Edited by: James L. Gulley, National Cancer Institute, USA
Reviewed by: Lokesh Jain, Food and Drug Administration, USA; Jacalyn Rosenblatt, Harvard Medical School, USA
*Correspondence: Douglas G. McNeel, Department of Medicine, 7007 Wisconsin Institutes for Medical Research, University of Wisconsin Carbone Cancer Center, 1111 Highland Avenue, Madison, WI 53705, USA. e-mail:
This article was submitted to Frontiers in Cancer Molecular Targets and Therapeutics, a specialty of Frontiers in Oncology.
This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in other forums, provided the original authors and source are credited and subject to any copyright notices concerning any third-party graphics etc.
While immune monitoring of tumor immunotherapy often focuses on the generation of productive Th1-type inflammatory immune responses, the importance of regulatory immune responses is often overlooked, despite the well-documented effects of regulatory immune responses in suppressing anti-tumor immunity. In a variety of malignancies, the frequency of regulatory cell populations has been shown to correlate with disease progression and a poor prognosis, further emphasizing the importance of characterizing the effects of immunotherapy on these populations. This review focuses on the role of suppressive immune populations (regulatory T cells, myeloid-derived suppressor cells, and tumor-associated macrophages) in inhibiting anti-tumor immunity, how these populations have been used in the immune monitoring of clinical trials, the prognostic value of these responses, and how the monitoring of these regulatory responses can be improved in the future.
Recent years have seen several exciting advancements in the development of active immunotherapy for the treatment of cancer. Sipuleucel-T, an active immunotherapy comprised of autologous dendritic cells (DC) pulsed with a fusion protein composed of granulocyte macrophage colony-stimulating factor (GM-CSF) and prostatic acid phosphatase (PAP), was shown to provide a significant increase in overall survival in patients with metastatic prostate cancer (Kantoff et al.,
As one of the central goals of tumor immunotherapy is to elicit and/or augment cytotoxic T-cell responses that can recognize and lyse tumor cells, the development of interim biomarkers of immunotherapeutic efficacy have largely focused on assays that measure these inflammatory, Th1-type anti-tumor responses. This has led to the near universal use of assays such as enzyme-linked immunosorbent spot (ELISPOT) assays, intracellular cytokine staining (ICCS), and HLA-peptide multimer analysis. However, as our understanding of the nature of the relationship between the tumor and immune response has matured, tumor immunologists have come to appreciate that these effector responses are only one aspect of the immune system that can impact anti-tumor immunity. The immune system (and the tumor itself) is also able to mount suppressive immune responses that target effector responses and can lead to the amelioration of anti-tumor responses. These suppressive immune responses are predominantly composed of regulatory T cells, myeloid-derived suppressor cells (MDSCs), and tumor-associated macrophages (TAM), which are able to survey the tumor microenvironment for effector immune responses to inhibit, which leads to the avoidance of anti-tumor immunity and further tumor growth. The monitoring of changes in regulatory immune responses, consequently, could theoretically serve as an additional biomarker of response to immune therapies, particular in the case of immune-modulating therapies or whole tumor vaccines where a specific antigenic target is unknown.
While there has been interest in monitoring suppressive immune responses following immunotherapeutic intervention, this has been a challenge given the difficulty in defining a set of cellular surface markers that can be used to easily identify and quantify regulatory immune responses. Furthermore, when evaluating antigen-specific vaccine approaches, there has been a noticeable paucity in the evaluation of antigen-specific regulatory responses following immunization. In addition, the intrinsic plasticity in regulatory immune function further complicates this analysis, as data in preclinical models indicates that immune responses can gain and lose suppressive activity depending on the microenvironment, which is particular important in the case of lymphocytes that infiltrate the suppressive tumor microenvironment. In this review, we will describe these regulatory cell populations, how they can suppress anti-tumor immune responses, how they have been used in the immune monitoring of clinical trials, and challenges associated with the implementation of regulatory cell detection into clinical trial immune monitoring.
Regulatory T cells (Tregs) are a subset of T lymphocytes identified in the early 1970s (Gershon and Kondo,
These regulatory responses are broadly broken up into either “natural” or “adaptive/induced” Tregs. Natural Tregs (nTregs) are produced by the thymus and constitutively express CD25, CTLA-4, and Foxp3 (a transcription factor that helps mediate the suppressive activity of this regulatory population), and are able to suppress both adaptive and innate immune responses (Read et al.,
As opposed to nTregs, induced Tregs (iTregs) enter the periphery as naïve T cells. However, rather than gain an effector phenotype, these iTregs encounter their specific MHC-peptide complex under conditions that promote the development of a regulatory phenotype, such as high levels of suppressive cytokines (Lohr et al.,
Inducible Tregs are further subdivided into Tr1, Th3, and Tr35 cells, which are loosely divided based on their mechanisms of suppression. Tr1 cells rely largely on IL-10 secretion to mediate suppression, are developed in the presence of high doses of IL-10, and express very low or no Foxp3 and CD25 (Groux et al.,
While these natural and induced regulatory T cells are conventionally viewed as CD4+ T cells, some of the earliest work into suppressive T cells identified that CD8+ T cells could also mediate the suppression of immunity
While research characterizing regulatory cells has made significant progress, one of the challenges that have come to light from these studies is the difficulty in defining a phenotype that can be used to reliably identify a regulatory T cell. Common markers used to identify Tregs are CD25, Foxp3, CD39, CD122, CD127, CTLA-4, LAG-3, and GITR (Mougiakakos et al.,
While research studying regulatory T cells has focused on antigen non-specific populations, emerging evidence has also shown a role for antigen-specific regulation in cancer. These antigen-specific Tregs require their cognate antigen to activate their suppressive activity; however, once active, these cells can suppress in an antigen non-specific fashion, so-called “bystander suppression” (von Herrath and Harrison,
Myeloid-derived suppressor cells are a diverse population of myeloid cells which have been shown to have the ability to suppress the proliferation and effector function of T cells. MDSCs consist primarily of immature myeloid cells and myeloid progenitor cells, cells which have not finished their differentiation into DCs, macrophages, or granulocytes. In healthy individuals, MDSCs represent a very small fraction of total peripheral blood mononuclear cells (PBMC), as these immature cells rapidly differentiate into mature cells. In a variety of malignancies, however, this differentiation process is blocked, leading to the generation of a sizable fraction of MDSCs. This is true in patients with many types of cancer, including lung, breast, colon, and melanoma, where patients have an increased frequency of peripheral and tumor-infiltrating MDSCs, and in some cases these frequencies correlate with disease grade (reviewed in Montero et al.,
Studies in mouse models have identified two categories of CD11b+ MDSCs based on their expression of the myeloid differentiation antigen Gr1 (which recognizes the Ly6G and Ly6C epitopes) – granulocytic MDSCs (CD11b+Ly6G+Ly6Clow) and monocytic MDSCs (CD11b+Ly6G−Ly6Chi). However, as humans lack a Gr1 homolog, the phenotypic characterization of human MDSCs has proved more complicated. While several surface molecules have been used to delineate MDSC subpopulations, common markers used to identify these subtypes include CD14 and CD15, with human granulocytic MDSCs being CD11b+CD33+CD14−CD15+ and monocytic MDSCs being CD11b+CD33+CD14+CD15−. Other markers that can be used in combination to identify MDSCs include CD13+, CD34+, IL-4Rα+, and HLA-DR− (Peranzoni et al.,
Myeloid-derived suppressor cells can mediate the suppression of effector immune responses using a variety of mechanisms (Gabrilovich and Nagaraj,
Macrophages are closely linked with the development of cancer-related inflammation. In the context of cancer, these cells are divided into either type 1 or type 2 macrophages. Type 1 macrophages (M1) have the ability to present antigens and activate T-cell responses, as well as being able to directly kill tumor cells. However, in the presence of Th2-biased cytokines such as IL-10, macrophages can be diverted to gain a type 2 phenotype. These immunosuppressive type 2 macrophages (M2) are marked by the expression of CD163 (the scavenger receptor) and CD206 (the mannose receptor), as well as traditional monocyte markers such as CD14, HLA-DR, and CD11b (Mantovani et al.,
Given the importance of regulatory immune responses in the development and progression of a variety of cancers, there has been interest in characterizing the effects of immunotherapies on the frequency of suppressive immune populations. Reports to date have predominantly focused on evaluating the effects of these therapies on regulatory T cells (particularly CD4+CD25+Foxp3+Tregs), though the effects of immunotherapies on MDSC frequency has begun to be implemented in clinical trial analyses. Most reports have focused on enumerating the frequency of Tregs following immunotherapy, with several immunotherapeutic approaches being shown to decrease the frequency of peripheral Tregs. This includes a report by Pohla et al. (
While many immunotherapies have been shown to alter the frequency of regulatory cells, these changes alone do not provide information on the immune and clinical efficacy of these therapies. However, as immunotherapeutic clinical trials have begun to elicit immune and clinical efficacy, it has become possible to determine how changes in regulatory populations correlate with the efficacy of these therapies, as shown in Table
Disease type | Immunotherapy | Cell population | Effects of immunotherapy on regulatory cells and responses | Reference |
---|---|---|---|---|
Glioblastoma | DC vaccine | Treg (CD4+CD25+CD127lo) | Decreased frequency of Tregs correlated with enhanced survival | Fong et al. ( |
CD4+CTLA-4+T cells CD8+CTLA-4+T cells | Decrease in CTLA-4 expression on CD4+ and CD8+T cells correlated with enhanced survival | |||
Malignant glioma | DC vaccine | Treg (CD4+CD25+CD127lo) | Decreases in Treg frequency correlate with increased survival | Prins et al. ( |
B-cell chronic lymphocytic leukemia | DC vaccine | Treg (CD4+CD25+Foxp3+) | Patients with clinical responses had a significant decrease in Treg frequency | Hus et al. ( |
Non-Hodgkin lymphoma | DC vaccine | Treg (CD4+CD25+Foxp3+) | Decrease in Treg frequency correlated with clinical responses | Di Nicola et al. ( |
Renal | DC vaccine+therapy | Treg (CD4+CD25+Foxp3+) | Non-responding patients had significantly higher expansion of Tregs compared to responding patients. | Schwarzer et al. ( |
Sarcoma | DC vaccine + irradiation | MDSC (CD11b+CD14− CD33+) | Higher frequencies of MDSC in non-responders | Finkelstein et al. ( |
Treg (CD4+CD25+Foxp3+) | No correlation between changes in Tregs and responder status | |||
Melanoma | Neoadjuvant ipilimumab | Treg (CD4+CD25hiFoxp3+) | Higher frequencies of Tregs correlated with enhanced progression-free survival | Tarhini et al. ( |
Monocytic MDSC (HLA-DRloCD14+) | No correlation between changes in MDSC and survival | |||
Melanoma | DC Vaccine + IL-2 | Treg (CD4+CD25hi) | Significant decrease in Tregs in patients with clinical responses | Bjoern et al. ( |
Treg (CD4+CD25hiFoxp3+) | No correlation between changes in Treg and clinical responses | |||
Melanoma | APC vaccines | Treg (D4+CD25+) | Expansion of Tregs correlated with decrease in CTL frequency | Chakraborty et al. ( |
Prostate | Viral vaccine | Treg (CD4+CD25hiFoxp3+) | Decrease in Treg function post-immunization correlated with enhanced prognosis, and increased Treg function correlated with poor prognosis | Gulley et al. ( |
Prostate | Viral Vaccine | Treg (CD4+CD25hiFoxp3+) | Decrease in Treg function post-immunization correlated with increased overall survival | Vergati et al. ( |
Effector:Treg ratio (CD4+CD25−: CD4+CD25+CD127-Foxp3+CTLA-4+) | Increased effector:Treg ratio post-immunization correlated with enhanced prognosis | |||
Prostate | Tumor cell vaccine + ipilimumab | Treg (CD4+CD25hiFoxp3+) | Increases in frequency of Tregs correlated with decreased overall survival | Santegoets et al. ( |
Effector:Treg ratio (CD4+CD45RO+: CD4+CD25hiFoxp3+) | Increases in effector: regulatory T cell ratio correlated with enhanced survival | |||
Lung, colorectal, gastric, breast, uterine, and renal cancer | Low-dose IL-2 | Treg (CD4+CD25+) | Patients with controlled disease have a decline in number of Treg cells | Lissoni et al. ( |
Effector:Treg ratio (CD4+:CD4+CD25+) | Patients with controlled disease have an increase in effector:Treg ratio | |||
Breast | Peptide vaccine | Treg (CD4+CD25+Foxp3+) | Decrease in Tregs correlated with enhanced effector immune responses | Gates et al. ( |
The observation that decreased regulatory cells following immunotherapy correlates with enhanced clinical responses is not wholly unexpected; as Tregs have been shown to correlate with more advanced disease and poorer prognosis in many disease types, it would be logical to conclude that a decrease in these Tregs would result in a better disease outcome following immunotherapy. However, this trend is not uniform; in a clinical trial report evaluating neoadjuvant ipilimumab in melanoma patients, the authors found that this treatment resulted in an increase in circulating Treg (both CD4+CD25hiFoxp3+ and CD4+CD25hiCD39+ T cells), and that increases in these Tregs correlated with enhanced progression-free survival (Tarhini et al.,
Another challenge associated with regulatory cell immune monitoring is that the techniques used to identify these populations can affect results. This is exemplified in the results from three studies evaluating ipilimumab in early stage clinical trials, each of which suggested different effects of ipilimumab on Treg frequency. In a Phase I trial in prostate cancer patients, treatment with ipilimumab was found to increase Treg frequency [as measured by circulating CD4+Foxp3+ T cells (Kavanagh et al.,
While an increase in circulating regulatory cells following immunotherapy is usually associated with a poor prognosis, and increased frequencies of regulatory cells in untreated individuals portends poor prognosis, the characterization of these populations as a prospective biomarker of immunotherapeutic efficacy remains relatively untested. Some reports have found that pre-existing Treg frequency does not correlate with vaccine efficacy one way or the other (Gulley et al.,
While this report found that elevated levels of CD4+CD25hiFoxp3+ Tregs predicted for poor prognosis, other clinical trials have found opposing results. In a report by Correale et al. (
While the enumeration of suppressive cell populations has been the central focus of regulatory immune monitoring, quantification alone may not be sufficient, as illustrated above. Immune monitoring efforts aimed at analyzing effector immune responses do not rely solely on quantifying the frequency of effector T cells – rather, they include functional analysis to determine the activity of these responses with respect to proliferation, cytokine expression, expression of cell surface molecules, and cytolytic activity. Similarly, immune monitoring of regulatory immunity should also include functional analysis of suppressive activity. This is particularly relevant to the analysis of regulatory immune responses, given the lack of distinct phenotypic markers that can be used to identify regulatory cells. The importance of analyzing regulatory function is clearly illustrated by the previously described reports evaluating a viral vaccine in prostate cancer patients, where immunization did not induce changes in Treg frequency that were associated with clinical responses, but decreases in Treg function were associated with an enhanced prognosis (Gulley et al.,
To evaluate the function of Tregs, most of the studies to date have largely focused on one of two aspects of Treg activity: suppression of T-cell proliferation and Treg expression of immunosuppressive cytokines. To measure Treg suppression, peripheral blood cells are sorted to isolate a purified Treg population (usually based on CD4+CD25+ expression), and these cells are then co-incubated with autologous CD4+CD25− conventional T cells that are non-specifically activated. T-cell proliferation can then be measured using standard techniques, including thymidine uptake, or by dilution of cell-labeling dyes such as carboxyfluorescein succinimidyl ester or PKH26, in CD4+CD25− cells. These assays can also be combined with assays measuring the effect on expression of cytokines by the effector T cells, collecting supernatants and measuring cytokine secretion by enzyme-linked immunosorbent assay (ELISA).
In addition to measuring the effect of Treg on the proliferation and cytokine secretion of effector T cells, another common assay of Treg function is measuring the cytokine expression by Tregs themselves. This has typically been performed by ELISA (where isolated Treg populations are stimulated non-specifically and cytokine release into the supernatant is measured) or ICCS. Intracellular cytokine staining has the benefit of not requiring isolation of Tregs, as whole PBMC can be isolated and stimulated, following by surface staining to identify the population of interest. Furthermore, ICCS also has the benefit of permitting concurrent staining for Foxp3, helping to identify Treg populations. While suppression assays and cytokine expression are most commonly used to evaluate Treg function, other methods employed to monitor Treg function in clinical trials include evaluating serum cytokine levels, evaluating tumor biopsies for expression of suppressive factors such as IDO, or evaluating the methylation status of the Foxp3 promoter as a surrogate for Treg activity (Polansky et al.,
A fairly comprehensive clinical characterization of regulatory T cell activity was reported by François and colleagues, in which melanoma patients were immunized with a MHC class II peptide (Francois et al.,
As a measure of antigen-specific regulation not requiring
The tvDTH assay is also useful in that it can be used to detect effector responses that are suppressed by concurrent regulatory responses. As we have described in a report evaluating a DNA vaccine in patients with prostate cancer, we found that we were not able to detect antigen-specific effector responses in peripheral blood samples from multiple patients when the antigen of interest was injected into the footpads of SCID mice alone. However, when we blocked regulatory responses using antibodies specific to CTLA-4, we were able to uncover antigen-specific effector responses that were otherwise undetectable (Olson et al.,
As immunotherapies for the treatment of cancer begin to show clinical benefit and become approved for use in the clinic, it is of crucial importance to identify biomarkers of efficacy that can also be incorporated into the clinic, both to identify which patients may optimally respond to therapy, as well as to determine whether individual patients are responding to therapy. This is particularly relevant because most immunotherapy clinical trials have relied on measuring overall survival as a primary clinical endpoint, and while overall survival remains the gold-standard for determining clinical efficacy, it is an impractical endpoint for making clinical treatment decisions. In addition, many treatments rely on repetitive administration (vaccines, for example), and it is conceivable that tolerance/regulation may be elicited that can prevent the generation of productive anti-tumor effector responses. As such, it is critical to monitor for these regulatory responses to determine if/when additional booster immunizations or other immune-modulating agents should be employed.
As our understanding of immune suppression has expanded from the identification of regulatory cell populations to current research aimed at elucidating the plastic nature of immune function and the interplay between effector and regulatory immunity within the tumor microenvironment, it has become evident that regulatory cells play a central role in the development and progression of cancer, and can influence the outcome of tumor immunotherapies. Therefore, it is important to include analysis of these regulatory populations in the immune monitoring of clinical trials, as it can complete the picture of how immune responses are affected by immunotherapeutic intervention. Furthermore, as our understanding of how these regulatory responses are affected by immunization develops, it will be possible to design more optimal combinatorial approaches that seek to activate effector responses as well as inhibit or deplete these suppressive cells.
However, as our understanding of regulatory cells continues to expand, it is important that immune monitoring efforts tracking these cell populations continue to grow as well to address the current challenges associated with the monitoring of regulatory populations. This includes identifying combinations of phenotypic markers that can be used to more reliably track suppressive populations, or alternatively using multiple definitions of regulatory cells to confirm results obtained by examining a single phenotype. It will also be important to incorporate the use of functional analysis of regulatory cell function as is done with effector cells, as this can provide a more complete picture of both the quantity and quality of suppressive responses. Additionally, as our knowledge of the plastic nature of suppressive activity in nominally non-regulatory immune cells expands, it will be important to incorporate this information into immune monitoring, which can help expand this monitoring from defining a single moment in time to generating a more complete understanding of the suppressive potential of the immune response following immunotherapeutic intervention. Equally important, it will be crucial to gain a better understanding for how pre-existing regulatory responses affect the ability to respond to immunotherapy, as this can be used to prospectively identify individuals who are most likely to respond to therapy.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
This work was supported for Brian M. Olson and Douglas G. McNeel by NIH (R01 CA142608), by the US Army Medical Research and Materiel Command Prostate Cancer Research Program (W81XWH-11-1-0196), and by the University of Wisconsin Carbone Cancer Center.
CTLA-4, cytotoxic T lymphocyte-associated antigen 4; DC, dendritic cell; ELISA, enzyme-linked immunosorbent assay; ELISPOT, enzyme-linked immunosorbent spot; GM-CSF, granulocyte macrophage colony-stimulating factor; ICCS, intracellular cytokine staining; IDO, indoleamine 2,3-dioxygenase; IL, interleukin; iNOS, inducible nitric oxide synthase; iTreg, induced Treg; MDSC, myeloid-derived suppressor cell; NO, nitric oxide; nTreg, natural Treg; PAP, prostatic acid phosphatase; PBMC, peripheral blood mononuclear cells; ROS, reactive oxygen species; TAM, tumor-associated macrophage; TGF, Transforming growth factor; Treg, Regulatory T cell; tvDTH, trans vivo delayed-type hypersensitivity.