Europe’s hub-and-spoke biotechs balance risk, harness AI

An Expert View from Cumulus Oncology about its innovative business model, which is all about improving the efficiency of drug R&D. 

Improving the efficiency of drug R&D has become more urgent as populations age and health systems are squeezed. The tools and technologies exist but figuring out how best to harness and organize them for new drug discovery and development remains a work in progress. Pharma is too big and slow to adapt; many biotechs are too small, under-resourced and risky.

Hub-and-spoke style biotechs have proliferated over the last decade to address this challenge. They assemble a suite of asset- or therapy area-focused subsidiaries (spokes) around a central operational hub, combining small-firm innovation with larger-firm resources and resilience. With a portfolio of programs controlled – to a greater or lesser extent – by central command, these models offer investors several shots on goal, curated by a tried-and-tested core team. Each subsidiary focuses exclusively on its asset or assets, tapping into central leadership, expertise and functions as required. The lean, nimble off-shoots are set up to fail fast and move on, sparing the expense of repeatedly building company infrastructure.

High-profile examples like listed Roivant (Nasdaq: ROIV), BridgeBio (Nasdaq: BBIO) and PureTech Health (LON: PRTC) appear to validate the model: these groups are collectively worth about $15 billion and have between them generated commercial products and a substantial late-stage pipeline. Roivant continues a profitable trade in ‘vant’ subsidiaries, with the recent sale of Dermavant to Organon (NYSE: OGN) and 2023’s sale to Roche (ROG: SIX) of inflammation and fibrosis-focused Telavant for $7.1 billion. BridgeBio this year created two new offshoots, BridgeBio Oncology Therapeutics and rare diseases-focused GondolaBio, each raising hundreds of millions of dollars in fresh funding. Private Nimbus Therapeutics had a golden 2023, selling its TYK2 inhibitor psoriasis asset (in a sub-co wrapper) to Takeda Pharmaceuticals (TYO: 4502) for $4 billion up-front, and raising a fresh $210 million to advance next-generation assets spanning cancer, auto-immune and metabolic diseases.

Several flavors of hub-and-spoke biotechs have sprung up over the last 15 years. Nimbus emerged in 2009 on the back of a partnership with computational drug discovery firm Schrödinger (Nasdaq: SDGR); Germany’s TVM Capital has had a long-term strategic partnership since 2012 with Eli Lilly (NYSE: LLY) to fund lean, single-asset focused companies designed for a rapid exit. Roivant was founded by entrepreneur Vivek Ramaswamy in 2014 to find de-prioritized assets. Swiss-headquartered Medicxi launched Centessa in 2021 by assembling ten portfolio biotechs with programs spanning several therapy areas and stages of development. Others are more focused: PureTech’s spokes initially targeted the brain-immune-gut axis; newcomer Alys Pharmaceuticals is in immuno-dermatology. 

The model offers investors a choice of entry points: at the ‘TopCo’ level, and/or in subsidiaries as they mature and secure their own financing.

Tailored to Europe

The portfolio approach is particularly well-suited to Europe, with fewer experienced biotech executives than the USA, and where investors tend to be more risk-averse. The biggest hub-and-spoke biotechs are USA-based (though Boston-headquartered PureTech maintains a London listing). Yet smaller, still-private European hub-and-spoke biotechs are leveraging their leaders’ knowledge and fund-raising experience to turn local science into attractive, risk-mitigated opportunities for follow-on investors and pharma buyers.

Their structure allows even the smallest R&D programs to access potentially transformative technologies like artificial intelligence (AI). Few standalone early stage European biotechs could hope to identify and access the most suitable AI while pushing their programs and fund-raising – particularly during 2024’s prolonged biotech market malaise.

Cumulus Oncology, set up in 2017 by entrepreneur Clare Wareing, aims to accelerate promising, cancer-focused projects from academia or industry through to pre-clinical validation or into early clinical studies. Many Phase III trial failures result from poor decisions earlier on – including over-hasty Phase I-to-Phase II transitions, notes Dr Wareing. Cumulus’ model “enables us to fail early” she says, filtering out all but the best projects. That suits today’s buyers’ market, in which pharma licensors can afford to demand better data.

Cumulus raised a £9 million ($11.7 million) seed round in December 2023 and has founded and nurtured four companies so far, including Nodus Oncology and GIO Therapeutics. Nodus targets DNA damage response pathways. It licensed a discovery-stage small molecule inhibitor of the poly (ADP ribose) glycohydrolase (PARG) enzyme from Basilea (SWX: BSLN) in 2022. PARG, which works alongside well-validated PARP to repair damaged DNA, isn’t a new target - IDEAYA Pharma (Nasdaq: IDYA) and Switzerland’s FoRx have, respectively, early clinical and IND-stage candidates. But fast-follower programs carry less risk.

That means they’re more likely to appeal to what pharma buyers (and later-stage investors) are interested in, says Dr Wareing, who founded contract research organisation Nexus Oncology, sold to Ockham in 2012. Her team, which has collectively taken ten programs to NDA stage, knows the asset profiles and data packages that pharma wants. The hub-and-spoke model extracts maximum leverage from this ability to select programs, identify value inflection points aligned to partners’ goals, and reach those milestones rapidly and cost-effectively. Cumulus is de-risking five further projects that may seed future subsidiaries. One is an isoform of a target that is “attracting lots of M&A,” says Dr Wareing.

AI à la carte

The ‘fail early’ philosophy is supported by AI/ML tools scouted out by Cumulus’ Chief Data Scientist Phil Chapman. “We’re building a central [AI] capability that can support the companies as they grow,” he says. Mr Chapman’s task is to find the best tools for the jobs at hand within the Cumulus family. Given the pace of AI development, most of those tools are likely accessed externally rather than built in-house.

Switzerland’s Cureteq is also capitalizing on AI to streamline assets’ path to proof-of-concept. The firm – which describes itself as part asset manager, part biotech - is hub to Oncoteq and a second “teq” subsidiary still in stealth mode. Cureteq itself is also a spoke within the broader portfolio of investor Partex, sitting alongside Partex-backed Innoplexus, whose AI technologies and expertise “we tap into as required,” explains Cureteq CEO Mads Dalsgaard. As at Cumulus, Cureteq’s plug and play approach applies specific AI solutions to specific challenges along the R&D chain, including asset-selection. “AI can surface completely hidden assets that would not have emerged from standard database searches or partnering events,” says Dr Dalsgaard.

Attracting talent

Hub-and-spoke biotechs also have advantages in attracting talent, often a challenge for small biotechs. The model lowers risk for prospective employees: with multiple programs running in parallel across the organisation, failure in one corner needn’t be the end. “You can move to another company” within the group, says Kirsty McCarthy, CEO of Oxford, UK-based Molecule to Medicine (MTM), which employs more than 80 people across the six companies it has co-created to date. The structure also allows less experienced teams to learn from executives in the central hub, for example around how to engage with investors during fund-raising. More senior colleagues may work across more than one subsidiary, with significant autonomy as these subsidiaries mature and seek independent funding.

MTM’s spokes include oncology and auto-immune disease-focused Greywolf Therapeutics, immuno-oncology firm Pathios Therapeutics and GI-focused ThirtyFive Bio, all based in Oxford, UK. They have collectively raised $250 million from blue-chip investors including Canaan Partners, ICG Life Sciences, Andera Partners and Pfizer Ventures. This success comes part thanks to the track record of MTM founders Kirsty and Tom McCarthy who sold CNS-focused Spinifex to Novartis (NOVN: VX) in 2015.

Yet as MTM’s eco-system grows, the entrepreneur couple is determined to retain the model’s nimbleness. Growing beyond three subsidiaries runs the risk of hierarchies and bureaucracy creeping in. So, instead of simply adding new firms to a single hub, they create additional ‘pods’ in other locations, each focused on a different therapy area. Oxford houses the immuno-oncology pod, London supports an emerging cardio-metabolic pod, and inflammation is in Brisbane, Australia. “Co-location is important,” says Ms McCarthy. 

The success of biotech eco-systems like Boston, US or Cambridge and Oxford in the UK prove that tight, concentrated networks spawn ideas. Pathios emerged from “chatting with an MD/PhD at Oxford”, illustrates Ms McCarthy; internal discussions and data analysis helped spawn ThirtyFive Bio. Data and AI is also a big focus across MTM, says Ms McCarthy, whose background is in technology. The Oxford pod alone houses 12 bioinformaticians, she adds.

Flexible funding

Hub-and-spoke biotechs offer investors a choice of entry points. Those seeking to spread risk across several projects may fund the TopCo (hub). “We operate on fund principles,” says Cumulus’ Dr Wareing, with a portfolio mix spanning different stages of development and risk-levels, and concrete milestones. Cureteq operates within a budget from investor Partex, but one of the pluses of the hub-and-spoke model “is the ability to raise capital at different levels”, says Dr Dalsgaard. For now, Cureteq doesn’t need additional outside funding, but the team is currently raising capital for Oncoteq. “It’s a means of validating the model externally.”

MTM’s middle-tier “pods” – another variation on the hub-and-spoke philosophy – may offer a further investor entry point: “We’re having discussions” about pod-level investments, says Ms McCarthy, adding that there are already several common shareholders across the eco-system. As investors grow more confident in the approach and in the core team, they’re more likely to engage with future subsidiaries.

Challenges

For all their strengths, hub-and-spoke biotechs cannot fully overcome the sector’s wider funding challenges. Paris, France-based Home Biosciences, founded in 2019 to create biotech ventures from attractive academic science, has struggled since its 2021 seed financing from Redmile Group and Sofinnova Partners. MTM’s Ms McCarthy and her team also felt the downturn “very significantly” and took longer to raise money.

Nor can the model escape the risks inherent in drug R&D, though it does its best to mitigate them. Roivant and BridgeBio have both faced high-profile setbacks – most recently, BridgeBio’s gene therapy for adrenal gland disorder – though these larger firms often spring back with new or re-focused subsidiaries. “Hub and spoke are still subject to the risks of any single project,” notes one Europe-based investor.

Having too many different spokes risks spreading central expertise and resources too thinly. As programs mature, maintaining the appropriate balance of sub-co autonomy and central control, and appropriately sharing central resources, can be a challenge. So can management incentives. “One reason the traditional start-up model is successful is that management have skin in the game, and significant up-side potential,” says one biotech investor. “The team isn’t looking after a whole portfolio” of companies, he adds.

One or two hub-and-spokes have back-pedalled. Cullinan Oncology (Nasdaq: CGEM), founded in 2016, gradually unwound its sub-co model. But it remains agile: the company in April 2024 rebranded as Cullinan Therapeutics as it expanded its focus beyond cancer into auto-immune diseases, led by a bispecific T-cell engager program pivoting from hematology to auto-immune disease. The company also raised $280 million in a private placement.

More to come?

The multiple hub-and-spoke experiments underway will help identify the optimal structures for accelerating and streamlining drug R&D. The model’s characteristics suit today’s uncertain financing climate, proliferating modalities and technologies, and shifting big pharma strategies. Momentum will likely continue.

Indeed, MTM’s expansion has happened faster than expected, says Ms McCarthy, who plans to consolidate current assets through the end of 2024, and thereafter build more pods in other locations, such as Switzerland. Cumulus is similarly ambitious, currently targeting US investors for a £50 million funding round.

Not everyone is convinced. “I don’t think there’s any good evidence that this is a superior approach,” says one VC. But for others, hub-and-spokes are an article of faith: Centessa founders Medicxi created dermatology-focused Alys from six portfolio biotechs in February 2024, assembling a ready-made asset suite for potential buyers.

“We definitely compete with VCs,” concedes MTM’s Ms McCarthy, in the sense of building and investing in a diverse portfolio of assets and companies. But the approaches are also complementary – both support progress from discovery to drug. Plus, flourishing innovation means “there’s room for all of us.”