Long-term outcomes and circulating tumor DNA analysis from a phase I/II study of lenalidomide and obinutuzumab with CHOP for newly diagnosed diffuse large B-cell lymphoma.

7553 Background: Diffuse large B-cell lymphoma (DLBCL) can be cured with rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (RCHOP), but a third of patients (pts) experience treatment failure. Randomized trials comparing RCHOP with modified regimens, including replacing R with obinutuzumab (O, CD20 antibody, Vitolo JCO 2017) and adding lenalidomide (L) to RCHOP (Nowakowski JCO 2021), have not significantly improved outcomes. The combination of L and O may have synergy via an enhanced NK-cell mediated antibody dependent cellular cytotoxicity. We hypothesized combining L and O with CHOP (LOCHOP) would be an effective and well tolerated treatment of DLBCL. Methods: In this single center phase Ib/II study (NCT02529852) pts with untreated CD20+ DLBCL received 6 cycles of CHOP with O 1000mg IV on days (D) 1, 8, and 15 during cycle 1 and D1 during cycles 2-6, and L 15 mg orally on D1-14 of each cycle. Growth factor and thromboembolism prophylaxis were mandated. The primary objectives were to determine the maximum tolerated dose (MTD) of L and efficacy of LOCHOP. Plasma samples were collected for circulating tumor DNA (ctDNA) analysis. Results: Fifty-three pts were enrolled, 6 in phase Ib and 47 in phase II. No dose limiting toxicities were experienced in phase Ib at an MTD of 15 mg of L. Median age was 62 years, 35 (66%) pts had stage III/IV disease, 12 (23%) an IPI > 2, and 26 (49%) and 22 (42%) were germinal center (GC) and non-GC subtype (5 not classified) by Hans algorithm. At end of therapy, 50 pts were evaluable, overall response: 49 (98%), complete response: 45 (90%). Median follow up was 4.5 years and 4-year progression free and overall survival rates were 87.4% and 91.3%. No characteristics were associated with differences in outcomes. Any grade and grade 3-4 adverse events (AEs) were experienced by 100% and 70% of pts. Grade 3-4 AEs included neutropenia (38%), thrombocytopenia (17%), fatigue (13%), neutropenic fever (13%), and infection (9%). Four (8%) pts developed venous thromboembolism. At study entry, 31/33 (94%) profiled pts had detectable ctDNA with CAPP-Seq, 24% had high ctDNA ( > log 2.5 hGe/ml) and 24/31 (77%) were evaluable by LymphGen classifier: molecular subtypes included EZB (5, 21%), ST2 (5, 21%), MCD (4, 17%), and other (9, 38%). All 5 EZB and 4/5 ST2 pts were GC subtype while 2/4 MCD pts were non-GC (1 GC, 1 not classified). With LOCHOP, 13/18 (72%) and 11/15 (73%) pts achieved early and major molecular response (Kurtz JCO 2018). By PhasED-Seq (Kurtz Nat Biotechnol 2021), 16/18 (89%) had no detectable ctDNA after ≥5 cycles. Conclusions: LOCHOP demonstrates high efficacy and tolerability in newly diagnosed DLBCL, leading to a high rate of undetectable minimal residual disease by ctDNA. Noninvasive molecular subtyping is feasible as a supplement to tissue diagnosis and should be incorporated in future studies aiming to improve on RCHOP. Clinical trial information: NCT02529852.