of 904 U/L. A cystoscopy and bladder biopsy were performed with findings of a lymphoid infiltrate that was diffusely and strongly positive by immunohistochemistry (IHC) for CD3 and CD4, weakly positive for BCL6, CD20, and PAX5 in a small subset of cells, and negative for ALK and Epstein-Barr virus by EBER-ISH. The patient was diagnosed with a mature T-cell lymphoma (TCL) although subsequent polymerase chain reaction for a T-cell receptor gene rearrangement performed at an outside hospital was negative. PET CT demonstrated metabolically active small and medium size lymph nodes above and below the diaphragm and several extranodal sites including right pleura, perirenal fascia, mesentery, gastric antrum, and bladder wall (maximum SUV 54.73). The patient received 6 cycles of CHOP (cyclophosphamide, hydroxydaunorubicin, oncovin, and prednisone) chemotherapy in South Africa and presented to our institution’s emergency department 6 months later with right arm weakness and eyelid droop. Workup revealed a new diagnosis of AIDS (peripheral blood CD4-positive T-cell count: 59/mm³) and lymphomatous involvement of the central nervous system (CNS). Of note, HIV testing was not performed at the outside hospital. Imaging studies showed perirenal nodularity, mesenteric and retroperitoneal lymphadenopathy, and a persistent bladder mass.

Materials and methods

Tissue samples of the bladder and bone marrow were fixed in 10% neutral buffered formalin, processed, and embedded in paraffin. Bone samples were decalcified in 12% formic acid. Cerebrospinal fluid (CSF) cytology specimens were collected in Cytospin collection fluid and processed on a Cytospin 4 centrifuge (Thermo Fisher Scientific, Waltham, MA). IHC was performed on a Ventana Benchmark ULTRA IHC/in situ hybridization (IHC/ISH) system (Tucson, AZ). A Ventana Tris-buffered alkaline antigen retrieval protocol was performed, and Roche ultraView and OptiView antigen detection kits were then used with appropriate positive and negative controls (Ventana, Tucson, AZ). Flow cytometry was performed on a 10-color flow cytometer (BD FACSCanto, Waters Biosciences [formerly BD Biosciences], San Jose, CA) and analyzed on FCS Express 6 (De Novo Software, Pasadena, CA). The CD3 probe is an SK7 monoclonal antibody (BD CD3 PE-Cy7, Waters Biosciences, San Jose, CA, catalog # 341101). The CD4 probe is an SK3 monoclonal antibody (BD CD4 PE, Waters Biosciences, San Jose, CA, catalog # 340670). Formalin-fixed, paraffin-embedded tissues were sent for high-grade/Large B-cell Lymphoma Panel FISH (NeoGenomics, Fort Myers, FL).

Results

A repeat transurethral resection of bladder tumor (TURBT) was performed, and histology showed an unremarkable urothelium with an underlying diffuse lymphoid infiltration of the lamina propria composed of intermediate- to large-sized lymphoid cells with occasional small nucleoli, irregular nuclear contours, scant cytoplasm, abundant mitotic figures, apoptotic figures, and tingible body macrophages creating a “starry sky” pattern. On IHC, the neoplastic cells had mildly to moderately downregulated pan-B-cell markers, namely patchy positive CD20 and intermediate to strong expression of PAX5. The cells were strongly and diffusely positive for CD10 and CD79a, and weakly positive for BCL6. However, the tumor was also positive for CD3 and CD4, though the tumoral cells had scant cytoplasm, so localization could not be definitively determined by IHC alone. Additionally, tumor cells were positive for CD99, CD43, and MUM1, had variable diffuse c-MYC positivity ranging from weak to strong, and were negative for CD34, CD5, CD7, CD8, TdT, CD30, ALK D5F3, Bcl-2, Cyclin D1, CD1a, HHV-8, and EBER ISH. The Ki67 proliferative index was approaching 100% (Fig. 1).

Fig. 1.

Histology slides of TURBT sample at our institution. H&E sections (a, 4×, b 40×) show intermediate- to large-sized lymphoid cells with occasional small nucleoli, irregular nuclear contours, scant cytoplasm, abundant mitotic figures, apoptotic figures, and tingible body macrophages creating a “starry sky” pattern. Diff-Quik preparation of a CSF sample (c, 40×) shows similar intermediate- to large-sized cells with deeply basophilic cytoplasm and rare vacuolization. The tumor cells are strongly positive for CD10 (d, 20×), cytoplasmic and membranous CD3 (e, 40×), have patchy positivity for CD20 (f, 20×), and variable weak to strong positivity for c-MYC (g, 20×). The tumor cells are negative for BCL2 (h, 20×), weakly positive for BCL6 (i, 10×), and have a Ki67 proliferative index approaching 100% (j, 10×)

Concurrent flow cytometry performed on a CSF sample showed a clonal CD10 positive B-cell population positive for pan-B-cell markers including CD19, CD20, and cCD79a, with surface and cytoplasmic lambda light-chain restriction. Importantly, however, the neoplastic cells were positive for cytoplasmic CD3 expression, dimly positive for surface CD4 expression, and negative for surface CD3 expression (Fig. 2). FISH on the TURBT specimen showed a t(8;14) MYC::IgH translocation, no BCL2 or BCL6 rearrangements, and extra copies of BCL2 and BCL6 signals.

Fig. 2.

Flow cytometry immunophenotype of CSF sample. Analysis shows a population of CD45-dim, high side scatter cells (blue) distinct from the normal lymphocyte population (green), comprising 75.15% of events (a). These cells are positive for cytoplasmic CD79a and cytoplasmic CD3 (b). These tumor cells are also positive for surface CD4 and CD19, but are negative for surface CD3 and surface CD8 (c, d). APC-Cy7 = APC-cyanine 7, SSC = side scatter, PE = phycoerythrin, FITC = fluorescein isothiocyanate, A = area

The patient was initially urgently treated with methotrexate, cytarabine, thiotepa, and rituximab (MATRix) chemotherapy for CNS involvement. Staging bone marrow biopsy was negative for lymphoma by morphology, immunohistochemistry, and flow cytometry. However, despite an initial response, the patient developed a malignant pleural effusion, worsening CNS involvement, and an orbital involvement which did not completely respond to intrathecal methotrexate and orbital radiation therapy. The patient’s condition continued to deteriorate, and he was pronounced deceased 8 months after presentation to our hospital.

Discussion

Antigenic aberrancies are not rare in hematopathology and include unexpected gain and/or loss of marker expression. In rare cases, this can result in tumors expressing lineage-defining markers inconsistent with their cell line of origin. In this report, we show the first HGBCL-NOS with flow cytometry confirmed expression of cCD3 and coexpression of CD4 [3]. The difficulty in diagnosing this case was two-fold: the first challenge was elucidating tumor B-cell lineage, the second differentiating between high-grade B-cell neoplasms.

Some T-cell markers, such as CD5, are well known and may carry diagnostic and prognostic implications in BCLs such as DLBCL [7, 11]. Many aberrancies are, however, rare and may cause significant uncertainty in diagnosis. For T-cell antigens in BCLs, the most robust literature descriptions involve DLBCL. Though still rare, there are numerous descriptions of DLBCLs expressing CD2, 4, 5, 7, and 8 [7]. Critically, expression of CD3, a coreceptor of the T-cell receptor that is highly specific for T-cell lineage [12], remains exceptionally rare even in DLBCLs [8–10].

Furthermore, many literature descriptions of CD3 expression in BCLs were evaluated with IHC and not confirmed with flow cytometry. Subcellular localization by IHC alone can be incredibly difficult, especially in tumors with scant cytoplasm. These distinctions may be important as cCD3 is a more sensitive marker for T-cell lineage than surface CD3 and is a reliable lineage marker even if surface expression is absent [13]. Though there are no well-defined lineage assessment criteria in mature lymphomas, cCD3 is an accepted T-cell lineage-defining marker in the WHO for leukemias [3]. Thus, the expression of cCD3 is a particularly rare aberrancy in BCLs and requires careful evaluation of cell lineage.

Differentiation of lineage-ambiguous neoplasms often requires molecular tools. In our case, aggressive lymphoma FISH demonstrated a t(8;14)(q24;q32) MYC::IgH translocation, a specific molecular alteration that has only been described in BCLs. Overall, this neoplasm that expresses multiple B-cell markers, surface and cytoplasmic lambda light-chain restriction, and a characteristic molecular alteration is best characterized as a BCL with aberrant expression of cCD3 and surface CD4. In our case, this initial incomplete lineage workup and misdiagnosis were immensely important, as they resulted in the patient receiving CHOP, an inappropriate chemotherapy regimen for an aggressive BCL [14, 15].

Though commonly present in BL and DHL/THL, t(8;14)(q24;q32) MYC::IgH translocations can also be seen in most tumors throughout the BL-HGBCL-DLBCL spectrum [3]. BL has strict morphologic criteria of a medium-sized, monomorphic cell population with squared off borders, round nuclei, finely clumped and dispersed chromatin and diagnosis of DHL/THL requires additional molecular signatures that are absent in this case; this leaves differential diagnoses of HGBCL-NOS and DLBCL, both of which can have a MYC-R [3]. Morphologic distinction between these two entities is notoriously difficult. Previous retrospective reviews have resulted in the recategorization of over 50% of diagnosed HGBCL-NOS, mostly as DLBCL. However, no differentially mutated genes were identified between the remaining HGBCL-NOS and those recategorized as DLBCL, and few recurrently mutated genes were shared between those remaining in the HGBCL-NOS category. This highlights the significant difficulty of this diagnosis and calls into question whether these are morphologically distinguishable entities [16].

This distinction may be important because HGBCL-NOS is more clinically aggressive than DLBCL and behaves more similarly to DHL/THL [5]. Often, dose-adjusted doxorubicin, etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, and rituximab (DA-EPOCH-R), a chemotherapy regimen used in other aggressive BCLs like BL, is the regimen of choice, though there is no definitively recommended regimen for HGBCL-NOS [17]. Given the difficulty in differentiating HGBCL-NOS and DLBCL, the primary dilemma is the uncertainty in whether rituximab and CHOP (R-CHOP), a common regimen in DLBCL, is sufficient for HGBCL-NOS. However, data on tumors with MYC-R is more clear, with multiple studies demonstrating R-CHOP is often insufficient [17–19] for HGBCL-NOS-MYC-R and DLBCL-MYC-R, and this is reflected in both of these entities having a similarly aggressive clinical course [5].

The problem remains that approximately half of HGBCL-NOS and approximately 85–90% of DLBCLs do not have MYC-R. There are no pathognomonic features that differentiate HGBCL-NOS and DLBCL, though there are reported statistically significant clinicopathologic differences that, in aggregate, may be diagnostically helpful [5]. Specifically, HGBCL-NOS is generally more aggressive (e.g., higher rates of bone marrow involvement and extranodal disease). While our case did not have definitive bone marrow involvement, it did demonstrate extranodal disease with extensive, persistent CNS involvement and rapid progression to pleural and orbital involvement. By IHC, HGBCL-NOS is statistically more likely to be positive for CD10 and c-MYC, negative for BCL2, and display a germinal-center phenotype by the Hans algorithm, all features seen in our case. Additionally, though both HGBCL-NOS and DLBCL can demonstrate MYC-R, it is significantly more common in HGBCL-NOS. Median reported Ki67 indices vary significantly, though Ki67 expression approaching 100%, as in our case, is not characteristic of DLBCL [5, 20]. HGBCL-NOS and DLBCL exist on a disease continuum, and their differences are active areas of research. We acknowledge that these are difficult entities to distinguish with significant overlaps. However, based on aggressive behavior and immunophenotype, we contend that our case is best described as a HGBCL-NOS-MYC-R and is the first reported case of HGBCL-NOS-MYC-R that aberrantly expresses cCD3 and CD4.

The patient’s diagnostic and treatment history highlights the importance of vigilance surrounding atypical antigen expression in lymphomas and the careful consideration of how to interpret and further work up lymphoid neoplasms with aberrant immunophenotypes. The patient’s initial biopsy had morphologic and immunohistochemical features that were interpreted to be consistent with a peripheral T-cell lymphoma based on a limited immunohistochemical workup. The presence of unusual mixed lineage immunophenotype in extranodal sites should however always give pause and require additional thorough workup for definitive lineage determination. This case adds a new phenotype to the catalog of BCLs with aberrant T-cell antigen expression and highlights the growing knowledge and appreciation of antigenic aberrancies in B-cell neoplasms that must always be kept in mind, especially when reviewing a case with unexpected findings.

Author contribution

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Vladimir Miltchev MD and Natalia Golardi MD. The first draft of the manuscript was written by Vladimir Miltchev MD and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

This study was not supported by any funding.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Declarations

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. For this type of study, the Baylor College of Medicine Institutional Review Board does not require approval.

Informed consent

For this type of study, informed consent is not required.

Consent for publication

For this type of study, consent for publication is not required.

Competing interests

The authors declare no competing interests.