The T cells of the immune system can destroy tumours, but their activation can be inefficient. Vaccines that exploit tumour mutations elicit robust T-cell responses to tumours, with potential clinical benefits. See Letters p.217 & p.222
A goal of treating cancer is the eradication of tumour cells while sparing healthy cells. Conventional therapeutic approaches such as chemotherapy or targeted drug treatments show such selectivity but are often toxic. The immune system can target tumour cells specifically, but immune responses are often suppressed within a tumour or its vicinity1. Although new clinical immunotherapy approaches can help to overcome this immunosuppression by targeting one of the pathways responsible in immune cells (the PD-1-receptor pathway), such treatment can be associated with considerable immunological side effects2. Mouse models3,4 and clinical5 studies have established the feasibility of increasing the immune system's response to tumours by using a vaccination strategy that targets mutant proteins present in tumour cells. In this issue, Ott et al.6 (page 217) and Sahin et al.7 (page 222) report the results of phase I clinical trials that assessed a personalized vaccine-based approach to treat people with skin cancer.
The vaccination strategy for triggering a tumour-specific response by immune cells called T cells can consist of the introduction into the body of peptide fragments known as antigens, which are presented to the immune system when bound by proteins of the major histocompatibility complex (MHC). To be successful, cancer vaccines need to meet several key criteria8. One of these is the incorporation of tumour-specific antigens that can induce potent immune responses from T cells. Antigens that arise in a tumour from a mutation and that are not normally present in healthy cells (neoantigens) fall into this category.
Another essential aspect of cancer vaccines is the selection of a suitable vaccination approach. Vaccines must effectively deliver materials, such as synthetically generated long peptides, DNA and RNA, that either provide a concentrated source of antigens or can be used by the body to generate antigens. Vaccines should also include a component known as an adjuvant, which provides a general immune stimulus, thereby enabling a more robust response against the tumour than would be generated naturally. For patients who have residual or recurrent tumours after therapy, a step to counteract the suppression of T cells in the cancer microenvironment is another important part of the vaccination strategy.
Ott et al. and Sahin et al. studied patients who had an advanced stage of a common type of skin cancer called melanoma. Many melanomas occur in skin that has been exposed to sunlight, and they are associated with a large number of random DNA mutations in cells known as melanocytes. Mutations in a melanoma that are causally involved in the growth of the tumour are rare — most mutations do not promote the tumour's growth9. But these mutations might generate abnormal protein sequences that are not present elsewhere in the body and are therefore potential targets for eliciting or boosting a tumour-specific immune response.
Ott et al. vaccinated six people who had previously undergone surgery to remove a tumour. To create personalized vaccines, the authors sequenced the DNA of tumour cells and healthy cells from each person to identify tumour-specific mutations and determine associated neoantigens. They then used an algorithm to predict which of the neoantigens would bind well to MHC proteins. Each study participant was vaccinated with synthetic long peptides representing up to 20 neoantigens that were specific to each patient's tumour. Such peptides are presented to T cells through the action of the antigen-presenting cells of the immune system.
Sahin et al. used a similar type of personalized approach to identify neoantigens with predicted good binding to MHC proteins in their study of 13 people who were undergoing treatment for melanoma. The authors selected up to ten mutations per person to create an RNA-based vaccine that was tailored to the individual's tumour. Sahin and colleagues had previously shown that such RNA molecules can be taken up by antigen-presenting cells10.
If an antigen bound to an MHC protein is recognized by a T-cell receptor (found on the surface of a T cell), the T cell can mount a response against any cell that contains the antigen. There are two types of MHC protein, and these present antigens to different types of T cell. MHC class I proteins present antigens to cells known as CD8+ killer T cells, which express the protein CD8 on their surfaces. These T cells can then mount an attack that directly kills cells expressing the specific antigen. Cells other than immune cells, including tumour cells, usually have MHC class I proteins on their surfaces.
MHC class II proteins are found on the surfaces of cells of the immune system. They present antigens to T cells that express the protein CD4 and are known as CD4+ helper T cells. Tumour-cell antigens can be presented to CD4+ helper T cells by antigen-presenting cells that ingest substances from dying cancer cells. Direct recognition of tumour cells by CD4+ helper T cells has been documented, and these cells are also associated with the direct killing of tumour cells11,12. CD4+ helper T cells have an additional role in optimizing the function of CD8+ T cells13, and in aiding the generation of CD8+ T cells with a 'memory' capacity14 that enables a more robust response on a subsequent encounter with a given antigen.
Crucially, the vaccines developed by Ott et al. and Sahin et al. generated both CD4+ and CD8+ T-cell responses, which probably resulted in the activation of CD4+ T cells, antigen-presenting cells and CD8+ T cells, as well as the necessary interactions between these cells to enable optimal T-cell action and memory function. The immune-response data in both studies revealed that the vaccines boosted the number of T cells engaging in an immune response that had previously been observed against certain neoantigens, and also generated T-cell responses against other neoantigens that had not been observed before.
Surprisingly, both studies found that the neoantigens represented in the vaccine that were recognized by CD4+ T cells generated a greater immune response, as assessed by analysis of molecular markers to determine immune cell activation, than was observed for the neoantigens in the vaccine that were recognized by CD8+ T cells. Yet Ott and colleagues' algorithm did not specifically search for neoantigens that were predicted to be presented by MHC class II proteins, and the algorithm used by Sahin et al. was better at predicting the antigens that bind to class I MHC proteins than those that bind to class II.
Of the 6 people who were vaccinated by Ott and colleagues, 4 showed no signs of tumour recurrence in a follow-up period of up to 32 months after vaccination. Tumours were still present in two participants after vaccination, but when these individuals received a treatment that blocked the immunosuppressive PD-1-receptor pathway, their tumours regressed. Of the 13 people who were vaccinated by Sahin and colleagues, 8 remained free of tumours throughout a follow-up period of 12–23 months. The remaining five had tumour relapses; however, complete regression of tumours occurred in one of these people who was given treatment to block the PD-1-receptor pathway. One of the people whose cancer relapsed had a mutation that abolished the expression of MHC class I proteins, indicating a possible mechanism for the tumour to escape from the T-cell response.
The two studies confirm the potential of this type of approach, and improvements in neoantigen prediction will probably allow the even more efficient and precise identification of neoantigens for use in therapeutic vaccines in the future. Although the numbers of people who were treated in these studies were small, both studies indicated potential benefits. For example, following vaccination, fewer episodes of tumour recurrence or tumour migration to other locations in the body (metastasis) occurred than might have been expected, given the clinical history of each patient. Sahin and colleagues noted that participants showed a significantly lower rate of metastasis after vaccination. In the cases in which tumours persisted or regrowth occurred, this could often be combated effectively by treatment that targeted the PD-1-receptor pathway (Fig. 1). Controlled, randomized phase II clinical trials with more participants are now needed to establish the efficacy of these vaccines in patients with any type of cancer that has enough mutations15 to provide sufficient neoantigen targets for this type of approach.
Pitt, J. M. et al. Ann. Oncol. 27, 1482–1492 (2016).
Naidoo, J. et al. Ann. Oncol. 26, 2375–2391 (2015).
Castle, J. C. et al. Cancer Res. 72, 1081–1091 (2012).
Gubin, M. M. et al. Nature 515, 577–581 (2014).
Carreno, B. M. et al. Science 348, 803–808 (2015).
Ott, P. et al. Nature 547, 217–221 (2017).
Sahin, U. et al. Nature 547, 222–226 (2017).
van der Burg, S. H., Arens, R., Ossendorp, F., van Hall, T. & Melief, C. J. Nature Rev. Cancer 16, 219–233 (2016).
Szczepaniak Sloane, R. A. et al. Cancer 123, 2130–2142 (2017).
Kreiter, S. et al. Cancer Res. 70, 9031–9040 (2010).
Hirschhorn-Cymerman, D. et al. J. Exp. Med. 209, 2113–2126 (2012).
Haabeth, O. A. et al. Leukemia 30, 1216–1220 (2016).
Schoenberger, S. P., Toes, R. E., van der Voort, E. I., Offringa, R. & Melief, C. J. Nature 393, 480–483 (1998).
Janssen, E. M. et al. Nature 421, 852–856 (2003).
Alexandrov, L. B. et al. Nature 500, 415–421 (2013).
The author declares competing financial interests. See go.nature.com/2sqtds6 for details.
Related links in Nature Research
About this article
Heterologous arenavirus vector prime-boost overrules self-tolerance for efficient tumor-specific CD8 T cell attack
Cell Reports Medicine (2021)
Immune-based mutation classification enables neoantigen prioritization and immune feature discovery in cancer immunotherapy
Nature Immunology (2021)
Biochimica et Biophysica Acta (BBA) - Reviews on Cancer (2020)
Expert Review of Vaccines (2020)