Faeth Bet $92 Million That Cancer’s Escape Route Can Be Blocked

The most useful way to understand Faeth Therapeutics is as a company built around a detour. Cancer cells are irritatingly good at finding one. Block a growth signal at one point and the biology can turn up another signal, mutate the target or activate feedback loops. The PI3K/AKT/mTOR pathway - mercifully shortened to PAM - has produced years of attractive diagrams and uneven medicines. Approved single-node therapies can help selected patients, but efficacy, resistance and toxicity have limited how broadly the class is used.

Faeth’s answer is not a new molecule. Its lead product, PIKTOR, combines two existing investigational oral medicines: serabelisib, which selectively inhibits PI3K-alpha, and sapanisertib, which inhibits mTORC1 and mTORC2. The bet is that simultaneous pressure upstream and downstream can suppress the pathway more completely without requiring an intolerable dose at any one node. It is vertical integration, but for cell signaling.

That sounds clean. Biology rarely is. PIKTOR is not approved, its safety and efficacy are not established, and the decisive evidence still lies ahead. Yet Faeth has managed to turn the thesis into two enrolling studies, FDA Fast Track designation in one endometrial-cancer setting, a Nasdaq listing and a cash balance large enough to run the experiment.

The meal-kit era was real

Faeth began in 2019 with a broader and stranger proposition. Independent research groups led by cancer biologists including Lewis Cantley, Greg Hannon and Karen Vousden had converged on the role of metabolism and nutrient availability in tumor growth. Anand Parikh, a former corporate lawyer, and cancer-metabolism researcher Oliver Maddocks built a company around that convergence with an unusually decorated scientific founding bench: Siddhartha Mukherjee, Scott Lowe, Marcus Goncalves, Benjamin Hopkins and Simon Knott were among those attached to the project.

The original product story paired drugs with precision nutrition and digital support. Faeth developed metabolically engineered meals intended to restrict nutrients a particular tumor depended on while supporting the patient. Parikh once offered a wonderfully concrete analogy: imagine HelloFresh designed to starve a tumor and make a drug work better. A machine-learning and functional-genomics platform called MetabOS mapped possible vulnerabilities by tumor genotype, organ and treatment. The company also studied non-essential amino acid restriction through a program called NEAAR.

That positioning solved an overlooked practical problem. Cancer patients routinely ask what they should eat, while rigorous tumor-specific dietary evidence remains thin. Faeth wanted nutrition prescribed with the precision and accountability of a medicine, not dispensed as internet folklore. Its immediate users were trial patients and oncologists; its theoretical customers were cancer centers, payers and patients receiving a reimbursed regimen.

But the current Faeth tells a more concentrated story. The public pipeline emphasizes PIKTOR in endometrial and breast cancer, plus exploratory sapanisertib work. MetabOS and NEAAR remain meaningful parts of the company’s history, but they are not the headline programs on today’s pipeline page. The mission did not abandon metabolism. The commercial object became a drug combination that regulators and oncologists already know how to evaluate.

What failed first

The first failure in this story belongs to the field, not neatly to Faeth. Single-node inhibition of the PAM pathway has repeatedly met feedback-driven resistance and dose-limiting toxicity. Hit PI3K and signaling can reappear downstream. Push harder and side effects can narrow the therapeutic window. In Faeth’s framing, the tumor learns the road closure before the drug can safely close every lane.

What changed the company’s level of conviction was a small investigator-initiated Phase 1b study. Nineteen heavily pretreated patients with advanced endometrial, breast or ovarian cancer enrolled; 15 were evaluable for response. PIKTOR with paclitaxel produced a reported 47 percent overall response rate. Among seven evaluable patients whose tumors had PAM-pathway alterations, five responded, including three complete responses across the response-evaluable population. Two endometrial-cancer complete responders had reported progression-free survival of 26.9 and 20 months.

These numbers are intriguing, not conclusive. The study was small, single-center, open-label and lacked a randomized control arm. Patients had different tumor types. Response-evaluable analysis is also narrower than intention-to-treat analysis. Faeth’s Phase 2 endometrial study exists precisely because early signals often get smaller when tested across more patients and sites.

What exactly they did - and what it cost

Faeth assembled PIKTOR through licensing and asset purchases rather than inventing both compounds from scratch. It acquired serabelisib-related rights and data in 2021 and obtained sapanisertib assets in 2023 through agreements connected to Ravenna Pharmaceuticals, Calithera Biosciences, Takeda, Millennium Takeda, Cornell and the University of California. One disclosed Ravenna transaction carried a $500,000 upfront fee. The less photogenic part of the bargain is a large contingent tab: company filings describe potential development, regulatory, launch and sales milestones running into hundreds of millions of dollars, plus royalties and license fees.

Capital arrived in stages. Faeth raised a $20 million seed round in January 2022, a $47 million Series A that June and another $25 million in October 2025 - $92 million privately. In February 2026, Sensei Biotherapeutics acquired Faeth in a reverse-merger transaction accompanied by a $200 million private placement. Faeth shareholders owned roughly 40.6 percent of the company immediately after the acquisition, before the preferred-stock conversion described in filings. In June, the combined company adopted Faeth’s name, put Parikh in the CEO chair and began trading as FTH.

This is Faeth’s business model in its honest form: spend investor capital on trials, manufacturing and regulatory work now; earn product revenue only if a candidate clears clinical and regulatory hurdles later. There are no approved products and therefore no disclosed commercial customers. The people receiving PIKTOR today are clinical-trial participants under investigator supervision, not app subscribers ordering an optimized supper.

A sharper place in the market

Faeth sits among oncology companies trying to make a heavily studied pathway more useful. Approved alternatives attack portions of it: alpelisib and inavolisib inhibit PI3K-alpha, capivasertib inhibits AKT, and several drugs inhibit mTOR. PIKTOR’s differentiation is the combination of selective PI3K-alpha inhibition with dual mTORC1/2 inhibition, all orally administered. In the endometrial study it is added to paclitaxel; in advanced HR-positive, HER2-negative breast cancer it is being studied with fulvestrant and potentially other therapies.

The endometrial trial selects patients whose tumors carry PAM-pathway alterations and whose disease advanced after platinum chemotherapy and an immune checkpoint inhibitor. The breast study targets a setting where PI3K activation can contribute to resistance after CDK4/6 inhibitors and endocrine therapy. That selection is not administrative trivia. It is Faeth’s attempt to place the combination where its mechanism should matter most.

FDA Fast Track designation, granted in August 2026 for PIKTOR plus paclitaxel in the specified endometrial population, may enable more frequent regulatory interaction and rolling review if future requirements are met. It is not an approval, a safety endorsement or proof that the regimen works. Faeth expects endometrial topline data by the end of 2026 and initial breast-cancer safety and efficacy data in 2027.

The financing move is copyable only under narrower conditions. A reverse merger can give a private biotech a listing and cash more quickly than a conventional IPO, but Faeth brought a clinical asset, human data, institutional investors and a public-company partner. A slide deck with a pathway diagram will not produce the same outcome. Nor does public capital fix weak biology; it simply finances a more expensive answer.

When the playbook breaks

There are four straightforward ways Faeth’s plan may not work:

• Tolerability. Blocking three nodes may still produce overlapping toxicities or doses too low for durable suppression.
• Signal shrinkage. The response rate from 15 evaluable patients may not reproduce in a larger, multicenter population.
• Wrong selector. PAM-pathway alterations may not enrich for benefit as cleanly as the early subgroup suggests.
• Another detour. Tumors may activate biology outside the pathway, making even multi-node blockade temporary.

Faeth’s culture language is unusually aware of that uncertainty. The company says it wants operators who have “built before, failed before, and know the difference,” and people who find constraints interesting. It is good copy because it describes the actual job. The company’s founding idea has already been narrowed, its assets carry complicated obligations and its lead evidence is preliminary. The work now is less about selling a new pillar of cancer care than about producing a clean result.

That focus is Faeth’s most transferable achievement. The company moved from an expansive story about food, software and drugs to a testable claim about two pills and three nodes. If the Phase 2 data hold, the old metabolism thesis will have found a conventional clinical vehicle. If they do not, Faeth will still have demonstrated the useful discipline of turning a grand theory into a number that can disappoint you.