Clinical boundaries in adult cases of large B cell lymphoma with IRF4 rearrangement

Large B‐cell lymphoma with IRF4 rearrangement (LBCL‐IRF4) is a recently described entity included in the revised 4th edition of the WHO Classification of Haematolymphoid Tumours (2017). Here we highlight the difficulties in classification of those cases which arise in adult patients with unusual clinical features.


Introduction
Large B-cell lymphoma with interferon regulatory factor 4 (IRF4) rearrangement (LBCL-IRF4) was included for the first time in the revised 4th edition of the WHO Classification of Haematolymphoid Tumours (2017); 1 it is a mature B-cell lymphoma with strong expression of IRF4, most commonly due to a chromosomal translocation of the IRF4 oncogene (6p25.3)adjacent to the IgH (immunoglobulin) locus. 2 Asymmetric tonsillar involvement or isolated cervical lymphadenopathy of Waldeyer's ring (WR) in children and young adults is the typical initial clinical presentation of this lesion, notably without Address for correspondence: J Maguire, Histopathology Department, St James's Hospital, James Street, Dublin 8 D08 NHY1, Ireland.e-mail: maguirekjessica@gmail.comThe purpose of our correspondence is to respond to an article by Hesius et al. published in Histopathology in February 2023 entitled 'Reclassification of diffuse large B cell lymphoma to large B cell lymphoma with IRF4 rearrangement in an adult population'.Our hope is to add to the collective knowledge on this entity only recently included in the WHO Classification of Tumours, by highlighting a possible additional population group susceptible to IRF4-rearranged large B-cell lymphoma that we have identified in our practice, namely immunosuppressed patients, who have not, to our knowledge, been described in this context in the literature.
2][3] The essential diagnostic criteria included in WHO-HAEM5 are 'intermediate or large cell morphology and follicular and/or diffuse growth pattern', 'mature B-cell phenotype with co-expression of BCL-6 and MUM1' and 'IRF4 translocation', with a desirable diagnostic criterion of 'absence of BCL2 and MYC gene rearrangement'. 2Berg et al. 4 suggest that the site and morphology differs in older adults.Streich et al. 5 suggest the locational predisposition to WR is less so in adults, and Hesius et al. 6 propose consideration of the diagnosis in older adults with presentation in WR or elsewhere in the correct context.
We wish to highlight the difficulties in classification of the three cases from our patient cohort, where IRF4 rearrangements were detected in cases of large B-cell lymphoma presenting in advanced stages of adult patients, each with a background history of immunosuppression.

Results and discussion
Patient A, a 47-year-old HIV-positive male, had a 2-year history of recurrent epigastric pain and a 2-month history of night sweats and weight loss.Imaging demonstrated stage IV disease with an F-18 fluorodeoxyglucose (FDG) avid gastric body mass invading into proximal jejunal loops of bowel and liver with loco-regional metastatic lymph nodes.At the time of diagnosis CD4 count was 716 and HIV viral load was undetectable.A biopsy of the ulcerated gastric mass was performed.
Patient B, a 30-year-old HIV positive male, presented with a 3-week history of a painful right inguinal mass and fatigue.Imaging demonstrated stage IV disease with FDG-avid soft tissue masses involving the medial right upper thigh and pelvic side-wall, associated bony destruction and loco-regional lymph nodes in addition to a FDG-avid right adrenal nodule.At the time of diagnosis CD4 count was 578 and HIV viral load was undetectable.The inguinal mass was sampled by needle core biopsy.
Patient C, a 45-year-old male with a history of renal transplant 10 years previously, on immunosuppressive therapy with tacrolimus and mycophenolate mofetil, developed abdominal pain over a 2-week period.Radiology was suggestive of acute appendicitis, with no adenopathy or splenomegaly.The appendix was removed and sent for histology.Positron emission tomography-computed tomography (PET-CT) imaging was consistent with stage IV disease with FDG-avid lymphadenopathy above and below the diaphragm, involvement of the caecum and multiple hepatic deposits.
The histology of all three patients showed an infiltrative population of poorly differentiated mediumlarge pleomorphic lymphoid cells, with prominent nucleoli, cytoplasmic clearing and abundant mitoses and necrosis (Figure 1).The neoplastic lymphoid populations were positive for CD45, CD20, CD79a, co-expressing CD10, B-cell lymphoma 6 (BCL6) and multiple myeloma oncogene 1 (MUM-1).The Ki-67 proliferation index in each case was 90-95%.The appendix of patient C also demonstrated nodular areas and scattered Hodgkin-like cells, positive for CD20, MUM-1, showing occasional CD30 and paired box 5 (PAX5) staining without CD15 positivity.Patient C's tumour was weakly positive for BCL2.The tumours were negative for CD3, CD5, CD23, cyclin D1, CD30 and Epstein-Barr encoding region (EBER), with no molecular evidence of MYC, BCL2 or BCL6 rearrangements on fluorescence in-situ hybridisation (FISH) analysis.A dual-specificity phosphatase 22 (DUSP22)-IRF break-apart rearrangement was identified in each of the three cases.None showed evidence of an 11q aberration on FISH analysis.
The morphological and immunohistochemical profile of each case was felt to closely align pathologically with a diagnosis of LBCL-IRF4; however, the unusual clinical presentations, namely the widespread stage IV disease at the time of presentation of all three patients with extensive nodal and extranodal involvement, is not a common presentation for this entity.These atypical clinical features possibly support the findings described by Berg et al. 4 and proposed by Streich et al. 5 that adult cases of LBCL-IRF4 that concomitantly express CD10, BCL6 and MUM1 (which prompted analysis for an IRF4 rearrangement as in our three cases) have a distinct clinical and regional profile when compared to the paediatric/young adult counterparts.This is supported by the Frauenfeld et al. study, 3 which genetically characterised cases of DLBCL (diffuse large B-cell lymphoma) that were found to have aberrant expression of CD10, BCL6 and MUM-1, and concluded that the adult DLBCL cases studied with IRF-4 gene rearrangements were genetically more complex than the paediatric LBCL counterparts.However, the boundaries defining LBCL-IRF4 arising in adult patients presenting in unusual sites has not been clearly defined, and such alternate categorisation of all three cases reported herein as DLBCL, not otherwise specified (NOS) with IRF4 rearrangements warrants consideration.
Patients A and B have completed treatment with R-EPOCH with complete metabolic response on PET-CT imaging in both cases.Patient C has commenced R-CHOP chemotherapy and interim PET-CT reimaging after four cycles is consistent with a complete metabolic response.
While LBCL is a well-described adverse outcome of solid organ transplant, only a single case of HIVassociated LBCL of the salivary gland with an IRF4 translocation has been described previously by Zhou et al., 7 with a post-treatment outcome that was favourable.While all the patients in this case series had a background history of immunosuppression, it should be noted that no viral causative link with immunosuppression was identified: EBER was negative in each case and the first two patients had long standing HIV with good viral control, as evidenced by their CD4 counts.As such, the link with LBCL-IRF4 and immunosuppression is debatable in these cases.WHO-HAEM5 proposes including the specific context of immune deficiency in the diagnostic report of lymphoproliferative diseases and lymphomas, 2 potentially classifying our three cases as 'LBCL-IRF4-, EBV-negative, arising in the setting of HIV infection' or 'LBCL-IRF4-, EBV-negative, post-renal transplant', for example, should the respective immunodeficient states be deemed to have contributed to disease development.

Conclusion
In conclusion, we describe three adult patients with background histories of immunodeficiency, presenting pathologically with LBCL-IRF4; however, the atypical clinical features make distinction from DLBCL, NOS challenging.We propose that the clinical boundaries of LBCL-IRF4 arising in adult patients need further delineation to allow distinction from cases of DLBCL, NOS.