C4X Discovery — Update 4 January 2016

C4XD is a Manchester-based leader in NMR-based drug discovery and design. Established as Conformetrix as a spin-out from the University of Manchester in January 2008 by the founders, Dr Andrew Almond and Dr Charles Blundell, it was renamed C4X Discovery in 2013. The company employs 21 people. To date the company has raised £14.5m, including £11m gross proceeds from its IPO on AIM in October 2014 (at 100p/share). The company’s proprietary NMR-based technology can be used to solve the 3-D conformations of biomolecules in solution. C4XD believes that this will enable data-driven rational design of superior drug candidates, on a significantly faster timescale than conventional techniques, as observed in the selection of a lead candidate from the Orexin programme. C4XD aims to become a highly efficient and productive discovery R&D engine, targeting c. 15-20 new development programmes within two years. Dr Clive Dix became executive chairman in November 2015, with a proven track record in licensing and M&A activity.

Assigning a fundamental valuation to C4XD necessitates consideration of the inherent value of the technology platform, pipeline candidates and potential future partnership deals. However, with a unique platform and multiple early-stage candidates, this is not appropriate at this stage. The experience with the candidate selection for the Orexin programme suggests the platform’s potential to significantly shorten the drug discovery stage, although this will require repetition against more than one target. Whether the technology will also lead to the selection of safer and more efficacious drugs for use in humans has yet to be proven. However, an illustrative valuation of the lead Orexin programme gives an rNPV of £21m, suggesting that this one programme alone largely supports the current share price, leaving significant room for upside when the earlier-stage programmes and the inherent value of the technology platform are considered.

C4XD is subject to the usual risks associated with drug development, including clinical development delays or failures, IP protection, regulatory risks, competitor successes, partnering setbacks, and financing and commercial risks. The biggest near-term sensitivity is the successful transition of the Orexin programme from preclinical to clinical development, and its subsequent partnering; this will provide validation of the drug discovery platform, influencing C4XD’s negotiating position with future partners for this programme and others in development. Likewise, repetition of the time and cost savings for lead candidate selection in the other programmes will confirm the findings with the Orexin programme, validating the platform’s premise that it can substantially accelerate the drug discovery stage of drug development. C4XD’s current development pipeline is focused on targets that have previously been validated clinically, therefore reducing the risk profile to a certain extent.

C4XD raised gross proceeds of £11m when it listed In AIM in October 2014. We believe net cash of £7.5m at end-July 2015 should be sufficient to fund operations through to end-FY16. We forecast £3m of illustrative financing included nominally as long-term debt on the balance sheet in 2017 (as per our policy). We forecast that collaborations will continue to provide stable, but modest revenues in FY16e and FY17e. Costs will increase as the Orexin programme enters early-stage clinical trials, earlier-stage programmes progress into preclinical and clinical development and discovery work ramps up. We forecast R&D spend of £3.8m in FY16e and £4.8m in FY17e (FY15: £3.2m). For G&A we forecast spend of £0.95m in FY16e and £1m in FY17e (FY15: £0.9m). Consequently, we forecast net loss will increase to £4.3m in 2016e (from £3.1m in 2015).

Management believes that its proprietary nuclear magnetic resonance (NMR)-based technology enables improved and accelerated drug discovery based on conformational design of solution structures. This potentially disruptive technology could transform small molecule drug discovery.

Management believes that its technology enables novel small molecule drug candidates to be generated on a significantly accelerated timeline. The company estimates that selection of the lead candidate for the Orexin programme was achieved in less than 50% of the time taken for traditional drug discovery techniques, with estimated overall cost-savings of up to 90% (c $10m). This data-based empirical approach can be used in combination with existing techniques such as x-ray crystallography, and also for targets where the use of x-ray crystallography is limited (eg certain receptor types). The conformational information it provides should enable the selection and rational design of candidates with superior properties compared with current ‘best-in-class’ alternatives. For example, developing a lead candidate for the Orexin programme that is over 10x more selective for the OX1 receptor than other described OX1-antagonists, could equate to a more efficacious compound with reduced side-effects. It stands to reason that this improved candidate selection could increase the chance of clinical success, resulting in further time and cost savings.

The fact that C4XD’s data so far indicate that the technology is able to significantly accelerate the drug discovery phase, at a substantial cost saving, is likely to be of considerable appeal to the global pharmaceutical industry as it battles to improve R&D productivity. If the approach is also shown to lead to the selection of safer and better novel leads, with a consequent reduction in R&D attrition rates, the appeal would grow further. C4XD is already collaborating with a number of pharmaceutical companies, including AstraZeneca.

C4XD’s investment in its platform has paid off, and it now has a database of 3D bioactive structures that can be used to increase the efficiency with which it maps new drug conformations. Five drug development programmes are in the early stages of development (Exhibit 1). The lead programme is targeting the Orexin system for the treatment of stress-related addictive disorders (initially binge-eating disorder and smoking cessation). It is now in formal preclinical development, including safety and toxicology studies, with plans to file a clinical trial application by end-2016. C4XD plans to add to this pipeline, aiming for at least 15-20 development programmes in a few years’ time; these will either be out-licensed after early clinical studies or C4XD will continue development in house.

Structure-based drug design (directed by the 3D structure of the drug target) is one tool used to improve the rate and success of drug development. Nuclear Magnetic Resonance (NMR) has been used to understand molecular structures for many years; data from the experiments can be used to provide 2D and some 3D structural information to aid rational drug design. C4XD has developed fundamentally new methods that exploit NMR data to determine not only the 3D shape of a drug molecule, but also the dynamic range of shapes that the molecule adopts in solution.

The shape, or conformation, of a drug molecule is a determinant of if, and how, it binds to a target, and therefore whether it will have the desired biological effect. Where the gold-standard X-ray crystallography enables the bioactive (or bound) 3D structures of molecules to be determined, C4XD’s platform can be used to provide additional information regarding the conformation change of the drug upon binding – a determinant of the binding affinity. Further, it is able to do this under the same physiological conditions as the drug molecules encounter in the human body (such as temperature and pH). This information can be used to design molecules with better affinity.1 This is important as drug molecules that are not highly optimised to bind to a particular target often interact with many different receptors or enzymes, which can give unpredictable side-effects and a narrow therapeutic window. Increasing the specificity, and so the affinity of a ligand for a particular target generally minimises off-target effects, so reducing side-effects and meaning that any toxicities only appear at doses well above the therapeutic level. Thus, an understanding of both the free and bound shapes of the drug molecule is vital to optimising the rational drug design.

C4XD believes it provides the only empirical method that can accurately measure the free dynamic 3D-structures of drug molecules, offering a unique approach. As far as we are aware, C4XD is the only company able to generate this level of experimentally-derived conformational data, and can do so in a matter of days (X-ray crystallography takes 6-18 months to generate). This information can then be used in conjunction with existing rational design technologies, and importantly, when X-ray crystallography for the drug target is not routinely available (eg G-protein coupled receptors, such as the OX1/2 receptors, and ion channels).

Drug development faces enormous challenges. It is time consuming, with developing a new medicine now taking on average 11 years, and high attrition rates mean that the probability of a Phase I product reaching the market is just 12% (and this does not even include the failure rate of the discovery and preclinical stages).2 As a consequence, it is extremely expensive: when costs of failed programmes are accounted for, the cost to get a drug to market has been estimated to be as high as $2.6bn.2 Thus, there is an obvious need, and desire, to improve R&D productivity.

R&D productivity can be considered to consist of two dimensions: efficiency and effectiveness.3 Using its technology in the lead Orexin programme, C4XD was able to improve R&D efficiency by significantly accelerating the ‘hit to lead’ and ‘lead optimisation’ stages of drug discovery, at a substantial cost saving (90% in the addiction programme). If this can be externally validated and repeated, it is likely to considerably interest the global pharmaceutical industry. However, if C4XD’s data-driven rational design leads to the selection of more selective, and therefore safer and better, novel leads, this could also improve R&D effectiveness by reducing attrition in both discovery and development stages, and the eventual approval of drugs with superior properties. Such a technology could potentially revolutionise drug discovery.

C4XD’s approach mainly involves pursuing existing, clinically validated, therapeutic targets. The technology can be used to solve the solution structures of the current best-in-class drug molecules for these targets, or other development-stage drug candidates, or indeed conduct more novel work on naturally occurring ‘endogenous’ ligands (such as peptide hormones), ie there are multiple potential starting points to develop safer and more effective compounds. This information is then used to design multiple molecules with better affinity and selectivity, from which a lead candidate, or indeed multiple candidates, can be selected. This approach results in the time savings, but is also de-risked compared to conventional drug discovery where the target may initially be unknown, and/or the clinical sequelae of acting on it are not apparent until clinical studies. The speed at which C4XD is able to deliver this capability means it has potential to become a sought-after R&D engine, producing multiple candidates against various targets, from which partners can then select for further development, all at a fraction of conventional costs. C4XD believes that selecting multiple candidates per target should enable the development of a portfolio of potentially complimentary therapies, as well as back-up options should the lead candidate falter.

C4XD plans to continue to build on the momentum of the last 12 months by adding to its pipeline of drug discovery programmes. Key to this is the recently announced collaboration with the University of Oxford’s Structural Genomics Consortium (SGC-Oxford). SGC-Oxford is part of the Nuffield Department of Clinical Medicine and consists of c 100 scientists who collaborate widely with major pharmaceutical companies and the worldwide academic network, combining world-class expertise in therapeutic target validation, protein expression, assay development and protein structural information. As part of the collaboration, C4XD will be granted access to structural, biological and therapeutic information that SGC-Oxford holds in relation to various therapeutic targets and related assays, as well as initial ‘hit’ molecules that SGC-Oxford has identified against these targets. C4XD’s technology will be used in combination with SGC-Oxford’s expertise in conventional techniques to identify new and improved hit molecules against the SGC-Oxford targets. However, C4XD will retain ownership of any new compounds it independently identifies as part of the collaboration, thus forming the starting point for C4XD’s future drug discovery projects. C4XD is aiming to have 15-20 projects, with multiple candidates per project, in development in the next couple of years; some of which it will develop itself and others it will partner.

Over the last three years, C4XD has forged fee-for-service collaborations with AstraZeneca, Takeda and Evotec; while this provides important external validation of the platform and a modest revenue stream, the longer-term goal is to seek more strategic collaborations with big pharma/biotech companies. Securing the right blend of partnerships to advance/fund certain programmes, while retaining and developing other projects in house, will be key toC4XD’s long-term success.

C4XD’s technology is also applicable throughout the life cycle of drugs. Polymorphism, the ability of a solid material to exist in more than one form, is important in the pharmaceutical industry for two reasons. Firstly, for IP reasons, it is important to identify all potential polymorphs of a drug. Secondly, although chemically identical, the different polymorphs can have significant differences in their physical properties, which may affect their pharmacological profile. In turn, this can affect development, manufacturing, product quality and product stability. C4XD’s technology can be used to predict and control the different polymorphs small molecules are able to adopt, information that is essential throughout the product life cycle and to ensure sufficient IP protection.4

C4XD is developing novel small molecules to act on drug targets in areas of unmet clinical need. It believes its technology will enable the selection of candidates that have substantially safer and better properties compared with current ‘best-in-class’ alternatives. The most advanced programme is in stress-related addiction disorders; earlier programmes are directed at validated targets in inflammation, diabetes and autoimmune diseases. C4XD intends to advance the pipeline into early clinical trials, after which it will seek partnerships for development and commercialisation.

C4XD is targeting the Orexin system for the treatment of stress-related addictive disorders. Orexins are neuropeptides produced in the lateral hypothalamus that regulate the sleep-wake cycle and appetite via the OX2 receptor. Belsomra (Merck & Co) is a dual OX1/OX2 antagonist approved for the treatment of insomnia. Via their action on the OX1 receptor, orexins also play a critical role in addiction and reward-related behaviours, making OX1 receptor a key target for the treatment of addictive disorders.5 Indeed, in a large number of preclinical studies, OX1 antagonists have effectively blocked addiction-related behaviours.6 However, a high degree of selectivity of the OX1 receptor is required. With the best reported literature selectivity for OX1 of 50-60-fold, off-target activity at the OX2 receptor may result in sedative side-effects, which may limit use.2,7

C4XD has used its technology to solve the structures of existing OX1 and OX2 inhibitors to identify the active groups required for OX1-selective inhibition. Once identified, multiple suitable cores were selected that work with the active groups, creating novel molecules with the desired conformation and selectivity. This has added IP benefits, whereby the ability to change cores enables C4XD to move outside of any existing patents.

C4XD has developed a number of distinct novel OX1 antagonists, which are over 1000-fold more selective for OX1 over OX2. Exhibit 2 illustrates how the C4XD candidate sits well within the target area (shaded in light green) of high OX1 potency and low OX2 potency, compared with three described OX1 antagonists. A lead candidate has been selected for clinical development. C4XD was able to achieve this in less than 50% of the time taken for the traditional drug discovery techniques (c three to five years); with estimated overall cost-savings of up to 90% (industry standard $10-14m).3

The lead candidate is currently in formal preclinical studies, with plans to file the clinical trial application by the end of 2016, and begin clinical development by mid-2017. C4XD is also working with Evotec to identify and develop follow-up compounds, aiming to create a complementary portfolio of agents to treat addiction. This will also increase the attractiveness of the addiction portfolio to any pharmaceutical companies that may wish to license or acquire it in the future. Evotec is providing resource support on a fee-for-service basis, with C4XD retaining complete ownership of any identified orexin compounds.

Addiction – a substantial market

C4XD is targeting the smoking cessation and binge eating disorder (BED) markets initially, believing that the increased selectivity and potency of its candidate will prove more effective and tolerable than existing therapies. Other potential indications may include drug and alcohol addiction.

Despite being largely genericised and losing market share to e-cigarettes, global sales of nicotine-replacement therapy (NRT) amounted to $2.4bn in 2013 (Bloomberg). Targeting OX1 would aim to modify addictive behaviour, and could potentially prove a more efficacious aid to cessation than NRT, providing an opportunity to make inroads into this market.

BED is characterised by repeated episodes of binge eating in the absence of compensatory behaviours to avoid weight gain; it is estimated to affect 1.5-2% of the general population.8 Vyvanse (Shire) became the first FDA-approved treatment for BED in January 2015. However, it carries a black-box warning due to cardiac and psychiatric risks; additionally it is a controlled substance due to the risk of addiction and abuse. OX1 receptor mechanisms have been shown to play a major role in the control of BED episodes7 and could potentially avoid these restrictive side-effects.

As far as we are aware, there are two other selective OX1 receptor antagonists in development, both preclinical. Sosei Group’s small molecule is also targeting BED and nicotine addiction; Eolas Therapeutics’ programme (partnered with AstraZeneca) is targeting smoking cessation.

C4XD has identified protein targets in the treatment of inflammation and diabetes, and is aiming to create best-in-class drugs in these blockbuster indications, where efficacy, safety and ease of use will be critical to obtaining significant market share. These are outlined in Exhibit 3.

C4XD is subject to the usual risks associated with drug development, including clinical development delays or failures, IP protection, regulatory risks, competitor successes, partnering setbacks, and financing and commercial risks. The biggest near-term sensitivity is the successful transition of the Orexin programme from preclinical to clinical development, and its subsequent partnering. In addition, as a novel technology, the platform requires validation. Although C4XD’s reported data to date has demonstrated the lead compound’s increased selectivity for the OX1 receptor, whether this will translate to a safe and efficacious drug in humans has yet to be established. C4XD’s current development pipeline is focused on targets that have previously been validated clinically, therefore reducing the risk profile to a certain extent.

Progress of the Orexin programme will provide validation of the drug discovery platform; influencing C4XD’s negotiating position with future partners for this programme and others in development. Likewise, repetition of the time- and cost-savings for lead candidate selection in the other discovery programmes will confirm the findings with the Orexin programme, validating the platform’s premise that it can substantially accelerate the drug discovery stage of drug development.

C4XD is in the process of pursuing patents for its method to determine the 3D structures of dynamic molecules; a patent has been granted in Japan with others filed in the US, Canada and Europe. C4XD must also obtain IP protection for its development candidates. Further competitor barriers arise from the significant degree of know-how required for the set-up of the NMR experiments and the use of the software.

Finally, C4XD’s stock is tightly held; poor liquidity can result in increased volatility.

Assigning a fundamental valuation to C4XD necessitates consideration of the inherent value of the technology platform, pipeline candidates and potential future partnership deals. However, with a unique platform and multiple early-stage candidates, this is not appropriate at this stage. The experience with the candidate selection for the Orexin programme suggests the platform’s potential to significantly shorten the drug discovery stage, although this will require repetition against more than one target. Whether the technology will also lead to the selection of safer and more efficacious drugs for use in humans has yet to be proven. However, an illustrative valuation of the lead Orexin programme gives an rNPV of £21m, suggesting that this one programme alone largely supports the current share price, leaving significant room for upside when the earlier-stage programmes and the inherent value of the technology platform are considered.

For illustrative purposes therefore, we applied a risk-adjusted NPV analysis of the Orexin programme. We arrive at an indicative rNPV of £21m based on a number of assumptions, including successful completion of Phase I trials in 2018, peak sales ($1.4bn), launch dates (2023), royalty rates (10%) and partnership deal metrics. Our indicative peak sales are based on the following assumptions:

Exhibit 4 illustrates the Orexin programme’s indicative rNPV. This programme alone largely supports the current share price (107.5p) and enterprise value (£25.5m), suggesting significant room for upside when one considers the value of the earlier-stage pipeline and the technology platform.

We note that in comparison to other drug discovery companies, C4XD appears comparatively undervalued, for instance Verseon (VSN), with a current market cap of £309m. Verseon’s small molecule drug discovery platform uses a computational modelling engine to design drug-like compounds. According to its website the platform allows for the systematic design, synthesis, and physical testing of multiple, high-quality drug candidates per programme, so ‘reducing the high failure rate that plagues the rest of the pharmaceutical industry’. Like C4XD, its pipeline has yet to enter clinical studies and therefore achieve clinical validation; it has with three preclinical programmes in development for anticoagulation, diabetic macular oedema, and solid tumours.

C4XD reported cash of £7.5m at end-July 2015. Our model suggests that this should be sufficient to fund operations through to the end of FY16. Revenues from collaborations amounted to £0.3m in 2015 (vs £0.6m in 2014). We forecast that collaborations will continue to provide stable, but modest revenues in FY16e (£0.5m) and FY17e (£0.7m). Although our valuation model includes risk-adjusted milestones from a partner for the Orexin programme, our financial forecasts do not include any such income. Hence our forecasts include £3m of illustrative financing included as long-term debt on the balance sheet in 2017.

Increased activity required to progress the discovery programme, along with the necessary rise in headcount following the IPO, resulted in R&D spend of £3.2m in 2015 (vs £1.2m in 2014). In keeping with this, we forecast £3.8m R&D spend in 2016 and £4.8m 2017, as the Orexin programme enters clinical development, and the earlier stage pipeline progresses. Beyond this, we assume that future R&D spend for the Orexin programme will be borne by a partner. G&A spend also increased (£0.9m in 2015 vs £0.6m in 2014), reflecting costs incurred following the IPO; we assume a gradual increase in the coming years. We forecast net loss of £4.3m in FY16 compared with £3.1m reported in 2015.