Addressing tumour tolerance to improve cancer immunotherapy
In May 2010, the Association for Cancer Immunotherapy (CIMT) held its 8th annual meeting in Mainz, Germany. More than 400 European and international participants attended this conference, where basic immunological findings as well as pre-clinical and clinical studies were discussed in light of their impact on tumour immunotherapy. This year's meeting was dominated by four major topics:
- (i)the growing insight on regulatory cells as well as on intracellular and intercellular mediators of tolerance,
- (ii)the improvement of cancer vaccines,
- (iii)the manipulation of cellular products for adoptive immunotherapy, and
- (iv)the formal requirements involved in translating basic and pre-clinical immunology into cancer therapy.
Strategies to overcome tumour-induced tolerance
Immune effector functions can be negatively regulated by many different pathways. One major suppressive mechanism is that mediated by Treg. Jon S. Bromberg (New York, NY, USA) reported the impact of natural Treg in a murine diabetes model in which allogeneic pancreas islet cells are transplanted beneath the kidney capsule. Allograft survival was observed if Treg were activated in the graft and subsequently entered the draining lymph nodes of the recipient. This chemokine-dependent migration of Treg was essential for maximum tolerance induction. In a heart allograft model, tolerance mediated by adaptive Treg involved several homing molecules and chemokine receptors (e.g. CD62L, CCR2, and CCL19). Dr. Bromberg also presented data on the epigenetic regulation of Foxp3, an important transcription factor of Treg. Controlling this transcription factor may help to develop strategies for specific modulation of Treg function. Recent efforts focus on an enhancer sequence upstream of the Foxp3 promoter that is repressed by methylation in naïve CD4+ T cells, effector CD4+ T cells, and TGF-β-induced adaptive Treg. On the contrary, the upstream enhancing sequence is completely demethylated in natural Treg. IL-6 can induce methylation of this domain, thereby converting natural Treg to their adaptive counterparts. Interestingly, DNA methyltransferase inhibitors can change the Foxp3 upstream methylation in adaptive Treg into the acetylated phenotype of natural Treg. Together, epigenetic studies of Treg offer important information on how Treg function can be controlled and how this may lead to Treg-based immunotherapies 1.
Malignant brain tumours that are infiltrated by high numbers of Treg are interesting models for studying Treg function 2. In a murine GL261 brain tumour model, Gosse J. Adema (Nijmegen, The Netherlands) observed that depletion of Treg by anti-CD25 antibody treatment and vaccination with DC loaded with tumour lysate led to lower tumour burden and prolonged survival. A similar approach of tumour-specific vaccination along with Treg depletion might be a very attractive strategy for human tumour immunotherapy. Studies in melanoma patients showed that the humanized anti-CD25 receptor α antibody daclizumab efficiently depleted CD25high cells, including Foxp3+ cells in the blood, for several weeks. However, the treatment also decreased B- and T-cell responses towards the vaccine, showing that daclizumab can adversely affect the vaccine-induced immune response. The results highlight that in order to achieve immunotherapeutic benefit by transient Treg depletion, timing and dosing as well as Treg specificity of the depleting compound are extremely important.
Isabel Poschke (Stockholm, Sweden) directed the attention of the audience to the role of myeloid-derived suppressor cells (MDSC) in the anti-melanoma immune response. MDSC inhibit T- and NK-cell functions by several different mechanisms including the production of immunosuppressive cytokines and enzymes (e.g. IL-10, TGF-β, ROS/RNOS, arginase, and iNOS), recruitment of Treg, and by nutrient starvation. A variety of cytokines (e.g. VEGF, M-CSF, GM-CSF, IL-6, IL-1β, IL-10, and PGE2) are able to enhance MDSC suppressive activity (e.g. IFN-γ, IL-13, and IL-4). In addition, MDSC numbers were found to correlate with tumour progression. Dr. Poschke demonstrated that STAT3 is overexpressed in MDSC in human melanoma and that targeting this factor by the inhibitor AG490 can revert MDSC-mediated T-cell suppression in vitro3.
Immunosuppression of the tumour micromilieu is not only mediated by cellular suppressors such as Treg and MDSC, but also by metabolic dysregulation. Andreas Mackensen (Erlangen, Germany) observed that tumour-derived lactic acid severely impaired the proliferation, IFN-γ production, and cytolytic activity of melanoma-reactive CTL in vitro. On the contrary, CD4+CD25+ Treg appeared less sensitive to this immunosuppressive effect of lactic acid. Analysis of renal-cell carcinoma specimen showed that this tumour type overexpressed glucose transporter GLUT-1 and lactate dehydrogenase A at the mRNA level. GLUT-1 expression correlated inversely with the number of tumour-infiltrating lymphocytes. Furthermore, renal-cell carcinoma cell lines produced high levels of methylthioadenosine that inhibited CTL proliferation and effector functions by an apoptosis-independent mechanism. Dr. Mackensen concluded that tumour cells have developed several strategies including the depletion of amino acids (e.g. tryptophane and arginine) as well as the production of metabolites (e.g. lactic acid, GLUT-1, and methylthioadenosine), which contribute to the aggravated immunosuppression at the tumour site 4.
It would be very attractive not only to revert tumour-induced immunosuppression but also to promote the development of efficient and long-lasting anti-tumour T-cell immunity. Protul A. Shrikant (Buffalo, NY, USA) underlined the importance of understanding how long-lasting CD8+ T-cell memory can be achieved. His group has demonstrated that the mammalian target of rapamycin (mTOR) determines the fate of activated CD8+ T cells towards effector or memory function by regulating the transcription factors T-bet and Eomesodermin 5. Inhibition of mTOR by rapamycin enhanced T-cell responses to recall antigens. In addition, Dr. Shrikant emphasized that homoeostatic T-cell proliferation (HP) plays an important role in establishing a long-lasting anti-tumour T-cell response. Data from a murine model showed that rapamycin programs IL-15-independent CD8+ T-cell memory in the context of HP. Blocking mTOR, therefore, resembles an interesting strategy for supporting tumour-specific immunotherapies. The fact that rapamycin also controls GVH disease (GVHD) in the context of allogeneic haematopoietic stem cell transplantation (HSCT) suggests that mTOR inhibitors might improve the risk-benefit profile of adoptively transferred allogeneic T cells and donor lymphocyte infusions.
A further approach that enhances the effect of tumour vaccines is to block negative regulatory checkpoints imposed by the B7-CD28 family of molecules via antibody blockade 6. This strategy was pioneered by James Allison (New York, NY, USA), who presented most recent results of clinical trials using the CTLA-4 antibody ipilimumab in cancer patients. More than 4000 patients have been treated so far, showing objective clinical responses according to RECIST criteria in melanoma, renal-cell carcinoma, prostate cancer, and ovarian cancer. Even with ipilimumab monotherapy, complete responses and a significant survival benefit have been observed 7. The major response pattern was similar to that of other immunostimulatory agents (e.g. IL-2), with early tumour progression followed by sustained tumour remission occurring several months upon initiation of treatment. Frequently observed side effects (10–15%) include rash, colitis, hypophysitis, hepatitis, which resolved after immunosuppressive treatment in most patients. Nevertheless, a few patients died due to immune-related adverse events, emphasizing the importance of negative regulation. Dr. Allison suggested the future of CTLA-4 antibodies as a combination partner for tumour vaccines and T-lymphocyte infusion products. He also proposed synergistic effects when combined with conventional cancer treatment approaches such as chemotherapy, irradiation, androgen blockade, and surgery. In addition to CTLA-4 blockade, the community will soon see the clinical results of agents interfering with other members of the extended B7-CD28 family such as PD1 (ligands: PD-L1 and PD-L2) and ICOS (ligand: B7H).
In 2010, the first ever therapeutic vaccine against cancer was approved by the American Food and Drug Administration (FDA). David L. Urdal (Seattle, WA, USA), Chief Scientific Officer of Dendreon, presented data on the development of Sipuleucel-T, an individualized cellular product generated from leukapheresis-derived autologous DC loaded with a fusion protein consisting of GM-CSF and prostatic acid phosphatase (PAP). Sipuleucel-T is used as an active therapeutic vaccine against metastatic hormone-refractory prostate cancer. In animal models, vaccination with PAP-loaded DC led to prostatitis, thereby proving the induction of a tissue-specific T-cell response. FDA approval was granted based on the data of three phase III trials (D9901, D9902A, and DD9902B) 8. The vaccination was associated with the increase and class switch of PAP-specific antibodies. Specific T cells were detected with a maximum response after three vaccinations. Clinical benefit of this cellular vaccine was demonstrated by an increased overall survival in the vaccine group compared with placebo controls (21.7 versus 25.8 months). The results obtained in patients with advanced cancer call for new trials in earlier stages and in the adjuvant setting. This is the first approved active tumour-vaccination strategy in the therapeutic setting. Dr. Urdal announced further clinical trials in cancers of the male and female genitourinary system, colon, breast, and lung.
An interesting approach of vaccination with tumour antigen-encoding DNA by multiple intradermal injections (“tattooing”) has been introduced by John Haanen (Amsterdam, The Netherlands). His study is based on the pre-clinical data in mice and non-human primates, where tattooing led to superior antigen-specific CD8+ T-cell responses compared with intramuscular administration 9, 10. Dr. Haanen also presented a human skin-explant model that was successfully used to optimize the local antigen expression by DNA tattooing. Based on the pre-clinical data, a clinical phase I trial was performed where stage IV melanoma patients were treated with DNA encoding melanoma-associated antigen recognized by T cells (MART)-1 linked to the carboxy-terminus of Tetanus toxoid fragment c (TTFc-MART-1). Cohorts of three patients were injected with increasing vaccine doses at days 0, 3, and 6 and boosted after 4 wk. Five of the six patients in the first cohorts of this ongoing trial showed enhanced frequencies of tetramer-positive T cells at 2 and 6 wk post-vaccination. Although the clinical response data are still pending, the data provided underline the importance of skin-derived antigen-presenting cells in the induction of antigen-specific T-cell responses. DNA vaccines for HPV16-encoded antigens E6/E7 are currently under way.
Chronic myeloid leukemia (CML) is a well-established model disease for cancer immunotherapy with adoptively transferred T cells in the context of allogeneic HSCT. After the approval of imatinib in 2001 (Europe) and 2003 (USA) and the introduction of other tyrosine kinase inhibitors (TKI) specifically blocking the activity of the BCR-ABL kinase, targeted therapy has replaced allogeneic HSCT as the first line therapy for CML; however, as pointed out by Katy Rezvani (London, UK), therapeutic strategies are still needed that target the leukemic stem cells and eradicate minimal residual disease. Dr. Rezvani and co-workers have detected T cells with specific reactivity to the leukemia-associated antigens Wilms-tumour gene 1, proteinase 3 11, and preferentially expressed antigen of melanoma 12 in CML patients. These antigens as well as neo-epitopes derived from the BCR-ABL fusion region are interesting candidates for CML-directed vaccination strategies. In the past, TKI have been accused of hampering T-cell function at various levels in vitro. A trial with 50 patients vaccinated against influenza and pneumococcal antigens in patients during ongoing TKI therapy induced vaccine-specific T-cell responses comparable to that found in healthy individuals. Studies on the effect of TKI on B-cell responses to vaccination are ongoing. Based on these results, a variety of vaccination studies in TKI-treated patients with CML and other myeloid malignancies are under way. Until the results of these trials demonstrate clinical efficacy, allogeneic HSCT remains the only curative therapy for patients in whom TKI treatment has failed.
Adoptive cellular immunotherapy
Allogeneic HSCT is hampered by unwanted allo-reactivity that causes GVHD. Depleting GVHD-mediating T cells from a stem cell graft or a preparation of donor lymphocytes has, therefore, been the focus of many research groups. John Barrett from the NIH (Bethesda, MD, USA) reported the clinical experiences with donor lymphocyte infusions that were specifically depleted of allo-reactive T cells by “photodynamic purging”. Donor T cells were labelled with a rhodamine-like phototoxic dye (TH9402) after stimulation with host-derived peripheral-blood mononuclear cells in a mixed-lymphocyte reaction in vitro. Activation-induced changes of the p-glycoprotein, the gene product of the MDR1 gene, resulted in an impaired efflux of the photosensitizer in T cells activated by the recipient cells. By exposure to visible light, these cells were selectively depleted. Dr. Barrett presented first results of a dose-escalation study, showing the safety and feasibility of the procedure in HSCT patients. Currently, this promising strategy for GVHD prophylaxis is tested in clinical trials of haploidentical transplantation in Europe as well as in North America.
Potentials and pitfalls of re-directing T cells by TCR gene transfer were discussed in many talks and posters. Mirjam Heemskerk from Leiden (The Netherlands) pointed towards the risks of unwanted allo-immune reactivity resulting from the mispairing of endogenous and transduced TCR chains 13. Dr. Heemskerk showed how the use of T cells with a defined specificity (e.g. for cytomegalovirus) as recipient cells for TCR transfer can reduce the likelihood of undetected TCR-heterodimer formation. In addition, the introduction of additional disulfide bonds between TCR-α and -β chains of the transgenic TCR decreased the risk of misdirected T-cell reactivity. Using these techniques, the Leiden group has prepared a clinical trial using T cells transduced with a TCR for the minor histocompatibility antigen HA-1 in the context of allogeneic HSCT. Another approach to limit the potential harm by mispaired TCR is to transiently transfect the T cells by in vitro-transcribed mRNA as presented by the group of Niels Schaft (Erlangen, Germany). They showed that antigen-reactive CD4+ as well as CD8+ T cells can be efficiently generated by this technique. This approach, however, is not without caveat: although limiting the risk of long-lasting unwanted allo- or auto-immunity, transient TCR expression calls for repeated application of transgenic T cells.
Beyond providing the platform for annual meetings and building bridges between basic scientists and clinical immunotherapists, the CIMT also addresses technical and regulatory aspects of translating immunology into immunotherapy. These issues are covered by the CIMT immunoguiding program, currently led by Sjoerd H. van der Burg (Leiden, The Netherlands) and the CIMT Regulatory Research Group chaired by Ulrich Kalinke (Hannover, Germany). During the short talk session of the immunoguiding program working group, Cedrik M. Britten (Mainz, Germany) gave an overview of the various activities and workshops. In the same session, Christian Ottensmeier (Southampton, UK) summarized past and future activities related to the ELISPOT assay and announced the harmonized response definition criteria for ELISPOT assays that were recently published 14. The actual HLA-multimer panels were introduced by Cecile Gouttefangeas (Tuebingen, Germany). The view was widened by a talk of Sylvia Janetzki (New York, NY, USA), who outlined the efforts of the Cancer Immunotherapy Consortium that in the past already have helped to improve the validity and reproducibility of immune-response monitoring with a focus on functional assays.
In this context of improving insight on the necessity for a comprehensive immune-response analysis, Mary L. Disis (Seattle, WA, USA) provided data from patients with advanced breast cancer who were vaccinated with HLA class II-restricted peptides of HER2/neu. In total, 66 women with HER2-positive breast cancer were vaccinated six times at monthly intervals using GM-CSF as an adjuvant. The vaccination group had a significant survival benefit compared with an age- and tumour stage-matched control group. In some patients, T cells specific for HER2 peptides that were not contained in the vaccine were detected after vaccination. In addition, some patients developed T-cell responses specific for tumour-associated antigens other than HER2/neu. The presence of this intramolecular and extramolecular epitope spreading was an independent predictor for improved survival. In addition to epitope spreading, a Th1 cytokine profile of the induced T-cell response also had a positive impact on survival. This finding supports the hypothesis that CD8+ T cells are the main mediators of successful anti-tumour immunity. The data presented by Dr. Disis illustrated the complexity of determining a clinically relevant immune response after vaccination.
Key note lecture
Mark M. Davis (Stanford, CA, USA) introduced a “Human Immunology Project” that aims at the identification of reliable biomarker signatures for the earlier diagnosis of diseases as well as for the prediction of treatment outcomes. The project will measure a broad panel of genetic and immunological parameters in patients and healthy individuals by an impressively systematic approach. A major part is the “Human Immune Monitoring Centre” at Stanford, where patient's samples are processed by very sensitive assays in a high-throughput manner. Novel tools recently developed for this purpose are used, e.g. a CyTOF Mass Cytometer combining flow cytometry and mass spectrometry that allows multi-parameter analysis at the single-cell level. Enormous data sets are produced that are processed by cutting-edge bioinformatics to identify reliable biosignatures associated with certain diseases and treatment outcomes. Dr. Davis plans to extend his visionary effort to a world-wide multi-centre approach in which centres of excellence join forces to reach the challenging goals.
In summary, very promising results in many fields of cancer immunotherapy were presented at the CIMT 2010 meeting. Current randomized phase III trials have successfully demonstrated an improved survival in cancer patients receiving therapeutic tumour vaccines or immunostimulatory antibodies. Future developments will surely benefit from the better definition of cellular and molecular mechanisms involved in tumour rejection and in break of tolerance. The correlation of comprehensive immune-response analysis with clinical results will help with understanding what distinguishes responders from non-responders in cancer immunotherapy. The apparent trend towards combined immunotherapies requires the selection of suitable partners capable of potentiating anti-tumour efficiency while ensuring minimal toxicity or at least, at a tolerable level. Clinical testing of the multitude of upcoming approaches depends on the close and positive interactions of researchers and clinicians with regulatory authorities and pharmaceutical companies. The next CIMT meeting will be held May 25–27, 2011 in Mainz to bring again together numerous international scientists contributing to the evolving success of cancer immunotherapy.
The authors declare no financial or commercial conflict of interest.