Cancer Immunology, Immunotherapy

, Volume 61, Issue 2, pp 239–247

Immunological monitoring of the tumor immunoenvironment for clinical trials

Authors

    • Laboratory of Cell-Mediated ImmunitySAIC-Frederick, Inc.
  • Susan L. Strobl
    • Laboratory of Cell-Mediated ImmunitySAIC-Frederick, Inc.
  • Michael R. Shurin
    • Department of Pathology, Division of Clinical ImmunopathologyUniversity of Pittsburgh Medical Center
Focussed Research Review

DOI: 10.1007/s00262-011-1148-6

Cite this article as:
Malyguine, A.M., Strobl, S.L. & Shurin, M.R. Cancer Immunol Immunother (2012) 61: 239. doi:10.1007/s00262-011-1148-6

Abstract

Monitoring of immunotherapeutic clinical trials has undergone a considerable change in the last decade resulting in a general agreement that immune monitoring should guide the development of cancer vaccines. The emphasis on immune cell functions and quantitation of antigen-specific T cells have been playing a major role in the attempts to establish meaningful correlations between therapy-induced alterations in immune responses and clinical endpoints. However, one significant unresolved issue in modern immunotherapy is that when a tumor-specific cellular immune response is observed following the course of immunotherapy, it does not always lead to clinically proven cancer regression. This disappointing lack of a correlation between the tumor-specific cytotoxic immune responses and the clinical efficacy of immunotherapy may be explained, among other reasons, by the notion that the analysis of any single immunological parameter is not sufficient to provide clinically feasible information about the complex interactions between different cell subsets in the peripheral blood and immune, tumor, and stromal cells in the tumor milieu. By contrast, a systemic approach is required for improving the quality of a serial monitoring to ensure that it adequately and reliably measures potential changes induced in patients by administered vaccines or immunomodulators. Comprehensive evaluation of the balance between the immunostimulatory and immunosuppressive compartments of the immune system could be critical for a better understanding of how a given immunotherapy works or does not work in a particular clinical trial. New approaches to characterize tumor-infiltrating leukocytes, their phenotypic, biochemical, and genetic characteristics within the tumor microenvironment need to be developed and validated and should complement current monitoring techniques. These immune-monitoring assays for the local tumor immunoenvironment should be developed, validated, and standardized for reliability and consistency in order to establish the overall performance standards.

Keywords

ImmunomonitoringVaccine clinical trialsTumor immunoenvironmentLeukocytesCITIM2011

Introduction

The last decades have been characterized by substantial progress in our understanding of the role of the immune system in tumor progression. There are now a large number of examples of how the immune system is able to recognize tumor antigens and eliminate or control tumor cell growth and spreading. As a result, we have learned how to manipulate the immune system to generate measurable tumor-specific immune responses. Recently, Sipuleucel-T vaccine (Provenge™, Dendreon Corporation, Seattle, WA) was approved by the FDA for the treatment of patients with castration-resistant metastatic prostate cancer as the first therapeutic cancer vaccine in humans. In addition, monoclonal antibody ipilimumab (Yervoy, Bristol-Myers Squibb) was approved earlier this year for the treatment of advanced melanoma as a second-line therapy. This raises high expectations among scientists and the general public that immunotherapy may provide further breakthroughs in cancer treatment.

Immune response profiling and monitoring are the key elements in the development of new biotherapies and a variety of assays that have been introduced for assessing different parameters of the immune status. However, one significant unresolved issue is the fact that when tumor-specific cellular immune responses are measured following the course of immunotherapy, the responses do not always correlate with clinically proven cancer regression. Comprehensive evaluation of the balance between the immunostimulatory and immunosuppressive compartments of the immune system in circulation and in the tumor immunoenvironment could bring new insights in our understanding of the prognostic and monitoring significance of accessible immune response associated with cancer therapy.

Systemic immunomonitoring: cell-mediated cytotoxicity

The choice of immune assays for a given clinical trial depends on the known immunomodulating mechanisms of a tested therapeutic modality. Therefore, selection of monitoring methods for an appropriate assessment of cell-mediated cytotoxicity, representing the key mechanism of the immune responses against various pathogens and tumor, is thought to be crucial for revealing potential correlations between the clinical and immunologic responses during and after immunotherapy. Cytotoxic CD8+ T cells (CTL) and natural killer (NK) cells employ two major contact-dependent cytotoxic pathways. The first is the exocytosis of lytic granules by cytotoxic effector cells, which comprise a pore-forming toxin (perforin) and pro-apoptotic serine proteases (granzymes), which synergistically kill target cells by activating various lytic pathways [1]. The second pathway is the production by the effector cells of molecules from the TNF family, such as TNF-α, FasL, and TRAIL, that induce multimerization of their cognate receptors on target cells resulting in apoptosis induction [2]. The most popular assay for evaluating cell-mediated cytotoxicity has been the 51Cr-release assay, which is considered to be the “gold standard”. However, while the assay benefits from being reproducible and relatively easy to perform, it has several drawbacks. It provides only semiquantitative data unless it incorporates a limiting dilution component, has a relatively low level of sensitivity, and some tumor-cell targets label poorly and produce high spontaneous release of isotope.

Assays that can monitor both CTL frequency and function, such as the IFN-γ enzyme-linked immunospot assay (ELISPOT), have gained increasing popularity for monitoring clinical trials and in basic research. Results from various clinical trials, including peptide and whole tumor cell vaccination and cytokine treatment, showed the suitability of the IFN-γ ELISPOT assay for monitoring T-cell responses [35]. However, the Granzyme B (GrB) ELISPOT assay may represent a more direct analysis of cell-mediated cytotoxicity compared with the IFN-γ ELISPOT, since GrB is a key mediator of target cell death via the granule-mediated pathway [6, 7]. In our study of melanoma patients vaccinated with gp100:209M peptide, we compared the activity of patients’ PBMC in the GrB ELISPOT with cell activity in the tetramer, IFN-γ ELISPOT, and 51Cr-release assays. Reactivity in the GrB ELISPOT was more closely associated with cytotoxicity in the 51Cr-release assay than the tetramer or IFN-γ ELISPOT assays. Moreover, the higher affinity g209-2M peptide elicited greater GrB secretion than the native g209 peptide, while this difference was not observed with IFN-γ secretion [8]. These results show that simultaneous use of the GrB ELISPOT assay with other immunological assays may provide important additional immunological insights into patient’s responses to cancer vaccines.

Obviously, assays that allow simultaneous estimation of several parameters are more useful for clinical immunomonitoring. In this respect, the flow cytometric assays that enable detection and enumeration of tumor-specific CTL and their specific effector functions provide new addition for the analysis of cell-mediated killing. Recently, we have developed a 3-color flow cytometric assay based on the detection of three groups of parameters: (1) degranulation marker of the activated effector cells, CD107a, (2) marker of early apoptosis in target cells, Annexin V binding, and (3) marker of cell membrane permeability, 7-AAD [9]. Furthermore, using both experimental models [10] and clinical samples from melanoma patients [11], we have demonstrated an excellent correlation between CD107a expression and GrB secretion, as well as between Annexin V positivity of target cells and specific cell lysis assessed in the 51Cr-release assay. Also after vaccination, an increase in both effector cell degranulation and target cell death could be determined by flow assay, when target cells were pulsed with g209. Thus, this flow-based assay detects the cytolytic effector cell activation, frequency, phenotype, and the target cell death in the same sample.

Although many novel tools are now available to detect immune responses against known and unknown tumor antigens, including MHC-tetramers, cytokine release/catch assays, ELISPOT assays, and flow cytometric assays, one significant problem remains unresolved. Clinical studies demonstrated that the therapy-induced tumor-specific immune responses do not always correlate with clinical responses regardless of the generation of tumor-specific cytotoxic lymphocytes recognizing and efficiently killing tumor cells ex vivo. For instance, our data revealed that 75% of peptide-vaccinated melanoma patients demonstrated specific immune response to vaccination [8]. However, according to the review of 1,306 vaccine treatments conducted in the Surgery Branch of NCI and several non-NCI trials, only 3.3% of patients demonstrated objective responses based on standard oncologic reporting criteria [12]. Recently, analyzing 936 patients with different types of solid cancers in vaccine trials, Klebanoff et al. [13] reported only 34, i.e., 3.6% objective responses. Another example is our recent analysis of more than 20 clinical trials, in which at least 396 patients with various cancers were treated with genetically engineered dendritic cells (DC) [14]. 272 patients were tested for tumor-specific CD8+ T-cell response to vaccination. Increased specific CD8+ T-cell frequency/activity was detected in 201 patients (74%), and increased NK activity was revealed in 19 of 24 evaluated patients (79%). However, even with these immunological responses, clinical results have been disappointing: from 333 analyzed patients, disease stabilization was observed only in 10.8% of patients and survival benefits were seen only in 7.5% of patients. Overall, only 8 from all 333 patients (2.4%) demonstrated a response by standard tumor measurement criteria. Other groups have shown that there was no difference in the levels of anti-tumor antigen-specific T cells in patients who recurred compared with those who remained disease-free, which suggests that the mere presence of profoundly expanded numbers of vaccine-induced antigen-specific T cells cannot by themselves be used as a “surrogate marker” for vaccine efficacy [15]. In a recent clinical trial with vaccinated glioma patients, 74% of patients demonstrated specific responses against targeted glioma-associated antigens, but only two from 19 patients (9%) had an objective clinical response, with both being non-responders in an ELISPOT assay [16]. Ribas et al. [17] reported that from 33 melanoma patients vaccinated with DC vaccine, only three patients (9%) had an objective clinical response, while 90% from 29 tested patients had detectable tumor antigen-specific T cells in the peripheral blood. In a multicenter study by Schwartzentruber et al. [18], 185 melanoma patients were treated with IL-2 or IL-2 and gp100 peptide vaccine, and no correlation between the anti-peptide reactivity and the objective clinical responses was found. These and other data demonstrated a lack of correlation between the tumor-specific cytotoxic immune responses and objective clinical responses to immunotherapy, suggesting that currently used immunomonitoring protocols require serious reconsiderations, improvements, or modifications.

Many reasons may be responsible for the lack of correlation between the immune and clinical response to cancer immunotherapy and have been extensively discussed [19, 20]. Among them are the following: (1) lack of agreement on measurable immunological parameters and assays, (2) inaccurate reflection of the in vivo immune responses by the immunological assays used, e.g., by cytotoxicity assays with lengthy in vitro stimulation of isolated T cells, (3) lack of uniform standards for immunological monitoring in clinical trials, (4) inability of common immunoassays to consider changes in T-cell differentiation, the antigenic profile of tumors and responding T cells, or “private” antigen responses, (5) failure to characterize the complexity of the immune responses in the tumor environment, and (6) apparent limitations of the assays utilizing only the peripheral blood components for immunomonitoring of the immune system in patients. Based on the current knowledge and a growing body of evidence, the last two problems should attract a special attention and interest as being readily doable and feasible for modern clinical immunology laboratories. First of all, the cytotoxic immune response to vaccination represents an extremely complex array of interactions between immune, tumor, and stromal cells, and therefore cannot be realistically evaluated based on a single immunological parameter or a cell type. Second, typically, only the peripheral blood components are accessible for serial analysis when in fact there is no convincing evidence that the peripheral blood immune responses are representative of those occurring at the site of the tumor.

Systemic immunomonitoring: regulatory cells

Immune responses against cancer, including those induced by vaccination, depend on a balance between functional activity of various subsets of effector and suppressor T cells. In an immunocompetent cancer patient, the immune system may suppress effector cell attack against tumor antigens, especially in the tumor microenvironment. The suppressive compartment of the immune system includes several heterogeneous subsets of immune cells, including regulatory T cells (Treg), myeloid-derived suppressor cells (MDSCs), alternatively activated (M2) or regulatory subsets of tumor-associated macrophages (TAMs), pro-tumorigenic neutrophils (N2), tolerogenic or regulatory tumor-associated DC (regDC), and regulatory B cells. While suppressor cells represent an important mechanism by which the immune system regulates specific immune responses, expansion of these cells in cancer patients interferes with the anti-tumor immunity and responses to therapy. Increased numbers and/or enhanced functionality of these cells have been detected in the peripheral blood, tumor mass, and tumor-draining lymph nodes of patients with hematologic malignancies and various types of solid tumors, suggesting that their evaluation can and should be included in immunomonitoring profiling of cancer patients. Recent data support the practicability and significance of this approach.

For instance, Gulley et al. [21] have tested the peripheral blood Treg cells in a multicenter randomized prostate cancer clinical trial, assessing Tregs as CD4+CD25highFoxP3+ T cells and confirming their immunosuppressive function in the in vitro assay. The Treg functional activity decreased following vaccination in patients with longer than predicted survival, while it increased in patients who survived less than predicted. On the other hand, though Schwartzentruber et al. [18] did not find a significant relationship between the development of anti-peptide CD8 T-cell reactivity and the objective clinical response in gp100 peptide melanoma vaccine trial, the post-treatment levels of CD4+FoxP3+ cells were higher in patients who had a clinical response to the treatment than in those who did not have a response. The authors hypothesize that the increased levels of Treg cells in patients who responded represent a counter-regulatory response after a strong anti-tumor immune reaction. The controversy may also arise due to the absence of definite markers for human Treg cells, a broad and changeable repertoire of functional Treg cell in the periphery, and different techniques of lymphocyte characterization in diverse patient populations utilized in many studies. Thus, although the prognostic significance of Treg cell immunomonitoring in vaccinated patients with cancer has not yet been established, many reports link their accumulations at the tumor site or peripheral blood to poor prognosis [22, 23].

Recently, a subpopulation of monocytic MDSCs phenotypically defined as CD14+HLA-DRneg/lo was shown to be significantly expanded in the peripheral blood of patients with metastatic melanoma [24], glioblastoma [25], and prostate cancer [26]. Circulating MDSCs were significantly increased in cancer patients of all stages relative to healthy volunteers. Importantly, the increased levels of circulating MDSCs have been correlated with the tumor stage, metastatic spread, and response to chemotherapy in different types of cancer [2729]. Increased circulating levels of MDSCs present in the blood of patients with breast cancer and patients with colorectal cancer have been shown to correlate with worse prognosis and radiographic progression [30].

Additional clinical studies are required to determine the significance of Treg and MDSC monitoring in the peripheral blood of cancer patients in different immunotherapeutic protocols. Analysis of these data in association with CTL reactivity should provide new insights into the importance of the monitoring of systemic immune responses in these patients.

Local immunomonitoring: rationale

Emerging data suggest that the main events that determine tumor fate in face of immune attack occur at the tumor site [31]. There is a significant number of identified mechanisms leading to immune unresponsiveness associated with the immunosuppressive tumor microenvironment. This includes down-regulation of antigen processing and MHC class I peptide complex expression on tumor cells, functional inhibition and apoptosis of immune effector cells (T cells, NK cells, conventional DC) induced by tumor cells, stromal elements, and immune regulatory cells, and tumor-induced polarization of immune cells into immune regulators, such as Treg cells, MDSC, M2 TAM, regDC, and N2 tumor-associated neutrophils (TAN), etc. Tumor cells, the surrounding stromal cells, and intratumoral immune cells support tumor angiogenesis, tumor cell invasiveness and spreading, and inhibit development of anti-tumor immunity. These cells produce a variety of soluble and membrane-bound molecules, including transforming growth factor β (TGF-β), interleukin (IL)-10, IL-8, IL-13, indoleamine-2,3-dioxygenase (IDO), inducible nitric oxide synthase (iNOS), prostaglandin E2 (PGE2), arginase-1 (ARG1), neuropeptides, gangliosides, B7-H1, B7-DC, CTLA-4 and other molecules [23, 32]. However, in spite of robust evidence demonstrating that within the tumor milieu, host immune cells with immunoregulatory function are responsible for the establishment of tumor antigen-specific tolerance and represent a significant hurdle to successful therapy for cancer, analysis of tumor-associated immune cells and factors has not been widely considered as an essential part of immunomonitoring in cancer clinical trials [33]. In addition to technical difficulties, limited accessibility of clinical materials and uncertain significance of these assays, the main problem that still limits incorporation of intratumoral immunomonitoring into clinical practice seem to be limited data available showing that number, phenotype or function of specific subsets of intratumoral leukocytes may correlate with the disease progression or a patient’s response to anti-cancer therapy.

Local immunomonitoring: tumor-infiltrating lymphocytes

The infiltration of primary or metastatic tumors by T lymphocytes can contribute both positively and negatively to the tumor growth, invasion, and patient outcomes.

In most studies, the infiltration of tumor mass with CD8+ T cells and the higher CD8+ to CD4+ T cell ratio correlated with the improved outcome [22, 23]; for example, a significant correlation was observed between the extent of CD8+ T cell infiltration, CD8+ T to CD4+ T cell ratio, and a high CD8+ T to Treg cell ratio in the patients who had tumors that failed to metastasize to the draining lymph nodes [34]. In a recent study of patients with stage IV non-small-cell lung cancer, it was reported that patients with a higher frequency of tumor-infiltrating CD8+ T cells in the tumor epithelium, as compared to the tumor stroma, had a significantly better survival [35]. Pages et al. [36] found that five-year survival in colorectal cancer patients with high densities of both CD8+ T and CD45RO+ T cells was 86.2% with only 4.8% of patients having tumor recurrence, whereas in the group with low densities of these cells, 75% of patients had tumor recurrence, and only 27.5% survived. In another study of metastatic melanoma, Murphy et al. [37] demonstrated that whereas pre-vaccination tumor biopsies failed to reveal significant infiltration by lymphocytes, biopsies obtained after vaccination were markedly infiltrated by T cells with CD8+ phenotype.

In contrast to tumor infiltration by CD8+ T cells, tumor infiltration by CD4+ Treg cells may have either a negative or positive impact on the clinical outcomes [23]. Regulatory T cells are usually characterized by concurrent expression of CD4, CD25, and FoxP3, an essential identifier of these cells in humans. Treg markers may also include cytotoxic T lymphocyte antigen 4 (CTLA-4), CD127, HLADR, CD45RA, glucocorticoid-induced TNFR-related protein (GITR), and lymphocyte activation gene 3 (LAG-3). Among CD4+ T cells present in the tumor, a subset of CD4+CD25highFoxP3+ Treg cells may constitute from 5% to 15% of CD4 T cells in the infiltrate. While FoxP3+CD4+ T cells have been shown to be increased in the peripheral blood of melanoma patients as compared to healthy donors, they could be even more enriched in the tumor-infiltrated lymph nodes and at the tumor sites [38]. In patients with ovarian cancer, recruitment of Treg cells supported tumor growth and was associated with reduced survival; at the later stages of disease, CD4+CD25+FoxP3+ cells were significantly accumulated at the tumor site with ~75% of Treg being in proximity to infiltrating CD8+ T cytotoxic cells [39]. Sinicrope et al. [40] have recently reported that a low intraepithelial effector CD3+/regulatory FoxP3+ T cell ratio can predict an adverse outcome in colon carcinoma patients, indicating the importance of an effector to Treg cell ratio in colon cancer prognosis. In contrast, analysis of T-cell infiltrates in 967 colorectal tumors showed that patients with higher FoxP3+ Treg density in tumor tissue had improved survival [41]. However, understanding a high heterogeneity of TILs, it is important to note that a recent study revealed that CD8+FoxP3+ TILs mark the presence of tumor-rejecting antigen-specific T cells, and their accumulation serves as a marker for an effective T-cell response [42].

An interesting analysis of a prognostic value of combined evaluation of intratumoral Treg and CD8+ CTL has been recently reported. Chen et al. [43] determined tumor infiltration by Tregs and CTLs in 141 hepatocellular carcinoma (HCC) patients after tumor resection. The density of intratumoral Tregs and peritumoral CTLs was an independent factor for overall survival (OS), but not for disease-free survival (DFS). The combined analysis of Tregs and CTLs demonstrated better prognostic values than analysis of either of them alone. Tumor infiltration by tumor-associated FoxP3+ Tregs and CD8+ cytotoxic T lymphocytes was also evaluated in 1270 cases of invasive breast carcinoma [44]. Within the tumor bed, increased infiltration of Tregs and CTLs was significantly more common in patients with unfavorable histological features; high density Treg infiltration was associated with Her2 overexpression and decreased overall and progression-free survival. In contrast, in the tumor surrounding tissue, high CTL/Treg ratio was associated with improved overall survival and progression-free survival.

Local immunomonitoring: tumor-associated macrophages

Up to 50% of a malignant tumor mass can be composed of TAMs. While normal macrophages (M1) uptake antigens and play an important role in control of infections, TAMs can be reprogrammed in the tumor microenvironment in M2 cells as a result of tumor-driven ‘alternative’ activation. M2 are able to inhibit functions of immune cells and promote tumor survival, progression, angiogenesis, and metastasis by releasing IL-10, PGE2, NO, high amounts of TGF-β, or reactive oxygen species (ROS) [19, 23]. Clinical studies of macrophage infiltration have suggested that high tumor infiltration by TAM often correlates with a poor prognosis in cancer patients. For instance, Leek et al. [45] found a positive correlation between macrophage infiltration of breast carcinoma and reduced overall survival. Analysis of available publications by Bingle et al. [46] revealed that over 80% of studies showed a positive correlation between macrophage density and a patient’s poor prognosis. However, in a few reports, macrophage infiltration was associated with good prognosis; for example, Shimura et al. [47] reported high TAM number to be an independent predictor of longer disease-free survival for prostate cancer patients. The contradiction between studies may reflect differences in the number, grade, stage, and size of tumors included in each study. Also, it should be noted that utilization of the wide variety of methods to assess TAM infiltration and unclear identification of M1 and M2 cells in tested specimens could account for inconsistency of the results. Unfortunately, there is a lack of information about alternations of TAM densities induced by cancer therapy, and more studies are required before the evaluation of TAM can be included in the list of feasible immunomonitoring procedures.

Local immunomonitoring: myeloid-derived suppressor cells

Myeloid-derived suppressor cells are a heterogeneous cell population composed mainly of myeloid progenitor cells that do not completely differentiate into mature macrophages, DCs, or granulocytes. Immature bone-marrow-derived myeloid cells (IMCs) represent less than 1% of PBMC in healthy individuals; in cancer, this subset of cells can be increased up to tenfold, due to partially blocked differentiation and acquisition of suppressive activity [48]. The tumor microenvironment effects the composition of cancer-induced MDSCs through the release of various tumor-derived factors, including cyclooxygenase 2, prostaglandins, granulocyte–macrophage colony-stimulating factor (GM-CSF), macrophage CSF (M-CSF), IL-6, IL-10, vascular endothelial growth factor (VEGF), stem cell factor, IL-3, FMS-related tyrosine kinase 3 (FLT3), and cell-expressed molecules (such as Notch) [49]. MDSCs are characterized by combinations of different phenotypic markers, such as CD11b, CD34, CD33, CD15, CD13, CD14, IL-4Rα, and HLA-DR [19, 22, 23] and can be divided into 2 major subsets: granulocytic PMN- and monocytic MO-MDSCs. Although many reports revealed augmentation of MDSCs in the peripheral blood of patients with cancer, analysis of tumor-associated MDSC subsets is limited by experimental animal studies, with the exception of a publication by Goedegebuure et al. [50], who have shown that the tumor microenvironment in pancreatic adenocarcinoma contains both monocytic MDSC (CD11b+CD14+) and granulocytic MDSC (CD11b+CD15+). At present, in spite of numerous data demonstrating that elimination of MDSCs may be beneficial for cancer patients [51], one can only speculate that evaluation of MDSCs in tumor specimens might be a useful monitoring tool to predict patients’ responses to therapy or confirm the efficacy of therapeutic approaches to cancer.

Local immunomonitoring: concluding remarks

We would like to speculate that assays characterizing and enumerating immune effectors and regulators within accessible tumor specimens could be beneficial from different perspectives when added to the battery of commonly available immunomonitoring assays. First, therapy-induced alterations of the local tumor immunoenvironment might better reflect the clinical efficacy of the therapy by showing critical changes in the immunosuppressive potential of the tumor, preservation of immune tolerance and formation and function of immune effectors. Second, the analysis of immune responses at the tumor site might help to select patients likely to benefit from systemic treatment as well as prevent patients unlikely to respond due to treatment-related side effects. Third, development of feasible monitoring of the local tumor immunoenvironment could aid in the development of a new staging system for advanced cancer.

In fact, tumor-infiltrating S100+ DC showed an inverse relationship with the systemic antigen-specific T-cell response, a positive correlation with regulatory T cells, and a positive association with survival in cancer patients [52]. The presence of intratumoral neutrophils was a poor prognostic factor for hepatocellular carcinoma after resection [53]. Numerous clinical studies concluded that tumor-infiltrating CD8+ T cells have anti-tumor activity as judged by their favorable effect in patients’ survival in melanoma, renal, ovarian, urethral, colorectal, esophageal, head and neck, breast, pancreatic, and lung cancers [5458]. Using microarrays, gene expression patterns of pretreatment biopsy specimens from patients with esophageal squamous cell carcinoma were analyzed to identify genes correlated with survival times [59]. The genes involved in the immune response were characteristically up-regulated in the long-term survivors, and an immunohistochemical staining confirmed an increased CD8+ T-cell number in the long-term survivors over that in the short-term survivors after chemoradiotherapy. These results additionally stress the importance of tumor analysis for predicting the prognosis of individual cancer patients to cancer therapy. Even more, cancer therapy has been shown to result in marked infiltration of CD3, CD4, CD8, B lymphocytes, DC, and NK cells in the treated breast lesions with significantly increased number of FasL+, granzyme+, and perforin+ TILs [60, 61]. Finally, new findings extend the impact of the local immune response on the clinical course of metastatic colorectal cancer by revealing a strong association of TIL number and function with chemotherapy efficacy and prognosis [62]. Gene expression profiling experiments also suggested that the presence of T cells in metastatic melanomas before vaccinating the patients with tumor antigens could be a biomarker for clinical benefit from the vaccines [63]. This additionally suggests the importance of developing a scoring system that can be used as a predictive tool for response to cancer therapy. Thus, because growing evidence demonstrates that both local and systemic immune responses play an important role in controlling progression of a variety of solid tumors, incorporation of existing methodologies that characterize the tumor immunoenvironment is highly justified, essential and appropriate for immunomonitoring of cancer patients.

Future directions

Monitoring of immunotherapeutic clinical trials has substantially evolved in the last decade to the broad agreement that it should guide the development of cancer vaccines. The emphasis on immune cell functions and quantitation of antigen-specific T cells has been playing a major role in the attempts to establish rational correlations between therapy-induced immune alterations and clinical responses. However, clinical data revealed that the tumor-specific cellular immune response after the course of immunotherapy is not often linked with clinically proven cancer regression, in spite of the generation of tumor-specific cytotoxic lymphocytes able to recognize and efficiently kill tumor cells ex vivo. This disappointing lack of correlation between the tumor-specific cytotoxic immune responses and the clinical efficacy of immunotherapy may be explained, at least in part, by inadequate attention to the analysis of the local tumor immunoenvironment, which reflects the balance between pro-tumorigenic and anti-tumorigenic forces operational in the tumor milieu. A comprehensive evaluation of the balance between the immunostimulatory and immunosuppressive compartments of the immune system could result in the more appropriate analysis of the efficacy of any given immunotherapeutic protocol. Although initial attempts to assess both effectors and regulatory arms of the anti-tumor immune response have been made, the data obtained are still controversial and limited to the systemic immunity in the peripheral blood. However, for tumor eradication, circulating effector cells have to infiltrate established and often hard to access tumor masses. At present, our ability to follow this process in the clinical setting is quite limited, and the evaluation of ‘in situ’ immune responses represents a significant technical challenge. New approaches to characterize the tumor-infiltrating leukocytes and intratumoral cytokine network are needed. Such technologies as in situ PCR, FISH analysis, scanning flow cytometry, and microarray methods should provide opportunities to study tumor-infiltrating immune cell populations in more detail. The development and validation of in situ monitoring techniques will allow researchers to combine the analysis of both systemic and local immune responses in order to accomplish a complete monitoring of patients in immunotherapeutic protocols. It is conceivable that a combination of in situ and systemic immune parameters may provide a missing correlation between the patient’s immune and clinical response. This approach is one of the ways to develop personalized immunotherapy and immunomonitoring in the future.

Acknowledgments

This project was supported by NIH RO1 CA154369 grant (to M.R.S.). This project has been funded in whole or in part with federal funds from the National Cancer Institute, National Institutes of Health, under contract no. HHSN261200800001E. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government.

Conflict of interest

The authors declare that they have no conflict of interest.

Copyright information

© Springer-Verlag (outside the USA) 2011