Summary
In a first-in-human phase 1 trial, the oral WRN inhibitor RO7589831 produced disease control in 74.2% of 66 evaluable patients with MSI cancers, with confirmed partial responses in seven. The study selected two doses for further testing, while tumour biopsies did not provide a conclusive measure of target engagement.
An oral inhibitor designed to disable the Werner syndrome helicase (WRN) produced disease control in 74.2% of evaluable patients with microsatellite instability (MSI) cancers in a first-in-human phase 1 trial. Seven of 66 patients had a confirmed partial response. The study selected two doses of the drug, RO7589831 (also called VVD-133214), for further clinical dose optimisation; it did not identify the maximum tolerated dose.
The results, from an ongoing trial of 88 people with advanced solid tumours, offer early clinical evidence for targeting WRN in MSI cancers. Most patients in the efficacy analysis had stable disease rather than tumour shrinkage, and tumour biopsy samples were insufficient to demonstrate WRN target engagement directly.
Contents
Why target WRN in MSI cancers
MSI can arise when cells have defects in DNA mismatch repair, the system that corrects errors made during DNA replication. WRN is a DNA-processing enzyme that helps maintain genome stability. Preclinical research cited by the trial authors found that MSI cancer cells depend on WRN to manage unusual DNA structures during replication, making them vulnerable to WRN inhibition. This kind of selective dependence is called synthetic lethality: disrupting a process that a particular cancer cell relies on can harm that cell more than cells without the same defect.
RO7589831 is a small molecule that binds WRN irreversibly at the Cys 727 residue and inhibits its activity. The phase 1 study tested the drug as an oral treatment in 21-day cycles in patients whose advanced MSI and/or mismatch-repair-deficient tumours had progressed after standard treatment. Of the 88 participants, 84 had previously received at least one immune-checkpoint inhibitor.
What the phase 1 trial found
The efficacy results were reported for 66 patients with MSI tumours who had suitable baseline and follow-up assessments. Disease control—defined as a complete or partial response, or stable disease lasting at least six weeks—was 74.2% (49 of 66). Seven patients had a confirmed partial response under RECIST version 1.1, an objective response rate of 10.6%; 42 had stable disease and 17 had progressive disease as their best response.
The seven responses occurred in patients with colorectal, endometrial or ovarian cancer. Five of the seven responders were still in response at the data cutoff of 9 September 2025; the estimated median duration of response was 10.2+ months. Median progression-free survival among the 66 patients was 6.7 months. These findings come from a dose-escalation phase 1 study, rather than a comparison with another treatment.
Treatment-emergent adverse events were reported in 87 of 88 patients. The most frequent were nausea (54.5%), diarrhoea (43.2%), fatigue (39.8%), anaemia (37.5%) and vomiting (30.7%). Grade 3 or higher anaemia occurred in 11.4%; no grade 5 treatment-emergent adverse events were reported. Three patients (3.4%) stopped RO7589831 because of a treatment-emergent adverse event. One dose-limiting toxicity—grade 2 nausea—occurred in a patient receiving 600 mg three times daily.
Dose selection and evidence limits
The trial did not reach its protocol-defined threshold for identifying a maximum tolerated dose. Researchers selected 150 mg twice daily and 600 mg twice daily as recommended phase 2 doses for dose optimisation. The study authors said the more frequent three-times-daily schedules did not provide a noticeable improvement in efficacy and added to the pill-taking burden.
Exploratory tests offered other signs of biological activity: 26 of 53 patients assessed with FDG-PET had a partial or complete metabolic response, and early circulating tumour DNA reductions were associated with radiological disease control. These measures are exploratory; the trial did not establish them as substitutes for clinical outcomes.
Direct target engagement in tumour tissue could not be demonstrated. Of 26 paired biopsy sets collected before and during treatment, only five had enough tumour cells for evaluation. The authors attributed the difficulty largely to mucin in the samples. A second part of the trial is enrolling patients with advanced MSI-high or mismatch-repair-deficient colorectal cancer to test two doses of RO7589831 with bevacizumab.