e-therapeutics — More to come in RNAi therapy discovery

HepNetTM, e-therapeutics’ proprietary computational biology platform, was built to expedite the identification of high-quality hepatocyte-expressed therapeutic targets. HepNet leverages large and complex datasets, proprietary computational capabilities and a hepatocyte-centric knowledgebase to create and analyse biological network models, providing a novel and mechanistic approach to target identification. Supported by HepNet, e-therapeutics has identified a number of therapeutic targets, giving rise to active preclinical programmes focused on cardiometabolic and metabolic disease, haemophilia and two further undisclosed indications. The ability to expedite effective target discovery is a critical consideration in the current highly inflationary environment, with limited funding for innovation and discovery.

e-therapeutics takes a completely computational approach to the design of siRNA constructs, using AI to predict siRNA efficacy. The company reports that its data demonstrate that the trained siRNA efficacy prediction model shows high accuracy and enables the identification of lead siRNA sequences at the in silico stage, significantly reducing the number of sequences that need to be screened in vitro. Ultimately, this may enable in vitro screening to be bypassed entirely, with predictions being taken straight to in vivo, dramatically reducing preclinical development timelines and costs. According to management, prior to the use of AI for siRNA efficacy prediction, the company would have identified and screened c 400 potential siRNA sequences for a single target, taking six months and costing $500k, as discussed in our initiation note. With the new AI approach, management aims to reduce timelines for identifying a high-quality clinical candidate to one month, and reduce the associated costs to c $50k, by screening fewer than 10 sequences; however, this is yet to be fully validated (Exhibit 1).

e-therapeutics believes it may offer differentiation within this market with its novel targets identified by HepNet. In the company’s H1 FY24 results, preclinical assets ETX-291 and ETX-148 were introduced. These have been designed to silence genes implicated in cardiometabolic disease and haemophilia, respectively. Management has reported that PoC data have been generated for these assets in H1 FY24 and the company indicated that preclinical studies are on track to be completed by the end of the year:

Management plans to present further details on these preclinical data packages in the near future.

e-therapeutics has reaffirmed its commitment to continue integrating large language models and generative AI into its processes in H1 FY24, which should further enhance its research efforts. Management attests that, as a smaller company, it has the competitive edge over its larger peers since it can better utilise generative AI at the centre of its operations. A key component of e-therapeutics’ AI strategy is the use of specialist ‘generative AI agents’, a proprietary technology that is being used to advance the capabilities of its HepNet platform (Exhibit 2). The company is training these specialist AI agents on specialised data, such as hepatocyte-specific data, siRNA sequences/constructs and scientific papers, enabling the AI agents to understand, reason and infer better within specialised domains. Additionally, the generative AI agents have live access to the internet and the ability to learn from feedback and to self-correct. With these capabilities to hand, and with domain experience in hepatocyte biology, management believes that the company is well positioned to realise the ambition to be ‘computing the future of medicine’.

As a case study, e-therapeutics has illustrated how its proprietary AI agents could be used to help formulate a patent strategy based on the analysis of unstructured patent data. For example, this might include extracting information from existing patents to find opportunities for freedom to operate (FTO) and new IP, as well as autogenerating new patent applications. According to management, there have been just under half a million RNAi-related patents filed since 2001. As each patent may contain hundreds of pages and include text, charts and chemical structures, this project was intended to compare how e-therapeutics’ AI agent might perform in comparison to a more traditional ‘manual’ approach across a number of tasks (Exhibit 3). Given the large number of patent documents, it was determined that some of the more advanced tasks, such as understanding the specialist syntax of patents, extracting and structuring data from tables, images and text, as well as inferring missing information are not feasible manually. However, management claims this is now possible at scale with the AI agent approach, though this is yet to be validated.

Post the reporting period, in November 2023, e-therapeutics announced a strategic collaboration with Arcturis Data. Arcturis is a UK-based company focused on using real-world data to support the discovery of new medicines. Its Real-World Evidence (RWE) platform uses high-quality, real-world data, as well as analytical capabilities derived from its unique access to patient-derived data. etherapeutics plans to utilise this RWE platform to enhance the capabilities of HepNet, with a focus on developing new therapeutic options for metabolic dysfunction and associated fatty liver disease, including non-alcoholic fatty liver disease. As part of this strategic collaboration, e-therapeutics will assess outputs from Arcturis’s RWE platform using HepNet and will have exclusive rights to nominate novel targets identified and to develop siRNA therapies using the GalOmic platform. In our view, the announcement marks a positive step forward for e-therapeutics, as it complements the company’s ongoing research efforts focused on metabolic diseases. We believe the external interest is encouraging for the company’s unique approach to accelerating and de-risking drug discovery.

During the first half of FY24, total revenue was recorded at £0.2m, which mainly comprised milestone payments related to the small molecule discovery agreement with iTeos Therapeutics, versus H1 FY23 revenue of £0.3m, which included the remaining milestone payments from the Galapagos collaboration. We note that the company does not generate a recurring revenue stream as it is still in the preclinical stages of drug development. The operating loss for the period increased to £7.0m, from £4.6m in H1 FY23, due to higher R&D expenses totalling £5.3m (H1 FY23: £3.1m). R&D expenses formed the majority (c 76%) of the total operating expenses. The increased R&D expenditure was attributed to higher outsourced contract research organisation costs associated with the development of the company’s preclinical assets, along with additional costs with respect to its RNAi platforms and related patent applications. Management expects a further increase in R&D costs and the cash burn rate in H2 FY24 as the company plans to continue the development of its preclinical assets and intends to expand its development pipeline across various therapeutic areas.

During H1 FY24, e-therapeutics reported an operating cash outflow of £6.9m (up from £4.7m in H1 FY23) and management anticipates this to increase further, in line with the expected higher R&D spend during the second half of FY24. The company ended H1 FY24 with a cash balance of £24.8m. However, management expects operating cash burn to increase further as the company ramps up R&D activities in line with its strategic priorities; this may affect the current runway.