Augmentation of Anti-tumor Immunity by Adoptive T-cell Transfer After Allogeneic Hematopoietic Stem Cell Transplantation

Marie Bleakley; Cameron J Turtle; Stanley R Riddell


Expert Rev Hematol. 2012;5(4):409-425. 

In This Article

Minimizing Toxicity of T-cell Transfer After Allogeneic HCT

The incorporation of strategies that employ gene transfer to derive more potent T cells in order to enhance anti-tumor activity suggests that toxicity may emerge as a significant problem both with autologous and allogeneic T-cell products.

On-target Toxicity

Serious adverse events have been observed after autologous T-cell therapy, primarily as a consequence of 'on-target' recognition of the tumor antigen by transferred T cells. Destruction of large tumor masses by autologous CD19-specific CAR-modified T cells has resulted in tumor lysis syndrome in patients with advanced CLL.[78,79] These patients also have prolonged and perhaps permanent depletion of normal B cells, the long-term consequences of which remain to be determined. Three patients developed severe colitis due to cognate interaction between transferred carcinoembryonic antigen TCR-modified autologous T cells and carcinoembryonic antigen expressed on colonic tissue,[122] and fatal pulmonary toxicity was reported in a patient in whom autologous ERBB2-specific CAR-modified T cells reacted with ERBB2 expressed on pulmonary epithelium.[123]

There is less experience transferring donor T cells after allogeneic HCT, however pulmonary toxicity was observed in three allogeneic HCT recipients after the infusion of minor H antigen-specific T cells. In one of these patients, the toxicity correlated with expression of the cognate antigen in lung epithelium.[51]

Off-target Toxicity

Transient, infusion-related toxicity that is independent of target antigen recognition is frequently reported after adoptive T-cell transfer, and typically consists of chills, fever and occasionally myalgias and transient hypoxia shortly after infusion. Infrequently, serious toxicities have been reported, including one death that was attributed to a culture-negative sepsis-like syndrome.[89,124] Transfer of T cells into allogeneic HCT recipients carries additional risks to those imposed by autologous T-cell transfer. Transfer of unselected polyclonal allogeneic T cells that express a diverse TCR repertoire carries the potentially serious risk of induction of severe GVHD. While GVHD is a possibility after any allogeneic T-cell infusion, the risk appears to be low when polyclonal T cells specific for viruses, such as EBV and CMV, are administered, in part due to their relatively restricted TCR repertoire compared with unselected TN or TM cells.[91,92] The strategy of selecting virus-specific T cells for transduction of a tumor-reactive CAR or TCR is feasible and should minimize the risk of GVHD due to allogeneic T-cell transfer when bone marrow or PBSC are the graft sources. Generation of gene-modified highly enriched virus-specific T-cell fractions from cord blood units, where the repertoire has not been primed to viral antigens, is likely to be more problematic.

Strategies to Eliminate Transferred T Cells

Toxicity after T-cell infusions and the potential risk of GVHD when T cells are used after allogeneic HCT suggest that strategies to eliminate the transferred T cells in the event of toxicity would be desirable. Conditional suicide genes that induce T-cell death in response to an appropriate trigger can be introduced into transferred T cells. A promising strategy involves modification of T cells to express human caspase-9 fused to a modified FK506-binding protein (FKBP). In vivo administration of a small molecule, AP1903, induces dimerization of the introduced caspase-9 (iCasp9), leading to activation of caspases-3, -6 and -7 and induction of apoptosis.[125] Proof of principle for this strategy was recently demonstrated in four patients who had undergone haploidentical HCT and received allodepleted donor T cells modified to express the iCasp9. AP1903 administration rapidly reversed grade I cutaneous GVHD that developed after HCT and normalized hyperbilirubinemia due to presumed hepatic GVHD within 24 h.[106] While these results are promising, it remains to be seen if more severe grade II–IV GVHD that develops after infusion of iCasp9-modified T cells can be efficiently abrogated by AP1903.

Herpes simplex type I viral thymidine kinase has also been used as a suicide gene; however, its use in allogeneic HCT recipients was problematic as depletion of T cells causing GVHD was incomplete.[126–128] This approach also precludes the administration of ganciclovir to treat or prevent CMV infection, and the transgene can be immunogenic.[129] A recently reported strategy for regulating T-cell survival after adoptive transfer involves incorporating in the transgene a truncated human EGF receptor (EGFRt), that lacks the intracellular signaling domain, and the portions of the extracellular domain for binding EGFRt ligands, but retains in the extracellular domain the epitope recognized by the pharmaceutical monoclonal antibody, Erbitux® that mediates both antibody-dependent cytotoxicity and complement-dependent cytotoxicity.[130] Thus, EGFRt enables antibody-mediated selection of transduced T cells in vitro and may be used for in vivo depletion of transduced T cells by administration of Erbitux. This strategy has not yet been subjected to testing in humans. The elimination of transferred T cells could be effective in treating toxicities such as GVHD that develop relatively slowly and may improve the safety profile of allogeneic T-cell infusions. However, some serious adverse events after adoptive T-cell therapy have occurred remarkably quickly, raising the concern that current suicide or antibody mediated depletion strategies may not act with sufficient rapidity to reverse toxicity.