The Chicago Cubs desperately needed a win. Not much has changed in 94 years, but to be clear we're talking about October 1932.

After losing the first two games of the World Series to the vaunted New York Yankees, the Cubbies went down by three runs in the opening frame before they faced their first pitch. Not a great start, but they recovered. By the end of the fourth inning the team managed to tie the score 4-4 to rejuvenate the 50,000 fans jammed into the seats at Wrigley Field.

Maybe a little too much. The raucous crowd and jeering from the dugout didn't sit well with this one guy in pinstripes.

In the top of the fifth inning, Babe Ruth had had enough. He stepped into the batter's box, pointed to center field, and then proceeded to launch a baseball over the fence. In center field. It became one of the most iconic moments in baseball history, now known simply as the "Called Shot." Little kids still imitate it on baseball diamonds across the country to this day.

The Cubs lost the World Series in four straight games after getting outscored 37-19. It wasn't even close. Fans would have to wait another 84 years to see their team win a World Series.

All right, I'm done picking on 'em.

The competitive landscape in vascular anomalies (VAs) is shaping up a lot like Ruth's Called Shot – and maybe even the 1932 World Series.

Relay Therapeutics had been telegraphing its confidence in zovegalisib's potential to reset clinical benchmarks for VAs. It set the tone in recent investor presentations, trademarked five variations of "VAntage" in an early effort to build commercial infrastructure, and even published an investor webinar walking through the nerdy details of the clinical obstacles and commercial opportunities. Not bad for a company with the public personality of a cardboard box.

Then on May 19, 2026, the company revealed nearly flawless preliminary data from the phase 1/2 ReInspire study. The middle dose evaluated had a 100% response rate, appeared well-suited for chronic use, and led to rapid improvements in quality of life (QoL) for patients with no treatment options. That's about as close as it gets to calling your shot in drug development.

That's bad news for Novartis, the only meaningful competition, which is taking its fourth crack at the commercial opportunity with its EPIK-P4 study of alpelisib (branded as Piqray in breast cancer and Vijoice in VAs). It earned an accelerated approval based on the phase 1/2 EPIK-P1 study, then flunked its first confirmatory EPIK-P2 study. The ongoing study is using an even higher dose that's unlikely to have an acceptable tolerability profile.

It's almost like the Cubs getting swept in four straight games. Okay, now I'm done.

Relay Therapeutics is now in an oh-so-sweet position to lock down a multi-billion dollar commercial opportunity years before challengers equipped with their own pan-mutant PI3K-alpha inhibitors arrive.

In fact, the protein motion pioneer is going to request an accelerated approval from the U.S. Food and Drug Administration (FDA). If granted, then zovegalisib might launch in VAs at some point during 2028 – months ahead of launching in 2L+ HR+/HER2- breast cancer. The company still needs to generate more data at an optimized dose and will be required to conduct a pivotal study to confirm the results, but that long-term study could be partially funded with cash flow made possible by an accelerated approval. That's quite the development.

While not guaranteed, the company's confidence in an accelerated approval is evident in its very early trademark filings (specifically relating to patient discovery and onboarding) and trio of capital raises in March, April, and May. That confidence could soon spill over with the potential unveiling of its next asset: a brain-penetrant PI3K-alpha molecule with the potential to treat the lone VA opportunity out of zovegalisib's reach.

What are Vascular Anomalies?

Vascular anomalies (VAs) are characterized by the abnormal growth of blood vessels, lymphatic vessels, and endothelium tissues. Overgrowth of these vascular systems and connective tissues can lead to painful swelling as blood or lymph fluids collect in excess tissues, the loss of proper blood supply to certain organs or areas of the body, or sometimes benign tumors.

There are many distinct kinds of VAs, which are grouped into two subcategories:

• Vascular malformations impact blood vessels (arteries that bring blood out of the heart, or veins that bring blood back to the heart) or lymphatic vessels (for transporting white blood cells as part of the immune system). Vascular malformations do not proliferate or spread throughout the body, but tend to grow with an individual into adulthood. These are permanent
• Vascular tumors result in the growth of benign tumors that do proliferate or spread throughout the body, although they can cycle through periods of growth and shrinkage.

Whereas cancerous tumors are often driven by acquired mutations in multiple genes, VAs are more likely to be caused by mutations in single genes. The most common gene implicated is PI3K-alpha, which plays a crucial role in cell growth by controlling glucose (energy) uptake of most cells in the body. Mutations in this gene are estimated to cause 30% to 50% of all VAs across subtypes. You might see nerds in lab coats refer to "somatic mosaic" mutations. That just means these mutations are not inherited, but occur before an individual is born.

Importantly, single-driver diseases caused by somatic mosaic mutations can be effectively treated by inhibiting the mutated gene, transcript, or protein. Zovegalisib is a pan-mutant PI3K-alpha inhibitor, which suggests it might be able to treat PI3K-driven vascular anomalies. If proven in clinical trials, then the molecule could become the first safe, effective, and well-tolerated treatment for the roughly 52,500 individuals in the United States impacted by moderate to severe PI3K-driven VAs.

For comparison, there are roughly 35,000 individuals in the United States with PI3K-alpha mutated HR+/HER2- breast cancer in the 1L and 2L+ settings. And entering that crowded competitive landscape means you're going to catch some elbows.

The phase 1/2 ReInspire study is focused on three major groups of VAs impacting roughly 85,000 total individuals in the United States. Of that total, an estimated 52,500 individuals would be potentially eligible for treatment with zovegalisib.

• (5,000 total U.S. patients) PI3K-related overgrowth spectrum (PROS) refers to specific types of vascular malformations typically affecting the skin, bones, fat, and vasculature systems. Many of these conditions are very rare and poorly understood. For example, fibro-adipose vascular anomaly (FAVA) was only recognized as a VA in 2025. Others include CLOVES syndrome, Klippel-Trenaunay syndrome, and macrocephalies.
• (60,000 total U.S. patients) Lymphatic malformations (LM) impact the lymph vessels. These can occur as standalone conditions or as part of other genetic disorders, such as Noonan syndrome, Turner syndrome, and Down syndrome.
• (20,000 total U.S. patients) Venous malformations (VeM) specifically impact veins, or the blood vessels that carry blood back to the heart. These are the most common type of vascular malformation, although they're less likely to be driven by PI3K-alpha mutations – the focus for zovegalisib.

Relay Therapeutics intends to initially focus commercial infrastructure and patient discovery on individuals who seek systemic treatment, or those with either moderate (32,000 individuals) or severe symptoms (20,500 individuals).

‍

You'll notice there's a fourth group of VAs called cerebral cavernous malformations (CCMs), which are characterized by overgrowths of blood vessels in the brain or spinal cord. Individuals with CCMs can experience no symptoms at all or very serious complications including paralysis, hearing loss, vision loss, seizures, and fatal hemorrhages.

Relay Therapeutics is not enrolling patients with CCMs in the phase 1/2 ReInspire study. That might be because zovegalisib cannot cross the blood-brain barrier, which is a common drawback of PI3K-alpha inhibitors. There are many variables to optimize for in drug discovery – liver toxicity, half-life, off-target selectivity, on-target selectivity, and so on – so it's easy for drug developers to sacrifice brain penetration.

That said, some molecules in the class can cross the blood-brain barrier, so it is possible. Relay Therapeutics does list a mystery PI3K-alpha inhibitor asset under the vascular anomalies section of its pipeline graphic in investor presentations. It's actually listed as being more advanced than the Fabry program, or roughly equivalent to the maturity of RLY-8161 (NRAS inhibitor) in early 2026. Maybe the company is quietly working on its own brain-penetrant molecule to address the roughly 37,500 individuals in the U.S. with moderate and severe CCMs.

Special Considerations for VAs vs. Breast Cancer

Similar to how Relay Therapeutics is using the phase 1/2 ReDiscover study to optimize zovegalisib as a potential treatment for HR+/HER2- breast cancer, the company is using the phase 1/2 ReInspire study to optimize zovegalisib as a potential treatment for VAs. Although the molecular target of mutant PI3K-alpha remains the same, there are several important differences between breast cancer and VAs.

• Patient age: The median patient in ReDiscover is roughly 60 years old. Makes sense, as cancer generally affects individuals in their fifth or sixth decade of life. By contrast, individuals are typically diagnosed with VAs in early childhood. That imparts unique optimization requirements for young children, adolescents, and adults. The ReInspire study is designed to evaluate three age brackets: 2 to 5 years old (weight-based dosing), 6 to 11 years old (weight-based dosing), and 12 years and older (fixed dosing).
• Patient health: Individuals with breast cancer can be frail and in poorer health, especially for later-line tumors. By contrast, individuals with VAs can be generally healthy depending on the specific type of VA and body region impacted. Healthier patients are generally easier to treat, better tolerate treatment, and have better responses.
• Combinations: Solid tumors in HR+/HER2- breast cancer are driven by multiple mutations and proliferate if left unchecked, which means optimal treatment requires combining different therapeutic agents. Aggressive diseases require aggressive treatments. By contrast, VAs are often driven by a single mutation (ex: PI3K-alpha, TIE2, RASA1) and don't spread throughout the body if left untreated. Although cases can be fatal depending on the body region and organs impacted, the optimal treatment for most patients would keep overgrowths in check and optimize patient quality of life (QoL).
• Resistance mechanisms: Similarly, cancer cells can acquire or mutate resistance to specific therapies over time, resulting in an arms race between tumors and treatments. By contrast, the somatic mosaic mutations in PI3K-driven VAs are unlikely to develop resistance to treatment. These are relatively stable genetic disorders, not cancer. That makes it more likely for zovegalisib monotherapy to be an acceptable chronic treatment for most patients – no combinations or treatment switching required. (Don't be surprised to see combinations attempted for specific subtypes though.)
• Efficacy measurements: In HR+/HER2- breast cancer, an objective response (OR) is achieved when the volume of a tumor is reduced by at least 30% since beginning treatment. In VAs, the apples-to-apples metric is called a volumetric response (VR), which is when the volume of an overgrowth is reduced by at least 20% since beginning treatment.
• Duration of treatment (DoT): Cancer is generally treated until remission, disease progression, or death. That means the duration of treatment generally has a median duration that can be measured. For 1L patients, the DoT could extend to 36 months or more for zovegalisib combinations. Pretty good in that context. VAs are genetic diseases, which will require chronic and potentially lifetime treatment. Dosing will likely be reduced over time once an individual responds with an acceptable VR, but is unlikely to be discontinued completely.
• Sex: Although biological males can have HR+/HER2- breast cancer, the overwhelming majority of patients are female. Women have two X chromosomes, which harbor more immune-related genes. That impacts responses to treatment, tolerability, and for PI3K-alpha inhibitors metabolic responses (women naturally have higher levels of body fat). By contrast, VAs occur in males and females at roughly equal rates. Because zovegalisib primarily has data in females from breast cancer studies, there could be some unknown unknowns lurking.

The ReInspire Study Data Readout, Explained

Relay Therapeutics will eventually share data from hundreds of patients treated through major timepoints – Week 24, Week 48 (one year), Week 96 (two years), Week >whatever zovegalisib resets the bar to< – in the next decade. We're not there yet.

The May 2026 data readout from the phase 1/2 ReInspire study included preliminary dose escalation data from 32 patients evaluable for tolerability metrics and 20 patients evaluable for efficacy metrics with a data cutoff of April 15, 2026.

As far as preliminary data readouts go – in which the dosing is being optimized and efficacy signals generally offer a glimpse of a molecule's potential – zovegalisib was nearly flawless.

• Dosing: Three doses were evaluated: 100 mg twice-daily, 300 mg twice-daily, and 400 mg twice-daily.
• Discontinuations: All 20 efficacy-evaluable patients remained on treatment with no discontinuations.
• Dose reductions: There were 12 total dose reductions, including one (1) at the 100 mg dose, four (4) at the 300 mg dose, and seven (7) at the 400 mg dose.
• Dose intensity: The dose intensity was 100% at the 100 mg dose, 99% at the 300 mg dose, and 77% at the 400 mg dose. Dose reductions impact the dose intensity.
• Efficacy: The overall volumetric reduction rate (VRR) was 60% (12/20) at Week 12. The highest ever achieved in VAs was previously 32%, which required a timepoint of Week 24. All patients (13/13) who received 300 mg or 400 mg of zovegalisib experienced overgrowth shrinkage at every measured timepoint, although three patients at the higher dose didn't qualify as a VR.
• QoL improvements: Whether measured from investigator measured (89% improved), patient-reported overall benefit (79% improved), or patient-reported pain (71% improved); all doses achieved QoL improvements by Week 12. Almost all doses achieved improvements across all QoL metrics from Week 12 to Week 20. The lone discrepancy occurred in patient-reported responses at the 100 mg twice-daily dose, which changed from a score of 1.5 at Week 12 to 1.3 at Week 20.
• Dose optimization: The lowest dose of 100 mg will not be advanced due to being at the lower bound of the therapeutic window, while the highest dose of 400 mg will not be advanced due to tolerability issues. The middle dose of 300 mg twice-daily, which had a 100% response rate, will be evaluated in the dose expansion portion of the study. Relay Therapeutics will also evaluate a new dose of 400 mg once-daily, which could offer improved tolerability over the 300 mg twice-daily dose without sacrificing efficacy.

Demographics

The preliminary data readout from May 2026 had the following patient demographics for enrollment. There are two important metrics for investors to consider – one good, one to keep an eye on.

First, patients who have previously been treated with PI3K-pathway inhibitors Vijoice or Rapamune are allowed to enroll in ReInspire. They represent 72% (23/32) of the patients in the May 2026 data readout. That increases the confidence in zovegalisib to drive responses where existing limited treatments or experimental therapies have failed.

Second, Relay Therapeutics didn't do a great job matching the real-world prevalence of VAs to the enrollment for this preliminary data readout. Of the 32 patients enrolled, 69% (22/32) had PROS, 25% (8/32) had LM, and 6% (2/32) at VeM. According to my modeling, the real-world prevalence of VAs presenting with moderate to severe symptoms for PROS is 6%, LMs is 76%, and VeMs is 17%.

This isn't too alarming. There's only so much matching you can do with 32 patients who can present with over one dozen unique types of VA. Matching enrollment to real-world prevalence is a little easier in breast cancer – patient discovery isn't a problem, and patient diversity isn't as complex.

The important thing is that zovegalisib drove excellent responses across patients with PROS and LMs, which represent 82% of the overall market opportunity. Only a single efficacy-evaluable patient (1/20) had VeM. By Week 12, that individual experienced a volume reduction of 3.5%.

This is a reminder of the unique difficulties ahead regarding study enrollment in VAs. It could also make the FDA balk at an accelerated approval if Relay Therapeutics doesn't address the enrollment imbalance before collecting dose expansion data.

‍

Tolerability profile

Zovegalisib has an unfair advantage over existing treatment options, which include non-selective PI3K-pathway inhibitors Vijoice (alpelisib, a PI3K inhibitor) and Rapamune (sirolimus, an mTOR inhibitor). The competitive landscape could shift as other pan-mutant PI3K-alpha inhibitors advance in the clinic, but for now there's a wide-open first-mover advantage.

For comparison, at the EPIK-P4 study dose of alpelisib, Novartis has previously reported an overall Grade 3+ adverse event rate of 71%. That included a 73% rate of hyperglycemia (Grade 3+ of 33%), diarrhea rate of 51% (Grade 3+ of 2%), rash rate of 42% (Grade 3+ of 21%), and stomatitis rate of 38% (Grade 3+ of 8%).

Relay Therapeutics reported a total Grade 3+ rate of just 18% at the 300 mg twice-daily rate, which could be lower at the 400 mg once-daily dose. Zovegalisib didn't trigger any cases of rash or stomatitis at any dose in the ReInspire study as of the May 2026 data readout.

‍

Efficacy

The early efficacy signals from the May 2026 data readout were very promising.

Zovegalisib drove an overall volumetric response rate (VRR) – analogous to an objective response rate (ORR) in cancer – of 60% across three dose levels. A volumetric response (VR) – analogous to a partial response (PR) in cancer – requires a 20% reduction in overgrowth volume from baseline, or the time a patient begins treatment.

The VRR was 29% (2/7) in the 100 mg dose group, 100% (6/6) in the 300 mg dose group, and 57% (4/7) in the 400 mg dose group. Three patients in the 100 mg dose group had lesion reductions of 18.2% to 19.8% -- jussst below the level required to qualify as a VR. One of these patients converted to a VR after the April 15, 2026 data cutoff. If the other two patients also convert, then the 100 mg dose group will have a VRR of 71% (5/7).

Although the 100 mg twice-daily dose level will not be advanced in the dose expansion portion of the study for patients aged 12 years and older, it might represent a future maintenance therapy dose level for chronic treatment. The strong responses and clean tolerability profile are very encouraging.

‍

These preliminary data are very early, but zovegalisib drove deep responses no matter how investors want to slice up the patient demographics. It didn't seem to matter whether patients had different types of PI3K-alpha mutations, had previously been treated with a PI3K pathway inhibitor, or had PROS or LM.

‍

News Flow & Modeling Insights

(Updated May 21, 2026)

The current model now reflects contributions from zovegalisib in 2L+ HR+/HER2- breast cancer, zovegalisib in 1L endocrine sensitive HR+/HER2- breast cancer, and zovegalisib in three major types of vascular anomalies. The modeled fair valuation is $10.885 billion or $43.54 per share, assuming 250 million shares outstanding.

Vascular anomalies

Specifically for VAs, there's a long road ahead. Securing the commercial opportunity will require separate approvals in patients aged 2 to 5 years old, 6 to 11 years old, and 12 years or older.

The May 2026 data readout includes dose escalation data in patients aged 12 years and older. The protein motion pioneer is now exploring two optimized doses in a larger number of patients aged 12 years and older. Meanwhile, dose escalation is ongoing in a separate cohort of patients aged 6 to 11 years old. That second, younger cohort will also advance into dose optimization.

Relay Therapeutics thinks dose escalation data in patients aged 6 years and older could be sufficient to earn an accelerated approval from the FDA. That means zovegalisib could be granted a conditional approval before completing a pivotal phase 3 study to confirm its clinical hypothesis that pan-mutant PI3K-alpha inhibition is a viable treatment for VAs. It would still have to conduct larger, longer studies.

Although not guaranteed and still requiring feedback from the FDA, an accelerated approval and initial market launch in VAs in 2028 is now the base case for the current model.

Consider the market opportunity for U.S. patients with moderate and severe symptoms across PROS, LMs, and VeMs. The overall market opportunity is modeled at $7.466 billion with certain assumptions about patient discovery, response rates, treatment eligibility, and time to maintenance dose.

• By anomaly type: The market opportunity for PROS is $476 million, for LMs is $5.711 billion, and for VeMs is $1.279 billion.
• By age bracket: The market opportunity for patients aged 2 to 5 years old is $219 million, aged 6 to 11 years old is $658 million, and 12 years and older is $6.589 billion.
• By symptom severity: The market opportunity for moderate VAs is $4.554 billion and for severe VAs is $2.913 billion.

The revenue per patient varies across VA type, patient age, and duration of treatment. Investors can assume every 1,000 patients on treatment would generate roughly $200 million in gross product revenue, although the net impact from discounts and rebates depends on the data package from dose expansion data and cannot be estimated at this time.

As we sit here in May 2026, an accelerated approval and launch in VAs in 2028 contributes $4.111 billion to the company's valuation.

Margin of Safety & Conviction

(Increased.)

Relay Therapeutics is considered a Future Compounder position with the following Conviction rating.

• 1 = High
• 2 = Above Average
• 3 = Average
• 4 = Below Average

The estimated fair valuation based on my current model is below:

• Market close May 22: $13.53 per share
• Modeled Fair Valuation: $43.54 per share

Relay Therapeutics is expected to have roughly 219.805 million shares outstanding in August 2026. The modeled fair valuation above assumes 250 million shares outstanding, which is equivalent to expected dilution through the launch of zovegalisib in 2L+ HR+/HER2- breast cancer in 2029.

Further Reading

• May 2026 press release announcing preliminary dose escalation data from the phase 1/2 ReInspire study from patients aged 12 years and older