checkpoint inhibitors and exploiting neoantigens Flashcards
Inhibitory receptors?
T cells express two inhibitory surface receptors (CTLA-4 and PD-1), which act as negative regulators of T cell activation. These receptors are crucial for the maintenance of self-tolerance but are also understood to limit the activity of tumour-specific T cells, hence restricting the immunological elimination of tumours
CTLA-4 functions?
CTLA-4 mainly reduces T cell activation by outcompeting CD28 for binding to its ligand B7 (CD80/86), hence inhibiting the major co-stimulatory signal required for the initial activation of T cells in the lymph nodes. It is also thought to sequester B7 from the surface of antigen-presenting cells (APCs) by transcytosis, preventing APCs from activating further T cells
Also enhances suppressive function of Tregs
studies for CTLA-4
Riley et al (2002) = using Jurkat T cells and primary CD4 T cells. Cytotoxic T lymphocyte antigen 4 (CTLA-4) engagement selectively blocked augmentation of gene regulations by CD28-mediated costimulation, but did not ablate gene regulation induced by TCR triggering alone
Qureshi et al (2011) = cultured CTLA-4 expressing cells with cells expressing GFP-tagged CD86 flow cytometry and confocal microscopy revealed substantial transfer of CD86 into CTLA-4 cells
Wing et al (2008) = CTLA4 KO in Treg cells (crossed mice that expressed Cre under FOXP3 promoter, with mice in which CTLA4 was flanked with 2 LoxP sites) inhibits ability to regulate autoimmunity in mice systemic lymphoproliferation, T cell mediated disease, hyperproduction of IgE confirms that CTLA-4 enhances suppressive functions of Treg cells
Also produced potent enhancement of anti-tumour immunity
PD-1 functions
In contrast, PD-1 facilitates T cell inhibition within the tumour microenvironment itself by binding its ligands PD-L1 (expressed on all cells) and PD-L2 (expressed on APCs during inflammation). Activation of PD-1 results in the recruitment of intracellular SHP1/2 phosphatases to the cell membrane, where they dephosphorylate neighbouring receptors to inhibit TCR and CD28 signalling. Tumour cells often upregulate their surface expression of PD-L1 when under attack by T cells (mechanism shown in figure 1), helping them to escape immune killing.
Inflammation –> IFNy released by T cells –> binds to IFNYR on cancer cells –> JAK1/2, STAT signalling -> IRF-1 –> PD-L1 –> engages PD-1 –> inhibits T cell activity
first study showing anti-tumour effects of CI
The anti-tumour effects of checkpoint blockade were first demonstrated in a seminal paper by Leach et al (1996). In these experiments, they injected mice with 2 x 106 tumour cells from the murine colon carcinoma 51Blim10 before starting treatment with anti-CTLA-4 antibodies 7 days later, by which point almost all the mice had established tumours. They found that tumour growth was drastically reduced in the mice treated with anti-CTLA-4 compared to controls, and that 2 of the 5 treated mice remained tumour-free even at 30 days after tumour injection. They also found that ant-CTLA-4 antibodies protected mice against a secondary challenge with the same tumour, suggesting that blocking CTLA-4 encouraged the generation of immunological memory.
BUT no test of toxicity
neoantigens?
Neoantigens are novel peptides expressed exclusively on the surface of cancer cells. They are usually produced as a consequence of oncogenic mutations, though also include viral peptides expressed in virus-associated cancers, such as HPV-induced cervical carcinoma.
Other classes of neoantigens
• protein products of human endogenous retroviruses (HERVs)
- derived from exogenous retroviral infections through evolution of human genome
- transcription suppressed by hypermethylation and repressive histone modifications in normal cells global DNA hypomethylation often seen in tumour cells results in HERV expression
• ‘dark antigens’
- derived from parts of genome not normally expressed as protein, silenced in healthy cells but activated in tumours likely due to abnormalities in transcription
- often shared between patients = potential for ‘off-the-shelf’ therapies?
why target neoantigens?
There has been a growing interest in exploiting neoantigens for personalised cancer vaccines, which induce and expand populations of tumour-specific T cells to encourage cancer killing. Cancer vaccines may also provide long-lasting immunological memory against the target antigen, providing protection against tumour recurrence. Neoantigens are attractive targets for cancer vaccines as they exhibit very high tumour specificity. As a result, T cell responses mounted against them result only in the destruction of cancer cells and not healthy tissues, hence reducing the likelihood of side effects and autoimmunity. In addition, vaccines focused on neoantigens are usually more immunogenic than those focused on self-antigens that are unusually expressed or overexpressed in tumour cells, as T cells specific to neoantigens are not subjected to central or peripheral tolerance.
how to select neoantigens for vaccines?
A range of sophisticated techniques are used to select appropriate neoantigens for cancer vaccines. Comparing next generation sequencing (NGS) data from the patient’s tumour biopsy and a matched control tissue allows neoantigens to be identified quickly at low cost. RNA sequencing is then used to identify neoantigens that are highly expressed within the tumour, whilst computational algorithms are employed to predict the binding affinity of neoepitopes to MHC class I and II molecules. It is usually easier to predict whether a particular neoantigen will bind MHC class I than MHC class II. This is because the peptide binding groove of MHC class I has closed ends that enclose peptide epitopes for presentation, whilst that of MHC class II has open ends, allowing peptide regions flanking the epitope bound to the core binding groove to influence binding. As a result, vaccine studies have traditionally focused on identifying neoantigens with high affinity for MHC class I. However, evidence indicates that cancer vaccines often preferentially induce CD4+ T cell responses, perhaps because the open-ended nature of MHC class II provides scope for the presentation of a broader range of peptides; as a result, focus has shifted to developing more advanced algorithms to allow for better predictions of MHC class II binding.
RNA vaccine trial
The first clinical trial of a neoantigen vaccine was carried out by Sahin et al (2017)9. In this study, they used comparative exome sequencing to identify non-synonymous mutations in tumour biopsies of 13 patients with stage III or IV melanoma and a high risk of relapse. They then inoculated each patient with 2 personalised synthetic RNAs, together encoding ten 27mer peptides that contained 1 mutation each. IFNy ELISpot assays of T cells isolated from patient sera before and after vaccination revealed that 60% of the neoepitopes used were immunogenic, with 57% inducing CD4+ T cells, 17% inducing CD8+ T cells and 26% inducing both populations. Every patient developed T cell responses against at least 3 mutations, demonstrating that the RNA vaccines were able to induce polyspecific immunity. Crucially, the vaccines were associated with a significant reduction in recurrent metastatic events, with 8 patients remaining tumour-free for the entire follow-up period of 12 to 23 months. The remaining 5 patients relapsed, though 2 then developed vaccine-related objective clinical responses.
peptide vaccine trial
Similar success was achieved in a trial by Ott et al (2017)10, in which 6 melanoma patients were vaccinated not with RNA, but with pools of long peptides each containing up to 20 personalised neoepitopes. In this study, 4 patients had no disease recurrence during the 25-month follow-up period, whilst 2 relapsed but achieved tumour regression when treated with PD-1 inhibitors, indicating that combining cancer vaccines with checkpoint inhibitors may improve their efficacy in some patients. Although both these trials had very small sample sizes, with no blinding or control arm, they provide proof-of-concept that neoantigen-specific cancer vaccines may improve outcomes in cancer patients.
problems with vaccines
Despite their success in early trials, it is important not to overlook the possible disadvantages of neoantigen vaccines. Firstly, due to the random nature of mutations, neoantigens are often unique to individual tumours. As a result, therapies exploiting their expression must be personalised to each patient, making the process very expensive and time-consuming. The groups of Sahin and Ott both declared that the formulation and manufacture of their vaccines took around three months, which would be too long to delay treatment for many cancer types. Secondly, there is concern that cancer vaccines may drive immune escape, by which cancer cells downregulate the expression of target neoantigens to escape killing. This risk could be reduced by producing vaccines targeting lots of mutated peptides, or targeting neoantigens indispensable for oncogenesis and/or survival. Sahin et al (2017) also noted that one patient in their study did not respond to therapy because their tumour downregulated expression of both alleles of B2M and TRIM69 required for MHC class I expression during treatment. Thirdly, it is unclear whether neoantigen vaccines will be equally effective in tumours with low numbers of mutations; both studies used melanoma as their tumour of choice, which usually carries a high mutational burden due to its induction by UV light. Cancers associated with long-term exposure to mutagens (e.g. melanoma and lung cancer) typically exhibit high mutational loads Cancers that arise in younger people (e.g. leukaemia, lymphomas) typically possess fewer mutations = redirected T cell therapies would generally be more effective than neoantigen vaccines
Other possible pitfalls
• Different metastases, or even different parts of the same tumour, may not express the same neoantigens vaccines may not result in killing of all cancer cells
• Risk of cross-reactivity to self-tissues = not encountered in existing studies, so unlikely to be
neoantigens in adoptive T cell transfer
Neoantigens may also be exploited by T cell transfer therapy, in which neoantigen-specific T cells are isolated from a patient’s blood or tumour, expanded ex vivo and infused back into the patient, helping to enhance tumour-specific immunity. This proved efficacious in a case study by Tran et al (2016)11, in which they treated a patient with metastatic colorectal cancer with 1.48 x1011 CD8+ T cells specific to the neoantigen KRAS G12D, produced by in vitro expansion of specific tumour-infiltrating lymphocytes isolated from a lung lesion biopsy. This treatment was sufficient to induce regression of the patient’s 7 metastatic lung lesions by 40 days after treatment. However, the patient eventually relapsed 9 months later following loss of heterozygosity of HLA-C*08:02 in one of the lesions, allowing the tumour to evade KRAS G12D-specific T cells.
trials of CTLA-4 inhibitors
Monoclonal antibodies against CTLA-4, PD-1 and PD-L1 are now approved cancer therapies and have achieved remarkable success in the clinic against a variety of different cancers. The CTLA-4 inhibitor ipilimumab was the first to be licenced following a landmark phase III trial by Hodi et al (2010)2, in which 676 patients with unresectable, treatment-resistant stage III or IV melanoma were assigned to ipilimumab with or without a gp100 peptide vaccine, or to a gp100 vaccine alone (active control). The CTLA-4 blocker was shown to significantly improve long-term survival (median overall survival = 10.1 months with ipilimumab alone, 6.4 months with gp100 alone), and reduce the risk of disease progression. Grade 3 or 4 adverse effects noted in 10-15% patients treated
Pooled trial analysis by Schadendorf et al (2015)3 confirmed that responses to ipilimumab are durable in a subset of patients, with 22% of patients living for more than 3 years, and a small number surviving for 10 years or more. However, it is important to note that this study failed to present the long-term outcomes of patients included in the control arms of ipilimumab trials, making it difficult to judge the true benefit of ipilimumab on survival compared to standard medical treatment.
Combination therapy for CI
This was illustrated in a phase I trial by Wolchok et al (2013)4, which found that 53% of advanced melanoma patients treated with ipilumumab and nivolumab achieved an objective response, all exhibiting at least an 80% reduction in tumour size.
BUT combination therapy associated with significant increase in incidence of side effects
CTLA-4 mAbs effect on Tregs
• CTLA-4 antibodies also deplete regulatory T cells (Tregs), which constitutively express CTLA-4 (only expressed in other populations of T cells after activation) = perhaps contributes to side effects
• Depletion appears to be FcγR -dependent
• Simpson et al (2013)
- Found that CTLA-4 antibodies markedly reduced the accumulation of Treg cells within tumours in mice led to an increase in the intratumoral effector T cell/regulatory T cell ratio, and tumour rejecton
- Effect still seen in complement deficient C3 KO mice = depletion not dependent on complement
- No depletion or tumour clearance seen in γ-chain KO mice, lacking all three activating FcγRs depletion is dependent on FcγRs and ADCC
