Background: Epigenetic modifications including CpG island promoter methylation that favor transcriptionally repressed chromatin have been identified in a variety of B-cell malignancies, including CLL, mantle cell lymphoma, and diffuse large B-cell lymphoma (DLCL). Recent investigations by our group and others have demonstrated aberrant hypermethylation in a variety of genes critical for cell cycle regulation or apoptosis in NHL and CLL, including p15, p16, DERMO1, DAPK1, FOXE1, GRM7, and TWIST1. Targeting aberrant DNA hypermethylation through inhibition of DNA methyltransferase (DNMT) with low-dose decitabine has provided a new treatment strategy in acute myeloid leukemia (AML), and we postulated that a similar approach may be effective in CLL and B-cell NHL.Methods: Two separate phase I trials were conducted at the Ohio State University to determine the minimum effective pharmacologic dose (MEPD) of the DNMT inhibitor, decitabine, in the treatment of patients (pts) with relapsed or refractory CLL and aggressive NHL. In both trials, the MEPD was defined as the dose of decitabine where evidence of hypomethylation of selected genes (DAPK1, TWIST, and DERMO-1 for CLL; DAPK1, p15, and p16 for NHL) or selected gene re-expression was observed in 5 of 6 pts with ≤ 1 pt at each dose level experiencing dose limiting toxicity (DLT). Eligibility criteria included at least one prior therapy, ECOG performance status 0–2, and normal hepatic and renal function. Absolute neutrophil count ≥ 1500/mm3 and platelets ≥ 75,000/mm3 were required for all NHL pts. Twenty-eight days defined a cycle and hematologic recovery to ≥ 75% of baseline or ≤ grade (gr.) 2 by day 35 was necessary for decitabine retreatment in both trials.Results: Twenty pts (median age 69.5, range 32–82) with relapsed/refractory CLL (n=16) or DLCL (n=4) after a median of 4.5 prior therapies (range, 2–8) were enrolled. Only 1 of 20 pts had previously undergone stem cell transplantation, 2 years prior to study enrollment. Pts received 1–3 cycles of decitabine, and were removed from study due to DLT or disease progression. DLT was observed in 2 of 4 CLL and 2 of 2 DLCL pts receiving decitabine at 15 mg/m2/day over 1 hour for 10 days, consisting of gr. 4 thrombocytopenia (n=3) beyond day 35 and gr. 3 hyperbilirubinemia (n=1). As a result, 6 pts with CLL were treated with decitabine at 10 mg/m2/day for 10 days without DLT; however, sufficient evidence of hypomethylation was not observed. Therefore, due to the failure to obtain MEPD at decitabine doses of 10 mg/m2 over 10 days, a shorter 5-day infusion schedule was examined in both CLL and DLCL in order to permit maximal dose escalation. With decitabine at 15 mg/m2/day on days 1–5, DLT occurred in 2 of 6 CLL and 2 of 2 DLCL pts, consisting of persistent gr. 3–4 neutropenia or thrombocytopenia after day 35, gr. 3 gram positive sepsis, and gr. 4 neutropenic fever. No responses were observed. In 18 pts with circulating CLL cells or lymph nodes accessible to biopsy, LC-MS analysis of global DNA methylation demonstrated no significant changes in genome-wide methylation in samples collected pre-treatment and on days 3, 5, 15, and 22 of cycle 1. Plasma pharmacokinetics (PK) in 11 pts demonstrated a mean Cmax of 1.4 ± 1.0 μM, AUC of 61.2 ± 28.1min*μM, total clearance of 1.5 ± 1.2 ml/min/m2, and beta half-life 25.5 ± 6.9 min, similar to previously published findings in AML (Blum W, JCO 25:3884-3891, 2007). There was a trend of increase in PK parameters with increased dose, but no difference was found between the PK in CLL and NHL patients nor with repeat dosing.Conclusions: Dose-limiting myelosuppression and resultant infectious complications prevents dose escalation of decitabine to dose levels associated with hypomethylation in CLL and DLCL. Alternative agents targeting DNA methylation in B-cell malignancies should be explored.