TdT-positive high grade B-cell lymphoma transformed from grade 3B follicular lymphoma in an HIV-positive patient

Follicular lymphoma (FL) is graded as low (grades 1 and 2) and high (grades 3A and 3B) based on the number of centroblasts.1Swerdlow S.H. Campo E. Seto M. et al.Follicular lymphoma.in: Swerdlow S.H. Campo E. Harris N.L. World Health Organization Classification of Tumours of Haematopoietic and Lymphoid Tissues. Revised 4th ed. IARC Press, Lyon2017: 266-290Google Scholar Low grade FLs are known to be clinically indolent, but have a natural tendency to recur or transform into more aggressive and diffuse neoplasms, usually diffuse large B-cell lymphoma (DLBCL).2Lee J.T. Innes Jr., D.J. Williams M.E. Sequential bcl-2 and c-myc oncogene rearrangements associated with the clinical transformation of non-Hodgkin's lymphoma.J Clin Invest. 1989; 84: 1454-1459Crossref PubMed Scopus (102) Google Scholar Grade 3B FL is thought a different entity from grade 1-3A FL because its molecular, cytogenetic and immunophenotypic features resemble, in part, cases of DLBCL.3Horn H. Schmelter C. Leich E. et al.Follicular lymphoma grade 3B is a distinct neoplasm according to cytogenetic and immunohistochemical profiles.Haematologica. 2011; 96: 1327-1334Crossref PubMed Scopus (113) Google Scholar While progression from lower to higher grade is common in FL, blastoid transformation is rare.4Wagner-Johnston N.D. Link B.K. Byrtek M. et al.Outcomes of transformed follicular lymphoma in the modern era: a report from the National LymphoCare Study (NCLS).Blood. 2015; 126: 851-857Crossref PubMed Scopus (132) Google Scholar Human immunodeficiency virus (HIV) infection is associated with a higher risk of lymphoma development. HIV-associated NHLs are a heterogenous group, but are predominantly aggressive B-cell lymphomas.5Beral B. Peterman T. Berkelman R. et al.AIDS-associated non-Hodgkin lymphoma.Lancet. 1991; 337: 805-809Abstract PubMed Scopus (572) Google Scholar DLBCL and Burkitt lymphoma (BL) are the most common subtypes, whereas FLs and peripheral T-cell lymphomas are rarely associated with HIV infection.6Raphael M. Borisch B. Jaffe E. Lymphomas associated with infection by the human immune deficiency virus (HIV).in: Jaffe E. Harris N. Stein H. World Health Organization Classification of Tumours. Pathology and Genetics of Tumours of Haematopoietic and Lymphoid Tissues. IARC Press, Lyon2001: 260-263Google Scholar In this report, we describe a patient with HIV infection who had grade 3B FL associated with BCL6 rearrangement. The lymphoma underwent transformation into a Terminal deoxynucleotidyl transferase (TdT)+ high grade B-cell lymphoma with acquired MYC rearrangement during disease evolution. The patient was a 44-year-old man with a diagnosis of HIV infection in the 1990s, intermittently on highly active antiretroviral therapy (HAART), and presenting with cervical lymphadenopathy. Laboratory workup revealed a CD4+ T-cell count of 87/μL (normal 500–1500/μL). Computed tomography (CT) scan showed generalised lymphadenopathy and marked diffuse gastric wall thickening. The patient initially underwent an excision of the left cervical lymph node. Histological sections showed complete effacement of the nodal architecture by back-to-back follicles that lacked the usual polarisation and tingible-body macrophages (Fig. 1A). Sheets of large centroblasts were identified in the neoplastic follicles (Fig. 1B,C). Immunohistochemical analysis showed neoplastic cells positive for CD19, CD20, CD79a, PAX5, BCL2, BCL6 (weak) and MUM1/IRF4 (Fig. 1D–I), and negative for CD3, CD5, CD10, CD34, CD43, CD138, EMA, TdT and HHV8 (Fig. 1). In situ hybridisation (ISH) for EBV-encoded RNA (EBER) was negative. CD21 highlighted the residual follicular dendritic meshworks within follicles. The antibody specific for Ki-67 showed a proliferation activity of 60–70% (Fig. 1E). Concurrent flow cytometric immunophenotyping showed an aberrant B-cell population with monotypic kappa expression (Fig. 1J). The morphology combined with immunophenotype was diagnostic of a grade 3B FL. Fluorescence in situ hybridisation (FISH) studies demonstrated BCL6 rearrangement with no evidence of IGH/BCL2, IGH/MYC, or DUSP22-IRF4 rearrangement. Additional studies showed that the patient's cerebrospinal fluid (CSF) and bone marrow were negative for lymphoma. After the diagnosis of high grade follicular lymphoma, the patient received dose-adjusted rituximab, etoposide, prednisone, vincristine, cyclophosphamide and doxorubicin (DA-R-EPOCH) for 6 cycles. At the same time, the patient remained on HAART. Fluoro-deoxyglucose (FDG)-positron emission tomography (PET) scan after therapy showed minimal residual FDG uptake of multiple lymph nodes and gastric wall lesion, suggestive of a good response. Four months after DA-R-EPOCH treatment, the patient presented with headaches and blurry vision. His CD4+ T-cell count was 126/μL. Cytospin preparations of CSF were cellular, composed of numerous intermediate to large-sized malignant cells with a high nuclear-to-cytoplasmic ratio, fine chromatin and prominent nucleoli (Fig. 2A). Flow cytometry of the CSF fluid showed B lymphoid cells with TdT expression. FISH analysis revealed rearrangements of BCL6/3q in 89% and MYC/8q in 93% of the cells; and a gain of BCL2/18q in 71% of nuclei. Peripheral blood and bone marrow aspirate smears (Fig. 2B) showed neoplastic cells with morphology similar to that described in the CSF. Bone marrow trephine biopsy demonstrated a hypercellular (80–90%) marrow within which 20–30% of the neoplastic cells were arranged in an interstitial pattern (Fig. 2C,D). Immunohistochemical analysis of marrow biopsy showed that the neoplastic cells were positive for CD79a, PAX5, MUM1/IRF4, BCL6 and TdT, and negative for CD34 and CD10 (Fig. 2E–I). Flow cytometric analysis of marrow aspirate showed a neoplastic population positive for CD19, partial CD20 and TdT (Fig. 2J). Conventional cytogenetic analysis showed a complex male karyotype: 54,XY,+del(X)(q22q27),+1,add(1)(p12),+der(2; 5)(p10; q10),+3,add(3)(p12),t(3; 22) (q27; q11.2),+6,add(6)(q13),+7,add(7)(q32),+10,add(10)(q25),+12,add(12)(q24.3),t(14; 17) (q32; q25)[7]/46,XY[13]. FISH analysis was positive for MYC and BCL6 rearrangements. A diagnosis of TdT-positive high grade B-cell lymphoma with MYC and BCL6 rearrangements was rendered on both the CSF and bone marrow specimens. Following this new diagnosis, the patient was placed on salvage chemotherapy with rituximab, dexamethasone, high-dose Ara-C-cytarabine and cisplatin (R-DHAP) and intrathecal methotrexate. He had significant clinical improvement after two cycles of R-DHAP and intrathecal methotrexate. We assessed the neoplasms for IGH and IGK rearrangements using polymerase chain reaction (PCR)-based methods to determine whether the FL in the initial lymph node specimen and the TdT-positive high grade B-cell lymphoma in the bone marrow were clonally related. In the initial left cervical lymph node specimen, we identified a clonal IGK rearrangement with no evidence of clonal IGH rearrangement. The bone marrow specimen showed both IGH and IGK clonal rearrangement. The clonal IGK rearrangement in the initial bone marrow specimen was similar to that of cervical lymph node specimen, indicating both neoplasms were likely derived from the same clone. Our case is an HIV-positive patient with a grade 3B t(14;18)-negative FL associated with BCL6 rearrangement who subsequently developed TdT+ blastoid transformation with an acquired MYC rearrangement. There are rare reported cases of patients with low grade FL who developed subsequent TdT+ B-cell acute lymphoblastic leukaemia with MYC and BCL2 translocations.7Agarwal A.M. Agarwal N. Glenn M.J. Blastic transformation of low-grade follicular lymphoma.J Clin Oncol. 2007; 25: 2326-2328Google Scholar However, our case appears to be unique as this transformed TdT+ high grade lymphoma was associated with BCL6 and MYC rearrangements. During the patient's clinical course, his lymphoma underwent clonal evolution to acquire blastoid features, MYC translocation and TdT expression. The mechanisms to explain this rare form of transformation in the patient we report are unknown, but it seems likely that HIV infection and the severe immunosuppression may have facilitated this unusual sequence of events. There is currently no existing consensus to determine the best terminology for the neoplasm in this patient. Given the morphology, immunophenotype, and cytogenetic changes, it seems reasonable to propose four possible designations: (1) high grade B-cell lymphoma (HGBCL) with MYC and BCL6 rearrangement; (2) B lymphoblastic leukaemia/lymphoma; (3) lymphoblastic transformation of FL; and (4) TdT+ DLBCL. HGBCL with MYC and BCL2 and/or BCL6 rearrangement was introduced in the revised World Health Organization (WHO) classification published in book form in 2017.8Swerdlow S.H. Campo E. Seto M. et al.High-grade B-cell lymphoma.in: Swerdlow S.H. Campo E. Harris N.L. World Health Organization Classification of Tumours of Haematopoietic and Lymphoid Tissues. 4th ed. IARC Press, Lyon2017: 335-342Google Scholar These neoplasms are usually highly aggressive clinically and can be further subclassified into double or triple hit lymphoma depending on the number of gene rearrangements. Double hit lymphomas with MYC and BCL2 rearrangement represent 67% of cases, triple hit lymphomas with MYC, BCL2 and BCL6 rearrangement comprise 21%, and lastly, double hit lymphomas with MYC and BCL6 rearrangement account for 14% of all HGBCLs.9Landsburg D.J. Petrich A.M. Abramson J.S. et al.Impact of oncogene rearrangement patterns on outcomes in patients with double-hit non-Hodgkin lymphoma.Cancer. 2016; 122: 559-564Crossref PubMed Scopus (51) Google Scholar Cases of HGBCL can morphologically resemble DLBCL or they may have a blastoid or Burkitt-like appearance. A subset of cases of HGBCL with MYC and BCL2 and/or BCL6 may have a history of FL, but according to the WHO classification, FLs with a double or triple hit genetics are not included in this category. TdT expression, however, is not considered a part of this category in the revised WHO classification. Based on TdT expression, one might consider designating this neoplasm as B lymphoblastic leukaemia/lymphoma. Although the patient we present had CSF and bone marrow involvement at diagnosis, this classification is not a good fit due to his history. The patient age, history of HIV infection, and history of FL would be highly unusual in a patient with B lymphoblastic leukaemia/lymphoma. While possible, MYC, BCL2, and BCL6 rearrangements are also exceedingly rare in B lymphoblastic leukaemia/lymphoma.10Uchida A. Isobe Y. Uemura Y. et al.De novo acute lymphoblastic leukemia-like disease of high grade B-cell lymphoma with MYC and BCL2 and/or BCL6 rearrangements: a case report and literature review.BMC Clin Pathol. 2017; 17: 21-29Crossref PubMed Scopus (10) Google Scholar B-cell lymphoblastic transformation of FL is currently the preferred designation in the WHO classification and such cases have been reported previously by Geyer et al.11Geyer J.T. Subramaniyam S. Jiang Y. et al.Lymphoblastic transformation of follicular lymphoma: a clinicopathologic and molecular analysis of 7 patients.Hum Pathol. 2015; 46: 260-271Crossref PubMed Scopus (49) Google Scholar These patients present with FL, and after a variable but often long interval with many relapses, they acquire TdT expression. Although we have no specific criticisms of the underlying science of this approach, in our experience the change in terminology from FL to B-cell lymphoblastic transformation of FL can be a challenge as follows. In clinical protocols designed for patients with FL who undergo histological transformation and may be eligible for chimeric antigen receptor (CAR) T-cell therapy, typically the designation of diffuse large B-cell lymphoma is required. The last option is the designation of TdT+ DLBCL. Although this terminology is somewhat better than B lymphoblastic leukaemia/lymphoma, this term does not seem to be a good fit as has been discussed by Ok and colleagues.12Ok C.Y. Medeiros L.J. Thakral B. et al.High-grade B-cell lymphomas with Tdt expression: a diagnostic and classification dilemma.Mod Pathol. 2019; 32: 48-58Crossref PubMed Scopus (33) Google Scholar As discussed in their manuscript, patchy or variable TdT expression seems to be an insufficient reason to change the designation of a neoplasm from DLBCL of germinal centre B-cell type, to lymphoblastic transformation of FL. In summary, in our opinion it is most reasonable to include this transformed lymphoma as HGBCL with MYC and BCL6 rearrangement, transformed from FL, despite the expression of TdT. TdT expression in this context likely suggests a distinctive pathway for high grade transformation of FL. Regardless of the designation employed for this neoplasm, we believe reporting this case can be helpful in calling this issue to the attention of others, with the hope that additional cases will be reported and perhaps lead to a consensus for the optimal terminology for cases such as that described in this report. We acknowledge that Dr Guilin Tang of the University of Texas MD Anderson Cancer Center kindly helped us perform and interpret the FISH testing for DUSP22-IRF4 rearrangement. No research funding was obtained for this study. The authors state that there are no conflicts of interest to disclose.