Why This Study Matters
Checkpoint inhibitors targeting PD-1, PD-L1, and CTLA-4 have transformed outcomes in advanced melanoma, but a substantial share of patients eventually develop primary or secondary resistance. Once resistance sets in, options narrow quickly: published salvage regimens have reported objective response rates of roughly 4%-31% and median progression-free survival of only 2.6-5.0 months. Tumor-infiltrating lymphocyte therapy, approved for this setting in February 2024, is not available to every patient and carries a meaningful hematologic toxicity burden.
Ceralasertib, an oral inhibitor of the ATR (ataxia-telangiectasia and Rad3-related) DNA damage response kinase, was hypothesized to help resensitize checkpoint-refractory tumors to immunotherapy. MONETTE was designed to test that hypothesis in a global, randomized phase II population — and its result is an instructive negative one.
Study Design
MONETTE (NCT05061134) was a randomized (2:1), open-label, multicenter, international phase II study conducted across 75 sites in 11 countries spanning North America, Europe, and Asia, per the Methods section of the published report. (The article’s Acknowledgments section separately credits 58 sites in 11 countries for investigator and site-personnel contributions — a discrepancy with the Methods-section figure that we flag here rather than reconcile ourselves.) Eligible patients had unresectable or metastatic cutaneous, acral, or mucosal melanoma with confirmed progression on at least one prior anti-PD-(L)1 regimen, with or without anti-CTLA-4, and ECOG performance status 0-1.
The primary endpoint was objective response rate by blinded independent central review per RECIST v1.1, with a prespecified minimum threshold of 23% for the combination arm. Key secondary endpoints included progression-free survival, overall survival, and safety.
Patient Population
Between August 11, 2022, and the April 12, 2024 data cutoff, 223 patients were screened and 151 randomized. Baseline characteristics were largely balanced between arms: median age was 63.0 years, 57.6% of patients were male, and 78.1% were White. Melanoma subtype was cutaneous in 61.6% of patients, acral in 15.2%, and mucosal in 14.6%. Nearly all patients (98.7%) had metastatic disease, and all had received prior anti-PD-(L)1 therapy — 63.6% as monotherapy and 30.5% in combination with a CTLA-4 inhibitor. Liver and bone metastases were present in 27.2% and 13.2% of patients, respectively, and about a quarter (25.8%) carried BRAF V600 mutations.
Primary Endpoint Results
| Best Response (BICR) | Ceralasertib + Durvalumab (N=97) | Ceralasertib Monotherapy (N=52) |
|---|---|---|
| Complete response | 3 (3.1%) | 1 (1.9%) |
| Partial response | 6 (6.2%) | 2 (3.8%) |
| Stable disease | 18 (18.6%) | 8 (15.4%) |
| Progressive disease | 58 (59.8%) | 30 (57.7%) |
| Non-evaluable | 12 (12.4%) | 11 (21.1%) |
An investigator-assessed sensitivity analysis was directionally consistent, with an objective response rate of 11.3% (95% CI, 5.8-19.4) for the combination versus 7.7% (95% CI, 2.1-18.5) for monotherapy.
Subgroup Analyses
| Melanoma Subtype | Ceralasertib + Durvalumab | Ceralasertib Monotherapy |
|---|---|---|
| Acral | 14.3% | 0% |
| Mucosal | 15.4% | 11.1% |
No confidence intervals or p-values were reported for these descriptive subtype comparisons. Beyond clinical subtype, the study’s exploratory biomarker program is the more statistically substantive subgroup story. Higher baseline tumor CD8+ T-cell density showed a positive prognostic trend for overall survival in both treatment arms, based on a forest-plot analysis of tumor biomarkers; the source does not provide a single pooled hazard ratio for this trend, so none is reported here. No association was observed between baseline tumor PD-L1 expression and overall survival.
A biomarker-defined split by baseline circulating GDF-15 (growth differentiation factor-15) did report explicit survival figures: in the monotherapy cohort, median overall survival was 10.45 months for patients with high baseline GDF-15 versus not reached for those with low baseline GDF-15; in the combination cohort, median overall survival was 9.1 months for high baseline GDF-15 versus not reached for low baseline GDF-15.
Safety Profile
| Adverse Event | Ceralasertib + Durvalumab (N=97) | Ceralasertib Monotherapy (N=52) |
|---|---|---|
| Any all-causality AE | 92 (94.8%) | 45 (86.5%) |
| Maximum grade ≥3 AE | 34 (35.1%) | 16 (30.8%) |
| Serious AE | 19 (19.6%) | 5 (9.6%) |
| Leading to death | 0 (0%) | 0 (0%) |
| Leading to discontinuation | 1 (1.0%) | 1 (1.9%) |
The most commonly reported adverse events by preferred term were nausea, anemia, asthenia, fatigue, decreased appetite, and vomiting. The most common grade 3 or higher adverse events were anemia, decreased neutrophil count, hypertension, neutropenia, thrombocytopenia, nausea, and increased lipase. No adverse events led to death in either arm, and the authors characterized both regimens as well tolerated, with a safety profile broadly consistent with each drug’s known monotherapy profile.
Interpretation and Broader Context
MONETTE’s final analysis did not clear the bar the trial set for itself: a 9.3% objective response rate fell well short of the prespecified 23% minimum threshold for the combination arm, and the secondary comparison between arms was not statistically significant (OR 1.59; P = 0.492). Numerically, the combination outperformed monotherapy across response rate, progression-free survival, and overall survival, and roughly a third of patients in both arms were alive at 18 months — but none of these differences reached statistical significance in this study.
In a previous phase II trial of ceralasertib plus durvalumab conducted in 30 patients with immunotherapy-resistant metastatic melanoma, clinical activity was generally higher than observed with ceralasertib plus durvalumab in MONETTE — objective response rate 30% versus 9.3%, median progression-free survival 7.1 versus 2.0 months — though median overall survival appeared similar between the studies (14.2 versus 16.0 months).
— Discussion, MONETTE publication (Clin Cancer Res 2026;32:2921-33)
The authors attribute part of that gap to differences in trial design and conduct: MONETTE used blinded independent central review rather than investigator assessment, enrolled a larger and more geographically diverse population, and used a shorter 7-day ceralasertib dosing schedule informed by tolerability data from the phase II HUDSON umbrella study in non-small cell lung cancer.
Taken together, MONETTE indicates that this particular ceralasertib-based regimen, at this dose and schedule, does not deliver a clinically meaningful response rate for the broad population of anti-PD-(L)1-resistant melanoma. The exploratory biomarker findings — tumor CD8+ T-cell density and circulating GDF-15 as overall-survival-associated markers — are hypothesis-generating rather than practice-changing, and the authors themselves frame them as leads for future patient-selection strategies rather than confirmed predictive biomarkers.
In a randomized phase II trial of 151 patients with advanced melanoma resistant to PD-(L)1 blockade, adding the ATR inhibitor ceralasertib to durvalumab (or using ceralasertib alone) did not restore clinically meaningful responses: the combination’s 9.3% objective response rate missed the trial’s prespecified 23% threshold, and no efficacy comparison between arms reached statistical significance. Both regimens were reasonably well tolerated, with no treatment-related deaths. Exploratory biomarker signals — tumor CD8+ T-cell density and circulating GDF-15 — may help identify patients more likely to benefit, but the authors frame these as hypothesis-generating leads for future trials rather than a basis for current treatment decisions.