Merck Opens Phase 1b/2 Trial of MK-2750 in Advanced Solid Tumors
Merck is moving MK-2750 into a first-in-human style development path in adults with advanced solid tumors, launching a randomized, open-label phase 1b/2 study designed to establish whether the antibody-drug conjugate can be given safely and whether it shows early antitumor activity. The study is not yet recruiting and is built around a classic early oncology framework: dose comparison first, signal-finding expansion second.
The trial will enroll an estimated 300 participants and is set to begin in November 2026, with primary completion projected for April 2029 and final completion in May 2031. Investigators will monitor adverse events, treatment discontinuations due to adverse events, and objective response rate over roughly two years, making this a study that is meant to answer both tolerability and activity questions before the program moves toward more definitive development.
Why an Antibody-Drug Conjugate Matters Here
MK-2750 belongs to the class of antibody-drug conjugates, or ADCs. The therapeutic logic is straightforward: an antibody is used to seek out a target on cancer cells, and a linked cytotoxic payload is delivered into the tumor cell rather than spread broadly through the body. In practice, that design aims to widen the therapeutic window by concentrating cell-killing activity where it is needed most while limiting exposure to normal tissue.
That strategy has become central in modern oncology because advanced solid tumors often escape conventional chemotherapy through heterogeneity, resistance pathways, and the ability to survive despite repeated systemic treatment. An ADC can sometimes exploit a molecular vulnerability without demanding that the tumor be universally uniform. The tradeoff is familiar to oncologists: the drug must be potent enough to matter biologically, yet precise enough to avoid a toxicity profile that shuts down dosing before meaningful activity appears.
Broad Eligibility Spans Several Refractory Tumor Types
The enrollment criteria point to a study population with substantial unmet need. Eligible participants include adults with advanced or metastatic non-small cell lung cancer, advanced or metastatic pancreatic ductal adenocarcinoma, unresectable or metastatic colorectal adenocarcinoma, squamous cell carcinoma of the esophagus, advanced or metastatic adenocarcinoma of the esophagus, gastroesophageal junction, or stomach that is considered incurable by local therapies, and recurrent or metastatic head and neck squamous cell carcinoma involving the oral cavity, oropharynx, hypopharynx, or larynx.
This is not a low-risk population. These are diseases in which standard therapy has often already been used, exhausted, or expected to have limited durability. By selecting tumors that remain advanced, metastatic, unresectable, or incurable by local therapy, the trial is positioning MK-2750 in a setting where even a modest signal can be clinically meaningful if it is coupled with manageable toxicity and credible pharmacokinetics.
A single positive scan response in this setting can matter more than it would in earlier disease.
Safety Guardrails Reflect the Toxicity Burden of Early ADC Development
The exclusion criteria reveal where investigators expect the margins to be tight. Participants with active central nervous system metastases or carcinomatous meningitis are excluded, as are those with interstitial lung disease or pneumonitis history outside a narrow exception. Severe dry eye syndrome, severe meibomian gland disease, severe blepharitis, and severe corneal disease are also disqualifying conditions. Those ocular exclusions are especially notable because ADC programs have increasingly had to account for corneal or ocular surface toxicity in addition to the more familiar marrow, gastrointestinal, and hepatic risks.
Cardiovascular instability is another hard stop. The study excludes uncontrolled significant cardiovascular or cerebrovascular disease, including New York Heart Association class III or IV congestive heart failure, unstable angina, uncontrolled symptomatic arrhythmia, and corrected QT interval prolongation by Fridericia above 470 ms. Patients with uncontrolled pleural effusion, pericardial effusion, or ascites requiring frequent drainage are also excluded, as are those who have not recovered adequately from major surgery.
These criteria are not administrative fine print. They are a safety map for a drug class that can fail in early development if target engagement comes at the cost of cumulative tissue injury, poor tolerability, or treatment interruptions that blunt exposure.
Response Assessment Will Be Matched With Pharmacokinetics
The trial is not only measuring whether tumors shrink. It is also tracking the maximum plasma concentration of the free payload, a pharmacokinetic readout that can help determine how the drug behaves once the ADC has released its cytotoxic component. In ADC development, the relationship between circulating payload, exposure, and activity is often as important as the tumor response itself. If a molecule generates responses at exposures that remain tolerable, it has a path forward. If the payload leaks too early or accumulates in a way that drives toxicity before efficacy, the entire concept weakens.
The primary efficacy endpoint is objective response rate by RECIST 1.1, with complete response and partial response as defined by standard radiologic criteria. Secondary durability assessment will use duration of response among participants who achieve a confirmed response. Those metrics are the right tools for a phase 1b/2 oncology program, because they show not just whether the drug can start a response but whether that response lasts long enough to matter in a disease with aggressive natural history.
Trial Logistics Keep the Design Flexible but Still Disciplined
The study uses a randomized, parallel-assignment structure without masking. Participants in part 1 will enter a dose comparison cohort, while part 2 will move into signal-finding cohorts. Each arm includes MK-2750 administered by intravenous infusion, along with standard supportive medications: a 5-HT3 serotonin receptor antagonist, an NK-1 receptor antagonist, a corticosteroid, and a steroid mouthwash with dexamethasone or an equivalent agent as rescue medication.
That premedication package signals a protocol built to anticipate infusion-related and mucosal toxicities before they become dose-limiting. It also hints that the development team is trying to preserve the ability to escalate or explore activity without letting predictable supportive-care issues obscure the underlying signal.
The study accepts adults aged 18 years and older and includes all sexes. Participants with well-controlled HIV on antiretroviral therapy, treated hepatitis B with undetectable viral load, and prior hepatitis C infection with undetectable viral load may be eligible, which broadens access while still keeping viral reactivation risk under control through protocol safeguards.
What This Early Readout Could Clarify
The central question is whether MK-2750 can deliver enough biologic effect across several hard-to-treat solid tumors to justify larger development. Early ADC programs are often decided by a narrow set of facts: can the drug be dosed, can it be delivered repeatedly, does the payload behave as expected in plasma, and do objective responses emerge without a toxicity profile that collapses treatment continuity?
This study is built to answer those questions in a disciplined way. If adverse events remain manageable, discontinuations stay low, and response signals appear in one or more of the enrolled tumor types, MK-2750 could move from a broad early-phase hypothesis into a more focused development strategy. If toxicity dominates or responses remain sparse, the data will still define the ceiling of what this ADC can do in advanced disease.
For now, the trial represents a measured bet on ADC biology in tumors where standard options often run out before the disease does. The first decisive clinical answers will not arrive until the dose-comparison cohort and the signal-finding expansion have time to accumulate enough exposure and imaging follow-up to show whether the payload reaches its target with enough precision to matter.