OSE Immunotherapeutics — Immunotherapy research engine

OSE Immunotherapeutics is a biotechnology company based in Nantes and Paris, France and listed on the Euronext Paris exchange. The company was created in May 2016 through the merger of OSE Pharma and Effimune. It specialises in the identification of immunotherapy targets, the generation of product candidates and early- to mid-stage development before out-licensing for late-stage development and commercialisation. OSE is able to benefit from strong research partnerships, especially with Inserm in Nantes. OSE’s research headquarters are in Nantes and benefit from these extensive networks developed over the years. OSE’s lead asset in clinical development is Tedopi, a cancer vaccine in development for non-small cell lung cancer (NSCLC, Phase III) and pancreatic cancer (Phase II). Other interesting assets include Phase II-ready FR104, OSE-127 (partnered with Servier) and OSE-172 (in late preclinical studies, partnered with BI).

We value OSE based on a risk-adjusted NPV analysis using a 12.5% discount rate, including estimated net cash of €7.5m at end-2018. This results in a value of €171m or €11.7/share. We include four assets in our valuation, namely Tedopi in NSCLC, OSE-172, FR104 and OSE-127. We have not included Tedopi in pancreatic cancer until the commercial strategy is clearer. We assume a licensing deal for Tedopi in NSCLC after the Phase III study. The two near-term milestone payments from Servier and BI are dependent on the respective assets moving from late preclinical stage to Phase I. Typically, this is a low-risk transition but, to reflect any remaining R&D risk and the fact that ultimately this will be a third-party decision, we include the €12.0m payment from Servier and the €15.0m payment from BI due in 2019 with a 90% risk adjustment.

In its H118 financial results, OSE reported €20.6m in revenues (vs €2.8m in H117). This includes the upfront payment from BI for OSE-172 received in April 2018. Total operating expenses for H118 were €10.2m vs €7.3m in H117. R&D expenses are the company’s largest expense: €8.0m in H118 vs €7.9m in H117. The largest contributor is the Tedopi Atalante 1 Phase III study. We expect R&D costs to remain largely constant at €8m per six months of operation until the end of the trial in 2021. As of 30 June 2018, OSE had cash of €18.6m (includes ‘current financial assets’). With cash burn around €15m per year, cash is expected to last into H219. If both milestone payments totalling €27m are received in H119 (from Servier and BI), the cash runway should extend into early 2021.

OSE is subject to typical biotech company development risks. We see the largest part of its value coming from Tedopi in NSCLC due to the late stage of this project. Therefore, the outcome of the Atalante 1 Phase III trial will be important. The licensing deals with BI and Servier substantially diversify this risk, but the risk remains if any of those partnerships are discontinued. If OSE does not receive the expected licensing income in 2019 (up to €27m), it might have to raise money from elsewhere to fund its operations. The main patent of Tedopi expires in 2024 with potential to extend the protection for five years. OSE filed for fresh patents in 2014/15, which are yet to be granted. However, market exclusivity after the launch of Tedopi could be secured if OSE receives orphan drug designation in the US, and a biologicals licence in the US and Europe. Finally, Janssen decided to discontinue the licensing deal related to FR104 asset. Once OSE regains the rights and data, it will have to clarify its future development plans.

OSE Immunotherapeutics (OSE) is an immunotherapy company created in May 2016 through the merger of OSE Pharma and Effimune, as both companies saw potential synergies between their business models and asset portfolios:

After the merger, OSE Pharma held 71% of the shares of the merged entity, OSE Immunotherapeutics, which has been operating since 2016 and is listed on Euronext Paris. The founders of both OSE Pharma and Effimune are still involved, eg CEO Alexis Peyroles was CFO at OSE Pharma previously. Dominique Costantini, currently chairman of OSE Immunotherapeutics, co-founded OSE Pharma and was its CEO. Maryvonne Hiance, who was co-founder and CEO of Effimune, is now serving as vice chairman and director of strategy for OSE, and Bernard Vanhove, who was co-founder and COO of Effimune, is now COO of OSE. Nicolas Poirier, PhD, is CSO of OSE and was also previously at Effimune. Emile Loria, a former chairman of OSE, and Dominique Costantini are majority shareholders in OSE and have been working on the Tedopi asset. Emile Loria has been involved with Tedopi asset since 2001.

OSE’s business model involves the identification of immunotherapy targets, the generation of product candidates, early-stage development and, finally, out-licensing for late-stage development and commercialisation. OSE has a somewhat unique position in the immunotherapy landscape in that it can benefit from immunotherapy research in Nantes through its longstanding relationship with Inserm and other institutions.

OSE’s most advanced asset is Tedopi, a synthetic peptide cancer vaccine which was in-licensed from Takeda in 2013 by OSE Pharma and which OSE is developing internally for NSCLC (Phase III), and in collaboration with GERCOR for pancreatic cancer (Phase II). OSE has also demonstrated a track record of establishing partnerships with large pharma (Exhibit 3):

Milestone payments from commercial deals will help to fund the company’s R&D, the majority of which is the ongoing Tedopi Atalante 1 trial. Up to €27m could be generated from milestone payments in the near term, which includes €15m from BI for OSE-172 and €12m from Servier for OSE-127. OSE is in the final stages of preparing to submit the Phase I clinical trial application studying OSE-172, and expects to receive the milestone payment once the first patient has been enrolled in H119. OSE is also undertaking Phase I preparation work for OSE-127, which should be ready to move into a Phase I trial in H119, which is when the Servier option can be exercised. Current cash burn is around €15m/year, so these payments would add roughly two years of runway assuming OSE maintains a similar level of R&D work.

Overall, we see OSE’s R&D programmes as very cost-effective due to low out-of-pocket R&D costs, and well diversified, with different mechanisms of action and specific indications.

Tedopi is OSE’s lead asset, a cancer vaccine in development for NSCLC (Phase III) and pancreatic cancer (Phase II). In June 2017, recruitment for Atalante 1, the Phase III trial in NSCLC, was halted following a recommendation by the Independent Data Monitoring Committee, but later resumed in February/March 2018 after a redesign of the study to include only patients who have progressed after a IL/2L checkpoint inhibitor. OSE plans to complete its Phase III study in NSCLC then out-license for the final stage of development and commercialisation. The first readout from this trial is expected in 2019. Management believes that a positive outcome of this study will allow for at least conditional, or even full approval, in this patient group. Another key direction that management envisions for Tedopi is to test its use in combination with other immunoncology drugs, primarily checkpoint inhibitors. The next nearest opportunity to explore this is the investigator-led Phase II trial in pancreatic cancer patients in combination with nivolumab.

In a malignant process, cancer cells die and proteins/antigens are released (step 1 in Exhibit 3). These are then taken up by the patient’s own antigen-presenting cells, or dendritic cells (DCs) and in lymph nodes they present these antigens to T-cells (step 2). This leads to an activation and production of populations of T-cells, which can now recognise and destroy cancerous cells that display the same antigens as those previously presented (steps 3 to 7). However, this process is not perfect, which is why not every malignant process is stopped. Once a tumour develops, it often also has multiple ways to suppress the immune response and enable the tumour to ‘hide’ from the immune cells. The goal of cancer immunotherapies is to expose the tumour microenvironment as foreign to the patient’s immune system so the tumour is recognised and immunologically attacked (step 7), ie turn from ‘cold’ to ‘hot’.

Cancer vaccines are one such immunotherapy strategy – they aim to elicit an immune response that will lead to the selective destruction of tumour cells. So far, Dendreon’s Provenge (dendritic cell vaccine, approved in 2010) and Amgen’s Imlygic (oncolytic virus, approved in 2015) are the only approved cancer vaccines and there is no lack of research being conducted in this area. The relative lack of more successful examples, however, could be explained by variability in design of the vaccine, the selection of antigens and the understanding of the inhibitory tumour microenvironment. In addition, more and more preclinical studies are supporting the idea that the backbone strategy in immunoncology should be combination treatments exploiting different steps in the immunity cycle (Schlom and Gulley, 2018).

The most popular cancer vaccine technologies include peptide vaccines, oncolytic viruses and dendritic cell vaccines. Peptide vaccines are a type of cancer vaccine and work by presenting peptides to the patient’s immune system that are likely to be expressed by the patient’s tumour. Compared with other approaches such as dendritic cell vaccines and other ex vivo approaches, peptide vaccines are more straightforward.

Historically, vaccines were first developed against infective diseases (first virus, then bacteria) and were made from live attenuated pathogens, inactivated organisms or purified whole antigens. However, live vaccines can cause the disease itself, while non-living vaccines were effective in inducing humoral immunity (antibodies produced by B-cells), but not effective in inducing cell-mediated immunity (T-cells directly destroy ‘foreign’ cells), which are the two major types of adaptive immunity that generate an antigen-specific immune response.1 With advancing knowledge of antigen processing and presentation, an epitope-based vaccine emerged that demonstrated the potential to initiate a cell-mediated, antigen-specific immune response and quickly gained popularity in cancer vaccine research. This approach aims to develop vaccines from highly conserved epitopes that are present in large proportions of the population and would elicit a cell-mediated immune response.

Epitopes are the antigenic determinant sections of antigens, ie smaller peptide sequences of the proteins that can be recognized by the immune system as foreign, when presented on the surface of antigen-presenting cells (step 2 in Exhibit 3). The epitopes that are recognized by CD8+ T-cells (cytotoxic T-cells) are bound to major histocompatibility complex (MHC) class I molecules on the surface of APCs. CD4+ T-cell (‘helper’) epitopes are bound by the MHC class II molecules. In humans, MHC is also called human leukocyte antigen (HLA) and MHC class I molecules are encoded by HLA-A, -B and -C genes. There are many specific HLA types, but due to substantial overlap three peptide specificities HLA-A2, A3, and B7 cover around 90% of the population.1

Tedopi is a subcutaneously deliverable combination of 10 synthetic peptides/neoepitopes, nine of which have been optimised from five antigens commonly expressed on tumour cells, known as tumour-associated antigens (TAAs). These TAAs are CEA, p53, HER-2/neu, MAGE-2 and MAGE-3, and are part of HLA-A2, therefore restricted to HLA-A2 positive subjects, around 45% of total population. Tedopi includes a tenth peptide (pan-DR epitope or PADRE), which is a CD4+ T-cell (helper) epitope included to augment the magnitude and duration of CTL responses. 90% of malignant tumours express at least one of these five tumour antigens. The patient’s antigen presenting cells (eg dendritic cells) take up the antigens, process them and present them on their cell surface through HLA-A2. The dendritic cells then prime T-cells in the lymph node (both CD4+ “helper” cells and CD8+ “cytotoxic” cells). These T-cells can now recognise tumour cells as foreign and kill them. Several steps were taken during the development of the vaccine with expectations to increase its potency:

Lung cancer is the leading cause of cancer-related deaths globally (19.4% of total cancer-related deaths); 2.1 million new cases were reported worldwide in 2018. NSCLC is the most common type of lung cancer, accounting for 85-90% of all cases. Five-year survival rates for NSCLC remain poor despite progress over the last 15 years; only 15% of patients diagnosed with all stages of lung cancer survive more than five years. The more recent availability of new treatment options as described below has improved outcomes and survival for patients, but there is still a great need for more effective treatments across first-line (1L), second-line (2L) and third-line (3L) settings.

NSCLC is not a single entity but a number of pathologies with different molecular abnormalities; subsets of NSCLC can be further defined at the molecular level by identification of driver mutation that occurs across multiple oncogenes (Exhibit 4). Therefore, in the clinical setting, treatment of NSCLC depends on the stage at which it is diagnosed and the molecular status of the tumour. Common treatable oncogene mutations in NSCLC include the EGFR mutation (19%, according to GlobalData) and ALK translocation (5%, GlobalData). Recently introduced targeted therapies against these mutations (eg crizotinib, gefitinib, erlotinib, afatinib) have improved progression-free survival in patients to 10-12 months versus six months on platinum doublet treatment in a clinical trial setting.

A standard of care has emerged based on the molecular status of a patients’ tumour in the US and Europe for 1L treatment of advanced NSCLC (Exhibit 4). Particularly the advent of immunotherapy has changed the treatment of NSCLC. The first- and second-line space for these patients is currently dominated by targeted therapies in ALK or EGFR mutated patients or PD-1/PD-L1 inhibitors in ALK/EGFR negative patients. Opdivo (nivolumab, Bristol Myers Squibb) and Keytruda (pembrolizumab, Merck) are approved for the treatment of second-line NSCLC patients who have progressed after platinum-based chemotherapy. Additionally, Keytruda is indicated for first-line patients with metastatic NSCLC whose tumours have high PD-L1 expression (≥ 50%). More recently, Tecentriq (atezolizumab, Roche) was approved for second-line NSCLC patients who have progressed after platinum-based chemotherapy.

However, a large proportion of patients do not respond to currently available checkpoint inhibitors (c 80%) and patients that do respond will acquire drug resistance and experience disease progression. According to the American Society of Clinical Oncology Clinical Practice Guideline Update (2017), the only recommended treatment after treatment with a checkpoint inhibitor and chemotherapy is single-agent pemetrexed or docetaxel as a salvage therapy. Initially, Tedopi is positioned as a therapy for this patient group. This is based on its Phase III Atalante 1 trial, where it is being studied as a 2L/3L treatment for patients with stage IIIB or IV NSCLC, with disease recurrence or progression after treatment with a checkpoint inhibitor and platinum-based chemotherapy and who are HLA-A2 positive. The trial excludes patients with EGFRmt, or ALK rearrangements. From our calculations, we estimate this addressable population for Tedopi to be 32,000 patients in the US and Europe (combined), which we use in our model:

Several other drugs are in late-stage trials for this patient population (Exhibit 5) including Opdivo, which is being studied in four Phase III trials. Many other earlier Phase I and II stage trials explore various combinations of approved therapies, but also novel drugs targeting new mutations including FGFR tyrosine kinase, RET, MET and ROS1.

The Atalante 1 trial was initiated at the beginning of 2016 by OSE. It was a multi-centre Phase III clinical trial for patients diagnosed with advanced NSCLC who are HLA-A2 positive and who had failed on at least the first line of therapy. The trial planned to enrol 500 patients in 70 centres in Europe and the US and was expected to complete in 2018. The first review in January 2017 by the Independent Data Monitoring Committee (IDMC) recommended that OSE continued the trial without modification. However, following the second review in June 2017, recruitment was halted following a recommendation by the IDMC, but later resumed in February/March 2018 after the study was redesigned. The trial has an updated protocol and is now focusing on a subgroup of patients in the original design (Exhibit 7). These patients are post-checkpoint inhibitor failures, ie patients who progressed after 1L/2L treatments with EGFR- or ALK-targeted therapies, or checkpoint inhibitors. This new focus could represent an interesting opportunity since no drug is currently registered for post-checkpoint failure and it represents an area of unmet need.

The new patients recruited into the trial will only be checkpoint inhibitor failure patients, but OSE will continue to treat patients who have already started treatment from other groups. Exhibit 7 shows the new design graphically. According to management, 100 patients will be recruited in addition to the patients already recruited from the checkpoint failure group (step 1). If a certain percentage (required by the IDMC) are still alive after 12 months, this will be regarded as clinically meaningful then the trial will be allowed to enrol an extra 225 patients to a total of 325, at which point OS will be measured (step 2). This pause of approximately nine months led to a delay in the expected data readout (previously 2018, now 2021). The readout from step 1 mentioned above is expected end of 2019 and these data will be published. The final results from step 2 are expected in 2021. Management believes that a positive outcome of this study will allow for at least conditional or even full approval in this patient group.

A Phase II study was carried out by Epimmune, a previous developer of Tedopi, between 2004 and 2007. The study design is shown in Exhibit 8.

This study was designed as an open-label, single-arm that enrolled 64 HLA-A2 positive patients, who received Tedopi for two years. For comparison, patients who were screened for this trial but had HLA-A2 negative status were followed for one year. Therefore, only short-term survival comparison data ares available. The Tedopi group was followed for a longer period. Results included:

As discussed above, increasing knowledge of immunotherapeutics suggests there is a strong case for combining treatment methods that affect different parts of the cancer immunity cycle. Outside the Atalante 1 trial, OSE sees combinations with checkpoint inhibitors as another key direction for Tedopi. OSE was approached by GERCOR, an association of physicians, to carry out an exploratory Phase II study with Tedopi in combination with nivolumab in pancreatic cancer. Since GERCOR is sponsoring the study and OSE only contributes limited financial support, this represents a cost-effective opportunity for Tedopi to be studied in combination. The trial protocol has not been finalised yet, but GERCOR plans to study Tedopi in combination with a checkpoint inhibitor and initiate the trial in H119. The preliminary trial design is shown in Exhibit 11. Management will evaluate options for further development and commercialisation depending on results from this study, which is when we will also reassess the opportunity. The insights from this trial could lead to further development for pancreatic cancer or a Phase III trial in NSCLC in combination with a checkpoint inhibitor.

The main patent protecting Tedopi was filed in 2004 (PCT application WO2004094454); the expiry date is therefore 2024 with potential to extend by five years. OSE has since filed new patent applications – two PCTs filed in 2014 (PCT/EP14/73975 for brain metastasis) and 2015 (PCT/EP2015/064746 method for inducing early T memory response with a short peptides antitumour vaccine). Tedopi was granted orphan drug designation in 2013 by the FDA for the treatment of NSCLC in patients expressing HLA-A2. This means that if Tedopi reaches the market before similar products, this could provide seven years of exclusivity from the date of approval. In addition, the biologicals licence in the US and Europe would give 12 and 10 years of market exclusivity respectively.

OSE-172 is a first-in-class SIRPα antagonist and is currently in preclinical stage with a Phase I study in collaboration with BI expected to start in H218 (fully funded by BI). The drug is expected to work in a similar way to T-cell immune checkpoint inhibitors in the tumour microenvironment, but instead of T-cells it inhibits the checkpoints between tumour cells and myeloid cells (Exhibit 12): myeloid-derived suppressor cells (MDSCs) and tumour-associated macrophages (TAMs). It binds to signal regulatory peptide alpha (SIRPα) on myeloid cells, which inhibits the SIRPα/CD47 interaction (CD47 on the surface of cancer cells). CD47 acts as ‘a don't eat me’ signal to macrophages of the immune system, so when blocked this increases the likelihood that the myeloid cell recognises the cancer cell as foreign, then attacks and digests the cancer cell. Phagocytosis leads to presentation of cancer antigens on the surface, which stimulates the immune system.

OSE licensed OSE-172 to BI in April 2018. Under the terms of the agreement, BI will fund all future development including the Phase I trial, which is expected to initiate in 2018. This Phase I trial (design not released yet) will study OSE-172 in multiple solid tumour indications. BI will take over development and commercialisation completely after Phase I.

This is OSE’s largest deal in terms of total deal value (€1.1bn potential development, regulatory and sales milestones plus royalties). According to management, this is broken down roughly as 30% development milestones and 70% sales milestones, plus a 10% royalty on worldwide net sales. For context, there have been other preclinical SIRPα/CD47 inhibitor deals in the past few years:

Assets in the CD47 and SIRPα pathway landscape are early stage (Exhibit 13). CD47, which is present in tumour cells, is the more studied target and six companies are studying CD47 monoclonal antibodies in Phase I/II clinical trials. Forty Seven is leading the way with its asset Hu5F9-G4, which is currently being studied in six Phase I/II trials including monotherapy and combinations with targeted therapies or T-cell checkpoint inhibitors. Its most advanced trial is a Phase II trial in non-Hodgkin lymphoma in combination with rituximab in 72 patients and, according to clinicaltrials.gov, has a primary completion date of March 2020. Forty Seven is also developing a SIRPα inhibitor FSI-189, which is in preclinical stage. The only SIRPα inhibitor which is in clinical stage is ALX148, so the data from ALX Oncology will be relevant for OSE.

OSE has accumulated proof-of-concept data in various in vivo models, which were recently presented at the AACR conference in April 2018. OSE’s SIRPα antagonist antibody demonstrated:

The next step for OSE-172 is a first-in-human study. This Phase I study, for which OSE will be reimbursed by BI, is expected to start in early 2019. The study design and timelines have not been made public yet. A milestone payment will be triggered on Phase I initiation.

FR104 is a fragment antigen-binding (Fab) region of an antibody that acts as CD28 antagonist. Currently it is in development for rheumatoid arthritis but could be used in other immune disorders. CD28 acts as co-receptor in the T cell receptor (TCR) and delivers stimulatory signals from antigen presenting cells to T cells. FR104 has a similar mechanism of action to Orencia (abatacept), which blocks T-cell binding to antigen-presenting cells via blocking CD80/86 on antigen-presenting cells, preventing it from binding to the CD28 receptor on the T-cell. In a similar way, Orencia normalises inflammatory mediators in rheumatoid arthritis patients. FR104 blocks the other side, ie the CD28 receptor. The differentiating aspect of FR104 is that, while abatacept blocks both stimulatory and inhibitory signals via CD28 and CTLA-4 respectively, FR104 selectively blocks only the stimulatory signal via CD28 (Exhibit 15).

Orencia is marketed by Bristol-Myers Squibb for RA and peak sales are expected to be $2.8bn in 2020 (source: EvaluatePharma). It is also marketed for juvenile idiopathic arthritis and psoriatic arthritis, and in clinical trials for various other immune disorders. Orencia was the first FDA-approved agent for RA that blocks activation of T-cells, and there are still no other drugs targeting T-cells; nor biosimilars yet for abatacept. We believe there is room for another T-cell co-stimulation inhibitor in RA, and so FR104 has potential. Belatacept is a marketed CD28 antagonist marketed for kidney transplantation in the US with $100m in sales in 2017 (EvaluatePharma). A private Russian biopharmaceutical company TheraMAB is developing an anti-CD28 antibody, albeit it has been described as having an agonistic action. The product is in early stages of development.

Rheumatoid arthritis is a chronic autoimmune disease characterised by inflammation in the joints. It affects approximately 1.8 million patients in the US and 1.4 million in Europe (EU14). Patients are treated according to disease severity: mild (23%), moderate (45%) and severe (32%) (GlobalData). Patients with mild RA are usually treated with disease-modifying, anti-rheumatic drugs (DMARDS) as monotherapy such as methotrexate plus steroid therapy if needed. Moderate disease is treated with DMARD monotherapy followed by biologics such as an anti-TNF or non-TNF biologic if required, and finally severe patients may be treated with DMARD combination therapy, anti-TNF or non-TNF biologics.2

Due to the chronic nature of the disease, high prevalence and introduction of higher priced biologic drugs, RA is a large market. According to EvaluatePharma, the RA market in 2017 was $26.0bn. Anti-TNFs have been the most successful class of RA drug in terms of revenues including Humira (AbbVie), which reached $6.2bn in sales in 2017; Enbrel (Amgen), which reached $4bn peak sales in 2016,3 and Remicade (J&J) which reached $2.5bn peak sales in 2013. Newer classes of drugs include JAK inhibitors, and anti-IL-6 MAbs.

However, the market is changing with the introduction of biosimilars, which suggests that the branded market is likely to remain flat and somewhat decline later on. Longer-term dynamics will depend on the novel drug launched, its health benefit and achieved pricing. A GlobalData survey of physicians carried out in 2018 found that biosimilars are already being prescribed extensively for RA, especially in Europe. 69% of physicians surveyed from EU5 countries reported prescribing biosimilars to at least half of their patients. In the US this is much lower at 24%. Marketed and late-stage biosimilars in RA are for anti-TNFs (ie Infliximab, adalimumab, etanercept). However, the introduction of biosimilars is causing pricing pressure in the whole market. For example, Orencia (BMS) costs around $50,000 per patient per year but has many discounts to remain competitive, so only bringing in around $10,000 per patient per year (EvaluatePharma).

So far, FR104 has been studied in Phase I in healthy subjects and OSE has not indicated which subpopulation will be targeted. We assume at this stage that FR104 would be most suitable for moderate/severe RA patients as a 2L/3L therapy similar to Orencia before its label extension in 2008.

In 2016, OSE completed a Phase I study with FR104, which showed a good safety profile. The study was a first-in-human, randomized, double-blind, placebo-controlled, single centre study evaluating single and multiple ascending intravenous doses of FR104 in healthy subjects. In total, 65 subjects were enrolled and received FR104 intravenously. The endpoints included safety and PK/PD analysis.

The study found that FR104 was well tolerated and the recommended Phase II dose was determined. There were no discontinuations and no deaths. There was only one serious adverse event, nephrolithiasis, which was not deemed to be related to FR104. There were 42 treatment emergent adverse events (TEAE, related to the treatment), of which 0 were categorised as severe, 15 were mild, 14 were moderate and 13 were categorised as ‘at least possibly treatment-related’.

In November 2018, Janssen returned the licensing rights to OSE. According to OSE’s press release, this was not related to the product itself, since no red flags had been identified, but rather the result of a strategic review and reprioritisation of its immunotherapy portfolio away from rheumatoid arthritis. OSE is now considering all options for further development of FR104, which include developing the asset itself (it has all the necessary data to start the Phase II study) and partnerships. Most likely, the company will try to find a partner as a first step so that it can focus on the Tedopi trial and on its research platform. OSE will receive data from Janssen by the end of 2018, and we therefore expect the company to initiate the business development activities with regard to this asset in early 2019, so an update on this programme could come in later in 2019.

OSE-127 is a humanised monoclonal antibody against IL-7Rα, specifically CD127, in late-stage preclinical development for ulcerative colitis and other autoimmune diseases. IL7 is a cytokine that controls the proliferation, apoptosis and activation of CD4 and CD8 T-cells in humans. OSE-127, a humanized monoclonal antibody, is an antagonist of the IL7 receptor (IL-7R) present on T-cells (CD127), thus down regulating the immune activity.

OSE has a commercial agreement with Servier, signed in December 2016. The potential deal value is €272m, including a €10.3m upfront payment (received 2016). The deal involves a two-step €30m option exercise, development/commercial milestones later on and double-digit royalties on sales. OSE is developing the asset until Phase II (each phase will trigger an options exercise fee). Servier will then continue the development of the asset. Servier plans to study OSE-127 in ulcerative colitis and potentially in Sjögren’s syndrome. A €12m option exercise fee (first step) is expected in H119.

Ulcerative colitis (UC) is a type of inflammatory bowel disease characterised by inflammation of the gastrointestinal tract (mainly large intestines) and associated symptoms including rectal bleeding and diarrhoea. Patients are treated according to whether they have active disease or are in remission, as well as in relation to the severity of disease, which can be mild, moderate or severe. Approximately 700,000 patients in the US and 600,000 patients in EU14 are treated every year (Global Data).

Exhibit 18 shows the UC market forecast. There are just a few of drugs that currently have blockbuster status in this market, which are Humira (AbbVie) with $2.2bn sales in 2017, and Entyvio (Takeda) with $0.9bn in sales in 2017. Biosimilars in this market are gaining market share in a similar way to RA. The company has not reached the stage at which it can decide on positioning in this market. However, we expect OSE-127 will most likely be positioned as a 2L/3L treatment after 5-ASAs, steroids, and anti-TNFs.

OSE is subject to typical biotech company development risks. The largest part of OSE’s value comes from Tedopi in NSCLC. Therefore, a significant risk is the outcome of the Atalante 1 Phase III trial and the success of subsequent activities. If the Phase III trial is successful, OSE plans to partner the asset, which is subject to execution and partnering risks. Our model assumes that Tedopi will be out-licensed and our valuation is therefore sensitive to potential licensing timing and actual deal terms. The main patent of Tedopi expires in 2024, with potential to extend the protection for five years. OSE filed for fresh patents in 2014/15, which are yet to be granted. However, market exclusivity after the launch of Tedopi could be secured if OSE receives orphan drug designation in the US, and the biologicals licence in the US and Europe.

Other significant risks are those relating to OSE’s existing commercial partnerships with BI and Servier. As with any licensing agreement, execution is dependent on both parties working together. There is always risk from a partner’s change in strategy, which could lead to deprioritisation of the asset. In this case, the asset could be returned to OSE. This was the case when Janssen returned the rights to FR104 to OSE, which is now faced with the task of re-partnering the asset or increasing investment in R&D in order to develop the asset.

OSE is due to receive up to €27m in licensing income in 2019. This is expected to finance the company to early 2021 but if these payments are not received, OSE will have to raise finance in another way to fund its operations.

We value OSE based on risk-adjusted NPV analysis using a 12.5% discount rate, including estimated net cash of €7.5m at end-2018. This results in a value of €171m or €11.7/share. We include four assets for four indications in our valuation, namely:

Exhibits 19 and 20 summarise our detailed assumptions for the risk-adjusted NPV valuation. The two near-term milestone payments from Servier and BI are dependent on the respective assets moving from late preclinical stage to Phase I. Typically, this is a low-risk transition but, to reflect any remaining R&D risk and the fact that ultimately this will be a third-party decision, we include the €12.0m payment from Servier and €15.0m payment from BI in 2019 with a 90% risk adjustment.

For Tedopi, we assume a licensing deal after this Phase III study for which the company may seek a commercial partner rather than building a sales organisation on its own. For FR104, we assume that OSE will not initiate the next Phase II trials on its own, but will focus on establishing a new partnership deal once the data package from Janssen is received (expected by end-2018). We assume better deal terms than those with Janssen, which were signed while the asset was still in the preclinical stage (currently finished Phase I, more details in Exhibit 19). For OSE-172 and OSE-127, we use the deal terms announced with BI and Servier, respectively.

We have discussed the target populations for each of the assets in the relevant sections above. Once launched, pricing of the products is benchmarked to the relevant drugs currently existing on the market. We use the US, the top five European countries, Benelux, Scandinavia, Austria and Switzerland as key initial markets.

In its H118 financial results, OSE reported €20.6m in revenues (vs €2.8m in H117). This includes the upfront payment from BI for OSE-172 received in April 2018. Total operating expenses for H118 were €10.2m vs €7.3m in H117. R&D expenses are the company’s largest expense: €8.0m in H118 vs €7.9m in H117. The largest contributor is the Tedopi Atalante 1 Phase III study. We expect R&D costs to remain largely constant at €8m per six months of operation until the end of the trial in 2021. This is because we expect other clinical trials to be fully financed by the partner and grants for the cost will be relatively low for the preclinical and some clinical work. We do not expect OSE to initiate its own Phase II study with FR104, recently returned from Janssen, so we do not factor in any costs relating to future studies in our financial model, as we expect the company to seek a partner. Overhead expenses were €1.73m in H118 vs €1.78m in H117. OSE’s operations are currently funded through its partnerships and Bpifrance grants, where some grants relate to a particular R&D project.

As of 30 June 2018, OSE had cash, cash equivalents and financial assets of €18.6m (includes ‘current financial assets’). Last reported (H118) results also include debt of €4.9m, which is mainly government loans. According to management, current cash burn is around €15m per year. Therefore, current cash is expected to last into H219. If both milestone payments totalling €27m are received in H119 (from Servier and BI), the cash runway should extend into early 2021. If these payments are not received, OSE may have to raise funds in another way or adjust its R&D activities. While we expect these milestones to be paid, we do not include them in our model currently.