Timing is everything: combining post-transplant adoptive cell therapy and tumor vaccines

Despite the potential for immune mediated graft-versus-tumor effects, relapse remains one of the main limitations of allogeneic hematopoietic stem cell transplantation (HCT). Several strategies targeting tumor antigens in the post HCT setting are under investigation in order to either prevent relapse or treat patients with early relapse. The Wilms tumor gene 1 encodes a zinc-finger transcription factor, Wilms tumor protein 1 (WT1) (1), which is important in cell growth and differentiation, and is broadly expressed in hematologic malignancies (2, 3), as well as a number of solid tumors. It has limited expression in normal adult tissues and therefore represents an interesting target for cancer immunotherapy. WT1 has also been identified as a prognostic factor and a marker for the detection of minimal residual disease in acute leukemia (4–6). Furthermore, graft-versus-leukemia (GVL) effects have been associated with detectable WT1-specific T cells after allogeneic HCT for acute lymphoblastic leukemia (ALL) (7). A number of investigational approaches are being used to target WT1 including peptide vaccines (8–10), dendritic cell (DC) vaccines (11, 12), ex vivo expanded WT1-specific T cells, gene modified T cells expressing a WT1-specific T cell receptor (TCR) (13), and TCR-like antibodies (14, 15). In this issue of the journal, Shah et al (12) report a pilot trial assessing the safety and feasibility of combining donor lymphocyte infusion (DLI) with WT1 peptide-loaded donor-derived DC vaccines in patients who relapsed after allogeneic HCT. DLI (dose at 1x10⁶ CD3/kg) was infused every 4 weeks for a total of three doses and a DC vaccine was injected once every 2 weeks, for a total of six vaccines. The DCs were loaded with three WT1 HLA A0201 restricted peptides (WT1 37–45, WT1 126–134, and WT1 187–195). Five patients with post-HCT relapse were enrolled and the three with ALL completed the full course of treatment. The other 2 (AML=1, Hodgkin=1) progressed during treatment and were taken off study. The treatment was well tolerated with only one case of grade 1-skin GVHD. The 3 patients who completed the treatment course demonstrated WT1-specific responses by ELISPOT, with two of them also having positive DTH responses to a WT1 peptide cocktail. All patients had progression of disease by the end of the study. This preliminary study demonstrates the safety and feasibility of post-HCT DLI and DC vaccination, with detection of immune responses in some of the patients, but no obvious clinical efficacy.

The lack of clinical efficacy raises some questions regarding potential future development of this approach. Many studies of peptide-loaded DC vaccines in the solid tumor setting have also shown variable clinical efficacy. There are potentially several mechanisms by which tumors can evade immune responses, including loss of specific antigen expression in the tumor, which is a risk when a limited number of epitopes are targeted (16). Another possibility is that the T cells generated by the DC-peptide vaccine, while showing functionality in vitro though interferon gamma expression, may lack polyfunctionality. Polyfunctional T cells, i.e. T cells that secrete multiple cytokines in response to antigens, have been shown to be critical in viral clearance in HIV infection, for example, and prior cancer vaccine studies in patients with melanoma have demonstrated the induction of polyfunctional T cells (17).

Another interesting aspect of the study by Shah is the generation of WT1-specific T cells after DLI. There is evidence, in at least some studies, that WT1-specific T cells can be detected pre-HCT and after DLI in the absence of specific immunization (3, 7). In particular, WT1-specific T cell responses were detected before T-cell depleted (TCD) allogeneic HCT in patients with multiple myeloma and increased in frequency after pre-emptive DLI beginning 5 months after HCT (3). Similar results have been reported in patients with ALL (7).

Finally, one of the critical questions for post HCT tumor immunization remains the identification of the ideal timing. It has long been thought that immune deficiency following HCT would prevent the generation of effective anti-tumor immunity. Nevertheless, preclinical studies as well as clinical trials have in fact shown that the lymphopenia early post HCT may actually provide a window for the recovering immune repertoire to be skewed towards tumor specific responses (18–21). In pre-clinical mouse models, sub-lethal irradiation followed by splenocyte infusion and DNA immunization with tumor vaccines resulted in significant increases in T cell responses to the tumor antigens and tumor eradication (18). This study also showed that optimal responses depended on both the timing of lymphocyte infusion and immunization of the irradiated host. Clinical trials in patients with myeloma undergoing autologous HCT followed by lymphocyte infusion and immunization support these pre-clinical findings (19, 20). In one study, patients with myeloma who were HLA-A0201 underwent immunization before and after autologous HCT transplantation with a multipeptide vaccine against human telomerase reverse transcriptase (hTERT) and survivin, in addition to receiving ex vivo stimulated autologous T cells two days after HCT (19). Immune responses to the tumor antigen vaccine were seen in 36% of patients. Additional studies of tumor immunization including DC vaccines are ongoing in patients with myeloma after autologous HCT. In the allogeneic HCT setting, studies of mouse preclinical models demonstrated that DLI followed by DNA immunization against tissue-restricted antigens after TCD HCT could induce potent antitumor responses without causing GVHD (21). The study by Shah is one of the first clinical studies in this setting (12).

Allogeneic HCT is an effective therapy in many patients with hematologic malignancies and graft-mediated anti-tumor effects can play a critical role in this setting. There is a growing body of pre-clinical and clinical data indicating that GVL effects may be further enhanced through adoptive cell therapy and immunization. The type of cell infused (unmodified DLI, gene-modified T cells, or antigen-specific expanded T cells) and vaccines (peptides, proteins, DNA, viruses, DCs), as well as the potential use of antibodies such as checkpoint inhibitors, present a large number of permutations that will need to be studied clinically (22). However, regardless of the combination selected, identifying the optimal timing for cell therapy and immunization in both autologous and allogeneic HCT will remain one of the critical elements to success.