Phase 1 Clinical Study of Adoptive Immunotherapy with Delayed Infusion of Alloanergized Donor T Cells to Improve Immune Reconstution after Haploidentical Stem Cell Transplantation.

Haploidentical related donors extend availability of hematopoietic stem-cell transplantation (HSCT) to patients (pts) lacking HLA-matched family donors, but profound T-cell depletion (TCD) is needed to prevent severe GvHD, thus delaying immune reconstitution and increasing infection. Adoptive transfer of alloanergized donor T cells is an attractive approach to reconstituting T cell number and function with concomitant GVHD abrogation. We established proof-of-principle in haploidentical bone marrow (BM) transplant trials using ex vivo induction of recipient alloantigen-specific anergy in T cells within donor BM by allostimulation with blockade of CD28-mediated costimulation. This strategy permitted infusion of large doses of haploidentical T cells with donor BM, resulting in rapid immune reconstitution without excess severe GvHD or chronic GVHD. However the optimal dose of alloanergized donor T cells and their impact on functional antigen-specific immune reconstitution were not determined in our prior studies. We now report the results to date of a new study evaluating delayed infusion of escalating doses of donor PBMC anergized to recipient alloantigens after haploidentical HSCT. Alloanergized PBMCs were generated by co-culture of irradiated stimulator PBMC from a second haploidentical related donor (or the pt) using clinical grade humanized monoclonal anti-B7.1 and –B7.2 antibodies. 7 adults (median age 41, range 22–50) and 4 children (median age 7.5, range 2–14) with high risk leukemia/MDS (8 AML (3 CR1 4 CR2, 1 persistent disease), 2 high-IPSS MDS and 1 ALL (CR2) have been treated. Pts received fractionated TBI (1200cGy, n=5) or melphalan (140mg/m², n=6), fludarabine, thiotepa and ATG followed by CliniMacs CD34-selected peripheral blood stem cells (PBSC) from haploidentical family donors without subsequent pharmacologic GVHD prophylaxis. Median infused cell doses (x 10⁶/kg) were 9.4 (CD34⁺) and 0.02 (CD3⁺). All pts engrafted and attained full donor chimerism, with rapid neutrophil and platelet recovery (median D+11 and D+12, respectively). Using novel Bayesian phase I/II adaptive design, 10 pts have received donor PBMC after alloanergization (which resulted in a median 6-fold reduction in alloresponses): Dose (Ds)1, 10³ CD3⁺ cells/kg (n=4), Ds2 10⁴/kg (n=3) and Ds3 10⁵/kg (n=3), infused on D+35 (n=8) or D+42 (n=2) without developing severe acute GvHD (Table). None of 5 evaluable pts developed chronic GvHD. At median follow-up of 8 months (range 1–23), 8/11 pts are alive. 6/11 pts are disease-free. 3 pts have died, from bacterial sepsis (Day +32), pulmonary veno-occlusive disease (D+59), and idiopathic pulmonary syndrome (D+78). Two AML pts have relapsed. Infusion of alloanergized donor PBMC appears to influence reconstitution of both CD4 T cell numbers and functional pathogen-specific T cells (Table). Normal SEB responses and functional CMV- and VZV-specific CD4 and CD8 T cells became detectable at 6–9 months in pts at Ds1, at 3 months in pts at Ds2 and at 2 months in the assessable pt at Ds3. Recovering T cells had a predominantly effector memory phenotype consistent with peripheral expansion of infused alloanergized donor T cells. These data suggest that delayed infusion of modest doses of alloanergized donor PBMC after haploidentical HSCT is not associated with significant GVHD, and may be associated with a dose-dependent improvement of quantitative and qualitative immune reconstitution. Ongoing recruitment of patients to higher alloanergized PBMC dose levels (up to 10⁷/kg if tolerated) will determine the optimal dose that benefits immune reconstitution without causing severe GVHD.