Phase I/II Study of Marrow Infiltrating Lymphocytes (MILs) Generates Measurable Myeloma-Specific Immunity in the Autologous Stem Cell Transplant (SCT) Setting

Abstract 997

Adoptive T cell therapy requires the infusion of highly tumor specific T cells capable of trafficking to the tumor site, killing the tumor and persisting over time. MILs (as compared to peripheral blood T cells) possess many of these features (Noonan, Ca, Res. 2006). We here report the initial results from our Phase I/II trial. 22 patients were transplanted with a Melphalan-200. 19 with a CD34-selected product, 1 unselected autologous stem cells and 2 auto-BMT. 12 underwent SCT as initial consolidation and 10 patients with relapsed disease that were heavily pre-treated with an average of 3.2 prior therapies. Mean ISS score was 2.0. Average age was 56 (29–71). MILs were expanded with anti-CD3/CD28 beads for 7 days (avg. fold expansion 48.5 and avg. cell dose 2.8×10e7 CD3/kg). MILs were infused on day +3. Patients in a CR were ineligible for this study. The best clinical response included 6CRs (27%), 10PRs (45%). Autologous GVHD which was observed in 32% of patients, was limited to the skin and required no treatment. Development of this syndome did not correlate with improved clinical outcomes. Lymphoid recovery was rapid with a mean absolute lymphocyte count (ALC) on day 15 of 886 cells/ul.

The laboratory immune monitoring studies were only performed in the bone marrow – the disease site and not in the peripheral blood. The percent CD3⁺ T cells in the BM was 13% at baseline with a CD4:CD8 ratio of 1:3. By day 60, BM T cell reconstitution was complete and showed an inversion of the CD4:8 ratio of 0.39 that persisted through day 360. At baseline, there were more CD4 effector memory (CD4_EM) (CD62L⁻/CD45RO⁺) than CD4 central memory (CD4_CM) (CD62L⁺/CD45RO⁺) (43% vs 27%). The CD4_CM population peaked on day 60 at 43% and persisted through 1 year, CD4_EM remained unchanged, and the CD4_Effector decreased from 19.8% to 5.6%. The CD8 subpopulations remained unchanged from pre-SCT to 1 year post-SCT. Treg numbers doubled from harvest to post-SCT consistent with previous studies showing a greater number of Tregs in healthy BM compared to MM-BM. Activated T cells (CD69⁺) doubled from pre- to post-SCT. IFNγ production in both CD4 and CD8 cells more than doubled compared to pre-SCT and was maintained at 1 year suggesting the persistence of the infused MILs.

The immune monitoring in the BM based on clinical responses revealed that patients achieving a CR/PR showed greater CD8 numbers at day +60 compared to stable disease (SD) or progressive disease (PD)(14.5% vs. 7.6%, respectively) and inversion of the CD4/CD8 ratio at 1 year (0.55 vs 1.26). In patients with SD or PD, the immune infiltrate in the BM was characterized by a large numbers of effector and effector-memory T cells and few CD8_CM at baseline. CR patients possessed the fewest CD8_Effector and the most CD8_CM with persistence of CD8_CM out to one year. These patients also showed a greater expansion in IFNγ cells as well as a greater amount of IFNγ production at each time point. Importantly, tumor-specific IFNγ production of CD3 cells in the BM was predictive of a clinical response. Patients achieving a CR showed twice the antigen-specific IFNγ production compared to all other groups at day 180 which persisted through 1 year. Analyses of additional immune subsets will be discussed in greater detail.

This is the first reported trial utilizing activated MILs as a source of tumor specific T cells. We have demonstrated the ability to effectively expand MILs ex vivo and provide an analysis of the parameters of immune reconstitution demonstrating that the clinical outcome correlated with the robustness of the immune response in the BM.