By David Orchard-Webb
Synthetic peptide vaccines are a nascent but highly promising therapeutic class.
The field unites cancer biology and infectious disease treatment principles. There is, however, a key difference, a cancer vaccine is not prophylactic. Even though infectious disease prophylactic vaccines, such as against HPV, can prevent cancer, personalized cancer vaccines are an active treatment of an existing tumor.
This whitepaper focuses on peptide cancer vaccines, although infectious disease vaccines are also touched upon. Several considerations are important such as the peptide selection, and strategies to overcome loss of MHC. Personalization opens up regulatory challenges and these are also discussed.
Synthetic vaccines are short or long synthetic peptides derived from viruses and bacteria, fungi, cancer, as well as innovative T cell and innate immune stimulating immunogens and fusion proteins. They fall into two broad categories, cancer vaccines which can be personalized and peptide vaccines for infectious diseases.
These vaccines could be comprised of peptides as the main immunomodulatory agent or the peptides could be an adjuvant to a standard vaccine.
Peptide vaccines can be synthesized as peptides, or encoded in a vector, usually of viral origin. The argument for vectorized peptides is enhanced immunogenicity, however adjuvant and excipient technologies are capable of increasing the immunogenicity of stand-alone peptides, with more specificity than vectorized peptides, in some formulations.
There are a number of vaccination strategies available against cancers that utilize peptides, namely, cancer-overexpressed peptides, cancer neoantigens (nAgs), tumour tagging with exogenous pathogen-associated molecular patterns (PAMPs) peptides via viral transduction or DNA transfection, and adjuvants and excipients that activate the innate immune system.
“Tumor tagging” is a promising strategy when the tumor has employed immune evasion via down-regulation of the MHC receptors, which would normally nullify adaptive immune stimulating vaccine strategies.
More recently the discovery of an evolutionarily ancient “fitness fingerprint”, consisting of combinations of Flower membrane proteins, presents an opportunity to recognize tumors irrespective of MHC. Proteins that indicate reduced fitness are called Flower-Lose, because they are expressed in cells marked to be eliminated. [Madan] Human Flower isoforms hFWE1 and hFWE3 behave as Flower-Lose proteins, whereas hFWE2 and hFWE4 behave as Flower-Win proteins. Cancer cells have increased Win isoform expression and proliferate in the presence of Lose-expressing stroma, which confers a competitive growth advantage on the cancer cells. Tagging tumor cells with Flower-Lose isoforms via viral transduction is one possible strategy.
The best tumor associated antigens are not normally expressed in somatic cells, such as testis antigen NY-ESO-1. NY-ESO-1 antigen is commonly expressed in 10-50% of melanoma, lung, liver, esophageal, breast, prostate, bladder, thyroid and ovarian cancer cases, 60% of multiple myeloma cases, and 70-80% of synovial cell sarcoma. There are currently over 50 active clinical trials targeting NY-ESO-1. However, even this excellent overexpressed protein vaccination target may carry a fertility risk in males.
Unique to tumor cells, nAgs are expressed as a result of mutations in coding genes, generating a non-self peptide, without pre-existing immune tolerance. Unlike overexpressed tumor antigens that may also be expressed at lower levels in normal cells, targeting nAgs carries a relatively low risk of generating autoimmunity.
The potential for combination of nAg vaccines with immune checkpoint inhibitors is highlighted by the correlation between tumor mutational burden (number of nAgs) and improved response to the treatment. Interestingly in patients with pancreatic cancer tumors rich in neoantigens derived from MUC16 and peptides that mimicked viral peptides there was an association with long term survival. [Balachandran]
Individual nAgs can be concatenated, which can sometimes increase immunogenicity and increase the breadth of immune response. A scoring method has been developed whereby candidate nAgs are selected from personal next generation sequencing of the patient’s tumor based on: (i) MHC-I and II predicted binding affinity, (ii) mutation allele frequency in the tumor DNA, and (iii) mRNA expression. [D’Alise] This system can be used to select the top 30 neoantigens for incorporation into a vaccine in an automated fashion.
Successful personalized nAg vaccination strategies are dependent upon accurate prediction of high affinity binding to MHC. This can be accomplished with machine learning and deep neural networks. An interesting approach called MHCSeqNet deployed neural network architectures developed for natural language processing to model the amino acid sequences of MHC alleles and nAg peptides as sentences with amino acids as individual words. This innovation allows MHCSeqNet to accept new MHC alleles as well as peptides of any length. More generally bioinformatic services such as ElliPro can predict epitope sites from a given structure, allowing rationale design of peptides for use in vaccines against infectious disease.
Several companies are developing personalized cancer neoantigen vaccines including, but not limited to, Neon Therapeutics, an immuno-oncology company developing personalized neoantigen vaccines using massively parallel sequencing for detection of all coding mutations within tumours, and machine learning approaches. [Ott]
Neon has an agreement with Elicio Therapeutics. Elicio Therapeutics (formerly Vedantra Pharmaceuticals, Inc.) is developing an amphiphile platform that enables precise targeting and delivery of peptide antigens directly to the lymphatic system by interacting with albumin. [Moynihan]
Gritstone Oncology is developing personized cancer vaccines in a similar vein to Neon Therapeutics; however, they are delivering their neoantigens in a prime boost regimen derived directly from infectious disease vaccine principles. They have created the EDGE machine learning model of antigen presentation for neoantigen prediction, which they believe could enable an improved ability to develop neoantigen-targeted immunotherapies for cancer patients. [Bulik-Sullivan]
Agenus Inc. is developing personalized cancer peptide vaccines and targeting the peptides to the immune system by conjugating them with heat shock proteins. Some methods include autologous peptide extraction from patient biopsies. [Bloch]
Flow Pharma Inc.’s FlowVax platform is optimized to simultaneously deliver multiple nAg peptide targets. Each of the chemical components used in the FlowVax platform are currently part of an FDA approved vaccine or pharmaceutical, simplifying the regulatory pathway. Immuneoprofiler™ software harneses the power of artificial intelligence (AI) to identify clinically relevant and immunogenic neoantigens from next generation sequencing data. [Clancy]
A handful of companies are developing mutant RAS vaccines. The NantWorks family of companies headed by Patrick Soon-Shiong are making use of peptide vaccines including mutant RAS neoantigens, which were developed and manufactured by GlobeImmune, in combination clinical trials. [Cohn] Targovax holds IP for mutant RAS vaccines, and positive clinical data, but has chosen not to pursue development further at this time. [Birkeli]
Several companies are developing cell penetrating or lytic peptides that activate the immune system. AMAL Therapeutics SA, which was recently acquired by Boehringer Ingelheim, is developing a self-adjuvanting (TRL activating) peptide platform, with a multiantigenic domain with room for four antigens, and a cell penetrating peptide domain for efficient antigen delivery to immune cells such as dendritic cells. [Belnoue]
GV1001 is a 16-amino acid fragment of the human telomerase reverse transcriptase catalytic subunit (hTERT) developed as a cancer vaccine by KAEL-GemVax and also found to be a cell penetrating peptide through interaction with heat shock proteins. [Kim] Lytix Biopharma develops proprietary oncolytic peptides. Lytix's lead candidate, LTX-315, is developed for intratumoral treatment of solid tumors. [Sveinbjørnsson]
Peptide vaccines can make use of autologous cells. ERYTECH Pharma has developed the ERYMMUNE platform that uses artificially aged red blood cells and targets the majority of them to the spleen, where they undergo erythrophagocytosis and present their encapsulated peptides to Antigen Presenting Cells (APC), thus activating CD4/CD8 T-cells. [Banz]
The platform is being further developed by SQZ Biotech. Marker Therapeutics is developing a MultiTAA peptide therapy that is designed to educate T cells grown from stem cells. [Marker] LEAPS (Ligand Epitope Antigen Presentation System) is a CEL-SCI patented platform technology. J‐LEAPS vaccines promote the maturation of precursor cells into a unique type of dendritic cell that produce interleukin 12, but not IL‐1 or tumour necrosis factor, and presents the antigenic peptide to T cells.
The formulation of a vaccine can have a major impact on its immunogenicity. One of the major considerations in peptide vaccination strategies is how to get the peptides to the lymph nodes, where they can be presented to the immune system.
In cancer diagnosis, dye labelled albumin is crucial for identifying sentinel lymph nodes. The most abundant protein in the blood, albumin, naturally accumulates at tumours and sites of inflammation, and preferentially accumulates in the lymph nodes. This property can be used to direct peptide vaccines to the lymph nodes. This could be accomplished via fusion proteins or taking advantage albumin’s lipid binding properties.
Adjuvant technology such as lipid-based adjuvants, oil-in-water adjuvants, and plant-derived adjuvants may help target peptide vaccines to the lymph nodes. Lipid moieties may confer binding to albumin as suggested by the multiple lipid binding sites present in human albumin. [Kawai] Novel lipopeptides, LP1 and LP2, which mimic the terminal structure of the native Helicobacter pylori HpaA activate TLR2 and protect against bacterial colonization in mice.
As albumin’s apparent preference for the lymph nodes may be based on little more than its molecular weight, suitably sized non-albumin nanoparticles can also deliver peptides to the lymph nodes for antigen presentation. Virus-Like Particles (VLPs) are biological nanoparticles having a size of around 20-100 nm. Viral capsid proteins have tendency to self-assemble independently of genetic material and can incorporate suitably designed peptides. These VLPs are fully non-replicative.
From a regulatory perspective the need to ensure the quality safety and efficacy of peptide vaccines remains paramount before marketing authorisation is granted. This regulatory consideration is designed to ensure that the risk:benefit ratio is favorable to the patients receiving the treatments.
Pharmacovigilance and continued clinical safety data collection post-authorisation, which also may include registry data or additional real-world-evidence, is important to ensure that the risk:benefit balance does not become less favorable once licensed, for example with the appearance of autoimmune disease. Pharmacovigilance allows the identification of adverse drug reactions which were not evident in the initial clinical data. This consideration could prove extremely important for cancer vaccines based on overexpressed proteins that are present in some normal tissues.
The regulatory requirements for peptide vaccines as a class of product do not differ from other forms of biologics, however developers of nAg vaccines may require close discussions with the FDA in order to ensure that the personalized nature of nAg peptides can be appropriately processed by the FDA regulatory structure.
The regulatory agencies for Europe, the EMA, and the United States, the FDA, regularly release guidelines for a range of quality, non-clinical and clinical topics. The FDA Guidance for Industry on clinical considerations for therapeutic cancer vaccines provides a detailed discussion including the development of companion diagnostics for patient selection in clinical trials.
This is an evolving space and newer personalized vaccination strategies that use patient specific nAgs may not require companion diagnositics per se, as the tumors are routinely next generation sequenced as part of the vaccine development protocol.
However, and similar to EU guidelines, patient population selection and appropriate selection of a feasible endpoint remain important considertations. Surrogates such as monitoring of the immune response is considered exploratory by the FDA and the utility of such measurements are seen as useful in proof of concept, dose finding and possible correlation with clinical efficacy.
The development of exploratory biomarkers for proof of concept is supported by the FDA. Some guidance on adjuvants and multi-antigen therapy is also available. Given that peptide vaccines are still a nascent field, there are a wide range of approaches being undertaken, which are impossible to fully account for in any guideline document. It is therefore recommended that the main clinical endpoints are clinically relevant and discussed with the FDA.
Breakthrough Therapy designation is a process designed to expedite the development and review of drugs that are intended to treat a serious condition and preliminary clinical evidence indicates that the drug may demonstrate substantial improvement over available therapy on a clinically significant endpoint.
Developers of novel peptide vaccines may seek this designation in order to have a closer working relationship with the FDA and expedite the approval process, to the extent that is regulatorily possible.
Personalized peptide vaccines for cancer exemplify a paradigm shift in cancer therapy from a one size fits all approach utilizing very large numbers of participants in clinical trials to achieve a statistically significant overall improvement in survival, which in many individual cases equates to no improvement, to a more tailored approach that in some instances is n of one.
This approach entails the need for automated selection of appropriate antigens and the development of ever better antigen efficacy prediction algorithms. Adjuvants and excipients that improve immunogenicity are crucial, especially ensuring targeting of the lymph nodes, where efficient antigen presentation can occur. The field is evolving and so are the regulatory considerations.
Moving towards something approaching n of one clinical trial requires close consultation with the FDA. Achieving therapeutic results warranting a breakthrough therapy designation can provide the access to the FDA that is essential in this endeavour.