Tebentafusp, Darovasertib, and Cell-Based Therapies Emerge as Pillars of an Evolving Uveal Melanoma Treatment Landscape

During a recent OncLive® Scientific Interchange and Workshop, experts gathered to examine how an expanding therapeutic toolkit anchored by tebentafusp-tebn (Kimmtrak), liver-directed therapy, and the emerging combination of darovasertib plus crizotinib (Xalkori) is reshaping the management of metastatic uveal melanoma.¹

“At the end of the day, what I hope we can say is that there’s been real progress in this disease and that there is a platform from which we can build an even brighter future,” Omid Hamid, MD, Chief of Translational Research and Immuno-Oncology at The Angeles Clinic and Research Institute, a Cedars-Sinai Affiliate, in Los Angeles, California, and moderator of the workshop, stated during the discussion.

How are biomarkers informing patient selection and monitoring in uveal melanoma?

The panelists began by discussing 2 distinct biomarker domains in uveal melanoma: genomic risk stratification for newly diagnosed patients and the evolving role of circulating tumor DNA (ctDNA) for treatment monitoring. Among the panelists, approaches to HLA-A*02:01 testing ranged from testing every patient at first contact to restricting testing to those with metastatic or high-risk disease. However, faculty agreed that early testing is a low-burden step that can identify patients who would benefit from tebentafusp, as HLA results return in approximately 4 days, which is often faster than tumor biopsy results.

ctDNA was the session’s most contested biomarker. The sharpest exchange centered on whether it is appropriate to order a test when no clearly superior treatment alternative exists at the time of progression.

“Always when I order my tests, I ask myself, what am I going to do with the result of this test?” Bartosz Chmielowski, MD, PhD, a professor at the University of California, Los Angeles, stated. “What if the scan shows stable disease, but ctDNA is going up? Am I going to stop? Do I have a better alternative after that? If I don’t, do I really want to order this test, which is not going to lead me to changes in therapy?”

The opposing view was that ctDNA catches what imaging misses, such as subclinical bone metastases prompting a confirmatory PET scan, and that the field’s obligation is to collect the data prospectively now so that future practice can become evidence-based. Faculty agreed on one clear use case for ctDNA: distinguishing true progression from the radiographic pseudoprogression and waxing-and-waning patterns characteristic of tebentafusp response. Emerging higher-sensitivity digital ctDNA platforms were seen as a meaningful step forward.

How should tebentafusp be optimally utilized in metastatic uveal melanoma?

Five-year overall survival (OS) data from the pivotal phase 3 IMCgp100-202 trial (NCT03070392), presented at the 2026 AACR Annual Meeting, demonstrated an OS rate of 16% with tebentafusp (n = 252) vs 8% with investigator's choice of therapy (n = 126) in patients with previously untreated HLA-A*02:01–positive metastatic uveal melanoma.² The median OS at 5 years was 21.6 months (95% CI, 19.0-24.3) in the tebentafusp group vs 16.9 months (95% CI, 12.9-19.5) in the investigator’s choice of therapy (ICT) arm (HR, 0.67; 95% CI, 0.54-0.85). This OS benefit with tebentafusp was maintained even without the effect from subsequent therapy (HR, 0.52; 95% CI, 0.31-0.83).

The panel’s enthusiasm for the drug was measured.¹ The comparator bar was viewed as low, deep responses are uncommon, and the OS curves converge over time. However, panelists generally viewed tebentafusp as a tolerable backbone onto which liver-directed and other therapies can be layered.

“What’s not captured in the 5-year tebentafusp data is what is happening in the real world,” Marlana Orloff, MD, an associate professor of medical oncology and Alexander and Johnston Family Endowed Clinical Director in Uveal Melanoma at Thomas Jefferson University Hospital in Philadelphia, Pennsylvania, noted. “I think we have a lot of patients who are doing much better than [those reflected in] the 5-year data because we keep tebentafusp on in the background, and you can add things. If we all look back in another year or two, we are going to see our OS on the whole being much better.”

Disease volume and line of therapy were identified as underrecognized modifiers of tebentafusp benefit, with faculty reporting better responses in frontline, lower-volume disease. This is consistent with trial data showing stronger responses in patients with liver lesions smaller than 3 cm and minimal benefit in heavily pretreated, high-volume settings.

Regarding cytokine release syndrome (CRS), experts noted that most events were grade 1 or 2 and manageable through step dosing. However, several faculty noted they still prefer inpatient monitoring for the initial doses, particularly in older patients with high disease burden. Faculty aligned on not using dual immune checkpoint inhibitor (ICI) therapy frontline outside a clinical trial and sequencing ICIs thoughtfully relative to liver-directed therapy and cell-based approaches.

“It's interesting [that] most places are doing outpatient CAR T-[cell therapy], but tarlatamab-dlle [Imdelltra] is [still administered as] outpatient [therapy],” Roxana Dronca, MD, director of the Mayo Clinic Comprehensive Cancer Center in Florida and professor of oncology at Mayo Clinic Florida in Jacksonville, said during the discussion. “This is one of the last treatments with a mandatory observation requirement."

Where does darovasertib plus crizotinib fit into the treatment paradigm?

Results from the phase 2/3 OptimUM-02 trial (NCT05987332) demonstrated a median progression-free survival (PFS) of 6.9 months with darovasertib plus crizotinib (n = 210) vs 3.1 months with ICT (n = 103; HR, 0.42; 95% CI, 0.30-0.59; P < .0001) in patients with HLA-A*02:01–negative metastatic uveal melanoma.³ The overall response rates were 37.1% vs 5.8%, respectively (P < .0001). Five complete responses (CRs) were reported in the darovasertib arm, whereas no patients in the control arm experienced a CR. The median duration of response (DOR) with darovasertib plus crizotinib was 6.8 months. OS data were not mature at the time of the topline analysis, though an early OS trend favored the darovasertib arm.

This is a clinically meaningful advance for a disease in which patients with HLA-A*02:01–negative disease have historically had no effective targeted option.¹ However, the panelists warned that although an oral regimen removes the referral-forcing function that intravenous therapies create, the combination carries toxicities such as nausea, vomiting, diarrhea, and edema. These toxicities often require sophisticated management, which low-volume community centers may not be able to provide, Chmielowski said.

To address this, Orloff proposed a mandatory link to an expert center with a prebuilt contingency plan, with community physicians managing maintenance. Meredith McKean, MD, MPH, the director of the Melanoma and Skin Cancer Research Program at Sarah Cannon Research Institute in Nashville, Tennessee, highlighted that the short half-life of darovasertib/crizotinib typically makes brief holds and dose reductions sufficient to keep patients on therapy when paired with dietary counseling. Faculty largely closed the door on combining darovasertib/crizotinib with tebentafusp upfront due to overlapping toxicity concerns.

Use of darovasertib in the neoadjuvant setting also generated discussion during the workshop. Experts discussed the phase 3 OptimUM-10 trial (NCT07015190), which is evaluating whether tumor-height reduction can spare patients standard enucleation or plaque brachytherapy.⁴

What is the potential role of emerging cell-based therapies?

Tumor-infiltrating lymphocyte (TIL) therapy and PRAME-directed T-cell receptor (TCR) therapy generated the most scientific optimism for durable benefit. Data from a phase 2 trial (NCT03467516) presented at the 2026 ASCO Annual Meeting showed that among 33 patients with heavily pretreated metastatic uveal melanoma who received a non-myeloablative lymphodepleting preparative regimen of fludarabine and cyclophosphamide followed by autologous TIL infusion and high-dose interleukin-12 (n = 32), the objective response rate (ORR) was 22% (95% CI, 9%-40%), with a median duration of response (DOR) of 10.8 months (range, 3.5-58+ months) and a maximum ongoing response of 58 months.⁵

Udai Kammula, MD, FACS, a professor of surgery and director of the Solid Tumor Cell Therapy Program at UPMC Hillman Cancer Center in Pittsburgh, Pennsylvania, spotlighted an in-situ transcriptomic biomarker—the Uveal Melanoma Immunogenic Score (UMIS)—that can predict reactive TIL growth and clinical response from a minimally invasive core biopsy performed locally, potentially doubling operational scale by sparing surgery for nonreactive patients. Access, restrictive trial eligibility, and absence of a commercialized dedicated facility remain the principal barriers for adoption.

Anzutresgene autoleuce (anzu-cel), a PRAME-directed TCR therapy, also drew strong interest. Phase 1/2 data from the IM203-101 trial (NCT03686124) presented at the 2026 ASCO Annual Meeting showed an ORR of 67% in 16 evaluable patients with uveal melanoma, a disease control rate of 88%, and a median duration of response of 11 months range, 4.2-38.2+ months) with the agent.⁶

Chmielowski noted that the relatively homogeneous, low-mutational-burden biology of uveal melanoma may favor single-antigen targeting. “I think it’s going to be better in uveal melanoma than in cutaneous melanoma. It’s going to be a breakthrough for everyone.” Orloff tempered this enthusiasm with practical caveats, including the lack of a clear dedicated uveal melanoma development pathway beyond the cutaneous-focused, randomized phase 3 SUPRAME trial (NCT06743126).

J. William Harbour, MD, a professor and chair of Ophthalmology at UT Southwestern Medical Center in Dallas, Texas, offered the session’s most pointed reframe, explaining how correlative data from a LAG-3 analysis suggest that treatment progression in uveal melanoma is not primarily a T-cell activation problem. Instead, melanoma-specific T cells are present at baseline and can clonally expand but are ultimately actively suppressed, likely by macrophages.

“In my mind, when we talk about adding another T-cell therapy and another T-cell therapy, that’s not really what the data [support],” Harbour said. “We need to activate T cells. But they’re being actively suppressed by something else. Until we start getting correlative data to understand why the therapies aren’t working, just adding more stuff onto it with no data, no evidence, is not going to lead us anywhere.”

References

1. Aligning evidence and practice in uveal melanoma. An OncLive® Scientific Interchange and Workshop. OncLive. May 29, 2026. Accessed June 23, 2026.
2. Nathan P. Piperno-Neumann S, Hassel JC, et al. Five-year survival with tebentafusp in previously untreated metastatic uveal melanoma in a phase 3 trial. Cancer Res. 2026;86(suppl 8):CT029.doi:10.1158/1538-7445.AM2026-CT029
3. IDEAYA Biosciences and Servier announce positive topline results from phase 2/3 registrational trial (OptimUM-02) of darovasertib in combination with crizotinib in first-line HLA-A*02:01-negative metastatic uveal melanoma. News release. IDEAYA Biosciences, Inc. April 13, 2026. Accessed June 23, 2026. https://ir.ideayabio.com/2026-04-13-IDEAYA-Biosciences-and-Servier-Announce-Positive-Topline-Results-from-Phase-2-3-Registrational-Trial-OptimUM-02-of-Darovasertib-in-Combination-with-Crizotinib-in-First-line-HLA-A-02-01-Negative-Metastatic-Uveal-Melanoma
4. Harbour JW, Reichstein DA, Angi M, et al. Neoadjuvant darovasertib in uveal melanoma patients undergoing primary local therapy (OptimUM-10): phase 3 trial. J Clin Oncol. 2026;44(suppl 16):TPS9608. doi:10.1200/JCO.2026.44.16_suppl.TPS9608
5. Leonard-Murali S, Bhaskarla C, Yadav GS, et al. Phase 2 trial of TIL therapy for metastatic uveal melanoma: evaluation of T cell potency and an in situ precision biomarker. J Clin Oncol. 2026;44(suppl 16):2512. doi:10.1200/JCO.2026.44.16_suppl.2512
6. Davar D, Patel SP, Hernandez-Aya, LF, et al. Patient-level clinical response dynamics in advanced melanoma with anzutresgene autoleucel (anzu-cel), a PRAME-directed T-cell receptor (TCR) T-cell therapy. J Clin Oncol. 2026;44(suppl 16):9508. doi:10.1200/JCO.2026.44.16_suppl.9508