Onxeo — Update 2 June 2016

With its new acquisition of DNA Therapeutics, Onxeo added a novel clinical-stage compound, AsiDNA (formerly known as DT01). AsiDNA is based on siDNA technology, which has already been tested in clinical trials by DNA Therapeutics and showed positive safety/tolerability results and preliminary anti-tumour activity in a Phase I trial with melanoma patients when administered locally. Following the closure of the deal, Onxeo is now optimising the manufacturing of the compound and plans to continue clinical development afterwards employing a systemic administration route.

First-in-class AsiDNA is based on signal-interfering DNA technology, which if introduced into a cell acts as a signal mimicking the damage of the cell’s own DNA. AsiDNA molecules are short double-strand DNA that mimic double-strand breaks in the cell’s DNA and are recognised as ‘damaged DNA’ by repair and signalling proteins. Therefore AsiDNA then activates a cascade of repair proteins, which are recruited to ‘repair the damage’, as a result of which the actual damage of a cell’s DNA (such as from radiotherapy or chemotherapy) remains unrepaired, leading to cell death. This false signalling of the DNA damage was first explored by DNA Therapeutics in combination with radiotherapy and delivered promising results in preclinical in vivo and in vitro models.1 Double-strand DNA molecules do not spontaneously enter tumour cells, however, this limitation was resolved using cholesterol as an agent enabling AsiDNA to pass the cell’s membrane.

Few products involving DNA repair inhibition are in advanced clinical stages and so far only one has been approved by the FDA and the EMA: olaparib (Lynparza) developed by Kudos Pharmaceuticals and Astra Zeneca. Olaparib inhibits poly ADP ribose polymerase (PARP), which is a downstream enzyme involved in the DNA repair process. Olaparib has been approved for the maintenance treatment of women with BRCA-mutated (BRCAm) ovarian cancer and AstraZeneca is further investigating the potential in other PARP-dependent tumours, such as gastric cancer, pancreatic cancer and metastatic BRCAm breast cancers. BRCAm tumour cells have increased reliance on PARP to repair their DNA and thus may be susceptible to treatment based on PARP inhibition. AsiDNA is, however, clearly differentiated, as it acts more upstream; it is not a specific enzyme inhibitor, but activates PARP (ie ‘distracts’, when olaparib inhibits) among other repair proteins. Therefore, it represents a new treatment approach with broader action than PARP inhibitors, targeting the entire DNA repair system by hijacking DNA repair signalling.2

The initial preclinical efficacy of siDNA technology was observed in combination with irradiation, ie cells treated with radiation have damaged DNA and siDNA prevents it from being repaired, which leads to tumours being more sensitive to radiation therapy. However, due to siDNA acting as a non-specific decoy activating a false DNA damage signal and thereby distracting repair proteins from cellular DNA damage, other treatment modalities that produce DNA damage were also found to be possible for use in combination with siDNA, such as chemotherapy (with alkylating agents, antimetabolites, topoisomerase inhibitor) and radio-ablation therapy (hyperthermia). In fact, existing preclinical data from a several tumour models with different DNA damage therapies and different AsiDNA administration routes show that AsiDNA is a very versatile technology (Exhibit 1).

AsiDNA (DT01) has already been tested in Phase I DRIIM with skin melanoma patients, where it was injected intratumourally or peritumourally alongside the radiation therapy. This was an open-label, non-randomised, multicentre, dose escalation study. In total, 23 patients received a full course of treatment and were evaluated for safety and pharmacokinetics, while 21 patients with a total of 76 skin melanoma lesions were evaluated for initial efficacy. Key headline results were presented at ASCO in May 2015 (full results yet to be published):

Onxeo completed the acquisition at the end of March and is optimising the manufacturing process while preparing the clinical development plan. Following this, the company will resume the clinical development, but the indication is yet to be selected. Based on preclinical data and insights from the DRIIM trial, Onxeo believes that AsiDNA’s systemic administration is feasible. As a result of this, a number of other opportunities may be more attractive that skin melanoma. As Onxeo indicated, two likely indications for further clinical development are triple-negative breast cancer and platinum-resistant ovarian cancer. Due to the changed route of administration, we believe that AsiDNA will have to undergo another safety/dose escalation study before moving into a Phase II proof-of-concept trial. However, because of the versatility of AsiDNA’s technology there is potential to expand into a number of different tumours later on, which implies a significant upside.

Our new valuation of Onxeo is €343m or €8.3/share, vs €329m or €8.1/share previously. The main changes to our model are the addition of the AsiDNA project with a TNBC indication, postponing the development timelines for Validive, rolling forward the model in time and adjusting for an updated cash positon at the end of Q1 2016. AsiDNA has clearly added value to Onxeo R&D pipeline and in order to capture that, we have focused on triple-negative breast cancer (TNBC), which is now added to our valuation. TNBC has a significant unmet need with no novel targeted treatment approved.

By pathological definition, TNBC lacks an expression of oestrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2). This type of cancer is typically more aggressive when compared to other types of breast cancer and is unresponsive to hormonal and monoclonal antibody therapies (eg trastuzumab). The standard initial treatment options are anthracycline and taxane-based combinations (eg doxorubicin, cyclophosphamide, docetaxel), but patients have a poorer prognosis and overall survival than in other types of breast cancer. A variety of new targeted therapies are in investigation, including PARP inhibitors, with Astra Zeneca also exploring its olaparib in Phase III in metastatic breast cancer with germline BRCA1/2 mutations and HER2 negative disease.3

TNBC accounts for around 15% of the total of new cases each year, which would imply c 111k patients in the US and Europe. In our valuation, we assume initiation of a Phase II study in 2018 and Phase III in 2020 and launch in 2024. Other assumptions include a price of $70,000 per patient in the US and 20% less in Europe, which is based on an average yearly cost of trastuzumab (Herceptin, Roche) used in other types of breast cancer. We also assume a 15% peak market penetration rate, although given the lack of options for these patients there is upside to that. Notably, the pricing and competitive environment are difficult to foresee given a number of other projects advancing in the field, with 34 projects in Phase III and 83 projects in Phase II in breast cancer overall (BioCentury). Our calculated peak sales are €1.1bn and we assign 15% probability to reach the market.

We assume that Onxeo will be able to partner AsiDNA after Phase II and the partner will cover all development and marketing costs from this point. Before that we include a cost of €4m for the Phase I trial (safety/tolerability) and €10m for the subsequent Phase II trial. Our partnering assumptions include a fairly typical deal structure, including an upfront payment, development and sales-related milestones, in addition to royalties on global sales. We assume a total deal value of $300m (€243m), which is below the $570m deal signed in 2015 between Medivation and BioMarin for a PARP inhibitor in Phase III with breast cancer patients. In another deal worth $255m with an undisclosed upfront payment, Clovis in-licensed a Phase I/II PARP asset being developed for solid tumours including breast cancer from Pfizer in 2011. In our model we include €24m as an upfront payment and €97m in development-related milestones, with the remainder as sales-related milestones, and tiered up to 15% royalty rates on global sales. The remaining details and our valuation are summarised in Exhibit 2.

Besides adding AsiDNA with a TNBC indication to our valuation, we have also made adjustments to the Validive project, as Onxeo will now focus on partnering the asset before moving to Phase III. As no timelines were disclosed, there is no visibility as to when this project could progress. Therefore we have postponed further development for two years and decreased the probability of success from 60% to 50% to reflect the additional risk of uncertainty of establishing a partner. This decreases Validive’s rNPV from €67m to €32m. If the Validive project is removed from our valuation entirely, Onxeo’s rNPV is €311m. The decrease in the Validive valuation is partially offset by the addition of the AsiDNA asset, which, although at an earlier stage of development (reflected in its probability of reaching the market), has a much higher (unadjusted for risk) NPV than Validive.

Validive’s key patent expires in 2029, which is still comfortable and, in our view, would not be a limiting issue in partnering. We have already included a deal for Validive in our NPV project (see our outlook report), assuming that Onxeo will commercialise alone in Europe and factoring in costs for the build-out of commercial infrastructure, with the partner covering other key regions such as the US in exchange for a royalty on sales. We leave this deal structure unchanged.

The DNA Therapeutics acquisition included an upfront payment of €1.7m from Onxeo paid by issuing 553,819 new shares (1.4% of the total). An additional payment of €1m is due once AsiDNA enter a Phase II trial (we assume in 2017). Onxeo will be able to choose whether this should be settled in shares or cash. For the purpose of our model, we include €1m in long-term debt to finance the additional payment. Future commercial royalty payments can total up to €25m per indication. In addition, DNA Therapeutics shareholders agreed to invest €1m in cash in Onxeo through a private placement. Our financials now reflect the acquisition and the investment.

On 28 April Onxeo provided a Q116 business update with key financials. Revenues were €782k versus €918k a year ago. The decrease was primarily due to lower non-recurring revenues of €27k compared to €157k in Q115. Recurring revenues were €755m, broadly in line with €761k in Q116, generated by royalty income from partner marketed products (Beleodaq, Sitavig and Oravig). Onxeo did not provide spilt per product, but we believe the bulk of this originated from Beleodaq US sales. Recurring FY15 revenues were €2.7m and our new 2016 estimate is €5.2m, which we revised downwards from €6.8m to reflect our more conservative growth expectations of the royalty income this year.

Onxeo reported cash and equivalents at end-Q116 of €24.4m and was virtually debt free at end-December 2015. Reported cash position does not include a €1m investment from the DNA Therapeutics shareholders. As Validive now is on hold, we believe that the resources will be redirected to the development of AsiDNA, however, at this stage there is little visibly about this project until Onxeo provides an update. Therefore, we continue to believe that current cash should be sufficient to fund operations well into 2017.