Introduction: Genetic subtypes of diffuse large B-cell lymphoma (DLBCL) have distinct oncogenic pathway feature, gene expression phenotype, and variable clinical outcome. We aimed to investigate whether adding novel targeted agents to R-CHOP (R-CHOP-X) based on genetic subtypes could improve clinical efficacy in newly diagnosed DLBCL. Methods: This was a randomized phase II study comparing R-CHOP-X with R-CHOP in newly diagnosed DLBCL (NCT04025593). Eligible patients were aged 18-80 years, and IPI risk of intermediate or high (IPI≥2). All patients were treated with standard R-CHOP for the first cycle, and were randomized 1:1 to R-CHOP-X or R-CHOP for the remaining 5 cycles. Targeted sequencing was performed on formalin-fixed paraffin-embedded tumor samples at diagnosis. We used a simplified method to classify patients into six genetic subtypes MCD, BN2, N1, EZB, TP53 and NOS, using mutations in 18 genes, BCL2 translocation, and BCL6 fusion. MCD and BN2 patients were treated with BTK inhibitor ibrutinib 420 mg/d, N1 and NOS patients with immunomodulatory agent lenalidomide 25mg d1-10, EZB patients with histone deacetylase inhibitor tucidinostat 20mg d1, 4, 8, 11, and TP53 patients with intravenous demethylating agent decitabine 10 mg/m2 d1-5 followed by standard R-CHOP. The primary endpoint was complete response rate. Secondary endpoints included progression-free survival (PFS), overall survival (OS), overall response rate, and safety. Results: A total of 128 patients were enrolled (n = 64 per arm) between July 17th, 2019 to December 29th, 2020. Baseline clinical characteristics were similar between arms, with a median age of 64 (range 25-74), 52% male, 83% ECOG 0-1, 77% stage III/IV, and 65% IPI≥3 for all patients. According to Hans algorithm, 62% patients were categorized as non-germinal B-cell subtype and 36% as MYC/BCL2 double expression. Regarding genetic subtypes, 20%, 18%, 4%, 2%, 16% and 39% patients were classified into MCD, BN2, N1, EZB, TP53, and NOS subtype, respectively. Until March 1st, 2021, all patients received at least 3 cycles of immunochemotherapy and 107 patients were available for response evaluation. The complete and overall response rates were 85% and 91% in R-CHOP-X arm, and 65% and 72% in R-CHOP arm, respectively. With a median follow-up of 14.1 months, 1-year PFS and OS rates were 96% and 98% in R-CHOP-X arm, and 79% and 94% in R-CHOP arm, respectively (PFS: HR 0.22, 95%CI 0.09-0.61; OS: HR 0.28, 95%CI 0.05-1.60). The most common grade 3/4 adverse events (AEs) for R-CHOP-X vs R-CHOP were neutropenia (81% vs 75%), thrombocytopenia (31% vs 11%), anemia (25% vs 20%), and febrile neutropenia (20% vs 11%). For non-hematological AEs, grade 3 infection occurred in 3% vs 2%, grade 3 gastrointestinal bleeding in 2% vs 3% in R-CHOP-X vs R-CHOP arm, respectively. No grade 4 non-hematological AEs were observed. Conclusion: Genetic subtype guided R-CHOP-X showed encouraging response and outcome in DLBCL. Keywords: Aggressive B-cell non-Hodgkin lymphoma, Molecular Targeted Therapies No conflicts of interests pertinent to the abstract.