A phase I/II minor histocompatibility antigen-loaded dendritic cell vaccination trial to safely improve the efficacy of donor lymphocyte infusions in myeloma.

Allogeneic stem cell transplantation (allo-SCT) with or without donor lymphocyte infusions (DLI) is the only curative option for several hematological malignancies. Unfortunately, allo-SCT is often associated with GvHD, and patients often relapse. We therefore aim to improve the graft-versus-tumor effect, without increasing the risk of GvHD, by targeting hematopoietic lineage-restricted and tumor-associated minor histocompatibility antigens using peptide-loaded dendritic cell (DC) vaccinations. In the present multicenter study, we report the feasibility, safety and efficacy of this concept. We treated nine multiple myeloma patients with persistent or relapsed disease after allo-SCT and a previous DLI, with donor monocyte-derived mHag-peptide-loaded DC vaccinations combined with a second DLI. Vaccinations were well tolerated and no occurrence of GvHD was observed. In five out of nine patients, we were able to show the induction of mHag-specific CD8+ T cells in peripheral blood. Five out of nine patients, of which four developed mHag-specific T cells, showed stable disease (SD) for 3.5-10 months. This study shows that mHag-based donor monocyte-derived DC vaccination combined with DLI is safe, feasible and capable of inducing objective mHag-specific T-cell responses. Future research should focus on further improvement of the vaccination strategy, toward translating the observed T-cell responses into robust clinical responses.

Efficacy of host-dendritic cell vaccinations with or without minor histocompatibility antigen loading, combined with donor lymphocyte infusion in multiple myeloma patients.

Donor lymphocyte infusions (DLI) can induce durable remissions in multiple myeloma (MM) patients, but this occurs rather infrequently. As the graft-versus-tumor (GvT) effect of DLI depends on the presence of host-dendritic cells (DCs), we tested in a phase I/II trial whether the efficacy of DLI could be improved by simultaneous vaccination with host-DCs. We also analyzed the possibility of further improving the GvT effect by loading the DCs with peptides of mismatched hematopoietic cell-specific minor histocompatibility antigens (mHags). Fifteen MM patients not responding to a first DLI were included. Eleven patients could be treated with a second equivalent dose DLI combined with DC vaccinations, generated from host monocytes (moDC). For four patients, the DC products did not meet the quality criteria. In four of the treated patients the DCs were loaded with host mHag peptides. Toxicity was limited and no acute GvHD occurred. Most patients developed objective anti-host T-cell responses and in one patient a distinct mHag-specific T-cell response accompanied a temporary clinical response. These findings confirm that DLI combined with host-DC vaccination, either unloaded or loaded with mHag peptides, is feasible, safe and capable of inducing host-specific T-cell responses. The limited clinical effects may be improved by developing more immunogenic DC products or by combining this therapy with immune potentiating modalities like checkpoint inhibitors.

The impact of circulating suppressor cells in multiple myeloma patients on clinical outcome of DLIs.

Allo-SCT followed by DLIs can establish long-term remissions in multiple myeloma (MM) patients. In many patients, however, the immunotherapeutic graft-versus-tumor (GVT) effect is moderate and not sustained, implying that immune suppression is mediated, among other factors, by regulatory T cells (Tregs) or myeloid-derived suppressor cells (MDSCs). Towards a better understanding and, eventually, manipulation of the immune-regulatory mechanisms in transplanted MM patients, we retrospectively sought a correlation between DLI outcome and circulating CD14(+) MDSCs, CD14(-) MDSCs and Tregs in 53 MM patients before their first DLI. We found significantly elevated frequencies of highly suppressive CD14(+) MDSCs, CD14(-) MDSCs and Tregs in pre-DLI samples from patients. Higher frequencies of Tregs, but not of MDSCs, were significantly associated with non-responsiveness to DLI. Furthermore, a lower frequency of Tregs predicted the development of chronic GVHD, which, in turn, displayed a high association with GVT. Elevated Treg frequencies before DLI were also associated with significantly shorter PFS and OS. Hence, our data reinforce the idea of active suppression of antitumor responses by Tregs in MM patients and therefore suggest that targeting patient Tregs before DLI may improve outcome of DLI.