Arecor: unlocking diabetes franchise value

Diabetes, alongside Speciality Hospital Products, is one of the two focus areas for Arecor’s in-house development pipeline. Recent news flow prompts us to revisit our valuation assumptions, notably the positive Phase I data from AT247 (ultra-rapid insulin) and more recently AT278 (ultra-rapid ultra-concentrated insulin). These encouraging data suggest suitably differentiated profiles that could capture larger market shares and open new market opportunities, thus providing greater value-creation opportunities for both products. The data so far provide evidence that Arecor’s formulation expertise can modulate the absorption profile of insulins selectively and consistently. This, in our view, should help not only in attracting pharma or device partners for the diabetes assets, but also other potential collaborators interested in the capabilities of the Arestat platforms. The signing of a fifth formulation development contract during 2021 attests to industry interest. Our upgraded Arecor valuation is £140.9m or 506p per share (£103.7m, or 374p/share previously).

Arecor employs its proprietary Arestat formulation platforms to develop novel formulations of existing products. These are specifically created to offer improved attributes ranging from a better shelf life due to greater stability, to increased patient convenience because of easier administration, and superior therapeutic profiles through tailored absorption characteristics. The formulation expertise was originally offered on a fee-for-service model but, over the past decade, there has been a notable shift towards creating, and retaining, more of the value added. Now technology partnerships and licensing deals involve more attractive success-based economics, including clinical and commercial milestones and net sales royalties or equivalent. Management is also developing a portfolio of in-house development programmes focussed on diabetes and specialty hospital products.

We detailed the various elements of Arecor’s investment case in our September 2021 Initiation, with emphasis on the diverse strands of the platforms’ development and the prospects for their successful commercialisation. Recent news flow, notably on Arecor’s diabetes programmes, has led us to revisit our assumptions and to update our valuation model. Despite still applying conservative assumptions throughout, the combination of better than expected AT278 Phase I data, and the subsequent potential commercial implications, coupled with the dosing of the first patient in the AT247 US three-day insulin pump study, sees our previous valuation of £103.7m, or 374p per share, increase to £140.9m, or 506p per share. As we highlighted in previous notes, continued clinical progress, greater visibility on partnered products (indications, market positioning, etc), and further licensing deals, would result in material uplifts in our valuation and this remains the case.

Arecor currently has four partnered products (two that emerged from technology partnerships, two from out-licencing internally developed formulations) that will generate development and commercial milestones, plus royalties or equivalent on sales (Exhibit 1). These are progressing well and should provide an attractive blend of medium- and longer-term revenue streams. However, it is the clinical results with AT278 (an ultra-concentrated ultra-rapid insulin) and AT247 (an ultra-rapid insulin) that demonstrate how a valuable, yet still low-risk, diabetes franchise is being fashioned. This clinical data provides supporting evidence that Arecor’s formulation expertise can modulate the absorption profile of insulins selectively and consistently.

In addition, Arecor is using its formulation skills to develop a portfolio of Specialty Hospital Products, both as partnered and in-house programmes. H121 results confirmed these are also progressing well. We continue to expect first launches of Specialty Hospital products by partners from 2024 onwards, with the first launches of other licenced partnered programmes from 2023.

Exhibit 1: A broad portfolio of de-risked and innovative assets

Source: Arecor Therapeutics

The diabetes franchise: addressing clear clinical needs

Arecor has three active diabetes programmes: two are the clinical stage specific rapid-acting insulins (AT247 and AT278), and the third is an insulin co-formulation (AT299) that is completing preclinical stages. These proprietary programmes will be taken to proof-of-concept Phase II studies before future development decisions are made. Typically, this is a value-inflection point, where the terms of an out-licencing agreement (upfront and milestone payments, plus royalties on net sales) are optimised against the costs of the more expensive Phase III clinical trials. However, we would argue the costs (and risks) of progressing such insulin formulations further, despite their novelty, is less than for conventional new drugs. We believe the eventual decisions will be driven by the quality of data from the clinical studies. Exhibit 2 provides an overview of current status and next milestones for the two lead assets.

The diabetes market is attractive not simply because of its growth prospects, due to well-documented shifts in demographics and lifestyles, but the clinical trends towards better monitoring and tighter glucose control are creating a demand for insulins that are faster acting and have better physiological characteristics. Possibly more importantly in our view is the rise of innovative delivery devices, initially CSII (continuous subcutaneous insulin infusion) pumps allied with digital technologies, where a fast and predictable onset of action is essential.

Exhibit 2: Arecor diabetes franchise development status and upcoming milestones

Source: Arecor Therapeutics  Note: As of 20/1/22, the AT247 US Phase I pump study is underway; the next milestone is data in H222.

Diabetes incidence: rises driven by Type II diabetics

The latest International Diabetes Federation (IDF) diabetes atlas fact sheets indicate that there are currently 537m people with diabetes globally, and this is expected to grow by 19.7% to 643m in 2030 and by 46.0% to 784m in 2045. In North America (including the Caribbean) there are 51m diabetics in 2021, with a forecast 11.8% increase to 57m in 2030 and a 23.5% rise to 63m in 2045. For Europe, the current figure is 61m, with a more modest 9.8% increase to 67m in 2030 and 13.1% rise to 69m in 2045. The real rises in the predicted growth are driven by increases in Africa (up 134% from 24m in 2021 to 55m in 2045), Middle East and North Africa (MENA) (up 87% from 73m to 136m) and South-East Asia (up 68% from 90m to 152m).

Type I diabetes affects c 9.5% of the global population, with 15 such diabetics per 100,000 people and growing by 2% to 5% per annum. It is caused by an autoimmune process that results in the body’s immune system attacking the insulin producing beta-cells of the pancreas so that very little or no insulin is produced. As a result, Type I diabetics need daily insulin injections to keep their blood glucose level within an appropriate range and a lack of insulin would result in death. Although it was originally known as juvenile diabetes, over half of patients are between the ages of 19 and 40 years at diagnosis. Type I diabetes currently accounts for around 10% of all diagnosed diabetics.

Type II diabetes accounts for around 90% of diagnosed diabetes globally, although the accuracy of incidence (the rate of new cases, usually annually) and prevalence (the number of cases per defined population) statistics is limited due to the difficulty in, and hence differing rates of, accurate diagnoses across geographies. Here the pancreas continues to produce insulin, but the body’s cells begin to respond less well (insulin resistance) and, over time, fail to maintain blood glucose levels in the normal range.

Most Type II diabetics can control their disease adequately through a combination of improved diet, more exercise, and the appropriate use of oral hypoglycaemics. However, as beta cell production wanes, an increasing number of patients (currently 7% to 8%) need to add in injected insulin. Again, estimates vary but around three to four times as many Type II diabetics are using insulin than there are Type I insulin dependent diabetics. It is estimated that better diagnosis and greater access would see this rise to over 15% of Type II diabetics. It is also important to note is that insulin-using Type II diabetics tend to require larger doses of insulin than the equivalent Type I diabetic.

Three players dominate a large and growing market

The overall drug therapy diabetic market is worth around $48bn, with oral hypoglycaemics the largest segment, accounting for c 44%; insulins account for some 40%; and GLP-1 (glucagon-like peptide-1) analogues most of the balancing 17%. Arecor is specifically targeting the prandial (mealtime) insulin market, which is valued at c $6.4bn. The overall insulin segment can be viewed as consisting of three elements (see later): human insulins, which are the lowest daily cost (less than $1) and have over 40 producers globally; “modern” insulins (typically analogues), priced at around 4x the daily price of human insulin and with some six producers; and “new-generation” insulins, priced at c 18x the human insulin daily price with two manufacturers. The higher pricing of later generation products is supported by extensive health economic data highlighting improved clinical outcomes and greater patient benefits.

The global insulin market can be considered as two segments: the Western regions (US and EU5) and the Low and Middle Income Countries (including China, the largest market by volume). Commercially the Western regions are currently the most important, for instance the US represents only c 15% of insulin-dependent patients but accounts for nearly half of insulin-related revenues. Three companies – Novo Nordisk, Eli Lilly, and Sanofi – account for around 90% of the global insulin market. The rest is largely divided among seven or so mainly regional manufacturers, although there are some 30 additional companies with some form of production capability.

With the caveat that estimates do vary, Novo Nordisk has 52% of the global market by volume and 41% by value, Eli Lilly has 23% by both volume and value, and Sanofi has 17% by volume and 32% by value. Interestingly, following the first VBP (Value Based Procurement) tender process in China (November 2021), the aggressive pricing of domestic companies like Gan & Lee Pharmaceuticals and Tonghua Dongbao Pharmaceutical should see them take sizeable share.

Product innovation: ultra-rapid insulins to the fore

The routine treatment of diabetes with insulin replacement is now a century old; the first commercially available insulin, Iletin a short-acting insulin derived from porcine pancreas, was introduced by Eli Lilly in 1923. It was sixty years before the next major breakthrough, when in 1982 the first recombinant human insulins, Humulin R and N, were approved (Genentech/Eli Lilly). The introduction of new insulin analogues represented the next innovative leap, notably the first rapid-acting analogue Humalog insulin lispro (Eli Lilly, approved in 1996), followed by Novolog insulin aspart (Novo Nordisk, in 2002), and Apidra insulin glulisine (Sanofi, in 2004). The more recent focus is on faster, more physiological insulins: the ultra-rapid insulins, such as Fiasp (Novo Nordisk, 2017) and Lyumjev (Eli Lilly, 2020), allowed for better postprandial glucose control. Interestingly, the first insulin lispro biosimilar, Admelog, was introduced by Sanofi in 2017.

The ultra-rapid insulins typically achieve their pharmacokinetic profiles through specific formulations, with Fiasp (faster aspart) containing niacinamide (vitamin B3) for faster absorption and L-arginine (a naturally occurring amino acid) to improve stability, whilst Lyumjev contains treprostinil (a prostacyclin analogue that improves absorption via local vasodilation) and citrate (increases local vascular permeability). These achieve a faster onset of action and a faster time to peak effect, typically being seven to 14 minutes quicker than an equivalent first-generation rapid insulin. Initially these insulins were indicated for patients who were not achieving optimal glycaemic control, particularly in the important postprandial period, where the accelerated kinetics can make a clinical difference; but while this improved profile is beneficial, it is still not ideal. As mentioned, it is the advent of sophisticated automated insulin delivery (AID), also colloquially known as artificial pancreas, that has brought ultra-rapid insulins to the fore.

Changing priorities: technology drives clinical practice

The use of insulin pens is almost universal in Western countries, with many claiming their ease of use and social acceptability has transformed insulin-dependent diabetics’ lives. Certainly, the advantages over a traditional vial and syringe are many, not least the simpler administration procedure, more accurate dosing, and greater patient compliance. However, we would argue the single greatest advance in the treatment of diabetes was the landmark Diabetes Control and Complications Trial (DCCT) in the 1990s. This seminal study demonstrated the importance of intensive insulin therapy to maintain tight glycaemic control. Keeping blood sugar levels as close to normal throughout the day (now known as time in range) helps prevent common, and debilitating, complications such as retinopathy (down by 76%), neuropathy (down by over 60%), nephropathy (effectively halved), and cardiovascular diseases (down by nearly a third).

The DCCT results changed clinical practice and led to Haemoglobin A1c (HbA1c) becoming the gold standard of glycaemic control and the self-monitoring of blood glucose (SMBG) being a cornerstone of care for insulin-dependent diabetics to verify their glucose variability (GV) on daily basis. The first blood glucose meter was invented in 1971 and, although celebrated as a marked improvement over urine glucose monitoring (Clinistix dipsticks), saw little clinical use. Subsequent improvements mean modern meters are highly compact, give results in five seconds and use just 3-5μL of blood. Advances in connectivity mean meters and pens can easily connect with smartphone apps that automatically log blood glucose values, identify trends, and share data with the treating clinician between clinic visits. These apps can also track food intake, exercise levels, and other relevant data.

Although such diagnostic and delivery devices are more than adequate for most insulin-dependent diabetics, a small, but significant, minority remain poorly controlled. Advances in miniaturisation and computing power saw the introduction of viable continuous glucose monitors (CGM). These allow patients, and clinicians, to assess trends, patterns, and time spent in range in real time. The sensor tests glucose every few minutes and sends an alarm if hypo- or hyper-glycaemia is threatened. Such devices have been transformative for some patients. In parallel, similar technological advances saw sophisticated, and reliable, wearable pumps developed. These pumps, known as continuous subcutaneous insulin infusion (CSII) therapy, have evolved rapidly and offer near-normal glucose control in previously uncontrolled diabetics. The advantage of eliminating multiple daily injections was initially particularly attractive for children and adolescents, however the improved short- and longer-term clinical outcomes has become a key uptake driver for many Type I and, increasingly, Type II diabetics.

“Artificial pancreas”: transformational shift in sight

The combination of CGM and CSII became the first iterations of an artificial pancreas, including DIY systems developed by technologically competent patients. These closed-loop insulin delivery systems use real time continuous glucose monitor data to automatically adjust, both up and down, the insulin pump’s basal insulin delivery. The earlier versions were not completely automated, requiring a manual input to manage bolus injections to cover meals. Advances in software, notably the use of artificial intelligence, has led to powerful algorithms that are able to “learn” the individual patient’s metabolic profile and predict insulin needs in most everyday situations. Whilst challenges remain, most commonly ensuring reliable inter-device connectivity, the real-world evidence is increasingly showing that the time in the appropriate glucose range is increasing for all patient age groups, with noticeable reductions in nocturnal hypoglycaemia.

Exhibit 3: The typical configuration of a closed-loop insulin delivery system

Source: Closed-loop insulin delivery. Hovorka, Nat Rev Endocrinol 7(7) 385–395

Advances in the flexibility, responsiveness, and predictive qualities of the algorithms has seen clinically meaningful improvements in time in the glucose target range, even in patients who were previously optimised on CGM and CSII regimens. The next software iterations will see algorithms that employ physiologic and/or activity sensors within a smartphone or incorporated into a dedicated wearable. These monitors can identify the quantity and intensity of any exercise or movement and assess the glycaemic response to it so insulin delivery can adjust appropriately. Examining behaviours will be used to “learn” routines and predict expected events, for instance whether certain meals are eaten at certain times (eg curry nights). Exhibit 3 shows we are nearing the commercial availability of reliable fully automated closed-loop systems.

Exhibit 4: Key milestones towards a truly artificial pancreas

Source: New closed-loop insulin systems. Boughton & Hovorka, Diabetologia 1007-1015 (2021)

Whilst these advances in miniaturising and integrating the various technologies have been impressive, major obstacles remain. For instance, the inherent delays in absorption of subcutaneous injected insulin compared with endogenous insulin production means that postprandial hyperglycaemia remains a challenge for these closed-loop systems. The pharmacokinetics and pharmacodynamics of current rapid-acting insulins are known to be sub-optimal. The new ultra-rapid acting insulins, which have faster onset and offset of action, have the potential to address this issue. Small studies with Fiasp (faster aspart) and Lyumjev (ultra-rapid lispro) have not shown the hoped-for conclusive results. It is for this clear clinical need that Arecor is developing AT278 (ultra-concentrated ultra-rapid insulin, applicable to miniaturised devices) and AT247 (ultra-rapid insulin, applicable to fully closed-loop delivery).

AT278: ultra-concentrated ultra-rapid insulin

AT278 is a novel formulation of insulin aspart with the focus on creating a highly concentrated, U-500 (500 units/ml), fast-acting insulin. Such high concentration insulins are expected to become increasingly in demand, reflecting the rising number of Type II and refractory Type I diabetics requiring higher daily dosing. Around 35% of Type II diabetics are already using over 60 units of insulin daily, with a growing number needing 200 units or more. The appeal is not simply to reduce the burden of daily therapy through fewer injections and lower injection volumes, but also to allow wider access to modern insulin pumps (where their smaller size often results in limited reservoir capacities), to next-generation miniaturised devices, or to longer-term infusion pumps where volumes are key. Importantly, as mentioned earlier, algorithm-driven devices require a really rapid acting insulin to optimise glycaemic control, with a fast response and short duration of action.

Type II diabetes is characterised by increasing insulin resistance and relative insulin insufficiency. A growing number of patients have transitioned from a variety of anti-hyperglycaemic agents to insulin treatment. Despite introduction of novel drug classes, such as GLP-1s, this trend appears set to continue. Over time, with progressive β-cell failure and rising resistance, the demand for higher daily insulin doses increases. The typical concentration of insulin products is U-100 (100 units/ml), with only a limited number of higher concentration products, such as Humulin R U-500 (human insulin, Eli Lilly), with the highest concentration rapid products available being U-200 (200 units/ml), such as Humalog U-200 (lispro, Eli Lilly). This necessitates a trade-off for patients of a rapid-acting insulin at typical concentrations (eg NovoRapid U-100) or a slower acting concentrated insulin with a more basal-type profile (eg Humulin R U-500).

AT278 is in early-stage clinical trials: top line results from the Phase I study in Type I diabetics were reported in September 2021. The trial met all primary and secondary endpoints, demonstrating a superior pharmacokinetic (PK) and pharmacodynamic (PD) profile to a comparable dose of lower concentration of NovoRapid (NovoNordisk’s gold standard rapid acting insulin). The trial evaluated 38 adults with Type I diabetes in an euglycemic clamp setting aiming to establish PK/PD equivalence between a subcutaneous dose of AT278 0.3 U/Kg (500 U/mL) that was five-fold more concentrated than the comparator 0.3 U/Kg NovoRapid (100 U/mL). AT278 matched or exceeded key measures such as glucose lowering, onset of action, and absorption profile, and there were no safety signals.

The outcomes are impressive and better than we expected. The top-line results showed AT278 has, despite the five-fold greater concentration, an absorption profile that does not simply match the rapid insulin criteria, but that PK/PD data justifies AT278 being labelled an ultra-rapid insulin. Full data are expected to be presented at a future diabetes conference, probably in H122. Next steps should see the initiation of the next clinical study in 2022. Assuming smooth progress through the clinical programmes and approval processes, first launch could happen as early as 2025 (we assume 2026 in our modelling).

The absorption profile demonstrated in this Phase I trial, if confirmed in larger studies, positions AT278 as a unique, disruptive, and highly desirable ultra-rapid acting ultra-concentrated insulin. To be clear, formulating a higher concentration rapid-acting insulin has been challenging as increases in concentration typically result in material reductions in absorption, slowing the onset of action and delaying the whole absorption curve. Such difficult PK/PD characteristics mean there are no high concentration rapid or ultra-rapid insulins available currently, nor do we know of any in clinical development. AT278’s profile suggests it could be ideally suited for the growing number of patients that require high daily doses of insulin (the majority of whom are Type II diabetics). This profile also offers the potential to enable the next generation of miniaturised insulin delivery systems.

Previously we had modelled AT278 use principally in patients with high daily insulin requirements (>200 units/day), typically Type II and refractory Type I diabetics. Here a high concentration rapid (not necessarily ultra-rapid) insulin has several benefits including the reduction of the injection burden (both volume and frequency). This segment is attractive and is expected to remain so, as even in the US, the market with highest pump penetration, users of insulin pens still outnumber insulin pump users by a factor of twelve.

The ultra-rapid profile means AT278 could address significant unmet needs in both the prandial (mealtime) and high insulin requiring diabetes markets. It also offers the potential to convert more Type II diabetics to insulin pump therapy. Estimating AT278’s potential in the pump segment is fraught with difficulties, not least understanding which patient groups will be early adopters and identifying the rate limiting factors in ensuring wider clinical adoption (each market has differing patient criteria, reimbursement rules, and educational/support infrastructure). Nonetheless, it is clear that over the longer term, pump usage will become a key element in our valuation.

Usually the sizeable, and robust, data from a Phase II clinical trial is the spur to out-licensing discussions with larger players who have the resources and funding to undertake the required Phase III registration trials. The obvious partner for AT278 would be one of the three global insulin companies (Novo Nordisk, Eli Lilly, and Sanofi); however, the insulin pump opportunity is such that a device manufacturer could easily seek to maximise its commercial positioning and integrate an optimised bespoke insulin into its product offering. Whilst a large device manufacturer, such as Medtronic, would be a suitable partner, we suspect it would likely be a smaller, and nimbler, player such as Tandem or Insulet who would be the ones to grasp the opportunity to differentiate their pump systems.

We expect AT278 to follow the FDA’s PHS 351(a) regulatory pathway, the traditional pathway for approval of biologics and innovator biologics, and not the 351(k) application employed for biosimilars (products that are highly similar to a reference or originator product). This reflects the expectation that AT278 will show “clinically meaningful differences” to existing products. Whilst PHS 351(a) is termed a “stand alone” application, the data burden is not expected to be onerous. For context, Eli Lilly’s Lyumjev approval was based on the results from the Phase III PRONTO clinical trial programme. This consisted of two 26-week studies: PRONTO-T1D enrolled 1,222 patients with Type I diabetes and PRONTO-T2D had 673 patients with Type II diabetes. In part this reflects that fact that Lyumjev (insulin lispro-aabc) is a reformulation of Humalog (insulin lispro).

Given this straightforward approval pathway, we would argue Arecor could consider maximising AT278’s value by retaining the option to progress AT278 through to US approval itself, and then either partnering with an existing diabetes player or, more creatively, striking a supply agreement with a pump manufacturer.

The number of possible commercialisation options mean that predicting the value as a single outcome is difficult. For our modelling we assume a typical out-licensing deal, with an upfront payment and development and commercial milestones, together with high single digit to low double-digit royalties on net sales. Given the early clinical stages we employ deliberately, arguably overly conservative assumptions, with the expectation of reviewing our models as further clinical data supports this attractive clinical profile. Increasing our success probability, peak sales expectation, and potential royalty rate lifts AT278’s contribution to our rNPV valuation model. This is now £35.0m, equivalent to 125.6p per share, vs £5.4m, equivalent to 19.4p per share, previously.

AT247: the leading ultra-rapid insulin programme

AT247 is a second-generation ultra-rapid prandial insulin analogue. Its formulation aims to materially accelerate absorption after injection, achieving a profile that closely approximates healthy (non-diabetic) physiological insulin secretion, giving more effective management of blood glucose levels. AT247 contains excipients that bind calcium ions and are hypothesised to cause a transient disruption of calcium-dependent cell adhesion through reversible interactions with the calcium-cadherin complex at the cell surfaces. The disruption at the injection site results in increased tissue permeability and the desired faster absorption. AT247 also contains a stabilising surfactant and standard preservatives (phenol and m-cresol).

AT247 was examined in a Phase I clinical trial that compared it against Novo Nordisk’s NovoRapid (IAsp) and Fiasp (faster IAsp). The double-blind study tested 19 Type I diabetics using a standard glucose clamp setting to determine the pharmacokinetic (PK), pharmacodynamic (PD), and safety characteristics of AT247. Full results were published in Diabetes Care February 2021, with AT247 having successfully met all study endpoints and suggesting a best-in-class profile. The relevant data are shown in the table and graphics (Exhibits 5 and 6).

Exhibit 5: AT247 Phase I PK/PD clinical study results

Source: Svehlikova et al. Diabetes Care 2021: 44: 448-455

These data show that AT247 has a superior onset of action and activity throughout the important 120 minutes after dosing vs both NovoRapid and Fiasp. For instance, AT247 was nine minutes faster than Fiasp for onset of action, achieved a three-fold increase in glucose lowering in the first 30 minutes and a two-fold increase in the first 60 minutes, yet was comparable over 480 minutes. As expected, AT247 was well tolerated with no safety concerns seen.

Exhibit 6: Phase I data on onset, offset and overall exposure and glucose-lowering effect

Source: Svehlikova et al. Diabetes Care 2021: 44: 448-455

A second Phase I study with c 24 Type I diabetics evaluating AT247 administered over three days through a continuous subcutaneous infusion via an insulin pump dosed the first patient in January 2022. The study is designed is a double blind, randomised, three-way crossover which will examine PK and PD, using a glucose clamp, against active controls (NovoRapid and Fiasp). Top-line results are anticipated in H222. A multi-centre Phase II study with c 42 diabetic patients will then explore AT247 against Fiasp when administered through an insulin pump over an extended period (around six weeks). AT247’s Phase II and Phase III programmes should be similar to those of AT278 and, assuming smooth progress, the timelines suggest first approvals, and commercialisation, could happen as early as 2025.

We also expect AT247 to follow the FDA’s PHS 351(a) regulatory pathway, not the 351(k) application employed for biosimilars, reflecting our expectation AT247 will show “clinically meaningful differences” to existing products, Again we believe the clinical data required should not be too onerous, with two modestly-sized pivotal Phase III trials sufficient. The appeal of AT247 is its demonstrated superior profile which suggests it has the potential to be the fastest acting most physiological insulin available to patients in what is a clearly growing and attractive marketplace. Its profile could also facilitate the development of a fully closed loop artificial pancreas.

Our modelling expectations for AT247 remain unchanged, except for a slightly higher potential royalty rate, and are, as always, based on conservative assumptions. We expect visibility to increase over the next 12-18 months as more study data confirms, or otherwise, the suggested clinical profile. AT247 now contributes £33.0m, equivalent to 118.4p per share, to our rNPV valuation model vs £23.9m, equivalent to 86.4p per share, previously.

Valuation

We value Arecor using an rNPV model, explicitly valuing the diabetes franchise, four partnered assets, and the in-house specialty hospital products research programme(s). Revisiting assumptions for the diabetes assets raise our valuation to £140.9m, equivalent to 506p/share, vs £103.7m or 374p/share previously.

Under this methodology, the rNPVs of the individual R&D projects are assessed and success probabilities adjusted for the clinical, commercial, and execution risks carried. These are summed and netted against operational costs and cash. Success probabilities, based on standard industry criteria for development stage, are flexed to reflect the inherent risks of the individual asset, indication, and development/ regulatory pathway. Although the current strategy envisages the out-licensing most of Arecor’s assets before the later costlier stages of clinical development, we allow for commercial and execution risks which are integral to any asset’s intrinsic value. We consciously apply conservative assumptions throughout.

Better than hoped for AT278 Phase I data prompts us to revise our assumptions. The data showed a superior efficacy profile that could support both more attractive pricing and greater market share capture for AT278; thus, we raise our success probability from 40% to 60% and our peak sales to $424m from $254m. This data could also justify improved licencing economics for both diabetes assets (assuming a single partner), reflected in a two percentage-point upgrade to our royalty rate. Collectively, the diabetes franchise now accounts for 48% of Arecor’s value. Exhibit 7 summarises model outputs and underlying assumptions.

Exhibit 7: Arecor rNPV valuation

Source: Trinity Delta   Note: AATD = Alpha-1 antitrypsin deficiency; assumptions include a 12.5% discount factor, £/$ FX rate of 1.30, and 10% taxation from 2026 (UK patent box).

In line with our philosophy, we take a conservative approach to our modelling. Key areas which could represent valuation upside include progress of the current pipeline and earlier-stage formulation development collaborations, while the Arestat technology platforms are a source of intangible value underpinning Arecor’s business model. Not only will development progress de-risk the pipeline but it is also likely to reveal more about the commercial terms of the partnered programmes as sensitivities around terms of these inevitably mean visibility is limited until later in the development process. For in-house assets that are yet to be partnered, more information on their profile could have a positive impact on the deal terms achieved. Hence, until our knowledge improves (especially with respect to the identity of some of the underlying programmes), we employ modest assumptions for launch timings, pricing, and market shares.

In addition, we only include very modest risk-adjusted development milestone assumptions for actual and potential licensing deals, with assumed royalty rates at the lower end of management guidance. The limited visibility on the formulation development collaborations means these are not included in our valuation until they convert to longer-term licence partnerships, and hopefully further disclosures are made with respect to key factors such as the identity of the underlying assets, indications to be pursued and likely development timeline. Management expectations for the licencing of one formulation development programme each year therefore provides upside.

Financials

Arecor’s H121 results reaffirmed its development and operational achievements during the six months to 30 June 2021 and post-period. The highlight was the June AIM IPO which raised £20m gross (£18.5m net), boosting end-June 2021 cash to £22.1m (vs £2.9m at end-December 2020). In its recent business update, the company confirmed an end-December 2021 cash balance of £18.3m and an intention to report FY21 results during the week commencing April 25. These cash resources, coupled with income from anticipated partnering/licencing opportunities – particularly from the in-house diabetes and/or specialty hospital products – as well as future milestones under current deals, provide funding through 2023, a comfortable runway to achieve multiple value inflection points, and upside potential.

H121 revenue of £460k was primarily derived from formulation development partnerships, which represent a relatively steady income stream with revenues recognised as performance obligations are satisfied over time. Higher revenue of £817k booked in H120 was boosted by a £339k milestone receipt from an existing licencing agreement. Revenues from licencing agreements are more variable and include upfront payments, milestones, and other payments; typically, the first two are recognised on licence grant/signature of the agreement and milestone achievement respectively, while other payments reflect performance obligations under the contract. Arecor’s other operating income in H121 of £51k reflected start-up activities under the £2.8m Innovate UK grant awarded in March 2021 for AT247 clinical development; In H120, £0.3m in other income was received under two separate Innovate UK grant projects that ended during 2020.

From FY21 onwards we expect modestly higher research income as infrastructure and headcount expansion increases capacity for formulation work, both in-house and with partners. We note that Arecor signed five formulation development contracts during 2021, which will both contribute to near term revenues and also offer conversion potential into future licences. However, our forecasts do not include any assumptions on potential conversion(s) of pre-licence technology partnerships to longer-term licence agreements (which bring potential for small upfront payments, plus future milestones and single-digit royalties).

The magnitude of licence derived income will be determined by development and commercial progress of licenced programmes, the timing and terms of new partnership deals (particularly for the in-house diabetes assets), and product launches. On this front, Arecor has provided some guidance with respect to its development agreement with Hikma, where the transfer of the final formulation of AT282 is anticipated in H122 and will trigger a milestone payment. Arecor’s four existing partnered products (two that emerged from technology partnerships, two from out-licencing internally developed formulations) are expected to generate development and commercial milestones, plus royalties or equivalent on sales from 2023 onwards following anticipated launches.

R&D investment in H121 rose to £1.9m (H120: £1.6m) with the increase due to costs of the AT278 Phase I trial. G&A expenses were also higher at £1.4m (H120: £0.7m), although this included £0.5m in non-recuring spend associated with the AIM IPO. Operating loss was £2.96m (H120: loss of £1.34m), with a pre-tax loss of £3.45m (H120: loss of £1.34m) and a net loss of £3.11m (H120: loss of £1.05m). Continued receipts of tax credits until profitability is expected.

Through its IPO, Arecor has the resources to progress its in-house diabetes and specialty hospital products to partnering inflection points, as well as expand its internal capabilities to support progression and growth in its earlier-stage formulation portfolio. Consequently, we anticipate a significant increase in R&D investment in future years. However, for FY21, we have lowered our R&D forecast to reflect phasing of R&D spend, mainly reflecting the Q122 start of the AT247 US Phase I pump study. We now expect R&D investment of £5.1m for FY21e and £11m for FY22e with the initiation of the Phase II diabetes trials. FY21 SG&A expenditure includes one-time IPO costs, so we expect this line item to fall to a run rate of c £2.5m from FY22e.

Changes to our estimates are presented in Exhibit 8, with a summary of our forecasts shown in Exhibit 9.

Exhibit 8: Summary of changes to estimates

Source: Trinity Delta

Exhibit 9: Summary of financials

Source: Company, Trinity Delta   Note: Due to subsequent restatement of accounts FY18 and FY19 relate to the 12 month periods ending 31 May 2018 and 31 May 2019 respectively.