The frequency of aggressive subtypes of B‐cell non‐Hodgkin lymphoma (B‐NHL), such as high‐grade B‐cell lymphomas (HGBL) with MYC and BCL2 and/or BCL6 rearrangement (HGBL‐DH/TH) or Burkitt‐like lymphoma (BL) with 11q aberration, is not well known in the HIV setting. We aimed to characterise HIV‐associated aggressive B‐NHL according to the 2017 WHO criteria, and to identify genotypic and phenotypic features with prognostic impact. Seventy‐five HIV‐associated aggressive B‐NHL were studied by immunohistochemistry (CD10, BCL2, BCL6, MUM1, MYC, and CD30), EBV‐encoded RNAs (EBERs), and fluorescence in situ hybridisation (FISH) to evaluate the status of the MYC, BCL2, and BCL6 genes and chromosome 11q. The 2017 WHO classification criteria and the Hans algorithm, for the cell‐of‐origin classification of diffuse large B‐cell lymphomas (DLBCL), were applied. In DLBCL cases, the frequencies of MYC and BCL6 rearrangements (14.9 and 27.7%, respectively) were similar to those described in HIV‐negative patients, but BCL2 rearrangements were infrequent (4.3%). MYC expression was identified in 23.4% of DLBCL cases, and coexpression of MYC and BCL2 in 13.0%, which was associated with a worse prognosis. As for BL cases, the expression of MUM1 (30.4%) conferred a worse prognosis. Finally, the prevalence of HGBL‐DH/TH and BL‐like with 11q aberration are reported in the HIV setting. The phenotypic and genotypic characteristics of HIV‐associated aggressive B‐NHL are similar to those of the general population, except for the low frequency of BCL2 rearrangements in DLBCL. MYC and BCL2 coexpression in DLBCL, and MUM‐1 expression in BL, have a negative prognostic impact on HIV‐infected individuals.

No abstract

Aggressive B-cell Lymphomas are the second most frequent AIDS-defining cancers. Few studies have compared the molecular characteristics of aggressive B-cell lymphomas in patients with and without HIV-infection; and to our knowledge, there are no reports comparing the incidence of gene rearrangements between the two groups and their impact on outcome in series treated with RCHOP. We retrospectively studied two series of patients with (N=32) and without HIV-infection (N=43) with diffuse large B-cell lymphomas (DLBCL) NOS (75% and 70%, respectively), T-rich DLCBL (13% and 5%), transformed DLBCL (3% and 14%) and double-hit (DH) DLCBL (9% and 11%) [defined by translocations affecting MYC (TMYC) concomitantly with translocation affecting BCL2 (TBLC2) and/or BCL6 (TBCL6)]. Tissue microarrays were constructed and translocations were studied by fluorescent in situ hybridization. The germinal center (GC) phenotype was defined according to the Hans' algorithm based on the expression of CD10, MUM1 and BCL6. Clinical data was retrieved from records. Continuous and categorical variables are presented using descriptive statistics and compared using Fisher's exact test, χ 2-test, and Mann-Whitney U-test. Survival analyses were performed using the Kaplan-Meier method. P-valuesof less than 0.05 were considered statistically significant. The median follow-up of HIV-infected patients was 6.9 years and of HIV-uninfected was 5 years. HIV-uninfected patients were older than HIV-infected patients, median age (range) 59 years (25-80) and 45 years (37-68) respectively; (P=0.002). There were differences between HIV-infected and HIV-uninfected patients regarding; male gender (81% vs. 54%, P=0.015), ECOG performance status higher than 2 (56% vs. 26%; P=0.018) and elevated ß-2 microglobulin (82% vs. 47%, P=0.005). On the other hand, the percentage of patients with III and IV Ann Arbor stages, two or more extranodal involvement, elevated LDH, 3 to 5 International Prognostic Index (IPI) scores, B-symptoms and bulky disease were similar in both series. The study of molecular features showed that more HIV-infected cases (57%) tended to have a GC phenotype than HIV-uninfected (35%); P=0.093. Regarding gene rearrangements, there was a trend for more HIV-infected patients (30%) to present TBCL2 than HIV-uninfected patients (11%); P=0.056. Of note, only 2 patients of the HIV-infected series were transformed from follicular lymphoma. The frequency of TMYC, TBCL6 and DH was similar in HIV-infected (24%, 28% and 9%, respectively) and HIV-uninfected (15%, 28% and 11%, respectively). HIV-infected patients were treated with RCHOP (N=27), intensive immunochemotherapy for Burkitt Lymphoma (IICT-BL) (N=4), CHOP (N=1) and HIV-uninfected patients with RCHOP (N=39), IICT-BL (N=2), RCOP (N=1), RESHAP (N=1). The complete response rate was not statistically different in HIV-infected and HIV-uninfected patients neither the relapse rate when considered all treatments given. Only patients treated with RCHOP were considered in survival analyses. The overall survival (OS) of HIV-uninfected patients with high IPI scores (3 to 5) (N=15) tended to be shorter than in patients with low IPI scores (0 to 2) (N=24), (5-y OS 60% [35%-85%] vs. 81% [64%-98%], P=0.089) and the progression free survival (PFS) was clearly inferior (5-y PFS 33% [9%-57%] vs. 73% [54%-92%], P=0.004). HIV-infected patients with high IPI scores presented similar OS and PFS than patients with low IPI scores. Contrary to other reports, TMYC had no statistically significant adverse impact on OS and PFS of HIV-uninfected patients, most probably because of small size of our series. The same reasoning could be applied for HIV-infected patients since both OS and PFS of patients with TMYC (N=5) was shorter than in those without TMYC (N=20) (5-year OS 40% [0%-83%] vs. 65% [44%-86%], P=0.365 and 5-year PFS 40% [0%-83%] vs. 60% [38%-82%], P=0.546). TBCL2, TBCL6 and GC phenotype had no impact on OS and on PFS of both HIV-infected and HIV-uninfected patients. Survival analyses of DH were not performed due to the small number of events. In summary, the frequency of TMYC and TBCL6 in aggressive B-cell lymphomas was similar in HIV-infected and HIV-uninfected patients but TBCL2 was less frequent in HIV-infected patients. Supported in part by grants EC11-041 from ISCIII, Spain. Disclosures No relevant conflicts of interest to declare.