CD4⁺ T helper cell responses to NY-ESO-1 tumor antigen in ovarian cancer resist perversion into immunosuppressive Tregs

Abstract

In a recent study, we have demonstrated that T helper type 1 (T_H1) cells specific for the tumor antigen NY-ESO-1 are amplified at ovarian tumor sites but are not “perverted” into immunosuppressive FOXP3⁺ regulatory T cells (Tregs). These findings encourage the development of protocols aiming to eliminate, or inactivate, FOXP3⁺ Tregs and reinforce Type I anticancer immunity, to improve clinical outcomes.

Among cancers exclusive to women, ovarian cancer has the highest mortality rate.¹ This malignant disease has been shown to be naturally immune reactive, being frequently infiltrated by immune cells, including T lymphocytes, but is also prone to develop a suppressive environment that blunts successful cancer immune eradication. Tumors are infiltrated by different T cell subsets, including helper and effector immune cells, as well as regulatory and suppressive immune cells, the balance of which may dictate opposing immune effects. Thus, whereas tumor infiltration by tumor antigen specific T helper type 1 (T_H1) and cytotoxic T lymphocytes (CTL) in ovarian cancer has been associated with favorable clinical outcomes, tumor infiltration by FOXP3⁺ CD4⁺ regulatory T cells (Tregs) is believed to dampen spontaneous anticancer immunity and to be associated with poor survival.^2,3 However, because the antigen specificity of tumor associated FOXP3⁺ Tregs is largely unknown, a major question is if, and to which extent, the specificity of FOXP3⁺ Tregs overlaps with that of spontaneous T_H cell responses to immunogenic tumor antigens.

In a previous study, we demonstrated that in ovarian cancer patients bearing antigen expressing tumors, circulating CD4⁺ T cells specific for the frequently expressed and highly immunogenic human tumor antigen NY-ESO-1 are prevalently of T_H1 type and not FOXP3⁺ Tregs.⁴ In another study aimed at characterizing ovarian cancer associated FOXP3⁺ Tregs, we have additionally shown that a major subset is represented by CXCR3⁺ T-bet⁺ Tregs that accumulate at tumor sites in proportions directly correlating with those of co-existing conventional T_H1 cells, both seemingly attracted by common chemokines.⁵ These results are consistent with the possibility that tumor antigen-specific T_H1 cells trafficking in tumors may be, at least partially, “perverted” into FOXP3⁺ “peripherally induced” Tregs, on site, due to the highly immunosuppressive environment generally present within these tumors. However, we did not favor this hypothesis, as we found that most ovarian cancer associated Tregs express Helios, a transcription factor that has been proposed to identify thymically-derived Tregs, called “natural”, rather than induced Treg. Nevertheless, without experimental evidence, we could not formally exclude the possibility of immunostimulatory T_H1 perversion into immunosuppressive Tregs.

In our recent study published in Cancer Immunology Research, we have used MHC Class II tetramers incorporating an immunodominant CD4⁺ T-cell epitope from NY-ESO-1 (MHC Class II/NY-ESO-1 tetramers) to directly address this question.⁶ The answer is indeed highly relevant to the development of strategies aimed at depletion of Treg and concomitant amplification of effector T cells. With the MHC Class II/NY-ESO-1 tetramers, that we previously developed and validated,⁷ we have, for the first time, simultaneously compared by direct ex vivo staining, quantitatively and phenotypically, NY-ESO-1 specific CD4⁺ T cells in circulating lymphocytes, tumor-associated ascites and solid tumor masses from ovarian cancer patients with spontaneous immune responses to the antigen (Fig. 1). We found that the frequency of NY-ESO-1 tetramer⁺ cells detected ex vivo was highly enriched in tumors, both in ascites (10-fold) and, to a greater extent (100-fold), in solid tumor masses, as compared with peripheral blood lymphocytes. At tumor sites, MHC Class II/NY-ESO-1 tetramer⁺ cells were detected concomitantly with high proportions (up to 50% of total CD4⁺ T cells) of FOXP3⁺ Tregs, but were clearly distinct from the latter and displayed characteristics of T_H1 effectors. Thus, even in the presence of high proportions of FOXP3⁺ Tregs, tumor antigen-specific T_H1 cells can accumulate in ovarian tumors, maintain an effector phenotype and do not seem to be easily “perverted” into Tregs. Since tumor-associated T_H1/effectors and FOXP3⁺ Tregs do not seem to have overlapping antigen specificities, our findings overall support the development of approaches aiming at eliminating or inactivating FOXP3⁺ Tregs at sites of ovarian tumors. Therefore, combinatorial treatment with Treg depleting agents, such as cyclophosphamide, anti-CD25 antibodies or denileukin diftitox, while simultaneously amplifying tumor antigen-specific T_H1 type responses, such as by adoptive transfer therapy or vaccination,⁸ should reinforce anticancer immunity and improve clinical outcomes.

Figure 1.

NY-ESO-1-specific CD4⁺ T cells accumulate at tumor sites and remain distinct from co-existing Tregs. Comparative analysis of CD4⁺ T cells, isolated from peripheral blood mononuclear cells (PBMC), tumor ascites and solid tumor masses of ovarian cancer patients, stained ex-vivo with NY-ESO-1/MHC Class II tetramers together with monoclonal antibodies (mAb) specific for CD25 and CD127 shows that tetramer⁺ cells accumulate at tumor sites but remain distinct from regulatory T cells (Tregs) present at high proportions in tumors.

Despite these encouraging data, numerous questions remain concerning the relevance of T-cell responses in ovarian cancer, particularly with respect to their impact on clinical outcomes. Both the initially reported correlation between the presence and frequency of CD3⁺ infiltrating T cells³ or CD4⁺ CD25⁺ Treg,² as well as that of intraepithelial CD8⁺ T cells⁹ and survival, have been questioned by more recent studies that have suggested that the ratio between tumor-infiltrating CD8⁺ T cells and CD4⁺ FOXP3⁺ or FOXP3⁻ T cells, rather than their absolute numbers, would better correlate with survival.¹⁰ Whether a certain subtype of CD4⁺ T_H cells, or other factors, are required for CD8⁺ T cells to efficiently accumulate in ovarian cancers remains as yet undefined. Thus, more studies are required to address the correlation, not only between clinical outcomes and the net quantity of T_H/effectors and FOXP3⁺ Tregs in ovarian cancers, but also take into account their subsets, the presence and frequency of other suppressor subsets and attempt to quantify and characterize the tumor-specific CD8⁺ T-cell components. In this respect, the study of spontaneous immune responses to well-defined, tumor-specific antigens will likely continue to be highly instrumental to clarify the role of cancer targeting T cells in the clinical outcome of human cancer, particularly ovarian carcinoma.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.