Depletion of alloreactive T cells: which cell?

Comment on Martins et al, page 3429

The bulk of alloreactivity in human hematopoietic stem cell products resides in a CD40^hi activation antigen–positive T-cell population, the specific depletion of which may abrogate graft-versus-host disease while retaining reactivity both to third-party antigens and, potentially, to the graft-versus-tumor effect.

Severe graft-versus-host disease (GVHD) remains the main limitation to mismatched donor stem cell transplants. While extensive depletion of donor T cells from the graft can completely eliminate GVHD, significant concerns remain over the consequent profound and long-lasting T-cell immunodeficiency after transplantation and potential loss of the graft-versus-tumor (GVT) effect. For over a decade, therefore, there has been increasing interest in depleting alloreactive T cells while retaining a significant number of functional T cells capable of reacting to third-party antigens. While several systems have been investigated, the optimal alloreactive T-cell target for these approaches remains poorly characterized. Most approaches have targeted all cells expressing activation antigens (eg, CD25 or CD69) following stimulation in mixed lymphocyte culture (MLC).^1-4  It remains uncertain, however, whether such strategies are optimal. It is possible that reactive T-cell populations capable of causing GVHD and expressing activation antigens other than CD25 or CD69 may escape depletion. Alternatively, depletion of all T cells expressing activation antigens may result in depletion of other T cells not involved in the initiation of GVHD. Finally,FIG1 concern still remains that depletion of all alloreactive cells will result in loss of GVT effects.

In this issue of Blood, Martins and colleagues present important data that significantly advance our knowledge of the phenotype of alloreactive T cells. Using cytokine flow cytometry assays to study the functional characteristics of cells responding to stimulation in MLCs, these investigators show that the bulk of human alloreactivity resides in the CD4⁺ T-cell population. Following prolonged allogeneic stimulation in MLCs, only a subpopulation of CD4⁺ T cells, with up-regulated expression of CD4 (CD4^hi T cells), demonstrates activation antigen expression and effector cytokine production. Selective depletion (≥ 1-2 logs) of CD4^hi T cells (coexpressing the activation antigen CD38) by flow sorting abolishes alloreactivity in secondary MLCs while retaining third-party antigen reactivity and preserving a diverse T-cell receptor repertoire in residual T cells. Importantly, depletion of CD4^hiCD38⁺ cells also results in loss of effector cytokine production by residual allospecific CD8⁺ T cells. The important implication of this work is that the selective depletion of alloreactive CD4⁺ T cells could be a more effective strategy than previous attempts that targeted all activated T cells following MLC; preservation of antigen-specific CD8⁺ T cells could also preserve a GVT effect. While indirect clinical evidence supports the hypothesis that CD4⁺ alloreactive T-cell depletion will be effective in eliminating GVHD,⁵  the possibility of preserving GVT effects by residual allospecific CD8⁺ cells remains to be tested.

In addition to defining the clinical significance of residual allospecific CD8⁺ T cells in the graft, other questions remain to be answered. For example, are there CD4^hi T-cell populations, coexpressing activation markers other than CD38, capable of mediating GVHD that should also be depleted? Is a 1-2 log depletion of alloreactive CD4^hiCD38⁺ T cells sufficient to eliminate clinical GVHD? What is the dose of nonalloreactive T cells required for consistent improvement in immune function after transplantation? If significantly more than a modest dose (10⁶cells/kg) is required, even high-speed flow sorting may not be a clinically feasible depletion method. Continued research into defining the optimal alloreactive T-cell population, such as that of Martins and colleagues, together with improvements in cell selection technology, will lead to the realization of safer transplants for many patients who do not have matched donors readily available.