DNA methylation inhibition in myeloma: Experience from a phase 1b study of low-dose continuous azacitidine in combination with lenalidomide and low-dose dexamethasone in relapsed or refractory multiple myeloma.

The DNA methyltransferase inhibitor azacytidine (aza) may reactivate pathways associated with plasma cell differentiation, cell cycle control, apoptosis, and immune recognition and thereby restore sensitivity to lenalidomide (len) and dexamethasone (dex) in relapsed and/or refractory multiple myeloma (RRMM). We aimed to develop an aza regimen that reaches epigenetically active levels 8 times in 28 days with less bone marrow toxicity than the myeloid malignancy standard of 7 consecutive doses to enable safe combination with len. Aza was escalated from 30 mg/m once a week up to a predefined maximum of 50 mg/m twice a week in combination with GFR-adjusted len (≥ 60 mL/min: 25 mg, 3059 mL/min: 10 mg) day 1 to 21 every 28 days and dex 40 mg once a week followed by a limited expansion study to a total N of 23 at the highest tolerated dose. Fifty-one patients (pts) with RRMM were screened, 42 were treated and 41 were evaluable for response based on at least 1 response assessment or progression after treatment start. The median number of prior lines of therapy was 5 (1-11) and 81% (34) were refractory to len and/or pomalidomide (pom). Two DLTs occurred in different cohorts, 1 neutropenic fever in 1/6 pts on the aza 40 mg/m twice a week GFR ≥ 60 mL/min cohort and 1 GGT elevation in 1/6 pts on the aza 50 mg/m GFR 30-59 mL/min cohort. An MTD was not reached and aza 50 mg/m SC twice a week was chosen for the expansion study. At least possibly related Grade 3/4 AEs occurred in 28 pts (67%) with the following in > 1 pt: neutropenia (N = 16, 38%), anemia (N = 6, 14%), lymphopenia (N = 5, 12%), thrombocytopenia (N = 4, 10%), leukopenia (N = 4, 10%), febrile neutropenia (N = 4, 10%), fatigue (N = 3, 7%), fever (N = 2, 5%), and infection (N = 2, 5%). At a median follow up time for alive pts of 60.2 months (range: 36.1-82.5 months), the overall response rate (≥ partial response) and clinical benefit response rate (≥ minor response) was 22 and 32%, respectively, with 4 very good partial responses (10%), 5 partial responses (12%), and 4 minor responses (10%). The median PFS was 3.1 months (95% confidence interval [CI]: 2.1-5.1 months), median TTP 2.7 months (95% CI: 2.1-7.5 months), and median OS 18.6 months (95% CI: 12.9-33.0 months). Achieving at least minor response and reaching TTP > 6 months was associated with approximately 35% lower median plasma levels of the enzyme that inactivates aza, plasma cytidine deaminase (CDA, P< .0001). Two of the len refractory pts achieved longer disease control than with any prior regimen and 1 responded immediately after progression on len, bortezomib, and prednisone. Analyses of the methylation state of over 480,000 CpG sites in purified myeloma cells at screening were possible in 11 pts and on day 28 in 8 of them. As in other studies, the majority of differentially methylated CpGs compared to normal plasma cells were hypomethylated in myeloma. Treatment decreased the number of CpGs that were differentially methylated in normal plasma cells by > 0.5% in 6 and by > 5% in 3 of the 8 pts, most pronounced in 2 pts with clinically convincing aza contribution who achieved a reduction in overall differentially methylated CpGs by 23 and 68%, respectively, associated with increased expression of immunoglobulin genes. The study demonstrated tolerability of twice a week SC aza at 50 mg/m with len and dex in RRMM and suggested aza may help overcome the len/pom refractory state, possibly by activating differentiation pathways. Relatively low response rates and association of clinical benefit with low plasma levels of the aza inactivating enzyme CDA suggest the aza regimen will need to be optimized further and pt selection may be required to maximize benefit.