Search This Blog

Tuesday, June 24, 2025

In latest AI Play, Sanofi Bets $630M+ for Formation Bio’s JAK/SYK Blocker

 

The star of Monday’s deal is gusacitinib, a small-molecule drug that Formation is developing for chronic hand eczema. Sanofi will explore additional indications for gusacitinib in a Phase I study.

Sanofi has licensed gusacitinib, an oral JAK/SYK dual inhibitor from the “AI-native” drug development company Formation Bio, marking the pharma’s latest investment in AI-focused drug development.

Terms of the deal, announced Monday, were sparse. The licensing agreement will involve upfront and milestone payments worth up to €545 million, or approximately $632 million, though the companies did not provide a specific breakdown of these numbers. Formation’s subsidiary, Libertas Bio, which is in charge of the drug’s development, will also be eligible for sales royalties in the low-to-mid-teen range.

The asset at the heart of the deal, gusacitinib, is an investigational small-molecule drug that blocks the JAK and SYK proteins, both of which are involved in immune and inflammatory signaling cascades. Formation itself acquired gusacitinib from Asana BioSciences in November 2022—and formed Libertas to take charge of its development—and has since taken the asset to Phase III studies for chronic hand eczema.

According to Monday’s announcement, however, Sanofi will chart a new path for gusacitinib and “explore its potential in a new indication not previously studied,” for which the pharma will run a Phase I trial. The companies did not reveal what this new indication is.

Sanofi and Formation are long-time partners. In October 2022, the French pharma signed a three-year contract with Formation—named TrialSpark at the time—to leverage the biotech’s “tech-enabled drug development capabilities” to advance novel therapies in “areas of high unmet patient needs,” though the companies did not reveal which indications they specifically planned to target.

Then, in May last year, Sanofi and Formation together teamed up with OpenAI to develop “AI-powered software.” In that partnership, Sanofi is providing its collaborators with access to its proprietary data, while Formation contributes engineering resources. OpenAI brings its AI capabilities to the table.

Again, the companies at the time did not give much detail: They did not provide the financial terms of the deal, nor did they reveal priority disease areas.

In recent years, Sanofi has been investing heavily in its AI capabilities. In 2022, for instance, the pharma opened an AI Center of Excellence in Toronto to “identify, develop and scale innovative solutions,” as per a press announcement at the time. A few months later, the company signed a research collaboration with the BioMed X Institute in Germany to develop a computational platform that can predict the efficacy of drug candidates.

More recently, in April of this year, Sanofi put more than $1.8 billion on the line, including a $125 million upfront commitment, in an AI-focused partnership with biologics specialist Earendil Labs. The companies will focus on advancing two bispecific antibodies for immunology indications.

https://www.biospace.com/drug-development/in-latest-ai-play-sanofi-bets-630m-for-formation-bios-jak-syk-blocker

TIGIT Turmoil: How Pharma Burned Billions on Failed Immuno-Oncology Projects

 

Leading companies spent $1.4 billion upfront on licensing deals and embarked on vast R&D programs. Clinical setbacks mean many companies are unlikely to ever recoup their investments.

TIGIT was once the hottest target in immuno-oncology. Just a few years later, the pipeline has thinned significantly as a who’s who of drugmakers have retreated from the mechanism in response to lackluster clinical data. Some R&D programs are ongoing, but the rapid retreat means the industry has already burnt billions of dollars on deals and development for failed candidates.

Bristol Myers Squibb, Gilead, GSK and Novartis paid $1.4 billion upfront between them to join Merck and Roche in the TIGIT race. The figure excludes the $175 million Gilead paid to enter into a multi-asset deal that gave it the right to license candidates including Arcus Biosciences’ domvanalimab, but includes the $275 million the big biotech handed over when it took up its option on the anti-TIGIT antibody.

The deals collectively included more than $5 billion in milestones. However, the overwhelming majority of them were unpaid when BMS, GSK and Novartis dropped their TIGIT programs. Gilead is still developing its asset and could owe Arcus up to $500 million in milestones tied to approvals in the U.S.

Even without the regulatory milestones, Gilead has bet more on TIGIT than some of its peers. The outlay includes $740 million in equity investments in Arcus. Gilead paid an average of $27.58 per share across the three purchases. Arcus currently trades below $10. The fate of domvanalimab will go a long way in determining what return, if any, Gilead gets on its investment in Arcus.

TIGIT Development Costs

TIGIT development costs are embedded in the vast R&D budgets of the biopharma companies advancing the candidates. Roche’s broad tiragolumab development program accounted for some of the almost $15 billion the Swiss drugmaker spent on R&D last year, but publicly available figures lack details of the company’s spending on TIGIT. Similarly opaque accounts obscure TIGIT spending at other companies.

Programs partnered with smaller biotechs shed some light on TIGIT development costs. GSK and iTeos Therapeutics agreed to spend $900 million on the global development plan for their TIGIT drug. The asset’s failure in Phase II means GSK and iTeos likely only spent a fraction of that amount, but the figure provides a ballpark estimate of what companies with large late-phase programs may have invested.

Roche has tested tiragolumab in 10 Phase III studies with a collective enrollment target of almost 4,600 patients. Merck’s Phase III vibostolimab program spanned five studies with a collective enrollment target of more than 4,600 patients. Gilead, Arcus and AstraZeneca, which is partnered on a domvanalimab trial, have sponsored three Phase III studies that are enrolling almost 3,000 patients among them.

If GSK and iTeos’ spending plan is indicative of R&D costs at other companies, Gilead, Merck and Roche may have spent billions of dollars between them taking their TIGIT assets through mid- and late-phase trials.

Gilead’s relationship with Arcus provides another datapoint. Arcus received between $161 million and $165 million from its collaboration partners as reimbursement for R&D costs in 2022, 2023 and 2024. Mechanisms other than TIGIT will account for some of the spending. Equally, the figures do not capture spending on Gilead-led activities, which Arcus said increased in 2024.

ITeos’ accounts provide another glimpse at the cost of TIGIT development. As of the end of March, the biotech had raised more than $1.1 billion across public and private financing rounds and its deal with GSK. ITeos, which is closing down over its TIGIT flop, had $624 million in cash and investments. The difference between the amount iTeos had raised and the amount it was holding was more than $500 million.

A History of Failed Oncology Bets

Spending on TIGIT adds to the mountain of money that drug developers have plowed into an alphabet soup of targets that at one time were seen as the next big thing in cancer immunotherapy. IL-2 attracted interest in 2018 and 2019, when BMS paid Nektar Therapeutics $1.85 billion to co-develop a molecule and Sanofi acquired Synthorx for $2.5 billion. Both programs ended in failure.

Other companies placed bets on CD47. Gilead acquired Forty Seven for $4.9 billion in 2020 and Pfizer snapped up Trillium Therapeutics for $2.3 billion the following year. Pfizer continues to list maplirpacept, a CD47 candidate acquired in the Trillium takeover, in its pipeline, but Gilead’s travails have dampened expectations. Gilead gave up on Forty Seven’s magrolimab last year after a series of setbacks.

TIM-3, 4-1BB, ICOS, OX40 and CD73 are among the other targets that have attracted heavy investment, including triple-digit upfront licensing fees, over the years. Active clinical development is continuing in some areas but the targets are yet to live up to the early excitement, such as the report that 4-1BB drug candidates were driving “pretty phenomenal responses” in cancer patients.

Work on LAG-3 delivered an approved drug, BMS’ Opdualag, but sales have been fairly modest by the standards of leading immuno-oncology products. Opdualag, a fixed-dose combination of PD-1 and LAG-3 blockers, generated sales of $928 million last year. BMS is on course to reach the blockbuster threshold this year, with sales in the first quarter coming in at $252 million.

As excitement about TIGIT and the other targets has faded, PD-1/L1xVEGF-A bispecifics have emerged as the industry’s next big hope. With China’s Akeso validating the mechanism, albeit without allaying concerns about overall survival and efficacy in global trials, leading biopharma companies have moved swiftly to secure riders in immuno-oncology’s new hottest race.

In the nine months since Akeso’s ivonescimab beat Keytruda, BioNTech, BMS, Merck and Pfizer have paid more than $4 billion upfront among them for rights to PD-1/L1xVEGF-A bispecifics. The outlay will rise by at least $2 billion as BMS makes noncontingent payments and could top $20 billion if the deals hit their milestones. The fallout from failure would make the TIGIT turmoil look like a tempest in a teapot.

https://www.biospace.com/business/tigit-turmoil-how-pharma-burned-billions-on-failed-immuno-oncology-projects

Scale AI used public Google Docs for confidential work with Meta, xAI after $14B investment: report

 The artificial intelligence start-up that recently clinched a $14 billion investment from Meta has an “incredibly janky” security system – using public Google Doc files to store confidential information on clients like Meta, Google and xAI, according to a report.

It was reported earlier this month that Meta agreed to take a 49% stake in Scale AI for $14.8 billion and bring the startup’s CEO Alexandr Wang over to lead a new “superintelligence” lab.

That shockingly steep price tag indicates that Meta believes Wang and his company are key to bringing the social media firm’s AI division to the next level.

Scale AI CEO Alexandr Wang will be leading Meta’s “superintelligence” lab as part of the deal.REUTERS

But the company has been strangely relaxed when it comes to its work with high-profile clients, leaving top-secret projects and sensitive information like email addresses and pay details in Google Docs accessible to anyone with a link, according to Business Insider.

“We are conducting a thorough investigation and have disabled any user’s ability to publicly share documents from Scale-managed systems,” a Scale AI spokesperson told The Post. 

“We remain committed to robust technical and policy safeguards to protect confidential information and are always working to strengthen our practices.”

While there is no indication the public files have led to a breach, they could leave the company susceptible to hacks, according to cybersecurity experts.

Google and xAI did not immediately respond to The Post’s requests for comment. Meta declined to comment.

Five current and former Scale AI contractors told BI that the use of Google Docs was widespread across the company.

“The whole Google Docs system always seemed incredibly janky,” one worker said.

BI said it was able to view thousands of pages of project documents across 85 Google Docs detailing Scale AI’s sensitive work with Big Tech clients, like how Google used OpenAI’s ChatGPT to fine-tune its own chatbot.

At least seven Google manuals marked “confidential” including recommendations to improve the chatbot, then-called Bard, were left accessible to the public, according to the report.

Public Google Doc files included details on Elon Musk’s “Project Xylophone,” like training documents with 700 conversation prompts to improve an AI chatbot’s conversation skills, the report said.

So-called “confidential” Meta training documents with audio clips of “good” and “bad” speech prompts to train its AI products were also left public.

While these secret projects were often given codenames, several Scale AI contractors said it was still easy to figure out which client they were working for.

Some documents tied to codenamed projects even mistakenly included the company’s logo, like a presentation that included a Google logo, according to BI.

When working with AI products, sometimes the chatbot would simply reveal the client when asked, the contractors said.

There were also publicly available Google Doc spreadsheets that listed the names and private email addresses of thousands of workers, the news outlet found.

One spreadsheet was frankly titled “Good and Bad Folks” and labelled dozens of workers as either “high quality” or “cheating,” BI said. Another document flagged workers with “suspicious behavior.”

Still more public documents laid out how much individual contractors were paid, and included detailed notes on pay disputes and discrepancies, according to the report.

https://nypost.com/2025/06/24/tech/scale-ai-used-public-google-docs-for-confidential-work-with-meta-xai-in-stunning-revelation-after-14b-investment-report/

Deutsche Bank outlines ’Pennsylvania Plan’ for US

 Deutsche Bank (ETR:DBKGn)’s Head of FX Research George Saravelos has introduced a new concept called the "Pennsylvania Plan" to address America’s twin deficit problem.

The plan aims to find new buyers for US debt as America faces what Saravelos describes as "existential macroeconomic constraints" due to large fiscal deficits combined with a large external deficit and negative net foreign asset position.

According to Saravelos, the US recently "experienced a sudden stop in capital inflow" which forced a policy reversal on trade. While tightening fiscal policy would be the clearest solution, he notes there is no political willingness to do so.

The proposed Pennsylvania Plan has two main components. First, it acknowledges the need to reduce reliance on foreign buyers of Treasuries, who currently hold record-high exposure to US sovereign duration risk. This includes promoting dollar stablecoins backed by short-dated Treasury bills to accommodate shifting demand preferences.

Second, the plan calls for increasing domestic absorption of US duration risk through financial incentives and potential financial repression. This includes regulatory carve-outs of US Treasuries from banks’ supplementary leverage ratio requirements, tax advantages for owning long-dated treasuries, and possibly mandating greater Treasury buying by retirement plans.

Saravelos explains the core aim is "to engineer a historic rotation of US duration risk from external investors to the domestic sector." While this won’t solve the underlying twin deficit problem, it could buy time by deploying domestic US savings.

The plan would likely put upward pressure on US term premia and increase fiscal dominance threats through heightened financial stability risks. It would also make foreign capital more sensitive to Federal Reserve policy, potentially leading to dollar weakness.

Saravelos concludes that without willingness to improve the US fiscal position, the path of least resistance is for the administration to seek greater domestic funding, resulting in "a weaker dollar, upward pressure on term premium and a persistent pressure on the Fed to stay easy."

https://ng.investing.com/news/economy-news/forget-the-maralago-accord-deutsche-bank-outlines-pennsylvania-plan-for-us-1980125

Jewish New Yorkers furious over bizarre pro-Mamdani texts evoking Oct. 7 attacks: ‘Disgusted’

 Jewish New Yorkers were “stunned” and “disgusted” in the run-up to Tuesday’s mayoral primary by bizarre pro-Zohran Mamdani text messages claiming to come from a friend of a slain Oct. 7, 2023 terror attack victim.

The shady sender told a registered Democratic voter — who shared the texts with The Post — that they were an Orthodox Jew supporting Mamdani, a vocal critic of Israel and its war in Gaza.

“I want to go a bit off script for a second because I’m Jewish, wear a kippah every day, and I support Zohran,” the text said.

Jewish New Yorkers were “stunned” and “disgusted” after an individual shared text messages in favor of voting for Zohran Mamdani allegedly from someone who lost a friend in the horrific Oct. 7, 2023, terror attack.LP Media

“My friend Idan was murdered on October 7 at the Re’im music festival. The war is personal to me, as it is to many in our community. Idan believed in peace, just like Zohran does,” the June 12 message continued, invoking the memory of a murdered Israeli.

“He believed everyone should have access to affordable housing, good-paying jobs, and a life free of discrimination, just like Zohran,” it said.

The gobsmacking messages came after the Jewish voter responded to an initial text tagged with “Zohran for NYC” by arguing that Mamdani wouldn’t receive any votes from Jewish New Yorkers like them.

The voter said they were “stunned” and “disgusted” by the texts seemingly trying to exploit the memory of murdered Jewish people to drum up votes for Mamdani — who appeared to be running neck-and-neck with former Gov. Andrew Cuomo in the heated Democratic primary.

“It’s trying to normalize that other Jews are voting for him, that he’s for the Jewish people – even though everything he says is the opposite of that,” said the voter, an Upper East Side Democrat who backed Cuomo and Whitney Tilson in the ranked-choice contest.

The unknown sender claimed to be Jewish — and to have had a friend murdered on Oct. 7.

Who sent the texts and whether they were a dirty political ploy launched by one of Mamdani’s campaign rivals was unclear — as was whether they’re from a text bot or a volunteer.

It also wasn’t known how many New York City voters had received the messages.

Another New Yorker who received a text claiming to be from the Mamdani campaign answered the political pitch with the same response, explaining that he’s Jewish and received the exact same initial talking point in a subsequent message, as seen by The Post.

Mamdani has supported the boycott, divest and sanctions movement against Israel, pledged to arrest Israeli Prime Minister Benjamin Netanyahu if he sets foot in New York City and recently refused to denounce the “globalize the intifada” rallying cry that many view as a call to violence against Jewish.

He has repeatedly contended he’s not antisemitic.

A Jewish watchdog group blasted the pro-Mamdani texts.

The unknown sender said to a registered Democratic voter that they are an Orthodox Jew supporting Mamdani, who has been seen as a vocal critic of the Israel-Hamas war.
The voter who received the text messages was “disgusted” by the texts he received and opted to vote for Andrew Cuomo and Whitney Tilson in the ranked-choice voting system.IDF
“It’s unthinkable that New York City, with one of the largest Jewish communities anywhere outside of Israel, would be led by an antisemite like Zohran Mamdani,” said Liora Rez, founder and Executive Director of StopAntisemitism.

“Many Jew-haters pretend to care about marginalized communities while targeting the most marginalized community in history, the Jews.”

The idea of Jewish voters backing Mamdani was also “baffling” for Ami Horowitz, a conservative filmmaker in New York City, who characterized the text chain as “an FU to Jews.”

Mamdani’s campaign did not respond to The Post’s requests for comment.

https://nypost.com/2025/06/24/us-news/jewish-new-yorkers-furious-over-bizarre-pro-zohran-mamdani-texts-evoking-oct-7-attacks-disgusted/