The Anti-CD25 Antibody Daclizumab Delays Treg Reconstitution, Promotes CD4 Memory, and Does Not Prevent Acute or Chronic Gvhd After Allogeneic Stem Cell Transplantation

Daclizumab, a humanized monoclonal antibody, binds CD25 and blocks formation of the high affinity IL-2 receptor on T cells. A multi-center, randomized, double-blind clinical trial using daclizumab to perturb IL-2 signaling as a way of reducing T cell mediated graft versus host disease (GVHD) after allogeneic unrelated bone marrow transplantation (BMT) was completed 18 years ago. We report long term outcomes and biological correlates.

Between April, 1993 and December, 1994, 209 adult or pediatric patients receiving marrow from an unrelated donor for hematologic malignancies or severe aplastic anemia at 12 sites in Europe and North America were randomized to one of three arms: 5 weekly doses of placebo (arm A, n=64); 0.3 mg/kg daclizumab (arm B, n=69); or 1.2 mg/kg daclizumab (arm C, n=76) to a maximum of 100 mg beginning on the day before BMT. Conditioning included total body irradiation (TBI) (1200–1440 cGy). All patients received methotrexate (15 mg/m2 day 1; 10 mg/m2 days 3, 6, 11) plus cyclosporine for at least 180 days. The study was designed to provide 80% power to detect a reduction in the incidence of acute GVHD (grade II-IV, requiring treatment with corticosteroids to day +100) from 80% to 55%. Randomization was stratified by donor HLA (matched vs. mismatched) and age (<20 vs. >=20). Arms were well balanced for HLA matching, age, disease risk, TBI dose, and diagnosis. Diagnoses were 31 ALL, 30 AML, 118 CML, 16 MDS, and 14 other. Median age was 31 years (range 1–54); 61% were male; 45% had high risk disease (14% AML with active disease or beyond 2nd remission, 20% CML beyond first chronic phase); 29% had a HLA-A, -B or -DR mismatched donor. T cells from a subset of patients (n=107) were collected at days 11–35, days 36–80, and days 81–100 and analyzed by flow cytometry for expression of total CD25, daclizumab binding, and free CD25 binding sites. Samples from arm A and arm C, including those from 1 year long term follow up (LTFU) were also evaluated for T cell phenotype.

The incidence of grade II-IV acute GVHD for arms A, B, and C were 34%, 42%, and 45%, and results did not differ between arm A vs. arm B (p=0.60) or arm C (p=0.38). Since other clinical outcomes did not differ significantly between arm B and arm C, the two daclizumab dose arms were combined for post-hoc exploratory analysis. There was a suggestion that the daclizumab arms showed decreased risk of relapse (HR 0.57, 95% CI: 0.32–1.00, p=0.06) and increased risk of chronic GVHD (HR 1.49, 95% CI: 0.96–2.31, p=0.07) compared to the placebo arm. Daclizumab did not increase OS (HR 0.89, 95% CI: 0.61–1.29, p=0.54) compared to placebo, and there was no interaction between disease risk and OS. Daclizumab administration reduced the total numbers of T cells expressing CD25 at days 11–35 (arm A=28% vs. arm B=16%, p<0.0001; or arm C=18%, p=0.0002); at days 36–80 (arm A=26% vs. arm B=19%, p=0.03; or arm C=19%, p=0.03); figures were equivalent at days 81–101. Daclizumab was bound to CD25 in vivo, at days 11–35 (arm A=3% vs. arm B=74%, p<0.0001; or arm C=83%, p<0.0001) and persisted at days 81–101 (arm A=4% vs. arm B=17%, p=0.05; or arm C=39%, p<0.0001). Daclizumab administration decreased the numbers of cells with free CD25 binding sites at days 11–35 (arm A=45% vs. arm B=7%, p<0.0001; or arm C=3%, p<0.0001); at days 36–80 (arm A=49% vs. arm B=25%, p<0.0001; or arm C=11%, p<0.0001); and was equivalent at days 81–101. Available samples from arm A (n=18) and arm C (n=40) were tested for Treg (CD4+ CD127- CD25+ FOXP3+) and central memory (CD4+ CD45RA- CCR7+) phenotype. Compared to placebo (arm A), daclizumab administration (arm C) decreased the proportion of CD4 cells that were Tregs at days 11–35 (12% arm A vs. 7% arm C; p=0.008), but not at days 81–101 and LTFU. Daclizumab increased the proportion of CD4 cells exhibiting central memory phenotype at LTFU (10% arm A vs. 21% arm C, p=0.02).

This randomized phase II/III study shows that daclizumab does not prevent GVHD. There was suggestion for increased chronic GVHD and decreased relapse. Daclizumab delayed Treg reconstitution while increasing CD4 central memory at LTFU. Based on these results, further trials should test whether anti-CD25 therapy can promote anti-tumor immunity by impairing Treg reconstitution after transplant, in the appropriate clinical setting.

Off Label Use: Daclizumab administration following hematopoietic cell transplantation. Walker:Baxter Corporation: Research Funding. Light:Bristol Myers Squibb: Employment.