P1207: MULTICENTER RANDOMIZED CONTROLLED (COMPARATIVE) OPEN PROSPECTIVE STUDY TO EVALUATE THE EFFICACY OF THE R-DA-EPOCH-21 AND R-MNHL-BFM-90 ± AUTO- HSCT PROGRAMS IN PATIENTS WITH DE NOVO DIFFUSE B-CELL LAR

Background: Short-pulse NHL-BFM-90 chemotherapy is highly effective in pediatric aggressive B-cell lymphomas. We modified R3 arm of NHL-BFM-90 program (named R-mNHL-BFM-90) for adults under 60 years with high-risk de novo DLBCL and conducted the randomized study compared to well-known R-DA-EPOCH-21 chemotherapy Aims: to evaluate R-mNHL-BFM-90 and R-DA-EPOCH-21 efficacy and toxicity and auto-HSCT reasonability as well in de novo adult DLBCL patients with 2 or more signs of poor prognosis Methods: Treatment protocol was registered as NCT02842931 at www.clinicaltrials.gov. Inclusion criteria were newly diagnosed DLBCL, not otherwise specified (NOS), according WHO, 2008 and 2017 criteria; two or more signs of poor prognosis (i.e. intermediate and high aaIPI risk); age 18-60. Complete remission (CR), partial remission (PR) or disease progression (DP) corresponded Lugano 2014 criteria. The study design intended randomization for four parallel arms. The first one was 6-cycle R-DA-EPOCH-21 as described earlier. The second, 6-cycle R-DA-EPOCH-21 plus auto-HSCT. The third, 6-cycle (A-B-A-B-A-B) R-mNHL-BFM-90. In addition, the fourth one was 6-cycle R-mNHL-BFM-90 plus auto-HSCT. If CR was not achieved after 6-cycle initiating treatment in every arm additionally 2-cycle R-DHAP chemotherapy were completed before auto-HSCT if intended. G-CSF after chemotherapy was administrated according clinical guides. A total of 140 patients from 13 medical Russia clinics were included from February 2015 to June 2021. Randomization was done according “intention-to-treat” as follows: R-DA-EPOCH-21 (35 patients); R-DA-EPOCH-21+auto-HSCT (30 patients); R-mNHL-BFM-90 (37 patients); R-mNHL-BFM-90+auto-HSCT (38 patients). Ten patients were excluded from the cohort due various reasons. Therefore, the first-step analysis was carried out for 130 patients according “intention-to-treat”. In cases of PR after 6 cycles or progression at any treatment stage or relapse (a total 14 patients) we performed the second tumor biopsy. In all cases, the final diagnosis was assessed as indolent lymphoma transformation into DLBCL or other “non-DLBCL” lymphomas. So, the second-step analysis was performed for 116 “proved” de novo DLBCL patients. Results: Evaluation of comparative efficacy was carried out in R-DA-EPOCH-21±auto-HSCT and R-mNHL-BFM-90±auto-HSCT cohorts. Detailed results are presented in Table 1. Hematological/non-hematological toxicity was acceptable in all groups. However, neutropenic fever and grade 3-4 thrombocytopenia differed statistically significantly (χ2 test, p<0.05) between R-DA-EPOCH-21 and R-mNHL-BFM-90 arms. The value of auto- auto-HSCT will be determined later. Image:Summary/Conclusion: Thus, de novo DLBCL NOS is potentially curable disease with no relapses. R-mNHL-BFM-90 program is highly effective in de novo DLBCL NOS adult patients. R-mNHL-BFM-90 therapy toxicity is acceptable. The estimated CR duration probability within 10 months after treatment completion in the high-risk group is 100% in R-mNHL-BFM-90 group compared to 57% in R-DA-EPOCH group (χ2 test, p=0.0047).