A Phase I, Open Label, Multicenter Trial of Oral Azacitidine (CC-486) Plus R-CHOP in Patients with High-Risk, Previously Untreated Diffuse Large B-Cell Lymphoma, Grade 3B Follicular Lymphoma, or Transformed Lymphoma

Introduction : Approximately 40% of patients (pts) with diffuse large B-cell lymphoma (DLBCL) do not achieve complete response (CR) or relapse after receiving R-CHOP. A proposed mechanism of chemoresistance is aberrant DNA methylation. Preclinical data show low doses of DNA methyltransferase inhibitors, such as azacitidine, suppress DLBCL tumor growth, while causing minimal DNA damage and enhancing chemosensitivity. In an earlier phase I clinical trial, subcutaneous (SC) azacitidine + R-CHOP showed 10/11 CRs in DLBCL pts with international prognostic index (IPI) score >2 (Clozel et al. Cancer Discovery 2013), providing a rationale for this study of oral azacytidine (CC-486) which is easier to administer and may achieve more continuous hypomethylation effects.

Methods: The CC-486-DLBCL-001 phase I trial examined 4 escalating doses (100, 150, 200, and 300 mg) of CC-486 combined with standard R-CHOP21 in pts with previously untreated DLBCL, grade 3B follicular lymphoma (FL), or transformed FL; IPI score ≥2; ECOG PS ≤2; and Ann Arbor stage II-IV disease (NCT02343536). Pts were enrolled sequentially starting at 100 mg CC-486 in a time-to-event continual reassessment method (TiTE-CRM) design. Priming with CC-486 was given for 7 days prior to R-CHOP initiation on day 1 of cycle 1. Thereafter, CC-486 was given for 14 days (days 8-21) before the next R-CHOP cycle. Patients standardly received G-CSF. Treatment was given for six, 21-day cycles. Primary endpoints were to determine safety, dose-limiting toxicity (DLT), and maximal administered dose (MAD) of CC-486 combined with R-CHOP (per NCI CTCAE v4.03). Secondary endpoints included PK and preliminary efficacy, including overall response rate (ORR) and CR per 2014 IWG criteria. Correlative analyses were performed at day -6 (pre-CC-486) versus day 1 of cycle 1 (post-CC-486) in matched tumor and circulating cell-free DNA (cf-DNA) samples for DNA bisulfite sequencing and gene expression, and in plasma for cytokine levels. The study design consisted of "dose escalation" assessing 4 dose levels, followed by a "dose expansion" at the recommended phase II dose (RP2D). The data reported here are from the completed dose escalation phase.

Results: A total of 33 pts were enrolled in dose escalation as of May 31, 2017. The median age was 65 years (range, 25-80), 67% were >60 years, and 55% were male. 32 (97%) pts had DLBCL (including 5 transformed from FL), and 1 patient had grade 3B FL. Thirty (91%) pts had stage III-IV disease, and 19 (58%) had IPI score ≥3. 91% of pts completed 6 full cycles of planned study treatment. CC-486 dose reduction occurred in 9 (27%) pts due to adverse events (AEs). Treatment with CC-486 (150 mg) was discontinued in 1 patient due to febrile neutropenia. The most common any-grade AEs included neutropenia (70%), nausea (64%), constipation (58%), fatigue (55%), vomiting (48%), and diarrhea (48%). Grade 3/4 treatment-related AEs occurred in 21 (64%) pts; most common (≥10%) were neutropenia (58%) and febrile neutropenia (21%). Serious AEs (SAEs) occurred in 13 (39%) pts; febrile neutropenia was the only SAE occurring in >1 pt (24%). There were no deaths related to the study. DLTs were seen in 2 pts, including 1 grade 4 febrile neutropenia (200 mg cohort) and 1 grade 4 neutropenia (300 mg cohort) . The MAD was 300 mg, which was identified as the RP2D for dose expansion. 32 of 33 efficacy-evaluable pts responded for 97% ORR, with 28 (85%) pts achieving a PET-negative CR. In 19 pts with IPI score ≥3, ORR was 100%, and CR was 84%. Correlative analyses for cytokine levels and DNA methylation pre- versus post-CC-486 showed significantly decreased IFN-α2a and increased IFN-β and IFN-λ in blood (n=24), and significant tumor and cf-DNA hypomethylation. There was a significant change in gene expression for IFN-related immune response and lymphoma microenvironment pathways (n=3). PK was similar to the known PK profile in AML and MDS patients and was unaffected by co-administration of R-CHOP.

Conclusions: CC-486 combined with R-CHOP showed promising preliminary efficacy in pts with high-risk, previously untreated DLBCL or grade 3B FL. Results from this study identified a RP2D of 300 mg for future studies of CC-486 plus R-CHOP in DLBCL pts. Adverse events were generally consistent with the known safety profile of azacitidine and toxicities associated with R-CHOP. Correlative analyses showed pharmacodynamic activity, the majority related to anti-lymphoma immune regulation.