Relay ‘full steam ahead’ on PI3K inhibitor fueled by first phase 2 data in blood vessel disorder

Relay Therapeutics’ R&D chief remains convinced the biotech’s PI3Kα inhibitor will carve out its own niche after reporting a slice of phase 2 data in patients with a type of blood vessel disorder.

The company has been evaluating various doses of the candidate, dubbed zovegalisib, in an ongoing phase 2 trial of patients with PIK3CA-driven vascular anomalies (VA)—a group of rare disorders that involve abnormal development of blood vessels, lymphatic vessels and the surrounding tissues.

VA affects around 170,000 people in the U.S., according to Relay, which said the condition can cause bleeding, lesions on the windpipe that may require a tracheostomy, difficulty walking or difficulty using arms and hands.

The phase 2 study is spread across three groups, split by age, and this morning’s data come from the group enrolling adults and children aged 12 years and older. As of an April 15 cut-off, 32 patients had been enrolled in this group, of whom 20 had already had their lesion response to the treatment assessed by MRI scan after 12 weeks.

Among these 20 patients, 60% demonstrated a volumetric response at their first MRI scan, the company said in its May 19 release. Even looking solely at the 13 patients who received the two lower doses, 100 mg and 300 mg, the response rate was 62%.

Meanwhile, a total of 95% of the 20 patients experienced some form of lesion reduction, according to Relay.

The study’s secondary endpoints include both investigator- and patient-reported impressions of symptom change, as well as an investigator-assessed measure of pain reduction. At Week 12, these measurements saw improvements of 89%, 79% and 71%, respectively.

Relay’s President of R&D, Don Bergstrom, M.D., Ph.D., hailed the results in a statement this morning as demonstrating “for the first time, the promise of PI3Kα mutant-selective inhibition for patients with vascular anomalies.”

“The combination of robust volumetric responses, symptomatic improvement and a safety profile that supports chronic dosing underscores the potential of zovegalisib to meaningfully change the treatment paradigm for this underserved population,” Bergstrom added in the release.

The diverse symptoms of VA have often made it difficult for physicians to diagnose, but in the past decade, researchers have connected the underlying biology to mutations of the PI3Kα protein.

“If you could potentially turn that mutated protein off, you might be able to effectively treat these patients who haven't had any type of systemic therapy to date,” Bergstrom told Fierce in an interview. 

“What we've shown for the first time is that if you design an exquisitely selective inhibitor of mutated PI3Kα, you can get efficacy much greater than has been seen in the past,” he added.

Treating PI3Kα mutations has historically been difficult because of toxicity, but it was one of the first targets Relay pursued when the company launched 10 years ago, because of the critical role it plays in breast cancer. PI3Kα is also important for maintaining skin integrity and regulating glucose, so inhibiting PI3Kα historically resulted in toxicity that limited effective treatment development.

Novartis has marketed its PI3K inhibitor Piqray in combination with AstraZeneca’s Faslodex for breast cancer since 2019, although hyperglycemia and other adverse events are a known problem for Piqray. In fact, one study of Piqray found that discontinuation due to adverse events (AEs) occurred in 25% of patients, with hyperglycemia as the most common cause.

Since then, Roche has launched Itovebi, while Eli Lilly axed LOXO-783 before returning to the PI3K space last year via the acquisition of Scorpion Therapeutics.

Of the three doses of zovegalisib that Relay has been assessing in its phase 2 VA study, the biotech said it has scrapped further work on the highest, 400 mg, twice-daily dose for this indication after it showed a “safety profile that was not optimal for this patient population.”

Instead, a 400 mg dose administered once a day is one of the options for an expansion portion of the trial.

Among the 22 patients who received a 100 mg or 300 mg dose twice a day, only two patients have so far experienced a treatment-related AE of grade 3 or above. The most common AEs were “low-grade, manageable and reversible,” said Relay, which noted that “no grade 3 hyperglycemia or diarrhea was observed” and no patients discontinued their treatment due to an AE.

Bergstrom pointed out to Fierce that zovegalisib appears to avoid negative side effects and that its safety profile “is unlike anything that's been seen in this space before.” 

“We would argue [patients] have no highly effective and safe drugs available for them today, but we think zovegalisib has the potential to change that,” he added.

The VA data come at the end of a rocky road for Relay, which laid off about 70 employees back in April 2025. That marked the biotech's third round of cuts in less than a year and represented a 75% reduction in Relay's annual research budget. 

But the Massachusetts-based biotech’s ambitions for zovegalisib remain undimmed, with a phase 3 study ongoing for a 400 mg dose alongside Faslodex for patients with PI3Kα-mutated, CDK4/6 pre-treated, HR+/HER2- advanced breast cancer. The company is also aiming to initiate another phase 3 breast cancer trial next year for zovegalisib alongside Pfizer’s investigational CDK inhibitor atirmociclib and an aromatase inhibitor.

Relay faces some competition in the form of Celcuity’s gedatolisib. The pan-PI3K/mTOR inhibitor—which was licensed from Pfizer—was shown to significantly extend progression-free survival when given alongside Faslodex compared to Faslodex alone. An FDA approval decision is expected in July, and could pave the way for one of the most highly anticipated drug launches of the year.

Despite the potential competition and dual indication strategy, Relay has no problem maintaining zovegalisib’s momentum in both breast cancer and VA at the same time, Bergstrom told Fierce.

“We're moving full steam ahead in both directions,” he said. “We've built the development infrastructure to be able to move these forward in both the breast cancer and vascular anomaly indications as quickly as possible.”