Abstract 2856 Background:For over 20 years, the purine analog cladribine has been the standard of care for the treatment of newly diagnosed and once-relapsed hairy cell leukemia (HCL), due to high rates of durable complete remission (CR). However, long term follow-up of patients after purine analog indicates relapse in ∼50% of patients by 20 years, without a plateau on the disease-free survival curve, and minimal residual disease (MRD) is present in a high proportion of patients in CR. The declining CR and partial response (PR) rates with each repeated course of purine analog suggest that improved therapy for early HCL might prevent or delay relapse. Rituximab begun 4 weeks after cladribine in 13 patients was reported associated with very low rates of MRD. To determine the value of rituximab in the early vs delayed setting, a randomized trial was undertaken. Patients enrolled were treated by their local physicians or at NIH. For patients with variant HCL (HCLv), the low response rates with initial cladribine (<5% CR, <50% PR) suggest that this group might benefit by addition of early rituximab. Methods:Patients with 0 (1st line, n=68) or 1 (2nd line, n=62) prior courses of cladribine are randomized 1 :1 to cladribine (0.15 mg/Kg/day×5) with immediate vs >6-month delayed rituximab (375 mg/m2/week×8). Delayed rituximab is begun for MRD in blood detected >6 months after cladribine, and a 2nd 8-dose course of delayed rituximab may be given to either arm if MRD in blood persists >6 months after beginning the 1st course of rituximab. The primary endpoint is MRD by all tests 6 months after cladribine. Secondary endpoints include comparing immediate vs delayed rituximab with respect to 1) CR and PR rates and progression-free survival, 2) MRD status at one month, 3) MRD-free survival after cladribine, and 4) long-term MRD-free survival enhanced if needed by up to 6 courses of rituximab. A non-randomized group of HCLv patients (n=20) are treated with immediate cladribine + rituximab. Patients were evaluable for response by 6 months after cladribine unless they achieved CR earlier. Results:So far, 60 patients were enrolled to the 5 arms. 100% of 48 evaluable patients responded. Of 33 HCL 1st line patients, responses in those evaluable included 15 (100%) CRs out of 15 with vs 9 (90%) CRs + 1 (10%) PR out of 10 without rituximab. MRD rates in those evaluable at 1 month include 5 (31%) of 16 with vs 15 (100%) of 15 without rituximab (p<0.0001). At 6 months, rates of MRD in blood were 0 of 10 with vs 6 (67%) of 9 without rituximab (0=0.003). The effect of rituximab on MRD was less in 19 HCL patients receiving 2nd-line cladribine. Six (67%) of 9 with vs 9 (100%) of 9 without rituximab were MRD+ at 1 month (p=0.2), and 1 (14%) of 7 with vs 3 (43%) of 7 without rituximab were MRD+ in blood at 6 months (p=0.6). Responses in 2nd-line patients included 8 (89%) CRs and 1 (11%) PR out of 9 with vs 7 (88%) CRs and 1 (13%) PR out of 8 without rituximab. 12 courses of delayed rituximab have been administered to 11 patients. One patient received 2 courses of delayed rituximab and cleared all MRD. Of the other 10 patients, MRD in blood cleared in 8 of 8 evaluable patients at 1 and 3 months and 7 of 7 at 6 months. Of 8 with HCLv, 6 (100%) of 6 so far evaluable achieved CR, all without MRD, and remain without MRD 8–28 (median 14) months after cladribine. Early rituximab was usually associated with infusional reactions and thrombocytopenia with the 1st but not subsequent doses, and infections were related more to baseline neutrophil count rather than to addition of rituximab. Conclusions:While CR rates are very high by 6 months after cladribine, without rituximab MRD is usually present even in blood, but can be cleared by delayed rituximab. Additional patients and follow-up will be needed to compare long-term outcomes of early vs delayed rituximab. Either option may be better than purine analog alone, given the relatively low response rates to single-agent rituximab after clinical relapse from purine analog. However, the correlation between MRD status and clinical relapse is not clear and is an important secondary endpoint of this trial. In HCLv, rituximab given concurrently with cladribine is markedly effective in achieving CR without MRD, and additional patients and follow-up will be important to determine response durability (Supported in part by NCI, intramural research program, NIH, and Genentech, Inc). Disclosures:Kreitman:Genentech: Research Funding. Off Label Use: Off-label use of rituximab for the treatment of HCL. Arons:Genentech: Research Funding.