CORDIS

Executive Summary:
The Glioma Actively Personalized VAccine Consortium (GAPVAC) was funded by the European Commission under the 7th Framework Programme. The project started on November 1, 2012 and lasted 51 months until January 31, 2017. GAPVAC was implemented by a consortium of 12 beneficiaries (9 academic centres, 3 SMEs) from 6 European countries and Israel as Europe-associated country.
The scientific-technological work programme of the project was structured into 9 work packages (WP1-9) designed to cover GAPVAC’s preclinical (WP1-4) as well as clinical (WP5-9) objectives. WPs related to project management (WP10) and to dissemination and exploitation (WP11) were also included in the work programme.
The GAPVAC consortium initiated a pioneering project with the aim to realize a biomarker-guided, actively personalised and highly specific immunotherapy with the ultimate goal to improve patient treatment. The key objectives of the project were to test the safety, feasibility and effectiveness (biological and early clinical) of the GAPVAC approach of patient-tailored immunotherapy in patients with glioblastoma (GB). In a pan-European multi-centre phase I clinical trial each patient received custom-made Actively Personalised Vaccines (APVACs). Each APVAC was unique and fully tailored to the patient’s individual tumour and immune system by a biomarker-guided process to maximize the anti-tumour immune responses. Patients were immunized with two different kinds of APVACs: APVAC1 vaccines were composed of peptides selected from a pre-defined warehouse according to the patient’s individual disease and immunological status. Peptides have been selected into the warehouse because of their high tumour association as observed in a set of GB samples analysed in the preclinical phase of the project (shared antigens). APVAC2 vaccines were composed of de novo synthesized peptides - mostly mutated neoantigens - based on analysis of tumour’s mutanome, gene expression profile and peptidome.
Overall, 16 patients have been enrolled in the GAPVAC-101 phase 1 trial. The APVAC selection processes including pathological review of the tumour specimen, peptidome, transcriptome and immunogenicity analysis as well as next generation sequencing worked very well for all patients. Both, APVAC1 and APVAC2 vaccines, were successfully selected for 16 and 15 patients, respectively. Importantly, all selected APVAC compositions were unique and showed a high level of personalization. There were no failures in the customized APVAC manufacturing; finally, 15 patients received APVAC1 vaccines, while 11 patients received APVAC2 vaccines. Thus, the feasibility of the GAPVAC concept has clearly been demonstrated. In March 2017, the regular treatment phase of the trial has been completed. However, some patients and their treating physicians have decided to continue vaccinations even beyond the regular schedule to fully exploit the potential benefit from the APVACs.
The treatment was well tolerated demonstrating the favourable safety profile of the vaccinations. Primary immune response analysis showed high biological activity: 90% of patients showed APVAC vaccine-induced, persistent immune responses. CD4 T cell responses were mainly of the TH1 phenotype.
First clinical data have already been presented at the CIMT Annual 2016 meeting in Mainz and during the AACR Annual Meeting 2016. Several scientific publications on the outcome of the GAPVAC project are planned during the next months.
Please visit the project webpage at www.GAPVAC.eu for further information.

Project Context and Objectives:
4.1.2.1 The concept
Glioma is a rare and fatal disease with 2-3 incidences in 100,000 persons. It is the most common and aggressive type of malignant brain cancer with a very high medical need; the 5-year survival rate is less than 6% with conventional therapy.
The relatively high prevalence in young patients, the high mortality rate, and the limited treatment options indicate the need for a concerted effort to develop new therapies. Currently available therapies usually affect tumour progression only temporarily, are accompanied by significant side effects and do not consider the individual aspects of each patient’s tumour and immune system.
Personalised cancer therapy today usually just means the biomarker-guided, stratified use of a conventional drug and does not fully account for the unique characteristics of each cancer. The GAPVAC project was designed to take personalisation to the next level by tailoring unique peptide cancer vaccines specifically for each patient’s disease. The employed concept of cancer vaccination aims at stimulating the patient’s immune system by activation of T cells against the cancer. These T cells are able to specifically recognize the tumour cells and to destroy them.
The core of the GAPVAC project was a European multi-centre phase I trial (GAPVAC-101 trial). Based on comprehensive biomarker analyses, patients suffering from glioblastoma (GB) were treated with a multi-peptide vaccines tailored to the individual genomic and peptidomic profile of the patients’ tumours. Two of those actively personalised vaccines (APVACs) were used in the GAPVAC project:
• APVAC1 vaccines: These vaccines comprise tumour-associated peptides frequently found overexpressed in the tumour tissue of GB patients (shared antigens). In the preclinical phase of the project, suitable peptides were identified, selected and pre-synthesized. For an individual patient in the GAPVAC-101 clinical trial peptides fitting best to the patient’s tumour and immune status are then readily available from this so-called “peptide warehouse” for inclusion into the individualized APVAC1 vaccine.
• APVAC2 vaccines: These vaccines comprise peptide antigens that are unique to the patient’s tumour and thus have to be synthesized de novo for the individual patient. A major source of APVAC2 antigens are peptides that bear tumour-specific mutations: Cancer genome instability and subsequent selective pressure lead to the accumulation of tumour mutations. Thus, mutated peptides are exclusively expressed on the HLA molecules in the tumour but not in healthy tissue. Mutations are in most cases highly specific to a given tumour, thus these peptides are not suitable for a warehouse concept, but are synthesized de novo for inclusion in APVAC2.
This two-pillared strategy of APVAC preparation is summarized in Figure 1. The preparation of an APVAC1 for an individual patient, from identification of all relevant peptidomic markers, up to the ready-to-use-GMP (good manufacturing practice) product takes less than 3 months. For APVAC2 mutation analyses and subsequent de novo synthesis of mutated peptides starts in parallel but takes about 3 additional months, thus they are ready for vaccination 6 months after patient enrolment.
Patients in the GAPVAC-101 trial are treated using a three-step therapeutic approach starting immediately after tumour resection and being applied concomitantly to standard treatment with conventional chemo-radiotherapy:
• In the first step, the patient’s tumour and blood samples are analysed to produce a comprehensive dataset of biomarkers enabling the design of the two completely individualised vaccines APVAC1 and APVAC2 for each patient. Adjuvant chemotherapy cycles with temozolomide (TMZ) continue as concurrent standard therapy.
• In the second step, vaccinations with APVAC1 and the immunomodulators GM-CSF and Hiltonol are initiated following the first adjuvant TMZ cycle.
• In the third step, vaccinations with APVAC2 start after the fourth adjuvant TMZ cycle, while APVAC1 vaccinations continue.

Figure 1: Two-pillared strategy of APVAC generation
GB tumours are analysed straight after surgery for expression of tumour-specific antigens, corresponding peptides are selected from the peptide warehouse (left) or are synthesized de novo (right). Patients are then vaccinated with two fully personalised vaccines APVAC1 and APVAC2 in combination with the immunomodulators GM-CSF and Hiltonol.

To date, the expression ‘personalised medicine’ has often been used for the stratified use of established drugs. GAPVAC takes personalised medicine to the next level by providing ‘true’ personalisation by a) developing a fundamentally novel approach of therapy and b) by creating a biomarker-guided custom drug product for each patient based on the individual molecular properties of the patient and his or her disease. A further important part of the project is the inclusion of a comprehensive T-cell immunomonitoring and biomarker program to describe the mechanism-of-action and support further clinical development. As an additional benefit, this enables the development of new tools to monitor and predict the clinical outcome of patients, and thus to identify patients that are more likely to respond to immunotherapy.
The long-term clinical effects of checkpoint inhibitors and their recent approval in many cancer indications clearly demonstrate that immunotherapy can be a powerful tool against cancer (recently reviewed by (Guo et al., 2017)). The GAPVAC project aims at combining the power and promising effects of immunotherapeutic modalities with a cutting-edge approach of biomarker-guided active personalization with the final goal to improve the therapy options for cancer patients.
4.1.2.2 The objectives
The main objective of GAPVAC was to assess the safety, feasibility and first signs of efficacy (by comprehensive monitoring of vaccine-induced immune responses) of a completely novel approach to treat cancer patients with actively personalised vaccines (APVACs).
Further GAPVAC objectives were:
• To establish a panel of biomarkers associated with induction of immune responses and clinical endpoints
• To compare the immunogenicity of mutated vs. non-mutated antigenic peptides
• To discover and validate tumour-associated peptides presented by soluble HLA molecules as an additional antigen source and as markers of tumour burden and relapse
• To standardise the collection, shipment and storage of biological samples to enable current and future assessments for centralised assays
• To harmonise sample collection and the conduct of in vitro assays for local immunomonitoring
• To establish a high-level European network for research on novel personalised immunotherapy for the treatment of glioma patients.
To achieve these objectives, the GAPVAC project was subdivided into a preparatory pre-clinical part (WP1-4) and a part focussed on the clinical trial (WP5-9). In addition, the project included two general work packages (WP10-11).
I. Pre-clinical/translational project phase (WP1-4)
In the pre-clinical/translational project phase, the consortium facilities were set-up and the GAPVAC-101 clinical trial was prepared. Already at this stage of the project, consortium members contributed to the tumour and plasma sample collection (WP1) and were involved scientifically to produce a definition of the concept of GAPVAC and the associated clinical trial (WP2-4). The research program focused on two key areas: the identification of laboratories for biomarker and immunomonitoring tasks (WP2) which were essential for evaluation of biological activity of APVACs during the GAPVAC-101 trial and the set-up of the APVAC1 warehouse of glioblastoma-associated peptides as basis for patient individualized vaccination (WP3). Simultaneously, preparation of the clinical documents and discussions of regulatory issues took place. In addition, scientific advice from regulatory authorities was obtained (WP2). ‘Preclinical dry runs’ for identification of mutated TUMAPs (= tumour associated antigens) with subsequent immunogenicity analyses (WP4) and GMP TUMAP manufacturing were performed in order to eliminate potential drawbacks, to streamline all associated process steps early in the project and to establish a functional collaboration and logistics structure between the consortium members.
II. Clinical project phase (WP5-9)
The central part of the GAPVAC project was a multi-centre phase I clinical trial (GAPVAC-101) assessing APVACs as a treatment for GB patients. Safety, feasibility and biological activity are the primary endpoints of the project. Moreover, a number of secondary endpoints should be investigated based on a comprehensive analysis of cellular and non-cellular biomarkers analysed from patients’ blood and tumour specimens and their association with the clinical outcome of the patients. The ultimate aim of all these analyses is 1) to gain a deeper understanding of the mechanisms of action of APVACs; 2) to identify parameters that could be improved in future trials following proof of the APVAC concept; and 3) to identify correlates for disease prognosis and benefit from immunotherapy.
Patients were recruited into GAPVAC-101, by the consortium’s clinical sites shortly after diagnosis of primary GB and prior to resection of the primary tumour (WP5). Tumour samples were analysed (including pathological review of tumour tissue, peptidome, transcriptome and immunogenicity analyses and next generation sequencing), and APVACs were individually selected for the patients specific tumour. The patients were then treated with standard chemo-radiotherapy. Their therapy was then supplemented with personalized APVAC1 and APVAC2 vaccinations (WP6). To ensure the successful and timely implementation of the GAPVAC-101 clinical trial, translational efforts covering regulatory aspects and the preparation of GAPVAC clinical documents, as part of WP3, were begun right from the start of project as part of the pre-clinical/translational work programme.
The biological activity of each APVAC was very closely monitored for each individual patient in a large biomarker and immunomonitoring programme (WP7) which included the local assessment of tumour-infiltrating lymphocytes (WP8) and the sHLA peptidome (WP9). All clinical work packages were performed in parallel and all members of the consortium collaborated closely in patient recruitment, APVAC manufacturing, treatment and monitoring. Safety and feasibility data were analysed continuously.
III. General provisions (WP10-11)
The overall management of the consortium was led by Immatics GmbH (IMM) as coordinator, and co-led by BioNTech AG (vice-coordinator, BIO). Members of the GAPVAC project management board (PMB) oversaw that GAPVAC met its objectives and time lines. This project management structure which provided a tight connection between the GAPVAC coordinator and the work package leaders was very successful. It ensured that potential project delays and bottlenecks were identified early and collaborative solutions were implemented using the members collective resources before the project was threatened. While internal revision of project milestones was performed by the PMB, certain administrative tasks (such as funding distribution to the consortium partners) was performed by the coordinator in order to facilitate fast and direct execution of these tasks. This approach also ensured maximum transparency to the project partners and reflects the overall project responsibility of the coordinator.
The work packages and their contents are summarized in Figure 2.

Figure 2: GAPVAC work packages
4.1.2.3 The partners
GAPVAC was led by two highly innovative biotechnology companies based in Europe: Immatics biotechnologies GmbH (IMM, Coordinator) and BioNTech AG (BIO, Vice Coordinator) who are dedicated to applying the best science for the benefit of patients. Ten academic centers in Europe recruited patients and collected scientific data for the GAPVAC-101 clinical trial. Academic centers were located in Barcelona (Spain), Copenhagen (Denmark), Geneva (Switzerland), Haifa (Israel), Heidelberg (Germany), Leiden (The Netherlands), Southampton (UK) and Tuebingen (Germany). The personalized vaccines were manufactured at BCN Peptides (Barcelona, Spain) and at the University of Tuebingen (Tuebingen, Germany). Moreover, first results from the study were released and distributed by the Association for Cancer Immunotherapy (CIMT) who will also contribute to the scientific analysis.
Hideho Okada from UCSF (San Francisco) is an associated member of the GAPVAC consortium and was a main contributor to the clinical concept and study protocol.

Project Results:

Overview
GAPVAC has been a highly successful project with numerous scientific results being generated. Below the GAPVAC work packages and major scientific results are summarized.

WP1 Tumour and plasma sample acquisition
Objectives:
Objectives of WP1 were i) to establish a protocol for a non-interventional study for biosampling and scientific analyses from GB patients, ii) to receive ethics committee approval to conduct this study and iii) to collect respective specimens (blood, plasma and tumour samples) from GB patients, which were essential for the successful implementation of WP2, WP3 and WP4.

Summary of progress towards objectives
A study protocol and respective patient informed consent forms were prepared for a non-interventional clinical study prior to the start of sample acquisition. Biosampling of blood, plasma and tumour samples of GB patients as well as the scientific analyses of the HLA status, the immune peptidome, the genome and the T-cell repertoire of patients was undertaken. After preparation of the trial documents for the non-interventional study, the research trial protocol was submitted to and was subsequently approved by all of the local ethics committees. Specimens from a total of 60 glioma patients were collected and these were used to set up a technical framework to translate the concept of an actively personalized vaccination into clinical development. This was an important prerequisite for successful initiation and conduct of the GAPVAC-101 clinical trial. All objectives of this work package were successfully accomplished on time.


WP2 Set-up of consortium facilities for the clinical trial GAPVAC-101
Objectives:
The aim of this work package was to set-up the consortium facilities for the clinical trial GAPVAC-101 including setting-up a laboratory network for PBMC (peripheral blood mononuclear cells) collection, APVAC development procedures, as well as development of a harmonized TIL isolation procedure. Further objectives of this work stream were to seek scientific advice from the European regulatory authorities and to prepare the clinical documents and operations necessary for the GAPVAC-101 trial.

Summary of progress towards objectives
The personalized approach in the GAPVAC-101 trial required the setup of several processes: a) collection of tumour, blood and leukapheresis samples, b) review of tumour sample quality, c) assessment of four different biomarkers (tumour HLA peptidome, tumour mRNA transcriptome, tumour mutanome, and an assessment of the patient’s immune response against the GAPVAC warehouse peptides), d) data integration and unbiased selection of the drug composition prior to e) drug manufacturing. These complex operations involved five partners situated at different locations across Europe. Thus, logistics, communications, data exchange, biomarker assessment and the personalized APVAC selection had to be thoroughly planned and well-coordinated. For GAPVAC, the standardised collection of high-quality human PBMCs from patients treated as part of the trial was essential for the performance of i) the immunogenicity testing during APVAC selection and ii) the immunomonitoring program described in WP7. Therefore, clearly defined procedures for the centralized (PBMCs) and decentralized immunomonitoring (TILs) operations for the clinical trial were setup. This included also the logistics of temperature controlled sample shipment.
PBMC laboratories at the various clinical sites were qualified and their lab staff was appropriately trained before first patient samples were collected.
Clinical documents (Clinical Study Protocol, Investigator Medicinal Product Dossier, Investigator Brochure and Informed Consent Forms) were prepared and then submitted initially to the Paul-Ehrlich-Institute, the national authority in Germany. Documents were reviewed by the authority and approved after minor modifications in August 2014.
Thereafter ethics committees and national authorities in Denmark, the Netherlands, Switzerland and Spain reviewed the GAPVAC-101 trial documents, which were approved after discussion on certain open questions and modification of the trial documents. Each agency had its own highly specific questions and the overlap was limited, reflecting the novelty of the APVAC approach also from a regulatory perspective. Significant efforts were taken to resolve the requests in each of the countries. Subsequent submission, discussion and resolution of requests, led to a step by step study approval in the participating countries. Working through the requirements of the regulatory authorities led to a few month delay of the study start (WP5) compared to the initial plan.
To setup a harmonized TIL procedure, a detailed survey was organized among GAPVAC clinical centres. Participants provided their individual protocols for TIL isolation and expansion (if available) as well as their experience with such protocols regarding feasibility, cell recovery and cell subsets analysis. An approach for a harmonized TIL isolation procedure was defined from the survey´s results and was agreed with the participating clinical centres. Following some further optimizations, as well as additional rounds of protocol harmonization, a harmonized protocol for tumour processing, TIL expansion and TIL screening was formulated and distributed to the GAPVAC clinical partners. For those labs unable to culture TILs, the protocol also contained re-commendations for freezing of fresh tumour samples, which were then sent to another GAPVAC lab for culture and test.
In summary, the objectives of WP2 were fully achieved in time with the minor exception of the delay in the transition to the start of the clinical trial (WP5). This was due to the regulatory challenges described earlier which were nevertheless successfully solved by the GAPVAC consortium.


WP3 Warehouse Set-up for APVAC
Objectives
The primary objective of this work package was the definition and GMP manufacturing of the GAPVAC warehouse composed of tumour-antigens relevant in GB and binding to the HLA alleles A*02 and A*24. The warehouse was the basis for the personalized APVAC1 vaccines in the GAPVAC-101 trial.

Summary of progress towards objectives
• Definition and GMP manufacturing of the antigen warehouses
The first step was the identification of TUMAPs from primary GB samples. Tumour association was determined by mass spectrometry (immunopeptidome level ) as well as gene expression profiling (mRNA level). 46 shock-frozen GB tumour tissue samples in total were analysed and HLA-associated peptides were identified by mass spectrometry (MS). These analyses resulted in 4321 and 1771 unique TUMAPs respectively being identified.
In order to identify the most suitable target candidates for the GAPVAC warehouse, screening for interesting TUMAPs was primarily based on quantitative peptide presentation data for the identified A*02 and A*24 binding peptides, being two of the most abundant HLA alleles in the European population. This focus was based on the fact that this immunopeptidome data reflects directly what is visible to T cells and thus to the active cancer-attacking cells of the immune system. In this first step, 276 HLA-A*02 and 208 HLA-A*24 interesting peptides were selected. Relative presentation levels on tumour samples were compared with presentation levels on a range of already available normal tissue reference samples to exclude antigens with too high levels on healthy organs and thus with an inherent risk to trigger autoimmunity.
Moreover, microarray experiments were performed on the GB samples as well as normal tissues in order to assess degree of mRNA over-expression of the source genes for all GAPVAC warehouse candidates in GB samples and to estimate how frequently the respective antigens may be relevant for different GB patients (the so-called antigen prevalence). Briefly, RNA was analysed from 26 GB specimens. A library of gene expression data was already available at IMM including data from a large number of different healthy tissues and organs. Based on this dataset, the number of interesting peptides for the GAPVAC warehouse was further reduced. For the remaining candidates for the GAPVAC warehouse, a comprehensive search of the published background literature and other available information was performed.
121 HLA-A*02 and 106 HLA-A*24 interesting TUMAPs were selected for non-GMP synthesis to assess their manufacturability and 88 HLA-A*02 and 87 HLA-A*24 TUMAP candidates were screened for in vitro immunogenicity using highly standardised in vitro assays. Briefly, CD8+ cells were enriched from PBMCs of leukapheresis products of healthy HLA-A*02+ or HLA-A*024+ donors. Cells were then stimulated with artificial antigen presenting cells coated with peptide-MHC complexes (pMHC) and anti-CD28 in the presence of interleukin-2 and interleukin-12. After three weeks of stimulation, the percentage of in vitro primed T cells was determined by MHC-multimer cytometric analysis, utilizing a two-dimensional combinatorial coding approach.
After integration of all data sets, top ranked TUMAPs were selected to define the GAPVAC warehouse. This consisted of 38 HLA-A*02-binding and 28 HLA-A*24-binding, immunogenic TUMAPs with high relevance for GB. Additionally 2 viral marker antigens and 3 pan HLA-DR binding antigens were selected for the warehouse. Selected peptides were synthesized in GMP quality by partner BCN Peptides to generate the physical GAPVAC warehouse that was the key repository for the GAPVAC-101 clinical trial.
The average prevalence of over-expression of GAPVAC warehouse antigens ranged from 54% to 100% for the single antigens. In addition, the degree of over-expression was generally high; for 62% of the GAPVAC warehouse peptides source genes a more than 5 fold mRNA over-expression was observed in GB samples compared with normal tissues. For all but one chosen warehouse peptides, specific T cell precursors were found, indicating that no general immune-tolerance existed for these peptides.
A total of 6 test runs for APVAC drug development were performed. Test runs 1-4 included biomarker-guided selection of APVACs while test runs 3-6 included de novo synthesis of peptides as active pharmaceutical ingredients for APVAC2 as well as formulation of APVAC drug products. APVAC selection was possible in each of the 4 test runs with no or only minor delays. However, the DryRuns were important steps especially for manufacturing as formulating the APVAC1 and APVAC2 vaccine cocktails experienced initial difficulties in DryRuns 3 and 4 including delays and production failures. To mitigate the risk of failures in APVAC production during the clinical trial, two additional test runs were implemented. Thereafter, a root cause analysis of all test run failures was performed and revealed that formulation failures were related to single antigens. These problematic antigens were removed from the antigen warehouse. The investigational medicinal drug dossier was also updated to allow the removal of such antigens by the manufacturer. As a result, no severe problems that could have led to an ultimate failure in the APVAC manufacturing process for future formulations for the clinical trial occurred. Following the removal of these antigens the warehouse finally consisted of 33 HLA-A*02-binding and 26 HLA-A*24-binding TUMAPs.

• Identification of TUMAPs from soluble HLA molecules isolated from patients’ plasma
During the preclinical phase of the project the soluble HLA (sHLA) peptidome from 27 samples of plasma from 17 GB patients was purified, analysed by mass spectrometry and compared to the sHLA peptidomes of plasma taken from 6 healthy donors. 9,420 peptides were identified that were derived from 4,617 different proteins. In addition, the sHLA peptidome of the plasma was compared to the membranal immunopeptidome of four glioblastoma cell lines. From the cell lines a total of 11,161 different peptides derived from 5,221 different proteins were identified. These peptides were bound to a variety of HLA types including A*02 in two cell lines and A*24 in one.
Analyses of soluble HLA-peptide repertoires analysed from blood samples taken from GB patients showed many differences between the peptidomes of the patients before and after surgery. This was both in the repertoires of the identified peptides and in the signal intensities of others suggesting that sHLA analysis may be a valuable tool for non-invasive assessment of tumour progression and early detection. During the clinical phase the data basis of sHLA peptidome was further expanded (refer to WP9).
In summary, the objectives of WP3 were fully met and on time and the GAPVAC warehouse was successfully established allowing to provide APVAC1 compositions to the GAPVAC-101 trial.


WP4 Identification of mutated TUMAPs for APVAC2
Objectives
The goal of this WP was to use the human samples collected as part of the WP1 to perform genomic mutation analysis by whole exome sequencing of GB material and autologous healthy tissue. This was followed by MS analyses to allow the identification of the mutated peptides presented by HLA molecules on the surface of corresponding tumour material. A further aim was to test the feasibility of mutation discovery in a real-life setting and to further optimize processes for the clinical trial GAPVAC-101. Moreover, the frequency of T-cell responses of healthy blood donors responding to mutated and native, shared TUMAPs were assessed by high throughput TUMAP validation.

Summary of progress towards objectives
• Identification of mutated peptides for APVAC2 vaccines
The consortium was able to demonstrate proof of concept for identifying mutations in pre-clinical model systems using next-generation sequencing (NGS), followed by identification of mutation-containing peptides on MHC molecules by mass spectrometry (MS). This included identification of mutations in pre-clinical tumour mouse models B16F10 and GL261. In addition, the consortium used NGS to profile additional pre-clinical samples and human samples.
Having established the processes and successfully tested them in the pre-clinical workflows, the consortium conducted test runs with GB clinical samples. The consortium was able to successfully execute the mutanome assessment by NGS and the immunopeptidome analysis using MS. It was also able to exchange the results securely, integrate data sets, prioritize targets, and identify targets for APVAC2 vaccine production. Thus, this proof-of-concept showed that assessment of private mutations in GB and the integration of the results into a personalized vaccine was feasible and could be applied in the GAPVAC clinical trial.
Weak points in the process were identified, discussed and improved. For example, different clinical sites measured nucleic acid concentrations and quantities differently, resulting in problems. Through the tests runs, the consortium was able to identify such issues and to agree and implement a harmonized procedure.
Furthermore, the consortium established processes for identifying and integrating APVAC2 targets from three sources for clinical application in the GAPVAC-101: a) mutation-containing peptides found on tumour HLA molecules, b) mutation-containing peptides predicted but not confirmed to be on tumour HLA molecules and c) tumour-specific, over expressed peptides without mutations found on tumour HLA molecules.

• Immunogenicity analyses of mutated vs. wild-type TUMAPs [IMM]
Four pairs of model peptides of a mutated and a non-mutated counterpart were analysed for their in vitro immunogenicity to assess whether there is a general rule for mutation-containing antigens vs. non-mutated and thus self-antigens with regard to immunogenicity. For two pairs, natural presentation of both forms (mutated and non-mutated peptide) was confirmed by mass spectrometry. For the other two pairs, the mutant peptides were predicted HLA epitopes without experimental confirmation of HLA presentation. Three out of the four tested peptide pairs were immunogenic. Comparative analyses of HLA-A*02 positive and HLA-A24 positive peptides contained in the GAPVAC warehouse showed that in vitro immunogenicity of the tested pairs of wildtype but also mutated peptides was comparable to the immunogenicity of the warehouse peptides.
In summary, the objectives of WP4 were fully achieved on time and thus paved the way to implement the APVAC2 approach into the clinical realization of the GAPVAC concept.


WP5 Multi centre open-label clinical phase I trial for personalised vaccination
Objectives
Objective of WP5 was to evaluate the GAPVAC concept of active personalization in a European multi-centre phase I trial with up to thirty glioma patients treated at multiple clinical centres across Europe. Primary endpoints of the clinical trial were i) feasibility of an actively personalized vaccination in GB patients, ii) safety of APVAC vaccinations, and iii) biological efficacy.

Summary of progress towards objectives
The GAPVAC-101 clinical trial achieved its first approval by the competent authority in Germany in August 2014. Regulatory approval in other participating countries was challenging (see WP2), but finally resulted in a total of six fully approved clinical sites in Germany (Heidelberg and Tübingen), Switzerland (Geneva), Spain (Barcelona), Denmark (Copenhagen) and the Netherlands (Leiden). Initiation visits were performed shortly after approval to ensure that the recruitment and treatment phase of the trial could start as quickly as possible. A total of 16 patients were eventually enrolled at all six clinical sites.
The first patient was enrolled in Heidelberg in December 2014, while the last patient was enrolled in Geneva in June 2016. Treatment with APVAC vaccines then continued and is still ongoing for a few patients that decided together with their treating physicians to enter so called “continued vaccination” to further benefit from their personalized APVAC vaccines. However, the regular treatment phase that defines also the main data collection period for the assessment of the objectives of the GAPVAC-101 clinical trial was terminated in March 2017.
The completion of the interventional phase of the GAPVAC trial (initially planned for January 2017) was delayed due to 1) challenges that occurred during the regulatory approval due to the novelty of the concept, 2) a transient drop in patient recruitment numbers at the end of 2015, and 3) a longer than initially expected treatment phase for enrolled patients. But importantly, all primary endpoints of the clinical trial were successfully met: The APVAC processes proved to be feasible in the real-life conditions of a clinical trial (primary endpoint 1) and patients generally tolerated the treatment very well (primary endpoint 2). Finally, data received from immune response analysis suggest high biological activity of APVAC vaccinations as sustainable immune responses were successfully induced by the treatment in 90% of evaluable patients (primary endpoint 3).

• APVAC design and manufacturing
GB patients were considered for the GAPVAC-101 trial shortly after diagnosis and even before their initial surgery to remove as much tumour mass as possible from the brain. If a patient then met all inclusion criteria, the individual APVAC selection process started shortly after surgery. Genomic and HLA peptidome data were gathered from the patient’s tumour material and immunogenicity data were generated from the patient’s blood. This dataset was used to identify the most suitable APVACs for each patient consisting of up to 10 TUMAPs from the warehouse (formulated into APVAC1) and up to 2 patient-unique (i.e. mutated) TUMAPs formulated into APVAC2. Selected patient-specific peptides from the warehouse were formulated into an APVAC1 drug product at the consortium’s GMP facility (TUE).
For APVAC2 the unique and patient-specific peptides had to be synthesized de novo at TUE (WP6) before APVAC2 drug products were formulated for each individual patient.
The biomarker-guided APVAC selection including peptidome analysis and NGS worked very well for all patients. During the clinical trial, no APVAC1 selection or its production failed and APVAC1 vaccine compositions were successfully selected for all 16 patients. APVAC2 vaccine compositions based on mutated and/ or over-presented peptides were successfully selected for 15 patients. For one patient, no APVAC2 could be selected as the poor quality of the tumour sample was insufficient to allow analysis, and for two patients, production of APVAC2 was stopped as they dropped-out early from the study. Some manufacturing issues occurred, but could be solved and then resulted in a delay of one APVAC1 and two APVAC2 productions. These delays were easily manageable due to the option for a delayed start of treatment implemented into the clinical study for such cases. The University of Tübingen reliably manufactured the individualized vaccines and has been able to cope with all production issues that emerged due to the uniqueness of each formulation.
In total, 15 patients received APVAC1 vaccines (one patient dropped-out before the start of vaccinations) and 11 patients received APVAC2 vaccines. The drop-out rates were slightly lower for APVAC1 (expected: 15%, real: 6%) and slightly higher for APVAC2 (expected: 25%, real: 31%) than anticipated from statistical considerations before the start of the trial. The last patient had his/her regular last study visit in March 2017, allowing the consortium to complete the analyses required for the clinical study report. Survival follow-up will continue until June 2018. Thereafter, an amendment to the study report will complete the clinical trial.
Further details on the APVAC production process and progress are described in the associated WP6 “APVAC preparation”.

• Tumour assessment and sampling schedule
Staging by MRI scans of the brain of patients was performed directly post initial surgery, before vaccinations started and then at least every 3 months or more often if clinically indicated or to rule out pseudoprogression. Physical examinations as well as routine lab work were done prior to each vaccine administration. Plasma samples were taken before and after the initial surgery for quantification of sHLA-peptides (WP9). Serum for biomarker analyses was collected at several time points (WP7) and blood samples for PBMC isolation with subsequent immunomonitoring were taken before vaccination and at several post-vaccination time points (WP7). If possible, TILs were isolated from the tumour samples taken at the time of the initial surgery and may be taken potentially during re-resection at tumour recurrence at later time points (WP8).

• Study management
IMM acted as the sponsor of the GAPVAC clinical trial coordinating and managing the operational set-up, as well as the execution and completion of the study. IMM is responsible for the preparation and management of all elements of the clinical trial documents such as the clinical study protocol, the investigators brochure, the case report forms (CRF), the submissions to regulatory authorities, drug supply, oversight of Contract research organization (CRO) services (e.g. site initiation and monitoring) and pharmacovigilance. This was done in close collaboration with the clinical sites as well as Prof. Wolfgang Wick (Heidelberg) as Lead Investigator and Prof Pierre-Yves Dietrich (Geneva) as Co-Lead Investigator. During the study the clinical documents were continuously updated and amended to respond to requirements of national authorities and ethics committees but also to scientifically driven decisions of the consortium.
In addition, IMM as the sponsor managed all of the activities of the CROs (clinical CRO and pharmacovigilance CRO) involved in the study. IMM also coordinated patient visits including the various corresponding samples and logistics. The individual clinical centres recruited and treated patients according to the study protocol; they monitored patient safety, performed radiological analyses, assessed patient survival, and completed the CRF documentation. A great deal of attention was given to the standardised collection of the biological samples needed for immunomonitoring and translational research.
The investigators asked the sponsor for confirmation of the availability of an APVAC drug production slot for 30 candidate patients. All 30 enrolment requests were processed by IMM and after planning and scheduling the pre-specified APVAC procedures (incl. all analysis steps and manufacturing), production slots were confirmed for 21 patients and the “green light” for further screening/enrolment was communicated to the respective sites. 16 patients were then finally enrolled by the clinical sites. It is important to note that this level of patient enrolment does not reflect fully the actual interest of patients to participate in the GAPVAC-101 trial, as many more slots were asked for by investigators informally: These respective patients were then not consented or enrolment was not officially requested, as lack of production capacities was already clear e.g. from previous email communication.
Of the patients who were enrolled in the study, one dropped out before receiving the first APVAC1 vaccination. All other patients entered the APVAC vaccination phase of the study. All patients completed the treatment phase, either regularly or prematurely (e.g. due to disease progression). Eight patients have chosen the option of continued vaccination (beyond the last regular vaccination); two of them are still continuing to receive APVAC vaccines and may receive further vaccinations until Q3 2017. Most of the remaining patients are in the follow-up phase, which is measuring overall survival. This will be continued until October 2018.
A scheme of the clinical trial design and the status of the enrolled patients is given below in Figure 3 (Status April 2017).


Figure 3: GAPVAC-101 trial design and status of enrolled patients
Upper part: Study scheme of GAPVAC clinical trial. Conventional therapy is comprised of initial chemo-radiotherapy followed by adjuvant TMZ cycles. After initial surgery and confirmation of diagnosis, peptidomic and genomic biomarker analyses are performed and the design and manufacturing of the individual APVACs starts. During the off phase of the first adjuvant TMZ patients will receive their first vaccinations with APVAC1. During the off phase of the fourth adjuvant TMZ cycle, APVAC2 vaccinations commence concurrently to adjuvant TMZ therapy cycles. Vaccination time points are indicated by arrows in the upper part of the chart above.
Lower part: Status GAPVAC-101 patients (April 2017)

• Safety and feasibility assessment
Primary endpoints of the clinical trial are the safety of the patients, the feasibility to perform a personalized approach of patient vaccination and the first information on biological activity.
Safety analyses show a favourable profile: As expected from the mechanism of action of peptide vaccines and the used immunomodulators, mild to moderate injection site reactions were the most frequent toxicities observed. Eleven SAEs have been reported in 9 patients; 3 have been assessed as at least possibly related to the study treatment. The authorities as well as to clinical sites were informed about these events as applicable and as required by law and regulations. During the treatment phase of the study no deaths were reported (four patients died during the follow-up period due to tumour progression). During four regular meetings the independent Data Safety Monitoring Board (DSMB) repeatedly recommended that the study could continue without modifications as no specific safety signals were observed.
Feasibility of patient treatment is dependent on timely sample analyses of the patients’ tumour material and blood, composition, manufacturing and delivery of personalized APVACs to the clinical sites. The complex logistics procedures were performed by professional logistics partners and were monitored and coordinated by exclusively dedicated personnel. Details about the feasibility of APVAC1 and APVAC2 vaccines of the 16 enrolled GAPVAC patients are given in WP6. None of the APVAC1 and APVAV2 selection processes or productions failed for technical reasons and the vaccines were provided for vaccinations according to the study protocol. A single failure of APVAC2 selection was due to a highly necrotic tumour samples (as already stated by central pathological review) and was not caused by a failure in the APVAC process chain itself.
Results for the third primary objective of the GAPVAC-101 trial (biological activity) are provided in WP7.


WP6 Preparation of the personalised vaccine (APVAC)
Objectives
Objective of this work package was the timely preparation of GMP-grade APVAC1 and APVAC2 vaccines for individual GB patients.

Summary of progress towards objectives
The feasibility of APVAC1 and APVAC2 drug development processes were clearly demonstrated during the preclinical phase of the project (WP3). In the clinical setting these processes were again successfully performed meeting the strict timelines and standards of this highly controlled and regulated environment. Overall, 16 APVAC1 and 15 APVAC2 vaccine cocktails were selected and 16 APVAC1 and 13 APVAC2 vaccine cocktails were produced.
Of the 16 enrolled patients, 12 were HLA-A*02:01 positive, 3 were HLA-A*24:02 positive, and 1 patient was positive for both HLA-alleles. For one patient, no MS and NGS analyses were possible, due to the low quality of the tumour material obtained after surgery. This meant that for this patient no APVAC2 selection could take place. For another patient, no reliable MS data could be acquired. The immunopeptidome of the analysed samples, as well as the over-expression of the GAPVAC warehouse TUMAP source genes as well as the immunogenicity of the warehouse peptides varied between the patient samples. This variation between the individual patients justified the personalized approach applied in GAPVAC.
Screening for somatic mutations by DNA sequencing detected an average of 42 non-synonymous mutations per tumour sample, including mutations in known onco-relevant genes (e.g. PIK3CA, IDH1, PTEN, TP53). The peptide candidates for APVAC2 cocktails were predefined by cancer immunotherapy experts from BIO and the final selection and ranking of four candidates for inclusion into APVAC2 was finally decided by the target selection committee (TSC), as scientific body of the Consortium that has been established for this purpose. The two highest ranked peptides which also met all CMC criteria for manufacturability were produced at TUE under GMP conditions.
In total, 24 GMP-grade Drug Substance batches (individual de novo peptides for APVAC2 production) as well as 16 APVAC1 and 12 APVAC2 ready-to-use GMP-grade vaccine batches (Drug Product) were successfully manufactured by the University of Tübingen.

• Selection of TUMAPs from the warehouse for APVAC1:
The APVAC1 vaccines were composed of TUMAPs selected from the pre-manufactured GAPVAC peptide “warehouse”. Using tumour material collected during the initial surgery, the mRNA expression profile of each tumour was analysed by microarray analysis and the immunopeptidome (also known as HLA ligandome or HLA peptidome) was assessed by MS. Furthermore, using peripheral blood mononuclear cells (PBMC), the immunogenicity to the warehouse peptides in the individual patient was pre-tested using an in vitro platform. The final APVAC1 selection was carried out for all 16 enrolled patients based on the integrated analysis of all the data obtained from the biomarker analyses of each individual patient.

• Tumour sample collection
Tumour samples from the 16 patients were collected and immediately frozen in liquid nitrogen during the initial surgery according to a standardised protocol. After enrolment of the patients, tumour samples were shipped on dry ice for a central pathology review at partner UKL-HD.

• Quality control of samples and enrichment of tumour cell content
To ensure that the tumour analyses required for the definition of an APVAC were performed from tumour tissue of high quality, a central pathology review (carried out by partner UKL-HD) of all tumour samples was made prior to the start of the APVAC analyses being performed. Tumour tissues were assessed by staining of cryosections applying standard hematoxylin and eosin (HE) staining. Tumours were analysed with regard to tumour cell content and necrosis and were sorted to enrich for high quality (HQ) tumour material. For 13 tumour samples the percentage of HQ tumour cell content was very high, for two samples it was low. One patient had only low-quality tumour material (low tumour cell content, very necrotic).

• HLA-peptidome (mass spectrometry) and transcriptome (mRNA microarray)
HQ tumour samples from 14 out of 16 patients were analysed according to established SOPs (Standard Operating Procedure) by IMM. As mentioned before, for one patient, no MS analysis was possible, due to the low quality of the tumour material obtained after surgery and from another patient, no reliable MS data could be captured.
HLA-peptide complexes from 10 HLA-A*02, 3 HLA-A*24 and one HLA-A*02 / -A*24 shock-frozen GB tumour tissue samples were purified and HLA-associated peptides were isolated and analysed by MS. From 11 HLA-A*02-positive patients, 950 to 3000 unique HLA-A*02 TUMAPs were identified including between 2 to 28 warehouse peptides per patient (Table 1).

Table 1: Detection of the HLA-A*02-restricted warehouse peptides on the tumours from HLA-A*02 positive patient samples in the GAPVAC-101 trial.
A*02 TUMAP-001 TUMAP-002 TUMAP-003 TUMAP-004 TUMAP-005 TUMAP-006 TUMAP-007 TUMAP-008 TUMAP-009 TUMAP-010 TUMAP-011 TUMAP-012 TUMAP-013 TUMAP-014 TUMAP-015 TUMAP-016 TUMAP-017 TUMAP-018 TUMAP-019 TUMAP-020 TUMAP-021 TUMAP-022 TUMAP-023 TUMAP-024 TUMAP-025 TUMAP-026 TUMAP-027 TUMAP-028 TUMAP-029 TUMAP-030 TUMAP-031 TUMAP-032 TUMAP-033
P1 x x x x x x x x x x x x x x x x x
P2 x x x x x x x x x x x x x x x x x x x x x x x x
P3 x x x x x x x x x x x x x x x x x x x x x x x x x x x x
P4 x x x x x x
P5* x x x x x x x x x x x x x x x x x x
P6 x x x x x x x x x x x x x x
P7 x x x x x x x x x x x x x x
P8 x x x x x x x x x x x x x x x x
P9 x x x x x x x x x x x x x x x x x x x x x x x x x
P10 x x x x x x x x x x x x
P11 x x

* Patient P5* is HLA-A*02 and A*24 positive.

From the four HLA-A*24-positive patients, 750 to 2000 unique HLA-A*24 TUMAPs were identified including 11 to 21 warehouse peptides per patient (Table 2).

Table 2: Detection of the HLA-A*24-restricted warehouse peptides on tumours from HLA-A*24 positive patient samples in the GAPVAC-101 trial.
A*24 TUMAP-034 TUMAP-035 TUMAP-036 TUMAP-037 TUMAP-038 TUMAP-039 TUMAP-040 TUMAP-041 TUMAP-042 TUMAP-043 TUMAP-044 TUMAP-045 TUMAP-046 TUMAP-047 TUMAP-048 TUMAP-049 TUMAP-050 TUMAP-051 TUMAP-052 TUMAP-053 TUMAP-054 TUMAP-055 TUMAP-056 TUMAP-057 TUMAP-058 TUMAP-059
P12 x x x x x x x x x x x
P13 x x x x x x x x x x x x x x x x x x x x
P5* x x x x x x x x x x x x x x x x x x x x x
P14 x x x x x x x x x x x x x x x x x x x
* Patient P5* is HLA-A*02 and A*24 positive.

Genomic DNA and mRNA were extracted for NGS and gene expression analyses, respectively. For all 16 patients, RNA was isolated and mRNA expression profiles were generated. A library of gene expression data was made available at IMM including data from several different healthy tissues and organs. Relative expression levels were obtained by comparing the mean expression values for patients’ GB tumours with the expression levels for normal tissues. Over-expression of the source genes of the warehouse peptides varied between the patients and showed a significant level of personalization.

• Immunogenicity testing
For 15 out of 16 patients, a leukapheresis sample was collected at the clinical site. Mononuclear cells were isolated and frozen (at or below -70 °C) by a local, trained laboratory in close proximity to the clinical study site. The cells were then shipped to IMM on dry ice. Using PBMCs, patient-specific immunogenicity to the HLA-A*02 and the HLA-A*24 warehouse peptides was pre-tested using an in vitro platform. Immunogenicity data showed a significant level of personalization with the presence of progenitor T-cells for the different warehouse peptides varying between patients.

• APVAC1 selection
Each warehouse peptide was analysed for a variety of characteristics based on patient-individual data, including peptide (over-)presentation and mRNA over-expression by the tumour compared with a panel of healthy tissues (from distinct individuals). Patient-individual results from in vitro immunogenicity pre-testing were likewise considered. Based on these data an algorithm was developed yielding an overall probabilistic score reflecting the likelihood of efficacy for each peptide in the given patient. This score then resulted in a personalized ranking of all warehouse peptides and finally determined the composition of each individual APVAC1 medicinal product (drug product = DP).
For each patient, one target selection committee (TSC) meeting was held to decide on their final APVAC1 drug product composition according to the ranking of the warehouse peptides.
For all 16 enrolled patients, APVAC1 selection was successful and the APVACs showed a high degree of individuality justifying the highly personalized approach. Usually the seven top-ranking warehouse peptides were selected for the APVAC1 composition. For the patient being HLA-A*02 and HLA-A*24 positive, the TSC decided to include the three top-ranking HLA-A*02 peptides and the four top-ranking HLA-A*24 peptides in the APVAC1 composition, as separate ranking lists were available for the two HLA alleles.

• Screening for somatic mutations by DNA sequencing
The APVAC2 process was defined by a standardised and controlled work flow that foresees participation of different departments/partners. The whole process was therefore divided into different working “blocks” (overview shown in Figure 4).
Tissue samples were prepared by IMM and were sent to BIO (Block I). Within Block II suitable patient-specific tumour mutations were identified, ranked and filtered to exclude potential higher risk targets via bioinformatics algorithms by BIO, resulting in the prioritized list of mutations. In Block IV confirmation of the TUMAPs by Sanger sequencing and mass spectrometry analyses were undertaken. In Block III, IMM searched for an alternative class of APVAC2 peptides via MS: non-mutated TUMAPs exclusively presented or over-expressed in the individual tumour compared to normal tissues.


Figure 4: Overview of the process of APVAC2 identification and selection for each patient
Patient- and tumour-specific somatic mutations were determined by NGS. Mutations were confirmed by Sanger sequencing or mutation containing tumour presented peptides were assessed via peptidome analysis. Moreover, non-mutated TUMAPs with high relevance for a given patient’s tumour were identified by mass spectrometry.

Tumour DNA and RNA of the 16 enrolled patients were extracted from the tumour specimens and germline DNA was obtained from peripheral blood mononuclear cells (PBMCs) using well established protocols. These nucleic acids were then sent to BIO for mutation analysis.
Samples (DNA and RNA) were checked for correct labelling, quantity and shipping conditions. Part of the germline DNA was shipped to an external service provider for 4-digit HLA typing. The remaining RNA and DNA was used to prepare sequencing libraries. DNA sequencing libraries were further enriched for the protein-coding exonic fraction (the “exome”). Exome and RNA sequencing libraries were used for high throughput NGS. NGS data were analysed using an algorithm for mutation detection to identify tumour-specific somatic mutations, allowing the comparison of the results from the tumour and germline DNA. In addition, the expression of the mutated genes and the frequency of the mutated allele were determined in the tumour-derived RNA sequencing reads. The output of this process is a list of expressed non-synonymous somatic mutations, ranked according to their predicted immunogenic potential. These mutations were confirmed by Sanger sequencing. The entire process was set up and standardised before the clinical trial by performing preclinical feasibility runs and was used routinely and in consistently form throughout the clinical trial.
In total, 15 patient samples were analysed and 15 APVAC2 cocktails were selected. Only 13 APVAC2 cocktails were produced because two patients dropped out of the study before the manufacturing of these personalized vaccines was completed.

• In silico determination of the APVAC2 composition
The consortium demonstrated the proof-of-concept by identifying mutations in pre-clinical mouse models B16F10 and GL261. This was done using NGS, followed by identification of mutation-containing peptides on MHC molecules by MS. In addition, the consortium has performed pre-clinical feasibility runs and successfully executed the NGS and MS profiling. It also established a complete process to securely exchange the results, integrate data sets, prioritize targets, and identify targets for vaccine production. Problematic steps were identified during the feasibility runs and were solved (refer to WP3). After these pre-clinical proof-of-concept and feasibility runs, the process was reliably and robustly applied during the GAPVAC-101 clinical study. This included the secure exchange of results and data, integration of the data sets, prioritization of targets, and finally the selection of targets for APVAC2 vaccine production.

• Identification of mutated TUMAPs
As shown in Figure 4 Block IV, mutation confirmation was performed by Sanger sequencing and mass spectrometry analysis. Peptides with over-presentation or exclusive presentation in the patient’s tumour were ranked and considered for inclusion in the APVAC2 vaccine. MS was not able to confirm any of the mutations identified by the NGS analysis – an important outcome of the study and of high relevance to the field of immune-oncology.

• Selection of APVAC2:
As mentioned before, APVAC2 drug products may contain 3 types of peptides: I) mutated peptides confirmed to be processed and presented on MHC molecules, II) peptides predicted from confirmed DNA mutations, III) individually over-presented non-mutated peptides not present in the warehouse.
Prior to the final selection of APVAC2 drug products, data from these three processes were evaluated and all candidates were ranked by cancer immunotherapy experts from BIO who combined significant medical, immunological and genomic expertise. The aim of the final selection was to create an APVAC2 drug product with the best possible benefit / risk profile for the patients.
The final selection and ranking of the four candidates was discussed and determined in the Target Selection Committee (TSC). The peptide production of all four candidates was performed under non-GMP conditions to check stability and ensure feasibility of GMP production. The two highest ranked peptides, which fulfilled all CMC criteria including manufacturability, were produced at TUE under GMP conditions.

• Manufacturing of ready-to-use GMP-grade APVACs [TUE]
Overall, 24 GMP-grade Drug Substance batches (individual de novo synthesized peptides for APVAC2 production) as well as 16 APVAC1 and 12 APVAC2 ready-to-use GMP-grade vaccine batches (Drug Product) were successfully manufactured by the University of Tübingen.
All clinical APVAC1 and APVAC2 batches were released by the responsible Qualified Person (QP). Related Certificates of Analysis (CoAs), Batch Certificates and TSE/BSE Certificates were issued.
A batch size of approximately 35 to 40 vials of each individual APVAC1 vaccine was produced. In contrast, the batch size of APVAC2 varied depending on the final yield of the individual peptide synthesized.
Test syntheses of the APVAC2 peptide candidates were performed to investigate the general manufacturability and suitability for the formulation process before initiating the expensive and time-consuming GMP process. The stability of the individual Drug Substances was also evaluated using the same solvent system as used for the production of the final APVAC2 vaccine. Both criteria – suitability of the peptides for solid phase synthesis and the stability of the peptides - were taken into account when deciding on the peptides for the final APVAC2 drug product.
However, some of the candidates were not included in the final APVAC2 vaccine cocktails because of unexpected issues during the GMP peptide synthesis or drug product production (e.g. peptide could not be purified due to very low raw peptide purity or very low yield after peptide purification). In these cases, the composition of the vaccine was modified according to the pre-defined process steps under control of the TSC, in order to achieve a suitable APVAC2 vaccine for that particular patient.


WP7 Centralized assessment of vaccine-induced immune responses and biomarkers
Objectives
These objectives of this work package were the standardised collection of high quality PBMC patient samples during the multi-centre GAPVAC-101 clinical trial and the determination of immunological efficacy (proof of concept) of APVAC vaccination by detecting and characterizing vaccine induced T-cell responses. In addition, the consortium was looking for new biomarkers which could be used as potential surrogates for the immunological efficacy of personalized vaccination and to identify patient populations with a better chance to respond clinically.

Summary of progress towards objectives
PBMC samples from all vaccinated GAPVAC patients have been centrally stored and immunomonitoring assays to assess APVAC-induced immune responses have been performed at IMM, which is functioning as the central immunomonitoring lab. A summary of the results is reported below.

• Standardised PBMC collection
PBMC collection was coordinated by IMM and was performed according to an established protocol that was assigned to the PBMC laboratories from the collaboration partners which were situated within or in close distance from the clinical sites. The requirement was to allow the venepuncture, shipment and PBMC isolations to take place within a total time of 8 h. PBMC operators trained and qualified (WP2) isolated PBMCs and cryopreserved these according to the relevant SOPs. All critical consumables were provided by IMM, including PBMC Kits containing components for PBMC isolation and freezing. Isolation and cryoconservation were documented during the PBMC process. The PBMC samples were stored short-term at -80°C or in liquid nitrogen cryostorage if storage for longer periods at the sites was appropriate. The PBMC samples were then sent to IMM on dry ice by a trained courier service. At IMM, frozen PBMC samples were transferred into a liquid nitrogen cryostorage system until immune response assays were performed.
Overall, 151 PBMC isolations have been performed at the PBMC laboratories and the samples have been transferred to IMM.

• T-cell Immunomonitoring and biomarker analyses
The central evaluation of the APVAC1 immune responses was performed using an established ex vivo workflow combining MHC class I multimer staining and MHC class II intracellular cytokine staining in one assay. This cutting edge immunomonitoring assay enables the simultaneous quantification, phenotyping and characterization of induced CD8+ and CD4+ T-cell responses without putative clonal alterations by in vitro pre-sensitization. The respective assays have been performed at IMM as central immunomonitoring lab.
TUMAP-specific vaccine-induced immune responses have been detected in 13 of 14 (93%) immune response evaluable patients while no responses to the not vaccinated negative control antigens have been detected. Immune responses seem to be sustainable and shifts of specific CD8+ T cells from a naïve to a memory T-cell phenotype after vaccination were observed. CD4+ T cell responses seem to be of the favourable TH1 phenotype. These results give the impression that APVAC-induced immune responses are more frequent, of increased amplitude and more sustainable as compared to previous trials. Moreover, this data impressively shows the high biological activity of APVAC vaccines and it is likely that the pre-immunogenicity testing implemented in the GAPVAC approach and the immunomodulator setting (GM-CSF plus poly-ICLC) contributed to this success. Additional immune readouts for APVAC2 immune responses showed that (most likely by cross-presentation) responses could be induced against APVAC2-contained class I antigens by vaccinating long mutation-containing APVAC2 peptides. Further biomarker analyses including additional APVAC2 response analyses are close to completion.


WP8 Local assessment of tumour-infiltrating lymphocytes (TILs)
Objectives
The objective of the WP8 was to establish an assay for the harmonized decentralized characterization of TILs at the GAPVAC clinical centres. Moreover, during the clinical trial, the TILs from GAPVAC patients should be isolated and characterized with regard to T-cell quality and antigen specificity. The quality of these TIL analyses should be controlled by co-testing reference samples.

Summary of progress towards objectives
Results from this work package have clearly shown that the well-established protocol for TIL isolation from melanoma could not be translated to glioma without modification. The protocol was optimized in order to isolate and expand TIL from glioma in a robust fashion. In addition, reporting forms for TIL expansion and TIL testing were prepared in conformity to the MIATA reporting framework. These revised reporting forms were distributed to the labs.
Altogether, TILs could be isolated and expanded from the surgically removed tumour material of 21 glioma patients before vaccination, 9 of which were eventually enrolled in the GAPVAC trial. Phenotyping could be performed on most of these TIL cultures and specifically for all TIL obtained from the vaccinated patients. The results showed that glioma tissues are predominantly infiltrated with classical TCRα/β+ T cells, both of the CD4+ and CD8+ phenotype with variable percentages per patient but in general there was a domination by CD4+ T cells.
The warehouse peptides were additionally tested for T-cell responses in GAPVAC TILs obtained from n=6 patients using an IFN-γ ELISPOT procedure earlier harmonized by CIMT (Britten et al., 2007). No spontaneous IFN-γ−gamma T-cell reactivity could be detected against any of the warehouse peptides.. This could be due to the fact that almost all APVAC1 peptides are targets for CD8+ T cells while TILs were dominated by CD4+ cells.

• Harmonization of TIL immunomonitoring assays
Three successive proficiency panels were organized, analysed and reported to the participating GAPVAC labs. Harmonized protocols for the HLA-multimer assay and intracellular cytokine assays have been established, and a TIL expansion protocol was harmonized.

• Immunomonitoring of APVAC-specific TILs at individual centres and central analysis
i) Standardised report forms
For TIL expansion and TIL testing, two report forms were prepared in conformity to the MIATA framework and distributed to GAPVAC labs. These forms are being used by the TIL labs to collect and report centrally TIL data to CIMT.
ii) TIL collection and phenotyping
Following the harmonization of TIL isolation (WP2), TIL cultures were established using tumour material collected during the clinical phase of the GAPVAC project. Between 1 to 6 individual TIL cultures per patient were established (3, 9, 5 and 4 patients at GEN, TUE, LUMC and CPH, respectively) of which 3, 3, 2 and 1, respectively were from patients finally enrolled in the trial. From these 9 patients, large numbers of TILs could be frozen for 5 patients (approx. 46x106 to 160x106 cells) whereas the number of cells that could be expanded was limited for 4 other patients (between 0.5x106 and 3x106).
Phenotyping could be performed on most of these TIL cultures and specifically for all TILs obtained from the vaccinated patients. Phenotyping included antibodies to assess T cell and NK cell populations (CD3, CD4, CD8, CD16, CD56, TCRα/β and TCRγ/δ). The results showed predominant expansion of TCRα/β+ T cells, with a variable CD4+/CD8+ ratio depending on the patients and the individual cultures; However, CD4+ T cells were predominant in most cultures.
No post-vaccination tumour samples have been obtained which was to some extend anticipated due to the low rate of re-resections in case of recurrence in glioblastoma. Hence, TIL testing was performed on pre-vaccination samples only.
iii) Harmonized testing strategy and reagent production.
Due to the delayed recruitment and/or the limited number of TILs for some patients, it was decided to coordinate the TIL testing with that of the patient PBMCs. Hence, the last step of TIL testing was initially postponed. When few TILs were available for testing, only APVAC peptides which were found to be recognized by autologous PBMCs were tested. Production and shipment of the reagents (peptides and HLA multimers/monomers) and of patient-individual testing instructions to the centres were coordinated by IMM, LUMC and TUE. Distribution was completed in January 2017.


WP9 Tumour-specific sHLA peptidome
Objectives
Aim of this work package was to quantify the soluble HLA-bound (sHLA) peptides from GB patients during the course of the disease. Therefore, the repertoires of tumour HLA peptides carried by soluble HLA molecules in the plasma of patients and their changes before and after surgery have been determined. These were compared to large number of healthy control plasma of blood samples performed in a parallel study by TEC. Furthermore, large-scale proteomics analyses of glioma samples have been conducted.

Summary of progress towards objectives
This part of the project included analysis of sHLA peptidomes and the tumour cells’ membranal HLA (mHLA) peptidomes as well as proteome analyses of multiple plasma and tumour samples. In addition correlations between 1) the plasma sHLA peptidomes and the tumour mHLA peptidomes, 2) between the plasma sHLA peptidomes, taken before and after surgery, and 3) of the sHLA and mHLA peptidomes with the proteomes of the same tissues/patients were assessed. The aim was to identify sHLA peptides that change their levels of presentation due to the removal of the mass of tumour tissue in order to select tumour-associated HLA peptides, which are released from the tumour cells with the soluble/secreted HLA molecules. The analysis of the proteomes of the same patient’s tumours allowed comparison of the lists of tumour proteins to the HLA peptides derived from them. Analysis of the membranal HLA peptidome allowed identification of further tumour antigen-derived HLA peptides and correlation with the sHLA peptidomes of the same patient.
During the GAPVAC project overall 118 plasma samples of GB patients and 10 different GB tumour samples of the same patients were analysed for their sHLA peptidomes and for their tumour cells’ mHLA peptidomes. Overall, the sHLA peptidome analyses resulted in the identification of 55,155 different HLA peptides, derived from 11,485 different source proteins with as many as 6,760 HLA peptides discovered from individual plasma samples. The mHLA peptidome analyses resulted in the identification of 40,116 different HLA peptides, derived from 9,715 different source proteins. Most of these peptides (52,400 of the sHLA peptidome (95%), and 37,558 of the mHLA peptidome (93.6%)) fitted the typical HLA class I ligands’ with a length between 8-14 amino acids. In addition, the above mentioned 10 tumour tissues were analysed for their proteome content. As many as 7,351 different proteins were identified by the proteome analyses, with at least two identified tryptic peptides per protein.
Importantly, the correlation between the plasma sHLA peptidomes and the tumour mHLA peptidomes was good, but not very high, as expected while the correlation between the plasma sHLA peptidomes, taken before and after surgery of the same patients, was very high. However, a small fraction of sHLA peptides were reduced after surgery. All of these HLA peptides were derived from genes/proteins from which these (or any other) sHLA peptides were not detected in any of the non-cancerous plasma samples.
The correlations of the sHLA and mHLA peptidomes with the proteomes of the same tissues were very poor (data not shown).
Large numbers of HLA peptides were identified from the tumour antigens. Selection of tumour antigens (TAs) was based on the cancer tumour database (CT gene database; http://www.cta.lncc.br/ ; (Almeida et al., 2009)) comprised of 277 different TAs (data accumulated between the years 2005–2009) and the tumour T-cell antigen database (TANTIGEN; http://cvc.dfci.harvard.edu/tadb ), comprised of 259 different TAs. We also define the Cancer Testis antigens (CTA)-derived HLA peptides as peptides whose source genes are expressed only in testis, placenta, embryo, or similar.
In summary, the vast majority of objectives of the work packages of the GAPVAC project were fully reached and the wealth of collected data will allow for several scientific publications from the GAPVAC project in the near future.

Potential Impact:
Impact for life science and medical use
Personalised medicine can be achieved by stratification of patients according to a biomarker/method predicting therapeutic benefit. Ideally, this would mean the grouping of patients to different kinds of therapies, with all of the patients benefitting from their respective treatments. In reality, treatment is often restricted to the patient population positive for a certain biomarker and patients, who are non-eligible for the available treatments, are neglected. Additionally, few disease-specific features are addressed by current biomarker-driven therapies. In malignancies the inclusion of tumour-specific features as a target for therapy is particularly interesting, given that tumours evolve differently in each patient. To our knowledge, therapeutic approaches incorporating several tumour-specific and well-defined molecular attributes relating to the individual patient (i.e. beyond sizeable subgroups of patient populations) are nearly completely unexplored not only in Europe, but also throughout the world.

Thus, GAPVAC-101 is to our knowledge the first multi-centre clinical trial where immune-therapeutics are actively tailored to the needs of the individual patient and therefore we believe that the GAPVAC consortium has advanced personalization to the next level. This is of special interest in glioblastoma where there are very limited therapeutic options and where the 5-year survival rates with current treatment modalities are very low. The successful completion of the safety and feasibility profiles of this trial has opened the door for a new class of immunotherapy applicable for a majority of European glioma patients and cancer patients in general. Importantly, GAPVAC gives rise to new hope for GB patients, affected by an up to now incurable disease, while the safety profile of the personalized vaccinations was overall very favourable. GAPVAC also significantly strengthened the direct cooperation between leading glioma centres all over Europe with regard to patient care and translational research. The creation and implementation of such a network has the potential to move Europe to a global leading position in the field of next-generation personalised medicine. Additionally, if proof-of-concept for personalised vaccines can be delivered for glioma, there is little doubt that this approach will be relevant for other cancers and potentially for the personalization of medicines more generally.

It is clear that GAPVAC has had a substantial medical and scientific impact on the European and international community in the field of clinical and translational development in rare cancer research and tumour immunology. This significant impact has also been reflected in the high interest into the GAPVAC project at important international conferences of the field (e.g. AACR, ESMO and CIMT Annual Meetings).
Impact on industry
The consortium supported close collaborations between three European SMEs, academia and a leading non-profit organization and took advantage of the expertise on all three sides. The design of the actively personalised vaccine made use of IMM’s peptide antigen identification and BIO’s genomic target identification expertise (plus proven expertise in performing biomarker studies in multi-centre trials). The manufacturing of the high number of warehouse peptides for a single project to GMP quality is probably an unprecedented effort and was reliably performed by BCN, again demonstrating the leading scientific role of the involved European SMEs involved in GAPVAC. High standards for GMP manufacturing combined with the enormous flexibility required for on-demand manufacturing were met by the GMP manufacturing facility at partner TUE, which was dedicated to the fast and small-scale synthesis required for the project’s personalised vaccine production. In addition to the medical and scientific impact of the results generated by GAPVAC, European SMEs are now in possession of most of the various intellectual property rights generated during the GAPVAC project. This is a prerequisite if the knowledge that has been gained can be directed towards the development of new therapies that are commercialized for the benefit of much larger group of patients. GAPVAC really benefitted from the assets and expertise that was already present at the SMEs. But in return, the SMEs gained new expertise in the field of personalized medicines which is expected to be of significant for many of their other development projects. For example, BIO further strengthened its significant expertise in the handling of personalised data, the identification of somatic tumour mutations and how they can be used to develop personalized immunotherapies. IMM gained international recognition by leading successfully all of the operations needed to implement the highly complex GAPVAC concept in a clinical setting. In addition, IMM further expanded its knowledge on identifying and characterizing valuable targets within the GB-associated immunopeptidome and their application in a personalized warehouse concept. BCN strengthened its position as internationally renowned and experienced supplier of GMP peptide products. Thus, the GAPVAC project provided very clear evidence that a highly innovative, multi-disciplinary translational research project can benefit from the leadership and active participation of SMEs. It also demonstrates that SMEs can gain significant benefits from such collaborative projects enabling them to translate cutting-edge science into future products and technologies.

The GAPVAC-101 trial was performed at European centres of excellence in the management of patients with glioma and was supported by internationally recognized investigators selected for their long-standing commitment in this field. The successful implementation of the GAPVAC phase I clinical trial demonstrated the capacity of the European network to execute a cutting edge personalized immunotherapy project. This success will not only attract further funding and private investments into the SMEs but also into follow-up clinical trials to further explore the GAPVAC concept. The SMEs IMM, BIO and BCN, are committed to develop actively personalised immunotherapies beyond this phase I study with the ultimate goal to commercialize this novel and attractive concept to the benefit of cancer patients globally. Given this very important objective, the success of the GAPVAC project is very important in attracting additional private investments into the consortium’s SMEs to support the further clinical development of APVACs in glioma and/or related actively personalized immunotherapies.
Impact for other areas of research
The competitiveness and productivity of the European research community and industry is expected to improve significantly as a result of the Lisbon strategy (innovation as the motor for the economic change). GAPVAC has had a very positive impact on the standing of the European Community in the field of clinical and translational R&D in rare cancer research, tumour immunology and especially in personalised therapy.

The GAPVAC project also strengthened the cooperation amongst the leading European clinical centres with regard to the development of individualized patient care based on translational research, including several cutting edge technologies. The successful implementation of the project using a network of European organizations was an important step in enhancing the global standing of European researchers in the field of GB and immunotherapy when compared to other areas of the world.

Most clinical trials are conducted by large pharmaceuticals companies and often led by US investigators, which gives US research groups a systematic advantage in the competition for research funding. Demonstration of a well-functioning European network that is able to successfully conduct a highly innovative and complex clinical trial in a rare cancer type will improve the standing of European translational research groups in the eyes of funding bodies around the world.

It is true that different regulations in the European member states were a challenge for the conduct of GAPVAC-101 as pan-European, multi-centre clinical trial. However, finally these challenges were overcome by the Consortium convincingly demonstrating that innovative clinical research is possible in Europe. Thus, GAPVAC has helped strengthen the collaborative framework of clinical development in Europe. This is, of course, true for any multi-centre trial, but particularly for GAPVAC, as this development – due to its complexity – required much stronger interaction between all the parties involved when compared to development of standard drug products.

In addition, the project included a T-cell immunomonitoring and biomarker program that were needed to comprehensively describe the mechanism-of-action of APVAC vaccines and to support their further development. This pioneering cooperation of European scientists and physicians also paved the way for the development of a regulatory framework which will guide the future development of highly personalized immunotherapies. Discussions with regulatory agencies about how to regulate such a new approach had already been started by CIMT before GAPVAC was initiated and first viable solutions based on existing guidance had been identified based on a detailed publication by CIMT and other partners of this consortium (Britten et al., 2013). This approach was been discussed with the European Medicines Agency’s (EMA) Innovation Task Force in January 2012 and resulted in the positive outcome that the regulatory hurdles to develop APVACs were manageable. With the approach to develop a standardised drug composition and manufacturing process that could results in a virtually infinitive number of different drug compositions, the GAPVAC consortium entered a new area of medicine and regulation. As both, target and lead structures for each individual patient were not known/defined prior to the recruitment of the patients, currently existing paradigms defining aspects of quality, preclinical testing and manufacturing were not feasible for GAPVACs. The GAPVAC consortium developed the first blueprint for the clinical development of highly variable drug products originating from a novel but standardised process. The principles guiding GAPVAC may be applied to any future approach in which molecular information from a patient (e.g. genotype, expression profile, disease stage, immune status) are used to define the parameters for a highly personalised and thus unique therapy. These implications reach far beyond therapeutic peptide-based cancer vaccines. Thus, GAPVAC has formed the basis for constructive discussions between partners of the consortium and the leading European regulatory authority EMA. This has resulted into a real successful clinical project which has further helped to refine and shape the regulatory framework for actively personalized therapies and to advance the overall regulatory approach to this complex type of therapeutic intervention.


Dissemination activities
GAPVAC is expected to have a major global impact on cancer immunotherapy and on the way personalized medicines are developed and perceived generally. This ground breaking outcome has led to the development of a comprehensive strategy for the publication/ promotion of the GAPVAC findings/results. CIMT, IMM and BIO are focused and will focus on all the key tasks related to the dissemination process; however, all partners have contributed and will continue to contribute to raising national and international awareness of the project in the networks and associations that they participate. The GAPVAC consortium is composed of academic partners with focus on translational research, clinical experts in treatment of glioblastoma patients and industry partners with significant experience in early to advanced stage clinical development of therapeutics to treat cancer. The consortium was and remains committed to disseminate the research findings from the GAPVAC project efficiently to the scientific community, including regulatory bodies, and to policy-makers, and the public at large. As technology transfer was the basis for the success of this complex project, close and strong contacts and collaborations with research groups from the different relevant areas were crucial to its success.

Management of intellectual property
In addition, the partners’ allocated commercially important knowledge and IP generated from the GAPVAC project. This was achieved by mutual definition of IP relationships as defined in the consortium agreement. The contracts guarantee that the respective interests of the partners are respected. Lastly, the SMEs of this consortium (led by IMM and BIO) are committed to proceed with development of actively personalized immunotherapies beyond the life-time of the project. Together with the above-mentioned IP dispositions, this will enable – in case of success – the commercialization of such strategies by the European SMEs and thus GAPVAC will contribute to the strengthening of the biotech sector within Europe.

Exploitation of project results
The GAPVAC consortium has implemented a range of activities to support the efficient dissemination of the GAPVAC research findings to ensure that they guarantee wide distribution of this knowledge, and globally increase the visibility of the project amongst scientists, regulators, drug developers, and policy makers. The public at large is another very key audience as it includes glioma/cancer patients who are the people that should finally profit form the outcomes of this project in the future (WP11).

The project results have been disseminated in the form of scientific presentations, website publications, press releases and reports sent regularly to the European Commission. Scientific publications include posters, lectures and workshops at scientific meetings as well as publications in scientific journals.

CIMT provides efficient access to the largest European scientific and networking platform with sole focus on Cancer Immunotherapy and thereby supports the dissemination of project results amongst peers, regulators and policy makers at an international level. CIMT – as a non-profit organization – is the ideal platform for dissemination to the scientific and regulatory community.

All dissemination activities and publications in the project acknowledge the Seventh Framework Program funding and the EU. Additionally, dissemination of information from the GAPVAC project is made via a dedicated web page (www.gapvac.eu).

According to these principles, GAPVAC has used a broad range of activities to publish and promote the GAPVAC concept and the clinical trial GAPVAC-101 in order to reach a wide and diverse audience. The various activities are summarized below.

1. Official project website
The official GAPVAC website, which has been available at www.gapvac.eu since March 2013, was set-up and coordinated by the two coordinating partners IMM and BIO. The technical development was done by “GENTHER/LEHR – Büro für Gestaltung” (Tübingen). The consortium appointed a communications officer who is the primary contact for members of the press or public seeking information about the project and is responsible for collecting and disseminating publishable information from the GAPVAC consortium that could be published. The objective of the website is to provide information regarding the project which is of value to the public, to patients and to the consortium partners. During the project lifetime, website visitors were informed about the project’s background, concept and latest events and activities. On a password protected area of the website more detailed information (e.g. deliverables and reports, protocols and other study documents, newsletter, forms and templates, research results) are available to the project partners. In addition, a large number of presentations related to or originating from the GAPVAC project are available in this restricted area.
2. Logo, PPT and project templates
A logo was created to provide the GAPVAC project with a clear and easily recognizable corporate identity. Standard templates for e.g. power point presentations, reports and deliverables were designed to ensure a unified graphic design to establish GAPVAC as a “brand” that can be easily recognized by others and to promote the corporate identity within the consortium.
3. Consortium meetings
During the GAPVAC project, a Kick-off meeting and four consortium meetings have been held where the current status of the different work packages and progress with the project as a whole were presented and discussed. Frequently, strategic discussions at the Consortium meetings led to important refinements of the GAPVAC concept. Noteworthy, an additional Annual Meeting has been scheduled for October 2017 to discuss follow-up projects on the basis of the final dataset from the GAPVAC-101 study

4. Conference presentations and publications
The GAPVAC consortium and the progress of the preclinical and the clinical part of the GAPVAC project were presented to the oncology and immunology community at several scientific meetings (e.g. ESMO, AACR, Brain Tumour Meeting) and to the European society by press releases and several cuttings in epub and print media (e.g. Washington post, The Parliament).

For the CIMT Annual Meetings in 2014, 2015, and 2016 special teasers for the GAPVAC project were prepared by CIMT, BIO and IMM. Moreover, the project per se as well as the progress of the project was presented as abstracts and posters at the CIMT Annual Meetings in 2014, 2015, and 2016.

In 2016, the progress of the GAPVAC clinical trial was also presented during oral presentations by Harpreet Singh during the CIMT Annual Meeting and at the AACR Annual Meeting in New Orleans. Noteworthy, the AACR is probably the most important international conference on translational science in the field of oncology with >18,000 participants each year.

A paper entitled “Predicting immunogenic tumour mutations by combining mass spectrometry and exome sequencing” in which IMM had a significant participation was published in November 2014 in Nature. Another paper involving GAPVAC partners BIO, IMM, LUMC, CIMT and TUE was published in April 2015 in Immunity (“Thinking Outside the Gate: Single-Cell Assessments in Multiple Dimensions”). A first manuscript from the GAPVAC project on the sHLA results has been completed for publication and will be submitted for publication during the next weeks by partner TEC.

In addition, GAPVAC partner CIMT and BIO published reports at the 12th, 13th and 14th annual meeting of the Association for Cancer immunotherapy in “Cancer Immunol Immunother” and in “Hum Vaccin Immunother”, respectively. Moreover, the GAPVAC project was presented by several partners at institutional meetings or within research networks on multiple occasions.

With the final data from the GAPVAC clinical trial now available, a number of additional publications in peer-review journals are planned. The Consortium is confident that it will be able to secure at least one GAPVAC publication in a high impact scientific/clinical journal.
5. Newsletters
Eight project newsletters have been written and distributed to the GAPVAC partners to keep them updated on the projects progress and to make them aware of e.g. pending tasks or upcoming problems. This strategy was supported by additional email communication on important or urgent topics as needed.

List of Websites:
www.gapvac.eu

The project website (www.gapvac.eu) was set up at the beginning of the project and has been updated continuously (Figure 5).
The website includes two parts: one open to the public and one restricted to the members of the Consortium. Peak page views per day were above 3000.


Coordination:
Sabrina Kuttruff-Coqui / Norbert Hilf
Immatics Biotechnologies GmbH
Paul-Ehrlich-Str. 15
72076 Tübingen
Germany
Phone: +49 (0)70 71 / 53 97 - 0
Fax: +49 (0)70 71 / 53 97 - 900

Judith Theelen
BioNTech AG
An der Goldgrube 12
55131 Mainz
Germany
Phone: +49 (0)61 31 / 90 84 - 0
Fax: +49 (0)61 31 / 90 84 - 390


List of participants
No Name Short name Country
1 Immatics Biotechnologies GmbH IMM Germany
2 University of Glasgow (left the consortium 30.09.2013) GLA UK
3 Region Hovedstaden CPH Denmark
4 BioNTech AG BIO Germany
5 University of Tuebingen TUE Germany
6 Université de Genève GEN Switzerland
7 University Hospital Heidelberg UKL HD Germany
8 Technion, Israel Institute for Technology TEC Israel
9 Leiden University Medical Centre LUMC Netherlands
10 Vall d’Hebron University Hospital, Institut Catala de la Salut HUVH Spain
11 Southampton Universities Hospital SOH UK
12 Association for Cancer Immunotherapy CIMT Europe
13 BCN Peptides BCN Spain