Neovacs — Update 1 August 2016

Neovacs is based on kinoid technology, a type of vaccine that aims to create a long-lasting immune response against the natural immune response signalling molecules driving particular medical conditions. Neovacs has rebuilt its pipeline, Exhibit 1, after the late 2014 failure of Tumour Necrosis Factor Kinoid (TNF-K) in rheumatoid arthritis. Interferon alpha kinoid (IFN-K) started an international, 178-patient, Phase IIb study, NCT02665364 in September 2015. Data are expected in mid-2017. This study now includes a 12-patient US arm. In 2015, Neovacs developed its relationship with a key supplier, Stellar Biotechnologies, and a manufacturing joint venture has been agreed with grant funding. In late December 2015, Neovacs partnered IFN-K with CKD, a Korean pharmaceutical company. If the current Phase II is successful, CKD will be able to market IFN-K in Korea on a registry basis from 2018. Neovacs is based in southern Paris. It has 24 employees

As IFN-K will be Phase III funded and sold by a partner, a 15% royalty is assumed with a €25m deal fee in 2018 and regulatory milestones of €100m. The IFN-K Phase IIb probability used is 30%, an upper Phase II level reflecting the IS strong positioning. US 12-year biological exclusivity ends in 2032; European protection ends in 2030. The French tax rate is set at 24% on the assumption that Neovacs will offset profits. This gives an indicative value of €1.29/share or €1.15 fully diluted. This is assessed before further funding needs, which are expected to be at least €8m in 2017. Over the next few years, Neovacs is well placed to develop products and do deals.

In FY15, Neovacs had an accounting gain of €4.3m as governmental loans for the failed TNF-Kinoid project became non-repayable. Operational costs in FY16 are expected to be €14m vs €12.5m in 2015, due to Phase IIb costs. FY15 cash use was €7.8m before funding. Year-end 2015 cash was €6.1m, with the €1m upfront fee from CKD due in Q116 and a tax credit of €1.8m also due. A June 2016 rights issue raised €8m gross at €0.85/share. This gives Neovacs sufficient cash for at least 2016. Additional funding of at least €8m may be needed in 2017 plus the next Korean milestone, due in H217 (management expects about €1.75m) if the Phase IIb is successful. We estimate a tax credit of €3.1m (30% of R&D) for 2016, paid in 2017. Further funding depends on partnering deals in 2018.

The hypothesis that the disease is exacerbated in about 80% of lupus patients due to overproduction of interferon alpha leading to an interferon signature (IS) has received support from two large Phase II AstraZeneca studies. Neovacs is using a combined IS and clinical co-primary endpoint in the current IFN-K Phase IIb. The use of a high IS score to select patients should reduce risk, as a high IS is seen in about 80% of patients and other trial data indicate better outcomes in this group (see Exhibit 5 below). However, the IS case is not yet fully established through a pivotal study. Neovacs intends to partner the project at Phase III using the Phase IIb data. The €5m Korean deal shows that the project can attract partners at a good value. Substantive Phase III trials will be required for the crucial US FDA approval and IFN-K is several years behind AstraZeneca’s anifrolumab, in Phase III with good Phase II data. IFN-K needs a different commercial strategy as it is a one-off course of injections, rather than a chronic monoclonal therapy with reoccurring monthly injections and a long-term income stream.

IFN-K had been in early clinical development (a 28-patient Phase I/II in 2010-11), but the former management gave the larger RA market targeted by TNF-K priority. TNK-K failed, apparently as the antibodies generated bound, but did not neutralise, TNF. Neovacs has data showing that IFNα-K generates a range of neutralising antibodies against IFN. Once immunised, patients may not need a booster course for a number of years. However, this makes the treatment, in effect, a one-time course with no repeat revenues. IFN-K- is likely to be much cheaper to manufacture than a monoclonal. This could mean higher margins.

The next clinical product is likely to be vascular endothelial growth factor (VEGF) Kinoid (VEGF-Kinoid) for cancer and eye disease (acute macular degeneration (AMD).

Malfunction of the immune system causes over secretion of the cytokines (proteins that act on cells to stimulate or inhibit their function), which play a key role in autoimmune diseases. Neovacs’s “kinoid” technology vaccinates patients to neutralise these cytokines and control the disease. Kinoids consist of a target protein, such as interferon alpha (IFNα) conjugated with a carrier immunogenic protein: keyhole limpet hemocyanin (KLH), Exhibit 2.

The patient is immunised by intramuscular injection of the kinoid with an emulsion adjuvant. The KLH stimulates the body’s immune system to produce polyclonal antibodies to neutralise and eliminate the target self-protein. A key observation is that the antibodies need to be neutralising against the cytokine. If not, the cytokine can still function and the disease may not be affected. Neovacs has stated that this was the reason TNF-K failed in clinical development as the antibodies were not neutralising. In IFN-K Phase II data, management observed that the antibodies were “strongly” neutralising. The immune response in some patients can last for at least four years. Polyclonal antibodies are a natural, diverse and more powerful response than monoclonal antibodies. By comparison, monoclonal antibodies need to be injected every four to eight weeks. IFN is produced in various subtypes. A monoclonal will only bind certain subtypes, but IFN-K generates a range of antibodies. Neovacs found that antibodies against all subtypes of INFα were generated.

IFN-K for lupus might be protected by a patent application (WO 2012136739 A1) filed on 4 April 2012. A patent WO 2004024189 A1 (granted in the US but not Europe) filed on 16 September 2003, covers IFNα liked to KLH and WO2002011759 (8 August 2001) claims IFN or VEGF linked to KLH.

Systemic lupus erythematosus (SLE) is the most common form of Lupus: an autoimmune disease of the connective tissues of various organs, Exhibit 3. There is no cure at present.

IFN-K is the lead product now in a 178-patient Phase IIb. This trial follows on from a Phase I/II in 28 patients (Lauwerys et al 2013) summarised in Exhibit 4. In the Phase I/II, most patients were tested for IS at baseline. Patients who produced higher levels of anti-IFN in response to IFN-K reduced their IS scores. The authors viewed the results as “promising” but not powered to show efficacy.

The Phase IIb (NCT02665364) started in September 2015. This trial has a 166-patient target for Europe, Russia and Latin America. It was extended in April 2016 to include 12 US patients under an FDA IND. The total will be 178 patients. Five doses of IFN-Kinoid are given: 240µg on days 0, 7 and 28 then a booster shot of 120µg in week 12 and week 24. The endpoint is determined at week 36. Data are due in June 2016, so recruitment needs to compete around the end of Q316.

The co-primary endpoint in the Phase IIb IFN-K study is a reduction in interferon signature (IS). Obermoser (2010) commented that: “the type I interferon cytokine family has been postulated to play a central role in SLE pathogenesis, by promoting feedback loops progressively disrupting peripheral immune tolerance and driving disease activity.” The development of the interferon signature theory and clinical data from AstraZeneca in the last few years offer a robust paradigm for patient selection. Exhibit 5 has details of IS and references.

Kennedy (2015) published a Genentech IFN biomarker study that found three signature genes marking out high and low populations. The high IFN group showed a clear serological response to various anti-IFNα therapies but the low IFN group did not. However, there was no correlation with mean disease activity assessed by BILAG or SLEDAI-SELENA scores. It would be natural for IFN responsive genes to show lower expression if IFN levels are lowered, but the disease only responds if IFN is the primary driver. Kennedy concluded that: “The ISM does not distinguish between high and low disease activity, but may have utility in identifying patients more likely to respond to treatment(s) targeting IFN-α”.

The other co-primary endpoint is the clinical response measured by the BICLA scoring system (see below). Various lupus systems have been developed over the years. These have continued to evolve. Mikdashi (2015) provides a summary of the latest iterations; Petri (2007) provides an overview and Lam and Petri (2005) give scoring sheets if much more detail is required. The BICLA composite score in the IFN-K Phase IIb trial combines BILAG and SLEDAI-2k.

Because lupus is so difficult to diagnose, estimates of its incidence (new cases per year diagnosed) and prevalence (patients with ongoing disease) vary widely. Exhibit 7 summarises data.

The incidence number is very low, so hard to measure. As lupus is an episodic chronic disease, many people live with the condition, so prevalence figures are more robust, though still variable. The most recent US estimates indicate about 160,000 US cases, Exhibit 8. Extrapolating from this, there may be about 170,000 cases in accessible European markets, but this is not a definitive figure and many might be undiagnosed. If an approved IFN-K is only indicated for high-IS patients, and probably only reimbursed for them, sophisticated genetic testing (21 genes) or a simpler, three gene test like that of Kennedy will need to be easily available to find the patients.

In December 2015, Chong Kun Dang (CKD) Pharmaceutical Corp, a leading Korean pharmaceutical business, licensed IFN-K for the South Korean market. It paid €1m upfront, with a further €4m due until 2018. A Koran IND has been granted and five centres in Korea are participating in the current Phase IIb study. The Korean system allows a conditional approval from 2018 if the current Phase IIb is successful. This means that IFN-K could be sold from 2018. If approved at Phase II, it is assumed that follow-up of patients (a registry) will be needed. Management expects that over 1,000 patients will be treated between 2018 and 2020.

The South Korean market has a population of about 50 million. Recent surveys suggest a lupus prevalence of 26.5 per 100,000 (Shim et al, 2014), so about 13,250. Ju et al (2014) found 9,000-11,000 treated patients in 2006 based on health insurance records. Neovacs management estimate 20,000 cases overall in the population. Treatment is in specialist centres. CKD was founded in 1941. It had 2014 sales of KRW544bn (€422m) and profits of KRW35bn (€29m).

Benlysta, a monoclonal antibody, is the only biological agent to be approved so far for SLE, in 2011. It reduces the production of autoantibodies. There are other therapies in development (Exhibit 9).

Benlysta sales have been hampered by a high price of around $35,000 per year, modest efficacy and a restrictive label. Worldwide sales for Benlysta in 2015 were £230m; growth was 25%. Of these sales, 90% were in the US. Of the development-stage products, most threatening might be AstraZeneca’s anifrolumab, which blocks the IFN receptor to reduce interferon effects; IFN-K neutralises IFN. Anifrolumab will have Phase III data by late 2018, so could be marketed from 2020.

Potential further indications for kinoid technology are autoimmune diseases like dermatomyositis. This a rare autoimmune condition related to lupus. There is no cure. It manifests as a skin complaint, but underlying muscle and connective tissues are attacked causing weakness. Neovacs plans to initiate a 15-patient Phase I trial starting in Q316, with data in late 2017, using the proceeds from the June rights issue.

Data on the market is very limited. Reeder (2010) estimated an incidence of up to 3,200 new cases a year, and a prevalence of 22/100,000: 70,000 cases based on a small sample. Neovacs has cited a similar incidence but prevalence at perhaps up to 32,000.

Neovacs has decided to enter the immunotherapy market with Vascular Epithelial Growth Factor Kinoid (VEGF-K; in preclinical development). Trials could start in Q316 with data in H217. VEGF is proven to be involved in the formation of new blood vessels in cancer growth, particularly of colorectal cancer. Stopping vascularisation slows tumour growth. VEGF is also involved in wet acute macular degeneration (AMD). This has developed into a major market for anti-VEGF antibodies, but these are expensive and need to be injected into the eye every month. A cheaper product that avoids the need for eye injection would have a considerable advantage, but it would need to be as effective to preserve sight. Markets are detailed in Exhibit 10.

The joint venture, based near Paris, will be owned 30% by Stellar and 70% by Neovacs and will operate as Neostell SAS. A CEO and board have been appointed and the company was legally established on 19 July 2016. Founding capital in Neostell, set at €400,000 will be made if the IFN-K Phase IIb study in Lupus is positive. Bpifrance, a French government agency, will provide non-dilutive funding for €5m until 2019 to finalise clinical development and to scale up manufacturing of IFN-K. Neostell will produce all Neovacs’s kinoid products (particularly IFN-K) and other conjugated therapeutic vaccines on behalf of third parties. Edison assumes that partners are supplied at a transfer price including a margin divided between the parties. Stellar Biotechnologies, a US company, manufactures keyhole limpet hemocyanin (KLH), the immunogenic carrier protein conjugated to the antigen to make kinoids like IFN-K. Stellar has supplied KLH to Neovacs for many years and will supply Neostell.

The hypothesis that the disease is exacerbated in about 80% of lupus patients due to over production of interferon alpha leading to an “interferon signature” (IS) has received support from two large Phase II AstraZeneca studies. Neovacs is using a combined IS and clinical co-primary endpoint in the current IFN-K Phase IIb. The use of IS should reduce risk as a high IS profile is seen in about 87% of patients and other trial data indicate better outcomes in this type of group. However, the IS case is not yet fully established through a pivotal study. To progress to Phase III, Neovacs intends to partner the project using the Phase IIb data. The €5m Korean deal shows that the project can attract partners at a good value. Substantive Phase III trials will be required for the crucial US FDA approval and IFN-K is several years behind AstraZeneca’s anifrolumab, in Phase III with good Phase II data. IFN-K needs a different commercial strategy as it is a one-off course of injections rather than a chronic therapy with reoccurring monthly injections and so income.

Most value rests on IFN-K in lupus. The other projects are small (like dermatomyositis) or lingering in preclinical like VEGF-K. Edison’s market forecast for lupus starts with the 80% of patients with IS; the IS figure is not definitive. This is derived from Exhibit 8, so 137,000 in the US and 146,000 in Europe. Exhibit 11 shows our US forecast and Exhibit 12 the European forecast. RoW is based on 20% of the European figure: US epidemiology is not typical. The fall in 2033 in the US and 2031 in Europe is due to the loss of biological exclusivity and assumed generic competition.

US pricing assumed by Edison is similar to Benlysta at €38,500 or €35,000. INF-K offers the advantage, in theory, of lasting immunity, so one 24-week course could last three to four years at least. The European price is assumed to be €20,000 as European state-backed healthcare is very price sensitive. Good pharmacoeconomic data will be required to justify the pricing in Europe. The sales decline as patients are only treated once and the incidence rate is very low. Market penetration rates of 15-25% three to five years after launch therefore erode the market. This gives the sales curve in Exhibit 13.

As IFN-K will be Phase III funded and sold by a partner, a 15% royalty is assumed. There is a further assumption of a €25m deal fee in 2018 and regulatory milestones of €100m, both probability adjusted, at 50% (Phase II deal) and 30% (regulatory milestone), respectively. The Korean fees potentially due (€4m) are 50% probability adjusted for Phase II outcome alone. The IFN-K probability of reaching the market is 30%, an upper Phase II level reflecting the IS strong positioning. This was formerly 20% and has been increased as the Phase II has started, there is validation from the Korean deal and we have evaluated data published by AstraZeneca on IS.

Neovacs intends to develop more Kinoid products so will incur ongoing costs. These are assumed by Edison to be no more than 50% of revenues. The model runs from 2021 launch to 2032 when US 12-year biological exclusivity ends; European protection ends in 2030. The French tax rate is set at 24% on the assumption that Neovacs will offset profits against new projects. This gives an indicative value of €1.29/share before dilution by warrants and options or €1.15/share fully diluted, Exhibit 14. This is assessed before further funding needs, which are expected to be at least €8m in 2017, with further funding needed thereafter depending on deal values and timings. The funding is additional to the current cash, which is being invested in products.

The previous value of Neovacs was €39.9m with 23.4m shares in issue at that time: €1.70 per share. The model and lupus market forecasts have been fully reassessed and restructured onto a partnering basis with the latest US epidemiology. The new, higher value of €54.9m has a lower per share value as there are now 42.59m shares in issue as of end June 2016.

Neovacs benefits also from the Neostell JV. This is assumed to supply the partner(s) with IFN-K with a profit margin. This is shared 70:30 between Neovacs and Stellar. A nominal value set at 10% of the IFN-K NPV has been included. Initial costs of these developments are included in expenses.

There are 2.18m remaining share options with various expiry dates up to 2025; this includes founder options that could convert to 20,500 shares. In addition, in 2015 Neovacs issued 7.5m warrants linked to the 2015 US investment, which can convert to 2.83m shares by 1 July 2020. This gives 5.01m potential new shares if all are exercised.

In FY15, Neovacs had a gain of €4.3m as governmental loans in relation to the previous TNF-kinoid project were deemed non-repayable as the project failed. Operational costs in FY16 are expected to be €14m, as against €12.5m in 2015, due to Phase IIb costs. The FY15 cash use was €7.8m before funding. Year-end cash was €6.1m but the €1m upfront deal fee from CKD was due in Q116 with a tax credit of €1.8m also due. A June 2016 rights issue raised €8m gross at €0.85/share issuing 9.5m new shares. The net cash from this issue is assumed by Edison to be €7.6m. This gives Neovacs sufficient cash for at least 2016. Additional funding of €8m in 2017 is assumed by Edison to be needed. The next Korean milestone could be paid in H217 if the Phase IIb is successful; management expects about €1.75m. We estimate a tax credit of €3.1m (30% of R&D) for 2016, paid in 2017. Further funding needs will vary depending on the completion and value of partnering deals in 2018.