Tumor-Specific Donor Lymphocyte Infusion for Tumor Relapse after MHC-Haploidentical Hematopoietic Stem Cell Transplantation

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is a curative therapy for refractory hematologic malignancies such as leukemia and myelodysplastic syndrome. Tumor relapse, graft-vs host disease (GVHD) and infections are, however, major obstacles to successful allo-HSCT. Donor lymphocyte infusion (DLI) is carried out in some cases of tumor relapse and infections after allo-HSCT. DLI potentially induces or aggravates GVHD due to allo-reactivity of the lymphocytes used for infusion and the efficacy of DLI for tumor relapse is actually limited. Therefore, development of the method of DLI with enhanced anti-tumor effects without induction or aggravation of GVHD is expected to improve the results of DLI in the context of tumor relapse. Tumor antigen-specific T cell receptor (TCR)-expressing cell infusion is one of the potentially effective immunotherapies for refractory tumors. We previously reported that tumor antigen-specific TCR-gene-transduced human lymphocytes engineered with a novel retrovirus vector silencing endogenous TCRs showed reduced allo-reactivity (ASH annual meeting 2014). Utilizing this technology, we may be able to enhance anti-tumor effects without induction or aggravation of GVHD. In this study, we conducted a mouse pre-clinical model that mimics tumor-specific TCR-engineered DLI for tumor relapse after MHC-haploidentical HSCT and explored the effect of DLI on tumor growth and GVHD.

We performed experimental study utilizing the following model: MHC-haploidentical BALB/c(H-2^d)→CB6F1(H-2^b/d) HSCT, BALB/c-derived sarcoma CMS5a and T cells from TCR-transgenic mouse DUC18. CMS5a has a unique mutated form of a mitogen-activated kinase ERK2, and a nonamer peptide, called 9m, incorporating the resulting amino acid substitution is presented on H-2K^d and recognized by the CD8⁺ CTL clone C18. DUC18 is a BALB/c-background TCR transgenic line that expresses the rearranged Vα10.1 and Vβ8.3 genes of the C18. CD8⁺ T cells of DUC18 mice can suppress CMS5a tumor growth in the syngeneic BALB/c hosts. Allo-reactivity of DUC18 T cells was reduced as compared to that of wild type BALB/c T cells. Based on the results of preliminary experiments, we set standard experimental conditions as follows: X-ray irradiation (10 Gy) for pre-conditioning, 5x10⁶ whole bone marrow and 0.25x10⁶ T cell transplant (BMT) for mild GVHD induction, 1x10⁶ CMS5a inoculation for progressive tumor growth in BMT recipients, and 4x10⁶ CD8⁺ T cells for observation of the effect of DLI on tumor growth and GVHD. In some experiment, CD4⁺ T cells were co-infused with CD8⁺ T cells. DLI was carried out 3 days after CMS5a inoculation at 2, 4 or 8 weeks after BMT.

We first used CD8⁺ T cells from naïve BALB/c or DUC18 mice for DLI at 8 weeks after BMT. No GVHD aggravation was observed in either BALB/c or DUC18 CD8⁺ T cell recipients and, as expected, suppression of tumor growth was observed only in the DUC18 CD8⁺ T cell recipients. Next, DLIs using CD8⁺ T cells only, or both CD8⁺ and CD4⁺ (CD8⁺/CD4⁺) T cells from naïve mice were carried out at 4 or 2 weeks after BMT. No apparent effect of CD4⁺ T cell co-infusion on tumor growth and GVHD was observed in both types of donor T cell recipients when DLI was carried out at 4 weeks after BMT. However, GVHD was aggravated in CD8⁺/CD4⁺ T cell recipients from both types of donors when DLI was carried out at 2 weeks after BMT, and no CD4-driven additional anti-tumor effect was observed in DUC18 T cell recipients. We then performed DLI using in vitro activated/expanded T cells including both CD8⁺ and CD4⁺ as a more clinically relevant model. BALB/c and DUC18 splenocytes were activated and expanded in vitro and infused (adjusted CD8⁺ T cell number to 4x10⁶/mouse) into recipients at 2 or 8 weeks after BMT. Aggravation of GVHD was observed in either BALB/c or DUC18 activated/expanded T cell recipients when DLI was carried out at 2 weeks after BMT. On the other hand, it was not observed when DLI was carried out at 8 weeks after BMT. Anti-tumor effect of DUC18 T cells was observed regardless of timings of DLI.

Taken together, infusion of tumor-specific donor lymphocytes with reduced allo-reactivity for tumor relapse after MHC-haploidentical HSCT will be promising strategy. Timing of DLI and condition of recipients such as inflammatory status may affect GVHD. Multiple model studies are required for development of more effective DLI strategies without aggravation of GVHD.