We have reported that treatment of CLL patients with weekly RTX doses of 375 mg/m2 depletes complement (C) and induces substantial loss (~90%) of CD20 on circulating cells which is not caused by B cell internalization or RTX masking of CD20 (Kennedy, J Immunol, 2004). This loss reduces the potential efficacy of RTX. CD20 loss may be caused by “shaving” of high density RTX-CD20 immune complexes from malignant B cells by fixed tissue macrophages of the mononuclear phagocytic system. To preserve C activity and reduce CD20 loss, we have investigated whether more frequent, but lower RTX doses might enhance therapeutic efficacy for CLL. Patients were treated 3 times/week (MWF) with either 20 mg/m2 or 60 mg/m2 RTX, for 4 weeks. During the first 2 weeks at least 6 blood samples were collected before, during, and after RTX infusions, and samples were analyzed for CBC, C activity, B cell expression of CD20, and degree of opsonization by C3 fragments. Findings for the first 3 patients enrolled in the 20 mg/m2 group include the following observations. Immediately after the first RTX infusion lymphocyte counts (cells/ul) decreased from 70000 ± 17000 to 9000 ± 7000, and then recrudesced to 21000 ± 1700, 2 hrs after RTX infusion. At this time B cell CD20 levels were reduced, but were still at ≥ 50 % of their initial CD20 values. After 2 days lymphocyte counts increased but were modestly reduced from starting values, and the second and subsequent RTX infusions induced similar patterns of cell clearance, recrudescence and partial CD20 loss. For example, CD20 levels on treatment day 6 started at ≥ 50% of their pre-treatment (day 1) values. These phenomena continued over the 4 week treatment course, except starting lymphocyte counts on each new treatment day continued to decrease: Lymphocyte counts at the start of the 12th treatment day averaged 10,000 ± 9,000 and 4 weeks after completion of treatment, lymphocyte count for patient #2 remained low (2,000). In all 3 patients infusion of RTX induced deposition of C3 fragments on circulating cells, but C activity remained substantial throughout 4 weeks. The first 2 patients enrolled in the 60 mg/m2 regimen demonstrated somewhat similar patterns, except CD20 levels showed greater decreases. CD20 on B cells of patient #4 decreased ~80 % after the first infusion, and continued to decrease throughout the treatment course, although lymphocyte counts were reduced from 86,000 (initially) to 21,000 at the start of treatment day 12. In patient #5, B cell CD20 deceased by ~ 75% during each RTX treatment, but partially recovered every two days, and was at ~ 30% of its initial value at the start of treatment day 6. Six patients will be enrolled in each treatment group. Our results to date suggest that frequent but low dose RTX regimens can promote slow but steady cytotoxic attack of RTX on CD20 cells while preserving C activity and maintaining CD20 levels so that additional RTX treatments continue to be effective.