Why This Study Matters
The NADINA and SWOG S1801 trials established neoadjuvant immune checkpoint inhibitors (ICIs) as the emerging standard of care for resectable stage III melanoma, but both enrolled patients without prior ICI exposure. A consistent proportion of these patients relapse after prior (neo)adjuvant ICI treatment and have never been thoroughly investigated in previous trials. Daromun is an intralesional immunocytokine that combines the T-cell–stimulating properties of IL-2 with the proinflammatory and vascular-disrupting effects of TNF.
Study Design
A total of 256 adult patients with locally advanced melanoma and at least one injectable skin or nodal lesion, eligible for complete surgical resection, were randomly assigned 1:1 to intralesional neoadjuvant daromun (n = 127; 13 million IU of L19IL2 and 400 μg of L19TNF, once weekly for up to 4 weeks) followed by surgery, or up-front surgery (n = 129). The primary outcome was recurrence-free survival (RFS) by blinded independent central review (BICR). Secondary end points were distant metastasis-free survival (DMFS), overall survival (OS) and safety; a post hoc analysis of event-free survival (EFS) was added. Data cutoff: November 28, 2025; median follow-up 36.8 months from random assignment. Sponsor: not specified in the source (the authors' affiliations include Philogen S.p.A. and Philochem AG). Number of sites: not specified in the source.
Patient Population
Patients were screened between June 2016 and May 2023 (364 screened). Thirty-four (13%) had newly diagnosed stage III disease and 222 (87%) had relapsed after surgery with or without radiotherapy and/or systemic treatment. Prior systemic therapy was reported in 34.6% (44 of 127) of the daromun arm and 32.6% (42 of 129) of the control arm; prior ICI in 21.3% and 15.5%. Post-surgery adjuvant treatment was given to 32.3% (41 of 127) and 41.8% (54 of 129).
Primary Endpoint Results
At the updated analysis, daromun followed by surgery improved RFS versus up-front surgery (HR, 0.55 [95% CI, 0.38 to 0.78]; P < .001). Median RFS was 23.8 months with daromun versus 6.5 months with surgery alone; 1-, 2- and 3-year RFS rates were 63.5%, 49.2% and 35.7% versus 38.6%, 25.5% and 16.6%. For comparison, the original analysis at a 21-month median follow-up reported HR 0.59; P = .005.
| Endpoint (ITT) | HR (95% CI) | P | Median, daromun + surgery | Median, surgery |
|---|---|---|---|---|
| RFS (primary) | 0.55 (0.38–0.78) | <.001 | 23.8 mo | 6.5 mo |
| DMFS | 0.53 (0.33–0.83) | .005 | 38.7 mo | 14.0 mo |
| EFS (post hoc) | 0.71 (0.51–0.98) | .034 | 16.1 mo | 6.1 mo |
OS data were still immature (37 of 104 events recorded).
Subgroup Analyses
The authors describe the subgroup analyses as exploratory. In the recurrent subgroup, a per-protocol analysis showed RFS HR 0.50 (95% CI, 0.34 to 0.75) and DMFS HR 0.43 (95% CI, 0.26 to 0.72); exploratory EFS favored neoadjuvant therapy (HR 0.68 [95% CI, 0.48 to 0.97]). By treatment history, EFS showed “only a strong trend without reaching statistical significance.”
| Subgroup | RFS HR (95% CI) | Median RFS (daromun vs surgery) | EFS HR (95% CI) | Median EFS (daromun vs surgery) |
|---|---|---|---|---|
| Recurrent (n = 222) | 0.50 (0.34–0.75) | 23.8 vs 6.1 mo | 0.68 (0.48–0.97) | 16.1 vs 6.1 mo |
| Prior surgery + systemic therapy (n = 86) | 0.53 (0.29–0.95) | 16.7 vs 5.9 mo | 0.68 (0.40–1.15) | 11.8 vs 5.9 mo |
| Prior surgery ± RT only (n = 136) | 0.46 (0.27–0.78) | 28.0 vs 8.4 mo | 0.65 (0.40–1.05) | 23.5 vs 7.6 mo |
Safety Profile
No new safety signals of concern were recorded.
— Hauschild et al., Journal of Clinical Oncology, 2026
| Event (safety-evaluable) | Daromun arm (n = 122) | Control arm (n = 124) |
|---|---|---|
| Grade ≥3 AEs | 33.6% (41) | 11.3% (14) |
| Grade 3 treatment-related AEs | 28.7% (35) | 7.3% (9) |
| Grade 3 injection site reaction | 12.3% (15) | Not specified in source |
There were no CTC grade >3 treatment-related AEs and no treatment-related deaths.
Interpretation and Broader Context
The authors conclude that the update confirms clinically and statistically meaningful improvements in RFS and DMFS in locally advanced recurrent melanoma, and that the findings extend neoadjuvant strategies to recurrent patients, whether or not they received prior systemic treatment. A parallel phase 3 neoadjuvant study (NeoDREAM; NCT03567889) is ongoing, and future work should evaluate predictive biomarkers and integration with checkpoint blockade and/or targeted therapy.
Limitations
The authors list the open-label design among the study's limitations. They describe the subgroup analyses as exploratory, and the EFS analysis was added post hoc rather than prespecified. OS data were still immature at this cutoff (37 of 104 events recorded), so the trial does not establish whether the RFS and DMFS gains translate into longer survival. Grade 3 or higher adverse events were more frequent with daromun (33.6% vs 11.3%). The number of sites is not specified in the source.