AI-driven drug discovery is having a moment in the digital world. As for the real world, results have been a little more difficult to come by.

Exscientia was one of the first companies to apply artificial intelligence to drug discovery. Unfortunately, it has frustratingly little to show for its efforts since being founded in 2012.

While investors don't have much data on which to pin their hopes and dreams for the stock, the development strategy seems right-sized for the company's size and cash runway. The U.K.-based drug developer's initial clinical focus includes a handful of high-value targets in oncology, while partnered programs span therapeutic areas from oncology to immunology to neuroscience. That's a refreshing balance between delusional ambition (making it difficult to deliver) and risk-avoidance (making it difficult to realize value).

A right-sized strategy enabled by a right-sized cost structure can go a long way. The combination provides more flexibility if pivots or program terminations are required later. And this is early-stage drug development. There will be pivots and/or program terminations.

But it doesn't mean much without clinical data.

Exscientia expects a preliminary phase 1 data readout for its lead drug candidate, a CDK7 inhibitor, by the end of 2024. That will provide the first meaningful safety data, early signals about efficacy (keywords are "early signals"), and a slew of nerdy metrics to help fine tune dosing levels. Meanwhile, the company expects to advance two wholly-owned programs into phase 1 clinical studies by the first half of 2025. That sets the stage for a busy schedule of meaningful data readouts in… 2026. Right before the cash runway is expected to run dry.

By the Numbers

The financial performance of early-stage drug developers must be analyzed with the proper perspective. Investors should still focus on operating efficiency – it's just defined a little differently when revenue is non-existent.

For Exscientia, investors want to see an operating structure that's right-sized for the maturity of the platform. A drug developer with a handful of programs doesn't need to spend hundreds of millions of dollars per quarter or enlist 1,000 employees working on the technology. Technology isn't that valuable by itself. A surge in headcount ahead of key expected developments could be justified, but investors simply need to hold management accountable when appropriate.

• The company's headcount has grown from 285 employees at the end of 2021 to 481 at the end of 2022. There were 483 employees at the end of last year.
• That surge in 2022 was a little ambitious, especially with meaningful data readouts several years away at the time. It also contradicts (or exposes…) the promise of using automation to reduce the labor part of R&D.
• In May 2024, the company announced plans to reduce headcount by 20% to 25% by the end of the year. The moves are expected to play a role in extending the cash runway into 2027 and save about $40 million per year beginning in 2025.

There's some fine print attached to the company's cash runway calculus.

Exscientia ended March 2024 with $417 million in cash and burned $39 million in cash from operating activities during the first quarter. That would suggest a cash runway for about 10 quarters, or into late 2026. But investors also need to account for higher expenses from a maturing pipeline. From summer 2024 to summer 2025, the number of active clinical trials will expand from one to three. That'll be expensive.

That's why the U.K.-based drug developer also includes expected milestones from partnered programs when determining its cash runway. That can get investors into trouble, but it does help that Exscientia's collaborations are with European drug developers falling behind their global peers. The increased pressure on them to keep up – or catch up – works in the company's favor.

There are three partnerships that will impact cash flow between now and the end of 2026:

Bristol Myers Squibb (autoimmune diseases)

For the math to pan out, Bristol Myers Squibb must continue advancing the PKC-theta inhibitor developed by Exscientia. The autoimmune target has tripped up over 10 drug developers to date, all of which have terminated their respective programs. The asset demonstrated super-early results that suggest it might be selective enough to overcome the industry's past challenges, but it must advance into a phase 2 study to trigger the next milestone payment.

Sanofi (autoimmune diseases)

A partnership with Sanofi, also focused on autoimmune diseases, is the next-most mature. It includes multiple programs, including one added in Q1 2024 that netted a $4 million milestone payment. Advancing programs from discovery to preclinical to clinical development can yield an additional $50 million to $100 million in milestones by the end of 2026.

This partnership has the most lucrative economics for Exscientia. Unlike the partnerships with Bristol Myers Squibb and Merck, the collaboration with Sanofi leverages the company's technology platform for both drug discovery and translational science. The latter represents an additional data layer and can help with drug design considerations, patient selection, and clinical trial design.

Merck (oncology, neuroscience)

In September 2023, Exscientia announced a collaboration with Merck that initially includes three programs across oncology and neuroscience. The company is eligible to receive up to $93 million in milestone payments for discovery-stage research, which could be realized by the end of 2026.

Other Partnerships

Exscientia also has active partnerships with Rallybio and GT Aperion Therapeutics ("Aperion") in which assets are co-owned. The former is structured to have Rallybio assume responsibility for clinical development, while the latter is structured to have a 50/50 split of development and commercialization costs.

The collaboration with Aperion is focused on discovering and developing next-generation CDK inhibitors, including the company's lead drug candidate, the CDK7 inhibitor named GTAEXS617.

Around the Horn

Exscientia's fundamental technology development goals are similar to those of Relay Therapeutics: combine experimental and computational data to design more selective drug candidates that might overcome key challenges of existing treatments. The two companies approach these goals differently, but both approaches can create value.

It helps to take a closer look at the pipeline, asset by asset.

GTAEXS617 (CDK7 inhibitor)

Exscientia is co-developing '617 with Aperion, although the smaller peer has been slowly shedding responsibility and economic rights for the asset – giving the AI-based drug designer a larger economic opportunity. The two originally synced up to develop a portfolio of CDK inhibitors, but the status of those ambitions is unclear.

CDK inhibitors are a well-established class of oncology drugs, especially in hormone-receptor positive (HR+) tumors such as breast and ovarian. These are often serving very large patient populations. In 2022, CDK inhibitors generated global revenue of $8.9 billion.

However, existing drug products are CDK4/6 inhibitors. It turns out two out of three patients develops resistance to existing treatments. As the understanding of the CDK family of proteins has evolved, scientists have realized that an ideal CDK inhibitor might instead target CDK2 (Relay Therapeutics) or CDK7 (Exscientia).

The most likely (or wisest anyway…) development strategy for '617 might be to position the asset as a first-line treatment for patients who have developed resistance to CDK4/6 inhibitors. That's a significant opportunity to start. Then, the company can conduct additional phase 3 studies to evaluate the asset's potential as a first-line CDK inhibitor.

Of course, there's a long way to go. A phase 1 clinical trial is still enrolling patients, but investors can expect preliminary results before the end of 2024.

EXS74539 (LSD1 inhibitor)

This wholly-owned preclinical asset is expected to begin a phase 1 clinical trial before the end of 2024. It will mark the first time Exscientia has advanced an asset for which it owns 100% of the economic rights into the clinic.

LSD1 is an emerging molecular target for treating advanced myeloid leukemia (AML), which is a type of blood cancer. The target is a bit unique, however, in that there's also potential to treat solid tumors such as small-cell lung cancer – a tumor type existing drug products struggle with.

The bar is pretty high in AML, as many patients can experience complete responses from initial treatment. But many patients relapse. The industry is trying to develop new treatments that ensure those complete responses are long-lasting, too.

The company designed '539 to have reversible binding to its target (improving clearance with once-daily dosing potential and lowering on-target toxicity), cross the blood-brain barrier (to potentially treat brain lesions), and have less selectivity for blood platelets (reducing off-target toxicity).

EXS73565 (MALT1 inhibitor)

This wholly-owned preclinical asset is expected to begin a phase 1 clinical trial in the first half of 2025.

MALT1 is an emerging molecular target for treating relapsed / refractory B-cell malignancies such as chronic lymphocytic leukemia (CLL), which are types of blood cancer. The idea is to combine MALT1 inhibitors with more established drug products, such as BTK inhibitors, to drive a greater response.

The primary challenge for MALT1 drug candidates studied to date is a lack of selectivity. Inhibitors often bind off-target to UGT1A1, which can drive significant liver toxicity, especially when combined with BTK inhibitors (which have well-known liver toxicities themselves).

Similar to AML, the bar is pretty high in B-cell liquid tumors, as many patients can experience complete responses from initial treatment. Similar to AML, many patients relapse. The industry is trying to develop new treatments that ensure those complete responses are long-lasting, too.

Margin of Safety & Allocation

Exscientia is considered a Growth (Speculative) position. The current modeled fair valuation for the company based on my 2024 model is below:

• Market close May 23: $4.73 per share
• Modeled Fair Valuation: $6.03 per share
• Allocation Range: Up to 5%

Exscientia reported 126.500 million shares outstanding as of March 31, 2024, according to the weighted average share count. The modeled fair valuation above assumes 139.150 million shares outstanding, which is equivalent to 10% dilution.

Further Reading

• May 2024 press release announcing Q1 2024 operating results

‍