Metastatic uveal melanoma has long occupied a grim category of its own in oncology: a cancer so resistant to the checkpoint inhibitors that revolutionized melanoma treatment elsewhere that a succession of clinical trials — pembrolizumab, ipilimumab, combination immunotherapy — came and went without meaningfully changing survival. The median overall survival for patients whose eye melanoma had spread to the liver was roughly 12 months. Two-year survival was uncommon; five-year survival was exceptional.

Data presented Friday at the SITC 2026 Annual Meeting in a poster from drug developer Immunocore — and reinforced by an oral presentation Saturday morning — may alter that calculus in a fundamental way. In Phase 1/2 trial NCT04262466 evaluating brenetafusp combined with the FDA-approved tebentafusp (KIMMTRAK), 70% of first-line patients carrying the required HLA-A*02:01 gene variant were alive at two years. Ten patients remain on the study. No patient discontinued therapy due to a treatment-related adverse event.

“The combination of two ImmTAC molecules targeting different tumor antigens has achieved something we have not seen before in this disease — durable, deepening responses in patients who historically would have had almost no good options,” said Dr. Omid Hamid, Chief of Research and Immuno-Oncology and Co-Director of the Cutaneous Malignancy Program at The Angeles Clinic and Research Institute, one of the study’s principal investigators.

The result is not simply a numerical milestone. It is evidence that a new mechanism — simultaneous redirection of T cells to two distinct intracellular tumor proteins — can unlock anti-tumor immunity in a cancer that has been stubbornly impervious to every other form of systemic immunotherapy.

Why Eye Melanoma Has Been So Hard to Treat

Uveal melanoma arises from pigment cells in the uvea — the eye’s inner lining of iris, choroid, and ciliary body — making it the most common primary eye cancer in adults, though it is rare by any other measure: roughly 2,500 Americans diagnosed annually. The primary tumor is usually manageable with radiation (plaque brachytherapy), proton therapy, or surgery. The problem is that roughly half of patients develop metastatic disease, almost always in the liver, and historically at that point the disease became a waiting game.

The most consequential structural fact about uveal melanoma is that it lacks the lymphatic drainage channels found in most tissues. Metastatic spread is exclusively through the bloodstream to the liver, and the immune environment in the liver is fundamentally immunosuppressive — which may help explain why checkpoint inhibitors that work by releasing existing T cell responses against tumors have consistently underperformed in this cancer.

The five-year survival rate for distant metastatic uveal melanoma is approximately 15%, and that figure has barely moved in decades. No chemotherapy, targeted therapy, or immunotherapy had produced a confirmed improvement in overall survival until tebentafusp’s approval in January 2022 — itself a narrow approval, applicable only to the HLA-A*02:01+ subset of patients (approximately 40–45% of people with European ancestry), and showing a median overall survival of 21.6 months versus 16.9 months for investigator’s choice. Long-term follow-up data presented at the 2026 AACR Annual Meeting found that tebentafusp doubled the five-year survival rate from 8% to 16% — still a low absolute number, but a meaningful relative advance over a disease where survival at five years had been extraordinary.

The SITC 2026 data describes what happens when tebentafusp is combined with a second, related molecule targeting a different tumor protein entirely.

How These Drugs Actually Work — and Why That Matters

To understand why the combination is mechanistically interesting, it helps to understand why tebentafusp and brenetafusp work so differently from every other cancer drug approved to date.

Conventional antibody-based therapies — including checkpoint inhibitors, antibody-drug conjugates, and most bispecific antibodies — can only recognize proteins displayed on the outer surface of a cancer cell. The vast majority of a cell’s proteins, including most of the proteins that make cancer cells cancerous, exist inside the cell, invisible to antibodies circulating in the bloodstream.

ImmTAC T cell engager molecules (Immune mobilizing monoclonal T cell receptors Against Cancer) solve this by exploiting the same system human immune cells normally use to survey the interior of other cells. Every cell constantly degrades its own proteins into short peptide fragments, loads them onto HLA class I molecules — HLA-A, HLA-B, or HLA-C — and displays this peptide-HLA complex (pHLA) on the cell surface. This is how CD8+ T cells normally “see” what is happening inside a cell; a T cell’s native T cell receptor (TCR) is shaped to recognize a specific pHLA combination. When a virus or mutation creates an abnormal protein fragment, the right T cell with the matching TCR will recognize the pHLA, become activated, and kill the infected or cancerous cell.

The limitation of this natural process in cancer is that the T cells with the right TCR may be absent, exhausted, or excluded from the tumor microenvironment. ImmTAC bypasses that limitation by design. It has two arms: one is an affinity-enhanced TCR, engineered to have roughly one million times greater binding strength than a natural TCR, allowing it to grip pHLA complexes even at the extremely low densities found on actual tumor cells — sometimes as few as 10 complexes per cell. The other arm is an anti-CD3 single-chain antibody that grabs CD3, a universal marker on all T cells. When the ImmTAC molecule bridges the tumor cell’s pHLA to a bystander T cell via CD3, the T cell activates and kills the tumor cell regardless of whether that T cell’s own TCR would ever have recognized the tumor protein.

Tebentafusp targets a gp100-derived peptide on HLA-A*02:01. gp100 (also called PMEL17) is a melanocyte differentiation protein expressed by both normal melanocytes and melanoma cells — including uveal melanoma, where gp100 is consistently expressed in both primary tumors and metastases.

Brenetafusp targets a PRAME-derived peptide (SLLQHLIGL) also displayed on HLA-A*02:01. PRAME (PReferentially expressed Antigen in MElanoma) is a cancer-testis antigen — a protein normally restricted to testis tissue that is aberrantly re-expressed at high levels in multiple tumor types, per 2024 research. In uveal melanoma specifically, high PRAME expression is a negative prognostic marker: patients whose tumors express more PRAME do worse. Targeting PRAME in this setting thus combines a mechanistically rational approach with a clinically meaningful subgroup.

Combining the two molecules means that tumor cells expressing gp100 or PRAME — or both — can be engaged and killed by T cells. In a cancer characterized by intratumoral heterogeneity and immune evasion, targeting two distinct intracellular proteins may help prevent the antigen-loss escape that can undermine single-antigen approaches over time.

Crucially, both molecules require HLA-A02:01 to function — the same HLA allele must present the peptide on the tumor cell surface for either TCR arm to bind. This requirement restricts both drugs to HLA-A02:01-positive patients, an eligibility criterion that affects approximately 40–45% of people with European ancestry and a smaller proportion of people with other ancestral backgrounds. All 48 patients in the trial — 22 receiving monotherapy and 26 receiving the combination — were confirmed HLA-A*02:01-positive.

The Numbers From SITC 2026

The trial enrolled patients in two arms. Twenty-two patients received brenetafusp monotherapy; of 16 evaluable patients, 3 (19%) had an objective tumor response, with a median duration of response of 14.5 months, median progression-free survival of 4.2 months, and median overall survival of 20.3 months.

Twenty-six patients received brenetafusp in combination with tebentafusp; of 17 evaluable first-line patients, 5 (29%) had an objective response, with a median duration of response of 24.1 months — nearly two years — and a median progression-free survival of 8.1 months. The median overall survival in the combination arm has not yet been reached; the two-year overall survival rate is 70%.

Ten patients in the combination arm remain on study as of the data cutoff (June 2026).

Molecular response data from circulating tumor DNA (ctDNA), an early indicator of tumor cell killing, showed that 8 of 9 evaluable combination-arm patients (89%) had a detectable molecular response, and 4 (44%) achieved ctDNA clearance, meaning the tumor DNA in their blood became undetectable. ctDNA clearance is associated with deeper tumor responses in multiple cancer types and may be a leading indicator of durable remission.

A historical comparison makes the numbers more vivid. Before tebentafusp’s approval, the standard-of-care options for metastatic uveal melanoma included chemotherapy (dacarbazine), liver-directed therapies, and investigator-discretion immunotherapy — none of which had demonstrated survival benefits in randomized trials. In the Phase 3 trial that led to tebentafusp’s approval, one-year overall survival was 73% for tebentafusp versus 59% for investigator’s choice. The 70% two-year survival in the combination arm compares favorably even against that improved baseline, noting that cross-trial comparisons should be interpreted cautiously.

Safety: Manageable, With a Class-Effect Profile

The combination was not entirely without side effects. The most common treatment-related adverse events reflected the expected immune activation from T cell-redirecting therapy: cytokine release syndrome occurred in 77% of combination patients and 73% of monotherapy patients. Rash occurred in 77% of combination patients and 50% of monotherapy patients. Grade 3/4 treatment-related adverse events (TRAEs) were reported in 38% of combination patients and 41% of monotherapy patients, primarily transient lymphocyte decreases.

No patient in the combination arm experienced Grade 3 or Grade 4 cytokine release syndrome — the severe form associated with hospitalization and organ stress — and no treatment-related deaths occurred in either arm. No patient discontinued therapy due to adverse events.

The absence of Grade ≥3 CRS in 26 combination patients is a notable finding. A theoretical concern entering any dual T cell engager combination was that two simultaneous activating molecules might produce additive or amplified immune activation, raising the risk of severe CRS beyond what either agent alone produces. This signal did not materialize at the doses tested.

Cytokine release syndrome in the context of ImmTAC molecules typically presents as fever and fatigue in the first few days of treatment, manageable with supportive care. The class-effect toxicity is well understood from tebentafusp’s commercial experience and was anticipated by the trial design; first doses are administered in clinical settings with monitoring protocols in place.

What’s Next: A Randomized Phase 2/3 Trial

Immunocore has announced plans to initiate DUET-UVEAL, a randomized Phase 2/3 trial that will formally test the brenetafusp-plus-tebentafusp combination against standard of care or tebentafusp monotherapy in HLA-A*02:01-positive patients with metastatic uveal melanoma in the first-line setting. The specific design, comparator arm, and primary endpoint have not yet been disclosed. DUET-UVEAL would be the first randomized Phase 2/3 trial of a dual-ImmTAC combination, and its results would determine whether the combination becomes a new treatment standard or requires further optimization.

The combination data at SITC 2026 also arrives in the context of two other large ongoing Immunocore trials:

TEBE-AM is a Phase 3 registrational trial testing tebentafusp monotherapy and tebentafusp in combination with pembrolizumab in previously treated patients with advanced melanoma beyond uveal melanoma. Enrollment of 540 patients was completed in September 2026. Topline overall survival data are expected in the second quarter of 2027.

PRISM-MEL-301 is a Phase 3 registrational trial testing brenetafusp in combination with nivolumab as a first-line therapy for advanced cutaneous melanoma — a substantially larger indication than uveal melanoma. This trial is currently enrolling.

The combined PRAME-expression profile across cancer types — over 90% of cutaneous melanomas, over 44% of lung cancers, over 63% of triple-negative breast cancers, and substantial expression in ovarian and endometrial cancers — means that if brenetafusp’s activity is validated in Phase 3, the platform could have applicability across far larger patient populations than the several hundred HLA-A*02:01-positive uveal melanoma patients diagnosed in the United States each year.

Commercial Context: KIMMTRAK Revenues as the Platform’s Foundation

Immunocore reported net revenues of $102.2 million in Q3 2026 from KIMMTRAK sales, bringing nine-month 2026 net revenues to $324.8 million, up approximately 10% year-over-year compared to $295.5 million for the same period in 2025. KIMMTRAK currently generates revenue solely from its approved indication in HLA-A*02:01-positive metastatic uveal melanoma in the United States and Europe.

The dual-ImmTAC combination data could, if validated in DUET-UVEAL, expand tebentafusp’s label into a combination indication in the same patient population — a commercial opportunity within the existing approved patient pool. The larger commercial opportunity from brenetafusp’s program in cutaneous melanoma (PRISM-MEL-301) would represent a substantially broader market.

Significance: Why This Isn’t Just a Rare Cancer Story

Uveal melanoma is rare enough that any advance within it affects a relatively small number of patients. But the significance of the SITC 2026 data is not primarily about those patients, as important as that is.

Metastatic uveal melanoma has served as a proof-of-concept battleground for the ImmTAC platform for a reason: it is a cancer where the tumor microenvironment is unusually immunosuppressive, checkpoint inhibitors have reliably failed, and there is no background noise from prior immune responses to confuse the signal from TCR-based T cell redirection. If ImmTAC technology can produce durable, deepening responses here — with 10 patients still on study and a median overall survival not yet reached — it demonstrates that bispecific T cell engagers built on affinity-enhanced TCRs can function in the most hostile solid tumor environments.

The combination of two ImmTAC molecules targeting two different intracellular antigens simultaneously has never been tested before in solid tumors at any scale. That it appears to work without producing catastrophic additive toxicity, while improving on the already-meaningful survival signal from tebentafusp monotherapy, is the kind of mechanistic validation that tends to reshape how a drug platform is understood — not just in the indication where it was tested, but across every cancer type where the same proteins are expressed and the same HLA allele is present.

For the roughly 1,200 Americans who will develop metastatic uveal melanoma this year and whose tumors carry HLA-A*02:01, the question posed by SITC 2026 is straightforward: will the randomized trial confirm what the Phase 1/2 signal has suggested? For the broader cancer immunotherapy field, the question is larger: does dual targeting of two intracellular tumor antigens with two bispecific T cell engagers represent a generalizable paradigm for immune activation in checkpoint inhibitor-resistant solid tumors?

DUET-UVEAL will begin to answer the first question. The answers to the second are likely years away.

HLA-A*02:01 prevalence estimates of 40–45% in European-ancestry populations are based on published allele frequency literature and Immunocore’s regulatory submissions. Prevalence varies significantly across populations.

Brenetafusp (IMC-F106C) is investigational and has not been approved by the FDA or any regulatory authority.