ABC Subtype, TP53 Mutation and BCL2 Overexpression are Associated with a Worse Outcome in Untreated Poor-Risk Young Diffuse Large B-Cell Lymphoma: Results from the Frontline Phase III BIO-DLBCL04 Study of the Fondazione Italiana Linfomi

Introduction. Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous disease characterized by cell of origin (COO) subgroups at different prognosis; in addition, the overexpression of MYC and BCL2 or the chromosomal rearrangements of MYC and BCL2 are associated with a worse outcome. The role of TP53 mutation is clear in chronic lymphocytic leukemia, but it is not yet established in DLBCL. The phase III randomized study FIL-DLCL04 in 399 young patients with untreated DLBCL at poor prognosis (age adjusted IPI, aa-IPI, 2-3), showed an advantage in term of failure-free survival (FFS) of an abbreviated Rituximab-dose-dense chemotherapy at two different dose-escalation (R-CHOP14 or R-MegaCHOP14) followed by an intensification with Rituximab-MAD+BEAM and ASCT (R-HDC+ASCT) compared to a full course of R-dose-dense chemotherapy alone (R-dose-dense). However, there was no difference in overall survival (OS). (Chiappella et al, Lancet Oncol 2017). Aims. The aim of BIO-DLCL04 was to correlate the COO profile and the presence of biomarkers (MYC and BCL2 expression, MYC and BCL2 translocations and TP53 mutation), with OS. Methods. COO subtyping was determined with immunohistochemistry (IHC) (germinal center (GCB) and non-GCB according to Hans' algorithm) and with gene expression profiling (GEP) using the NanoString Research Use Only Lymphoma Subtyping (NanoString Technologies, Inc., Seattle, WA, USA) identifying GCB, activated B-cell (ABC) and unclassified subgroups. Expression of BCL2, MYC and TP53 were studied in IHC; cases were deemed positive if at least 50%, 40% of lymphoma cells were stained with BCL2, MYC antibodies, respectively. BCL2 and MYC translocations, copy gains and aberrations were tested with fluorescent in situ hybridization (FISH) and TP53 anomalies by Sanger sequencing. Results. From 2005 to 2010, 399 DLBCL were enrolled in FIL-DLCL04 and randomized to receive R-HDC+ASCT in 199 and R-dose-dense in 200 (NCT00499018). Central histology revision was mandatory. OS was analyzed; an adjusted hazard ratio (aHR) for treatment (dose escalation and transplantation), gender, aa-IPI, performance status and bone marrow involvement, was calculated using a logistic regression model. Ninety-five DLBCL were analyzed for COO according to GEP with Nanostring and have complete data for all analysis: 55 (58%) were GCB, 25 (26%) ABC, and 15 (16%) unclassified. COO in IHC according to Hans algorithm was reported in 236 patients: 98 (42%) were GCB and 138 (58%) non-GCB. MYC and BCL2 expression were studied in 156 and 284 cases respectively and were deemed positive in 41 (26%) and 229 (81%); 39 cases (21%) were dual expressors. TP53 was studied in 115 cases and resulted mutated in 10 (9%) and wild type in 105 (91%). MYC and BCL2 translocations were tested in 123 and in 131 cases; 10 (8%) were MYC positive and 25 (19%) BCL2 positive; 12 cases (10%) were double hit. Clinical outcome was superimposable between the analyzed cases and the global FIL-DLCL04 study populations. At a median follow-up of 72 months, 5-years OS were reported in table 1 by biological characteristics. In particular, BCL2 overexpression had an aHR of 2.52 (95% CI: 0.99-6.40), p 0.05; dual expressors an aHR of 2.06 (95% CI: 1.00-4.28), p 0.051; TP53 mutation an aHR of 3.26 (95%CI: 0.98-10.90), p 0.05 and ABC an aHR of 4.59 (95%CI: 1.55-13.59), p 0.006. No significant differences by R-HDC+ASCT versus R-dose-dense were observed in the different biological subgroups (data not shown). Conclusions. In our prospective FIL-BIODLCL04 trial, with a long-term follow-up, COO assessed by GEP is able to discriminate groups at different prognosis. BCL2 overexpression, dual expressors, and ABC subgroup (assessed with GEP) significantly influenced OS. An important role is played by TP53 mutations, that represent a factor impacting on OS in DLBCL. The intensification with R-HDC+ASCT is not able to overcome the chemorefractoriness associated with TP53 disruption and the dismal prognosis of the unfavorable subgroups.