The Rise and Fall of Anti‑TIGIT Cancer Therapies

A detailed look at why dozens of billion‑dollar anti‑TIGIT programs collapsed, what the science behind the target was, and how the industry’s collective bet turned into a costly lesson for immuno‑oncology.

The promise that sparked a multi‑billion dollar race

TIGIT (T cell immunoreceptor with Ig and ITIM domains) was identified in the early 2010s as a second checkpoint protein that tumors could exploit to suppress T‑cell activity. The success of PD‑1 blockers such as Keytruda made any new checkpoint an instant headline. Early pre‑clinical work suggested that blocking TIGIT could release two inhibitory pathways and activate an additional co‑stimulatory signal, a combination that looked too good to ignore.

Roche was first to act. After publishing a 2014 paper on TIGIT’s role in CD8⁺ T‑cell function, the company launched tiragolumab, an IgG1 antibody that retained an active Fc region. A phase‑2 trial in metastatic non‑small‑cell lung cancer (NSCLC) reported a 31 % response rate versus 16 % for the control arm, prompting the FDA to grant breakthrough therapy designation in January 2021. The headline numbers convinced investors and executives that the next blockbuster checkpoint was on the horizon.

Funding and partnership waterfall

Company	Anti‑TIGIT candidate	Funding / deal size	Notable partners
Roche	tiragolumab	Billions spent across a 12‑trial “SKYSCRAPER” program (est. >$2 B)	None (in‑house)
Merck	vibostolimab	$1.2 B in internal R&D, plus $250 M for parallel trials	None
BeiGene	ociperlimab	$300 M upfront from Novartis (option later returned)	Novartis (2021‑2023)
Gilead / Arcus	domvanalimab (Fc‑silent)	$175 M upfront + $200 M equity (2020)	Gilead (co‑development)
iTeos / GSK	belrestotug (Fc‑silent)	$625 M upfront, up to $1.45 B milestones (2021)	GSK (2021‑2025)
AstraZeneca	rilvegostomig (Fc‑silent)	$1.1 B allocated to 11 phase‑3 studies (2023‑2026)	None

These figures illustrate how the market positioned anti‑TIGIT as the next “immune brake” to hit. The sheer number of concurrent trials—often twelve per company—was designed to lock in a lead indication before competitors could copy the approach.

The clinical reality check

Roche’s cascade of setbacks

SKYSCRAPER‑01 (first‑line NSCLC) – missed primary PFS endpoint; OS trend later proved non‑significant (Nov 2024).
SKYSCRAPER‑02 (small‑cell lung cancer) – failed PFS in May 2022.
SKYSCRAPER‑06 (interim analysis) – showed worse OS for tiragolumab arm (July 2024).
SKYSCRAPER‑08 (esophageal squamous‑cell carcinoma) – only trial with a statistically significant OS benefit, but published after Roche had already removed tiragolumab from its pipeline (early 2026).

Merck’s parallel collapse

Melanoma trial – halted May 2023 due to severe immune‑related adverse events.
Small‑cell lung cancer – stopped Aug 2024 for OS futility.
Multiple NSCLC studies – abandoned Dec 2024 and throughout 2025, culminating in a full program termination.

The Fc‑silent hypothesis tested and rejected

Both Arcus/Gilead (domvanalimab) and iTeos/GSK (belrestotug) pursued antibodies engineered to silence Fc‑mediated killing, hoping to avoid depletion of TIGIT‑expressing T‑cells. Phase‑3 trials in upper‑GI and other solid tumors failed to meet primary endpoints, and the programs were shuttered by the end of 2025.

AstraZeneca’s lone stand

Rilvegostomig remains in eleven phase‑3 studies, but the underlying premise—Fc silence alone will rescue efficacy—has yet to be proven. Analysts note that without a clear mechanistic breakthrough, the drug faces the same uphill battle as its predecessors.

Why did the biology not translate?

Dual expression dilemma – TIGIT is present on both suppressive myeloid cells and on activated cytotoxic T‑cells. An Fc‑active antibody can trigger ADCC against the very cells it aims to empower, potentially nullifying any net benefit.
Context‑dependent signaling – The tumor micro‑environment varies dramatically across cancer types. A checkpoint that is dominant in one setting may be irrelevant in another, making a one‑size‑fits‑all trial strategy risky.
Insufficient patient stratification – Most studies enrolled broad populations. If TIGIT blockade only helps a narrow genetic or immunologic subset, the signal gets drowned out in large, heterogeneous cohorts.
Trial design pressure – The “herding” effect described in a 2026 BMJ Oncology analysis (≈49 000 patients, >$3 B spent) meant companies rushed to replicate each other’s protocols rather than explore alternative dosing, combination, or biomarker strategies.

Lessons for the biotech ecosystem

Checkpoint hype can outpace mechanistic certainty. The early PD‑1 success created a bias toward assuming any new inhibitory receptor will behave similarly.
Parallel large‑scale programs amplify risk. When multiple firms chase the same target, the collective loss multiplies, as seen with the anti‑TIGIT wave.
Engineering the Fc region is not a panacea. While Fc‑silencing removes one source of toxicity, it does not guarantee therapeutic benefit if the target itself is not a driver of disease.
Adaptive trial designs could salvage value. Smaller, biomarker‑driven studies might have identified a responsive niche before committing billions to blanket phase‑3s.

Where does the story go from here?

AstraZeneca’s rilvegostomig will be the next data point to watch. If it finally demonstrates a clear OS advantage, the field may revive with a more nuanced view of TIGIT as a context‑specific modulator rather than a universal brake. If not, the anti‑TIGIT chapter will close, leaving behind a cautionary tale about how even well‑grounded science can falter when scaled without sufficient biological granularity.

For further reading: