Belrestotug (EOS-448): A Clinical and Scientific Post-Mortem of a Terminated Anti-TIGIT Immunotherapy

Section 1: Executive Summary

Belrestotug, also identified by its development codes EOS-448 and GSK4428859A, is a human immunoglobulin G1 ($IgG1$) monoclonal antibody developed to target the T-cell immunoreceptor with immunoglobulin and immunoreceptor tyrosine-based inhibitory motif domains (TIGIT).[1] Co-developed by iTeos Therapeutics and GlaxoSmithKline (GSK), Belrestotug was positioned as a next-generation immuno-oncology therapeutic designed to enhance anti-tumor immunity, particularly in patients with cancers resistant to existing programmed death receptor-1 (PD-1) or its ligand (PD-L1) inhibitors.[4]

The clinical development strategy for Belrestotug centered on its combination with dostarlimab (Jemperli), an anti-PD-1 antibody, in first-line treatment settings for cancers with high PD-L1 expression.[6] The primary target indications were non-small cell lung cancer (NSCLC) with a PD-L1 tumor proportion score (TPS) of $\ge50\%$ and head and neck squamous cell carcinoma (HNSCC) with a PD-L1 combined positive score (CPS) of $\ge1$.[8] This strategy was predicated on the hypothesis that dual blockade of the TIGIT and PD-1 pathways would produce synergistic anti-tumor effects and overcome mechanisms of immune escape.

The program's pivotal trial, the Phase 2 GALAXIES Lung-201 study, yielded a paradoxical and ultimately fatal set of results. While the combination of Belrestotug and dostarlimab demonstrated a statistically significant and clinically meaningful improvement in the study's primary endpoint of Objective Response Rate (ORR), it failed to show a corresponding benefit in the key secondary endpoint of Progression-Free Survival (PFS).[9] This discordance between initial tumor shrinkage and durable disease control, coupled with a substantially more toxic safety profile for the combination regimen, undermined the therapeutic rationale.

In May 2025, GSK and iTeos Therapeutics jointly announced the termination of the entire Belrestotug development program.[9] The decision was based on the totality of the data, including the lack of PFS benefit in the NSCLC trial and an unfavorable trend in ORR from an interim analysis of the GALAXIES H&N-202 trial in HNSCC, which failed to meet predefined efficacy criteria for continued development.[10]

This report provides a comprehensive analysis of the Belrestotug program, from its scientific foundation to its clinical demise. The failure of Belrestotug was not due to a lack of biological activity but rather to an inability to translate potent, initial anti-tumor effects into durable clinical benefit, compounded by an unfavorable risk-benefit profile. Its discontinuation adds to a growing list of late-stage failures for the TIGIT inhibitor class, prompting a critical re-evaluation of the therapeutic potential of this target and the prevailing strategies for developing next-generation immuno-oncology combinations.

Section 2: The TIGIT/CD226 Axis: Scientific Rationale for Belrestotug

The therapeutic hypothesis for Belrestotug is rooted in the complex interplay between the inhibitory TIGIT receptor and the activating CD226 receptor, a critical axis that governs the function of key anti-tumor immune cells.

2.1 TIGIT as an Immune Checkpoint

TIGIT is a co-inhibitory receptor and immune checkpoint that belongs to the immunoglobulin superfamily.[1] It is primarily expressed on the surface of activated immune cells critical for cancer immunosurveillance, including CD8+ cytotoxic T-lymphocytes, CD4+ helper T-cells, regulatory T-cells (Tregs), and Natural Killer (NK) cells.[1] The expression of TIGIT is often upregulated on tumor-infiltrating lymphocytes (TILs), particularly those that are exhausted and dysfunctional.[1] Engagement of TIGIT by its ligands initiates an intracellular signaling cascade through its immunoreceptor tyrosine-based inhibitory motif (ITIM) domain, leading to the suppression of T-cell and NK-cell proliferation, cytokine production, and cytotoxic activity.[1] Tumors exploit this pathway as a mechanism of immune evasion. Furthermore, TIGIT is frequently co-expressed with other major inhibitory receptors, most notably PD-1, on the most dysfunctional T-cells within the tumor microenvironment, providing a strong biological rationale for dual therapeutic blockade.[13]

2.2 The Competing CD226 Pathway

The TIGIT pathway does not operate in isolation but is part of a dynamic signaling nexus centered on a shared set of ligands. TIGIT directly competes with a co-stimulatory receptor, CD226 (also known as DNAX Accessory molecule-1 or DNAM-1), for binding to the ligands CD155 (also known as the poliovirus receptor, PVR) and CD112 (Nectin-2).[1] These ligands are often highly expressed on the surface of cancer cells.

The balance of signaling through this axis dictates the immune response. When CD226, expressed on NK cells and CD8+ T-cells, binds to CD155 or CD112, it delivers a potent activating signal that promotes immune-mediated tumor cell killing.[1] However, TIGIT binds to CD155 with a much higher affinity than CD226, allowing it to effectively outcompete the activating receptor and deliver a dominant inhibitory signal.[14] The therapeutic goal of an antagonistic anti-TIGIT antibody, therefore, is to block the TIGIT-ligand interaction, thereby shifting the competitive balance in favor of CD226. This is intended to "release the brakes" on the immune system and restore the potent, CD226-mediated anti-tumor response.[1]

2.3 Belrestotug's Dual Mechanism of Action

Belrestotug was specifically engineered as a high-affinity, antagonistic human $IgG1$ monoclonal antibody to leverage a multifaceted mechanism of action designed to maximize immune restoration.[2]

First, through checkpoint blockade, Belrestotug binds to TIGIT with high affinity (affinity constant of $6.401 \times 10^{-10}$ M), physically obstructing its interaction with CD155 and CD112.[2] This blockade is designed to prevent the transmission of inhibitory signals and, critically, to enhance the interaction of CD155 and CD112 with the CD226 receptor. This leads to CD226 dimerization and signaling, activating a T-cell-mediated immune response against cancer cells.[1]

Second, Belrestotug was designed with a fully functional $IgG1$ Fc region to enable Fc-gamma receptor (Fc$\gamma$R) engagement. This "Fc-active" design was a key differentiating feature, intended to mediate additional anti-tumor effects.[15] By binding to Fc$\gamma$Rs on innate immune cells like NK cells and macrophages, Belrestotug was hypothesized to trigger Antibody-Dependent Cellular Cytotoxicity (ADCC). This mechanism would lead to the preferential depletion of cells expressing high levels of TIGIT. Within the tumor microenvironment, the cells with the highest TIGIT expression are often the profoundly immunosuppressive Tregs and terminally exhausted T-cells.[1] The depletion of these cell populations was expected to further shift the balance toward an effective anti-tumor immune response.[1]

However, this Fc-active design may represent a double-edged sword. The potent pro-inflammatory signaling and cell depletion mediated by Fc$\gamma$R engagement, while intended to enhance efficacy, may have also driven the significantly increased systemic immunotoxicity observed in clinical trials. The high rates of immune-related adverse events (irAEs) and subsequent treatment discontinuations seen with the Belrestotug-dostarlimab combination suggest that the very mechanism designed to be a key advantage may have become a critical liability.[6] This potent, systemic immune activation may have contributed to an unfavorable risk-benefit profile, ultimately preventing the observation of a durable clinical benefit like PFS by forcing a high proportion of patients off treatment prematurely. This outcome raises a fundamental question for the TIGIT inhibitor class regarding the optimal molecular format: whether a potent, depleting antibody is superior to an Fc-silent format that provides pure checkpoint blockade without broad inflammatory effects.

Table 1: Key Molecular and Pharmacological Properties of Belrestotug

Property	Description	Source Snippet(s)
Generic Name	Belrestotug	[1, 17]
Synonyms/Code Names	EOS-448, EOS884448, GSK4428859A	[1, 2, 3, 18]
Drug Class	Antineoplastics, Immunotherapies, Monoclonal antibodies	[3]
Target	TIGIT (T-cell immunoreceptor with Ig and ITIM domains)	[1, 18]
Mechanism of Action	TIGIT protein inhibitor, T lymphocyte stimulant, Antibody-dependent cell cytotoxicity (ADCC)	[3]
Molecular Format	Human Immunoglobulin G1 ($IgG1$) kappa	[1, 2]
Binding Affinity ($K_D$)	$6.401 \times 10^{-10}$ M (as determined in BLI assay)	[2]
Originator/Developer	iTeos Therapeutics / GSK	[3, 4]

Section 3: The Belrestotug Clinical Development Program

The clinical development of Belrestotug was ambitious, employing a modern platform trial design to rapidly evaluate its potential in combination with PD-1 blockade in patient populations with a high unmet medical need.

3.1 Strategic Rationale and Target Indications

The overarching clinical strategy was to combine Belrestotug with the anti-PD-1 antibody dostarlimab to overcome resistance to PD-1 inhibitor monotherapy, which remains the standard of care for many patients with high PD-L1 expression but has a limited response rate.[6] The program specifically targeted two first-line settings:

1. Non-Small Cell Lung Cancer (NSCLC): Patients with previously untreated, advanced or metastatic NSCLC whose tumors exhibited high PD-L1 expression, defined as a Tumor Proportion Score (TPS) of $\ge50\%$, and who lacked targetable driver mutations such as in EGFR or ALK.[7]
2. Head and Neck Squamous Cell Carcinoma (HNSCC): Patients with recurrent or metastatic HNSCC whose tumors were PD-L1-positive, defined by a Combined Positive Score (CPS) of $\ge1$.[9]

The rationale was that in these PD-L1-high tumors, TIGIT expression on TILs represents a key parallel mechanism of immune suppression, and that dual blockade would lead to deeper and more durable responses than could be achieved with PD-1 inhibition alone.[6]

3.2 The GALAXIES Platform Study Design

To accelerate development, GSK and iTeos utilized an innovative "platform study" master protocol named GALAXIES.[6] This Phase 2, randomized, open-label design allowed for the simultaneous and efficient evaluation of multiple novel immunotherapy combinations against a shared control arm within a single trial infrastructure. This approach was intended to expedite decision-making and identify promising regimens for advancement into later-stage trials.[21]

3.3 Key Clinical Trials

The Belrestotug program was anchored by three key studies, two in Phase 2 and one that was initiated as a registrational Phase 3 trial before being halted.

• GALAXIES Lung-201 (NCT05565378): This was the cornerstone Phase 2 study that ultimately determined the fate of the program. It was designed to evaluate the efficacy and safety of Belrestotug in combination with dostarlimab in first-line, PD-L1-high NSCLC.[9] The trial included a dose-finding component, comparing three different doses of Belrestotug (100 mg, 400 mg, and 1000 mg) administered intravenously every three weeks, each in combination with dostarlimab, against a control arm of dostarlimab monotherapy.[6] The primary objective was to assess the ORR and select an optimal dose for Phase 3 development.[6]
• GALAXIES H&N-202 (NCT06062420): This parallel Phase 2 platform study evaluated Belrestotug in first-line, PD-L1-positive HNSCC.[2] It included cohorts testing the Belrestotug-dostarlimab doublet as well as a triplet combination that added nelistotug, an investigational anti-CD96 antibody, to the doublet.[24] The failure of this trial to show a promising efficacy signal contributed to the decision to terminate the program.[9]
• GALAXIES Lung-301 (NCT06472076): Based on encouraging interim ORR data from the Lung-201 study, GSK and iTeos made the strategic decision to initiate this global, randomized, double-blind Phase 3 registrational trial.[8] The study was designed to compare Belrestotug (at the 400 mg dose selected by the European Medicines Agency) plus dostarlimab against the established standard of care, pembrolizumab plus placebo, in first-line PD-L1-high NSCLC.[2] Enrollment in this trial was active but was abruptly halted following the negative PFS data from the Phase 2 study, representing a significant strategic and financial setback.[9]

Section 4: Analysis of Clinical Efficacy from Pivotal Trials

The clinical efficacy data from the GALAXIES program presented a challenging paradox: a strong signal of initial anti-tumor activity that failed to translate into durable disease control, ultimately leading to the program's termination.

4.1 GALAXIES Lung-201: The Paradox of Response Without Progression Benefit

The results from the GALAXIES Lung-201 trial in first-line, PD-L1-high NSCLC were the primary driver of the decision to halt development. The study enrolled patients with previously untreated, unresectable, locally advanced or metastatic disease with a PD-L1 TPS $\ge50\%$ and no actionable EGFR or ALK genomic alterations.[8]

The trial successfully met its primary endpoint. The combination of Belrestotug and dostarlimab demonstrated a consistent and clinically meaningful improvement in ORR across all three dose cohorts compared to dostarlimab monotherapy.[9] As shown in Table 2, the confirmed ORR (cORR) was approximately 60% in each of the combination arms, representing an absolute improvement of over 30% compared to the 28.1% cORR observed in the control arm.[23] This robust initial signal of tumor shrinkage was compelling enough to support the initiation of the Phase 3 GALAXIES Lung-301 trial.[26]

However, a subsequent, more mature interim analysis revealed the program's fatal flaw. Despite the impressive ORR, the combination of Belrestotug and dostarlimab did not demonstrate a clinically meaningful improvement in the secondary endpoint of Progression-Free Survival (PFS) when compared to dostarlimab monotherapy.[9] This critical finding indicated that while the combination could induce tumor responses more frequently, these responses were not durable enough to delay disease progression over the medium term. Additional analyses did show that the combination was associated with a greater depth of tumor reduction in waterfall plots and a more significant decrease in circulating tumor DNA (ctDNA), particularly at the 400 mg and 1000 mg doses.[6] This suggests a potent but ultimately transient biological effect.

This discordance between ORR and PFS is a pivotal finding that challenges the utility of ORR as a primary basis for late-stage development decisions in this therapeutic class. Several hypotheses can be formulated to explain this disconnect. The immune activation induced by the doublet may be potent but short-lived, failing to establish the long-term immunological memory required for durable disease control. Alternatively, the high toxicity and subsequent high rate of treatment discontinuation may have confounded the PFS analysis, with patients ceasing therapy for reasons other than disease progression, thereby artificially shortening the observed time to progression. Regardless of the underlying cause, the Belrestotug experience serves as a cautionary tale against advancing expensive registrational trials based solely on an early ORR signal without corroborating evidence of durable benefit.

Table 2: Summary of Efficacy Outcomes in the GALAXIES Lung-201 Trial (Substudy 1A)

Efficacy Endpoint	Dostarlimab Monotherapy (n=32)	Belrestotug 100mg + Dostarlimab (n=30)	Belrestotug 400mg + Dostarlimab (n=32)	Belrestotug 1000mg + Dostarlimab (n=30)
Investigator-Assessed ORR (95% CI)	37.5% (21.1%-56.3%)	63.3% (43.9%-80.1%)	65.6% (46.8%-81.4%)	76.7% (57.7%-90.1%)
Confirmed ORR (cORR) (95% CI)	28.1% (13.7%-46.7%)	60.0% (40.6%-77.3%)	59.4% (40.6%-76.3%)	63.3% (43.9%-80.1%)
Progression-Free Survival (PFS) Outcome	Baseline	No clinically meaningful improvement vs. monotherapy	No clinically meaningful improvement vs. monotherapy	No clinically meaningful improvement vs. monotherapy
Median Follow-up (months)	7.3	7.3	7.3	7.3

Data based on a cutoff date of June 7, 2024, as presented in interim analyses.[9]

4.2 GALAXIES H&N-202: A Failure to Show a Signal

Corroborating the negative findings from the lung cancer trial, an interim analysis of the GALAXIES H&N-202 study in first-line, PD-L1-positive HNSCC also yielded disappointing results. The data showed that the ORR trends in the Belrestotug combination cohorts fell "below the threshold of clinical significance" when compared with dostarlimab monotherapy.[9] The lack of a clear efficacy signal in a second distinct tumor type provided strong evidence that the issues observed in NSCLC were not indication-specific, reinforcing the conclusion that Belrestotug lacked a viable path forward.

Section 5: Comprehensive Safety and Tolerability Profile

The clinical development of Belrestotug was significantly hampered by a challenging safety profile, particularly in combination with dostarlimab. The data from GALAXIES Lung-201 clearly demonstrated that the addition of Belrestotug led to a substantial increase in the frequency and severity of adverse events compared to PD-1 inhibitor monotherapy, ultimately creating an unfavorable risk-benefit equation.

5.1 Overview of Adverse Events

As detailed in Table 3, the incidence of Treatment-Related Adverse Events (TRAEs) was markedly higher in the combination arms. While 59% of patients in the dostarlimab monotherapy arm experienced a TRAE of any grade, this figure rose to 80-97% in the Belrestotug combination arms.[6] More critically, the rate of severe (Grade $\ge3$) TRAEs was substantially elevated, occurring in 16% of the monotherapy group versus 22-43% of patients receiving the combination therapy.[6]

5.2 Treatment-Related and Immune-Related Adverse Events (TRAEs & irAEs)

The increased toxicity was primarily driven by an amplification of immune-mediated adverse reactions. The incidence of any-grade irAEs was more than threefold higher in the combination arms (56-73%) compared to the monotherapy arm (19%).[6] Similarly, Grade $\ge3$ irAEs were observed in 16-37% of patients on the combination regimens, compared to 13% with dostarlimab alone.[6]

The most frequently reported irAEs were skin and subcutaneous tissue disorders, including immune-mediated dermatitis, pruritus (itching), and rash, along with endocrine disorders such as hypothyroidism.[6] While these types of toxicities are known to occur with checkpoint inhibitors, their incidence and severity were significantly greater with the dual blockade. For instance, the rate of Grade $\ge2$ immune-mediated dermatitis was as high as 20% in one combination arm, while being 0% in the monotherapy arm.[6] These skin toxicities were described as generally manageable with steroids, but their high frequency contributed to the overall therapeutic burden.[6]

5.3 Treatment Discontinuation and Fatal AEs

The clinical consequence of this heightened toxicity was a dramatically higher rate of treatment discontinuation. TRAEs led to the cessation of therapy for 16% to 40% of patients across the different Belrestotug dose cohorts, a stark contrast to the 6% discontinuation rate in the dostarlimab monotherapy arm.[6] This high rate of discontinuation due to adverse events is a critical confounding factor that likely contributed to the trial's failure to demonstrate a PFS benefit.

Furthermore, the combination regimen was associated with fatal serious TRAEs, including cases of immune-mediated lung disease, immune-mediated hepatitis, and immune-mediated myocarditis.[6] These events, although rare, underscore the potential for severe and life-threatening toxicity when combining potent immunomodulatory agents and highlight the narrow therapeutic window of the Belrestotug-dostarlimab regimen.

Table 3: Comparative Incidence of Key Adverse Events in GALAXIES Lung-201

Adverse Event Category	Dostarlimab Monotherapy	Belrestotug 100mg + Dostarlimab	Belrestotug 400mg + Dostarlimab	Belrestotug 1000mg + Dostarlimab
Any Grade TRAEs (%)	59%	80%	84%	97%
Grade $\ge3$ TRAEs (%)	16%	33%	22%	43%
Serious TRAEs (%)	9%	33%	25%	37%
TRAEs Leading to Discontinuation (%)	6%	23%	16%	40%
Any Grade irAEs (%)	19%	67%	56%	73%
Grade $\ge3$ irAEs (%)	13%	30%	16%	37%
Immune-Mediated Dermatitis (Grade $\ge2$, %)	0%	17%	0%	20%
Immune-Mediated Hypothyroidism (Grade $\ge2$, %)	3%	3%	9%	13%

TRAE: Treatment-Related Adverse Event; irAE: immune-related Adverse Event. Data compiled from interim trial reports.[6]

Section 6: Discontinuation of the Belrestotug Program

The culmination of the challenging efficacy and safety data was the definitive and public termination of the entire Belrestotug clinical development program.

6.1 The Official Announcement

In May 2025, GSK and iTeos Therapeutics issued official press releases confirming their mutual decision to end the development program for Belrestotug and terminate their collaboration agreement.[9] The announcement stated that all ongoing clinical trial cohorts containing Belrestotug would be closed, and new enrollment into the Phase 3 GALAXIES Lung-301 study was being halted immediately.[9] Health authorities and clinical trial investigators were to be informed about the next steps for managing patients currently enrolled in the studies.[10]

6.2 The Stated Rationale

The companies provided a clear and data-driven rationale for the termination, based on the totality of evidence from the GALAXIES platform studies.[9] The primary reasons cited were:

• The failure of the GALAXIES Lung-201 study to meet the pre-established criteria for a clinically meaningful improvement in Progression-Free Survival (PFS) for the Belrestotug-dostarlimab combination compared to dostarlimab monotherapy in first-line NSCLC.[10]
• An interim analysis from the GALAXIES H&N-202 study showing that the trend for Objective Response Rate (ORR) in the Belrestotug combination arms fell below a meaningful clinical threshold compared to dostarlimab monotherapy in HNSCC.[9]
• The overarching conclusion that the overall efficacy data did not support continued development, despite the initially promising ORR signal in the lung cancer trial.[9]

6.3 Financial and Strategic Impact

The discontinuation of the program had significant financial and strategic repercussions for both partners, particularly for iTeos Therapeutics. The decision marked the end of a major collaboration that began in 2021 with a $625 million upfront payment from GSK and the potential for an additional $1.45 billion in milestone payments.[5]

For iTeos, the failure of Belrestotug was a substantial blow, as it was the company's most advanced clinical asset and the cornerstone of its pipeline.[28] The termination eliminated the prospect of future milestone payments and royalties from GSK. In response to this setback and what it termed "current market conditions," iTeos announced that it was taking immediate steps to preserve capital and had initiated a targeted strategic review to identify opportunities to maximize shareholder value.[11] This move signaled a necessary pivot to its earlier-stage assets and a fundamental reassessment of its corporate strategy.

Section 7: Expert Analysis and Concluding Remarks

The failure of the Belrestotug program is not an isolated incident but rather a significant event that reflects broader challenges within the TIGIT inhibitor class and the evolving landscape of immuno-oncology development.

7.1 The TIGIT Conundrum: A Pattern of High-Profile Failures

Belrestotug's discontinuation is the latest in a series of high-profile, late-stage failures for anti-TIGIT antibodies, a class once heralded as the next major breakthrough in cancer immunotherapy.[24] This pattern strongly suggests the presence of class-wide challenges that transcend the specifics of any single molecule.

The most direct parallel can be drawn with Roche's anti-TIGIT antibody, tiragolumab. In a strikingly similar narrative, tiragolumab, when combined with the anti-PD-L1 antibody atezolizumab, showed a promising improvement in ORR in the Phase 2 CITYSCAPE trial.[30] However, in the much larger, confirmatory Phase 3 SKYSCRAPER-01 trial conducted in a nearly identical first-line, PD-L1-high NSCLC population, the combination failed to demonstrate a statistically significant benefit in either PFS or Overall Survival (OS) compared to atezolizumab alone.[30] Roche ultimately scrapped the entire tiragolumab program after it failed across numerous indications, including lung, liver, and esophageal cancers, in trials involving nearly 5,000 patients.[31] The repeated observation that a promising ORR signal in Phase 2 does not translate to a durable clinical benefit in Phase 3 points to a fundamental issue with the TIGIT-PD-1/L1 combination strategy in these broad patient populations. Expert analysis suggests that the underlying biological hypothesis—that simply blocking TIGIT on top of PD-1 is sufficient to overcome immune resistance—may be overly simplistic or incomplete.[13]

7.2 Final Verdict on Belrestotug

The Belrestotug development program was ultimately terminated due to a fatal combination of insufficient durable efficacy and excessive toxicity. The combination with dostarlimab produced a higher rate of initial tumor responses, but these responses were not sustained long enough to improve Progression-Free Survival. This modest and transient efficacy could not justify the significantly higher rates of severe and life-threatening immune-related adverse events and the alarmingly high percentage of patients forced to discontinue treatment due to toxicity. The impressive ORR signal, which propelled the program into a Phase 3 trial, was ultimately a misleading indicator of clinical value, unable to overcome the regimen's fundamentally unfavorable risk-benefit balance.

7.3 Future Directions for TIGIT and Immuno-Oncology

The failures of Belrestotug and tiragolumab may signal a critical inflection point for immuno-oncology (IO) development. For years, the dominant paradigm has been an "add-on" strategy: combining a novel IO agent with an established anti-PD-1/L1 backbone. The repeated stumbles of the TIGIT class in large, well-conducted trials represent a major blow to this approach, suggesting that overcoming resistance to PD-1 blockade is a far more complex biological challenge. The tumor microenvironment likely possesses redundant immunosuppressive pathways that are not adequately addressed by this dual-blockade strategy. Furthermore, the escalating toxicity seen with these combinations highlights the inherent limitations of broadly unleashing the immune system.

The path forward for the remaining TIGIT assets, such as Arcus/Gilead's domvanalimab, may require a more nuanced strategy. Success could depend on identifying predictive biomarkers beyond PD-L1 to select patient populations with a specific biological dependency on the TIGIT pathway, or on exploring novel combination partners outside of the traditional IO-IO paradigm.

More broadly, the field may need to pivot away from simple additive combinations and toward more sophisticated therapeutic designs. This could include the development of multi-specific antibodies capable of engaging several targets simultaneously with a single molecule, the engineering of cell therapies (e.g., CAR-T) to be inherently resistant to checkpoint inhibition, or the combination of immunotherapies with entirely different modalities, such as antibody-drug conjugates (ADCs) or targeted small molecules. The era of testing broad IO-IO combinations in relatively unselected patient populations may be drawing to a close, supplanted by a more precise, biomarker-driven, and mechanistically sophisticated approach to cancer immunotherapy.

Works cited

1. Definition of belrestotug - NCI Drug Dictionary - National Cancer Institute, accessed October 30, 2025, https://www.cancer.gov/publications/dictionaries/cancer-drug/def/belrestotug
2. Belrestotug (EOS-448) | Anti-TIGIT Antibody - MedchemExpress.com, accessed October 30, 2025, https://www.medchemexpress.com/belrestotug.html
3. Belrestotug - iTeos Therapeutics - AdisInsight - Springer, accessed October 30, 2025, https://adisinsight.springer.com/drugs/800051067
4. belrestotug (EOS-448) / iTeos, GSK - Larvol Delta, accessed October 30, 2025, https://delta.larvol.com/Products/?ProductId=143aacd6-70bf-434c-9a65-eb6c6e5ba747
5. GSK, iTeos pull the plug on belrestotug as TIGIT stumbles again - FirstWord Pharma, accessed October 30, 2025, https://firstwordpharma.com/story/5960147
6. Frontline Belrestotug/Dostarlimab Boosts Responses Vs Dostarlimab Alone in PD-L1–High NSCLC | OncLive, accessed October 30, 2025, https://www.onclive.com/view/frontline-belrestotug-dostarlimab-boosts-responses-vs-dostarlimab-alone-in-pd-l1-high-nsclc
7. A trial looking at immunotherapy for non small cell lung cancer (Galaxies Lung 201), accessed October 30, 2025, https://www.cancerresearchuk.org/about-cancer/find-a-clinical-trial/a-trial-looking-at-immunotherapy-for-non-small-cell-lung-cancer-galaxies-lung-201
8. NCT06472076 | A Study of Belrestotug Plus Dostarlimab Compared With Placebo Plus Pembrolizumab in Previously Untreated Participants With Programmed Death Ligand 1 (PD-L1) High Non-small-cell Lung Cancer (NSCLC) | ClinicalTrials.gov, accessed October 30, 2025, https://clinicaltrials.gov/study/NCT06472076
9. Belrestotug Program Terminated After Mixed Phase 2 Results in ..., accessed October 30, 2025, https://www.targetedonc.com/view/belrestotug-program-terminated-after-mixed-phase-2-results-in-nsclc
10. iTeos, GSK end belrestotug program - Medthority, accessed October 30, 2025, https://www.medthority.com/news/2025/5/iteos-and-gsk-have-made-the-decision-to-terminate-the-belrestotug-development-program-and-end-the-collaboration/
11. GSK, iTeos Abandon TIGIT Therapy, Adding to Long List of Recent Class Failures, accessed October 30, 2025, https://www.biospace.com/drug-development/gsk-iteos-abandon-tigit-therapy-adding-to-long-list-of-recent-failures
12. GSK provides update on belrestotug development programme | GSK, accessed October 30, 2025, https://www.gsk.com/en-gb/media/press-releases/gsk-provides-update-on-belrestotug-development-programme/
13. TIGIT in cancer: from mechanism of action to promising ..., accessed October 30, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC12402281/
14. TIGIT/CD226 Axis Regulates Anti-Tumor Immunity - MDPI, accessed October 30, 2025, https://www.mdpi.com/1424-8247/14/3/200
15. 10-K - SEC.gov, accessed October 30, 2025, https://www.sec.gov/Archives/edgar/data/1808865/000095017024026866/itos-20231231.htm
16. iTeos Therapeutics and GSK Terminate Belrestotug Development Program Following GALAXIES Lung-201 Results | Nasdaq, accessed October 30, 2025, https://www.nasdaq.com/articles/iteos-therapeutics-and-gsk-terminate-belrestotug-development-program-following-galaxies
17. Evolving Precision First-Line Systemic Treatment for Patients with Unresectable Non-Small Cell Lung Cancer - MDPI, accessed October 30, 2025, https://www.mdpi.com/2072-6694/16/13/2350
18. Pembrolizumab as a First-Line Standard-of-Care Option in Advanced Non–Small-Cell Lung Cancer See also Value-Based Care, accessed October 30, 2025, https://www.ahdbonline.com/issues/2018/august-vol-11-special-issue-payers-perspectives-in-oncology-asco-2018-highlights/pembrolizumab-as-a-first-line-standard-of-care-option-in-advanced-non-small-cell-lung-cancer-see-also-value-based-care
19. NCT05565378 | A Platform Study of Novel Immunotherapy Combinations in Participants With Previously Untreated, Advanced/Metastatic Non-Small-Cell Lung Cancer | ClinicalTrials.gov, accessed October 30, 2025, https://clinicaltrials.gov/study/NCT05565378
20. Expert consensus on immune checkpoint inhibitors treatment for recurrent/metastatic head and neck squamous cell carcinoma (2024 edition) - China Oncology-, Volume Issue, accessed October 30, 2025, https://www.china-oncology.com/EN/10.19401/j.cnki.1007-3639.2024.04.010
21. UPDATED: iTeos-GSK's TIGIT combo shows 30% more tumor shrinkage than Jemperli, but safety signals scare investors - Fierce Biotech, accessed October 30, 2025, https://www.fiercebiotech.com/biotech/esmo-iteos-gsks-tigit-star-shows-30-more-tumor-shrinkage-jemperli-ph-2
22. Another TIGIT bites the dust | ApexOnco - Clinical Trials news and analysis, accessed October 30, 2025, https://www.oncologypipeline.com/apexonco/another-tigit-bites-dust
23. iTeos Reports Third Quarter 2024 Financial Results and Provides Business Updates, accessed October 30, 2025, https://firstwordpharma.com/story/5911692
24. A Study of Belrestotug Plus Dostarlimab Compared With Placebo Plus Pembrolizumab in Previously Untreated Participants With Programmed Death Ligand 1 (PD-L1) High Non-small-cell Lung Cancer (NSCLC) - larvol clin, accessed October 30, 2025, https://clin.larvol.com/trial-detail/NCT06472076
25. iTeos Announces Clinically Meaningful Objective Response Rate Observed at Every Dose in Follow-up Interim Analysis of GALAXIES Lung-201 Study of Belrestotug + Dostarlimab in First-Line, PD-L1 High Non-Small Cell Lung Cancer Patients - BioSpace, accessed October 30, 2025, https://www.biospace.com/press-releases/iteos-announces-clinically-meaningful-objective-response-rate-observed-at-every-dose-in-follow-up-interim-analysis-of-galaxies-lung-201-study-of-belrestotug-dostarlimab-in-first-line-pd-l1-high-non-small-cell-lung-cancer-patients
26. GSK and iTeos Terminate TIGIT Cancer Drug Development After Phase 2 Failure - MedPath, accessed October 30, 2025, https://trial.medpath.com/news/e651316cab379869/gsk-and-iteos-terminate-tigit-cancer-drug-development-after-phase-2-failure
27. Why has immune "checkpoint" therapy failed in most clinical trials? - PubMed, accessed October 30, 2025, https://pubmed.ncbi.nlm.nih.gov/40866292/
28. Surprise Failure in Phase 3 Lung Cancer Trial: Combo Regimen ..., accessed October 30, 2025, https://www.oncologynewscentral.com/nsclc/surprise-failure-in-phase-3-lung-cancer-trial-combo-regimen-misses-dual-endpoints
29. 5,000 patients later Roche scraps its TIGIT | ApexOnco - Clinical ..., accessed October 30, 2025, https://www.oncologypipeline.com/apexonco/5000-patients-later-roche-scraps-its-tigit