Perspectives on the Science
Neoadjuvant Intralesional Daromun in Resectable Melanoma: Interpreting the PIVOTAL Trial
2026
Melanoma
Neoadjuvant Therapy
Intralesional Therapy
Clinical Trial Design
This Perspectives on the Science article examines the phase III PIVOTAL trial of neoadjuvant intralesional daromun in resectable melanoma, focusing on the population studied, interpretation of trial endpoints, systemic activity of an injectable therapy, and the potential place of daromun in the contemporary melanoma treatment landscape.
Featured Article
Hauschild A, Hassel JC, Ziemer M, et al. Neoadjuvant Intralesional Daromun (L19IL2/L19TNF) in Resectable Locally Advanced Melanoma: An Update on the Efficacy and Safety Results of the PIVOTAL Phase III Trial. Journal of Clinical Oncology (2026).1
Introduction
On September 25, 2026, The Readout, the Society of Cutaneous Oncology’s Journal Club series, discussed the updated phase III PIVOTAL trial of neoadjuvant intralesional daromun (L19IL2/L19TNF) in resectable stage III melanoma1. Forty-one participants joined the session, including clinicians, investigators, trainees, and members of the study team.
Daromun combines two tumor-targeted immunocytokines, bifikafusp alfa (L19IL2) and onfekafusp alfa (L19TNF). The L19 antibody fragment recognizes the extra-domain B of fibronectin, an angiogenesis-associated extracellular-matrix isoform enriched around tumor blood vessels, and is fused to interleukin-2 or tumor necrosis factor-α. Intralesional administration is intended to concentrate these immune-active cytokines within the tumor microenvironment while limiting systemic exposure2.
PIVOTAL was initiated before neoadjuvant immunotherapy became established in melanoma. Since then, SWOG S1801 (neoadjuvant pembrolizumab) and NADINA (neoadjuvant ipilimumab plus nivolumab) have changed the management of resectable stage III disease3,4. PIVOTAL therefore enters a treatment landscape that differs substantially from the one in which it was designed.
The pre-meeting survey reflected uncertainty about how to interpret the study in that context. Among 50 respondents, 46% were unsure how compelling the evidence was or felt they needed to understand the data better. After the discussion, 62% of 21 respondents rated the evidence very compelling and 38% moderately compelling (Figure 1). The pre- and post-meeting surveys were aggregate and unpaired and therefore do not represent within-person change.
The discussion focused on the population studied, interpretation of the trial endpoints, the significance of distant disease control after an intralesional therapy, and where daromun might fit among contemporary options for recurrent melanoma. In the post-meeting survey, 67% of respondents said the PIVOTAL results should influence current clinical practice substantially or be practice-changing, while 29% selected moderately and 5% a little (Figure 2).
NoteFigures 1–2. The Readout survey responses
The PIVOTAL Trial
PIVOTAL randomized 256 patients with resectable stage III melanoma and injectable disease to intralesional daromun once weekly for up to four weeks followed by surgery or to upfront surgery1,5. Prior surgery, radiotherapy, and systemic therapy were permitted, and postoperative adjuvant therapy was allowed in both arms.
In the updated analysis at a median follow-up of 36.8 months, daromun followed by surgery improved recurrence-free survival (RFS) compared with surgery alone (HR 0.55, 95% CI 0.38–0.78; P<0.001), with median RFS of 23.8 versus 6.5 months1. Distant metastasis-free survival (DMFS) also favored daromun (HR 0.53, 95% CI 0.33–0.83; P=0.005), with median DMFS of 38.7 versus 14.0 months. A post hoc event-free survival (EFS) analysis was directionally consistent, with an HR of 0.71 (95% CI 0.51–0.98; P=0.034) and median EFS of 16.1 versus 6.1 months.
The treatment effect was also evident among the 222 patients with recurrent disease, in whom RFS and DMFS favored daromun with HRs of 0.50 and 0.43, respectively1. Overall survival remains immature.
NoteTable 1. Selected efficacy outcomes from the updated PIVOTAL analysis
| Outcome | Daromun followed by surgery | Surgery alone | Hazard ratio |
|---|---|---|---|
| RFS, median | 23.8 months | 6.5 months | 0.55 |
| DMFS, median | 38.7 months | 14.0 months | 0.53 |
| EFS, median, post hoc | 16.1 months | 6.1 months | 0.71 |
RFS, recurrence-free survival; DMFS, distant metastasis-free survival; EFS, event-free survival. EFS was analyzed post hoc.1
The randomized data establish that a short course of preoperative intralesional daromun can improve disease-control outcomes compared with immediate surgery. The principal questions now concern how the study population, endpoints, and treatment strategy relate to current melanoma practice.
Interpreting PIVOTAL in the Current Treatment Landscape
Population and contemporary treatment options
PIVOTAL was predominantly a trial of recurrent melanoma. Of 256 randomized patients, 222 (87%) had recurrent disease after prior surgery, with or without radiotherapy or systemic treatment; only 34 patients (13%) entered with de novo disease1. This distinguishes PIVOTAL from the major contemporary systemic neoadjuvant trials and makes the study particularly relevant to resectable locoregional recurrence. However, heterogeneity in prior treatment exposure and clinical context limits our ability to identify which patients are most likely to benefit from this strategy.
Recurrent disease in PIVOTAL, however, should not be equated with anti–PD-1-refractory disease. Prior immune checkpoint inhibitor exposure was reported in 47 patients, representing fewer than one in five participants1. The trial therefore provides randomized evidence in a predominantly recurrent population but substantially less evidence for the increasingly common patient whose melanoma recurs after adjuvant or neoadjuvant anti–PD-1 therapy.
The treatment landscape for such patients has also changed. Some patients with resectable recurrence may be considered for systemic neoadjuvant therapy rather than immediate surgery. For patients with unresectable injectable disease progressing after PD-1 therapy, intralesional RP1 (vusolimogene oderparepvec) plus systemic nivolumab is now another option. In August 2026, based on data from the IGNYTE trial, the U.S. Food and Drug Administration granted accelerated approval to the combination for adults with unresectable advanced cutaneous melanoma after progression on a PD-1–blocking regimen6.
IGNYTE is not a direct comparator for PIVOTAL. It evaluated a different intratumoral agent and treatment strategy, combining intratumoral RP1 with systemic nivolumab in a single-arm study of unresectable advanced melanoma7. Its relevance here is therefore contextual rather than comparative: intratumoral therapy now has an established role in the post–PD-1 melanoma treatment landscape. The boundary between resectable and unresectable injectable recurrence may therefore become increasingly important when considering where daromun might fit. Longer-term follow-up, including maturation of overall survival in PIVOTAL and randomized results from the phase III IGNYTE-3 study, should further define the role of these distinct approaches.
RFS and EFS differ by event definition, not simply by time origin
Endpoint interpretation was a major focus of the discussion. PIVOTAL was designed with RFS as its primary efficacy endpoint. Although RFS is commonly associated with the postoperative setting, the PIVOTAL protocol did not begin RFS at surgery.
The protocol defined RFS from randomization, designated Day 0 for both study arms, to the first local, regional, distant, or new primary melanoma recurrence or death2. Thus, the RFS clock began before neoadjuvant treatment and before surgery.
EFS was not included among the prespecified endpoints in the protocol. The updated JCO report introduced EFS as a post hoc analysis designed to capture preoperative events that could prevent definitive surgery1. The distinction between RFS and EFS in PIVOTAL therefore does not primarily concern when follow-up begins. Both are anchored to randomization. The difference is which events during the neoadjuvant interval are counted as treatment failure.
This distinction is clinically relevant. Progression or another event that prevents definitive surgery may represent failure of a neoadjuvant strategy even if it does not satisfy the protocol definition of recurrence. The post hoc EFS analysis attempts to incorporate these events. Its consistency with the RFS result supports the overall treatment effect, but its retrospective definition should remain clear.
The endpoint terminology also reflects the period in which PIVOTAL was developed. Contemporary neoadjuvant melanoma trials have generally adopted EFS to capture outcomes across the perioperative treatment strategy3,4. PIVOTAL instead used a randomization-based RFS endpoint and subsequently examined EFS as a sensitivity analysis.
This distinction also has regulatory relevance. FDA guidance distinguishes EFS from postoperative disease-free survival when randomization occurs before definitive surgery, because EFS can capture progression that precludes surgery in addition to recurrence or death8. In that sense, the post hoc EFS analysis places PIVOTAL in a framework more closely aligned with contemporary neoadjuvant drug development, while its retrospective introduction appropriately limits the inferential weight placed on that analysis.
Systemic activity and the problem of evaluating an injectable therapy
The DMFS result is particularly relevant when interpreting an intralesional treatment. Daromun is administered directly into accessible tumor deposits, yet the randomized reduction in distant metastatic events was similar in magnitude to the RFS effect1. Earlier studies also reported regression of noninjected lesions after intralesional L19IL2/L19TNF9.
The recent regulatory experience with RP1 illustrates a narrower methodological point: why evidence of activity beyond injected lesions matters when evaluating an intratumoral regimen. The published IGNYTE analysis reported a RECIST 1.1 objective response rate of 32.9% among 140 patients with anti–PD-1-failed melanoma, with responses in both injected and noninjected lesions7. FDA review emphasized that an overall response estimate can be difficult to interpret when directly injected lesions contribute to RECIST measurements10,11.
The IGNYTE population used for the subsequent FDA approval therefore differed from the full published cohort. FDA included 91 patients with at least one noninjected lesion in the efficacy-evaluable population and reported an ORR of 24.2% with a median duration of response of 14.1 months6. The published and regulatory estimates answer different questions rather than representing alternative calculations of the same estimand.
What does DMFS tell us that local response cannot?
This distinction is pertinent to PIVOTAL. Regression of an injected lesion is clinically meaningful, but it does not by itself demonstrate systemic activity. DMFS addresses that problem more directly: it measures development of distant metastatic disease and was evaluated in a randomized comparison. Interpretation is complicated, however, by postoperative systemic therapy, which was permitted at physician discretion and was not balanced between the treatment groups. Postoperative adjuvant therapy was received by 41 patients in the daromun arm and 54 in the surgery-alone arm, including immune checkpoint inhibitors in 28 and 38 patients, respectively1. Thus, subsequent therapy could have influenced DMFS. Notably, the imbalance favored greater use of postoperative immunotherapy in the surgery-alone arm, which would not readily explain the observed DMFS advantage with daromun and, if anything, could have attenuated it. Nevertheless, because postoperative treatment was not randomized and may have reflected clinical factors associated with recurrence risk, its effect cannot be determined from these data. The favorable DMFS result therefore supports disease control beyond the directly injected lesions, but it does not establish the biological mechanism responsible for that effect.
Safety and surgical implications
Grade 3 or higher adverse events occurred in 33.6% of safety-evaluable patients assigned to daromun and 11.3% assigned to surgery alone. Grade 3 treatment-related adverse events occurred in 28.7% and 7.3%, respectively, and grade 3 injection-site reactions occurred in 12.3% of daromun-treated patients1. No treatment-related deaths or treatment-related adverse events above grade 3 were reported.
This toxicity profile differs from that of systemic checkpoint blockade but should not be characterized simply as less toxic. Local inflammation, ulceration, and tissue breakdown may be relevant when treatment is administered in a field that will subsequently undergo surgery. At the same time, the clinical significance of these events depends on context. A grade 3 local reaction may be less consequential if it occurs within tissue already scheduled for definitive resection and is effectively resolved by the planned operation. Thus, conventional toxicity grading may not fully capture the practical burden of treatment in a perioperative setting. Conversely, limiting systemic exposure may be attractive for patients in whom additional systemic immune therapy is less desirable.
PIVOTAL was not designed to compare these treatment burdens directly. Surgical delay, wound complications, resection complexity, persistent immune toxicity, treatment duration, and patient experience are likely to be relevant when comparing local and systemic neoadjuvant approaches.
Clinical Placement and NeoDREAM
PIVOTAL does not establish that daromun should replace contemporary systemic neoadjuvant immunotherapy. It does, however, identify a clinically relevant setting in which local therapy may have value: resectable locoregional melanoma with injectable disease and surgery remaining part of curative-intent management.
Where could daromun fit now?
The post–PD-1 setting is of particular interest. RP1 plus nivolumab is now available for unresectable advanced cutaneous melanoma after progression on PD-1 therapy6. PIVOTAL, by contrast, addresses resectable disease but included relatively few patients previously exposed to checkpoint inhibition. Prospective evaluation of daromun in patients with resectable injectable recurrence after anti–PD-1 therapy would therefore address an increasingly important clinical gap.
NeoDREAM (NCT03567889) should provide additional information in a more contemporary perioperative setting. The study compares daromun followed by surgery and postoperative therapy with immediate surgery and postoperative therapy12. The current public registry identifies RFS as the primary endpoint and includes EFS as a prospectively specified time-to-event outcome. During the discussion, the study team emphasized the growing importance of EFS in the NeoDREAM development program following regulatory discussions, consistent with the broader shift toward endpoints that capture the entire neoadjuvant treatment strategy. By incorporating surgery and contemporary postoperative management in both arms, NeoDREAM should help isolate the incremental contribution of the neoadjuvant daromun phase within a modern perioperative treatment strategy.
Priorities for Further Study
Further study should focus on the clinical settings in which daromun could meaningfully alter a contemporary treatment decision.
The first is resectable recurrence after prior anti–PD-1 therapy or BRAF/MEK-targeted therapy. The recurrent subgroup in PIVOTAL is clinically relevant, but it cannot be considered a surrogate for a uniformly treatment-refractory population. This is an increasingly common clinical scenario: patients with injectable, surgically resectable recurrence after adjuvant or neoadjuvant checkpoint blockade or, for BRAF V600-mutant melanoma, after adjuvant BRAF/MEK inhibition. Prospective evaluation in these settings would help define whether daromun can provide meaningful disease control while preserving surgery as part of a curative-intent strategy.
The second is comparative net clinical benefit relative to contemporary systemic neoadjuvant immunotherapy. For patients with resectable injectable disease, the clinically relevant question is no longer simply whether daromun is superior to surgery alone, but how the overall treatment strategy compares with approaches established by SWOG S1801 and NADINA3,4. Cross-trial comparisons cannot establish relative efficacy because the populations, treatment strategies, and endpoints differ. Nevertheless, daromun raises an important hypothesis: whether a locally administered neoadjuvant therapy can achieve comparable disease control while reducing systemic immune toxicity and its potential for persistent morbidity. Future studies should therefore evaluate the total treatment package, including EFS or RFS, distant disease control, overall survival, acute and chronic toxicity, surgical morbidity, treatment duration, need for postoperative therapy, quality of life, and patient preference. The relevant comparison is ultimately not whether daromun produces fewer adverse events, but whether it provides a more favorable balance of efficacy and treatment burden.
The third is the relationship between local treatment and systemic disease control. The DMFS result is one of the more distinctive findings from PIVOTAL. Correlative studies of injected and noninjected lesions, tumor tissue, and blood may help determine which patients develop clinically meaningful systemic immune activity and whether that biology can be predicted before treatment. This question is particularly important for an intralesional therapy, for which local tumor regression and systemic antitumor activity should be considered separately.
A fourth consideration is how intratumoral therapy should be sequenced with systemic immunotherapy. Local treatment may alter the tumor microenvironment in ways that influence subsequent systemic immune responses, but whether such priming improves clinical outcomes remains uncertain. Differences in the activity of immunotherapy across treatment lines underscore the importance of treatment context but do not, by themselves, establish a sequencing effect. The relatively limited systemic exposure of intratumoral approaches such as daromun may provide flexibility to test rational sequencing and combination strategies prospectively. Future studies should determine whether intratumoral therapy is most effective before, during, or after systemic immunotherapy and whether prior treatment exposure modifies its benefit.
Finally, response to daromun may itself become clinically useful. Pathologic response has become relevant to postoperative decision-making in other neoadjuvant melanoma strategies4, but its role after intralesional immunocytokine therapy remains uncertain. Prospective studies should determine whether clinical or pathologic response can support less extensive surgery, de-escalation of postoperative systemic therapy, or identification of patients who require additional treatment. Such studies should also incorporate clearly defined perioperative endpoints and surgical outcomes, including the clinical consequences of local inflammatory toxicity in tissue ultimately removed at planned resection.
Conclusion
PIVOTAL demonstrated that neoadjuvant intralesional daromun followed by surgery improves RFS and DMFS compared with upfront surgery in patients with resectable stage III melanoma and injectable disease1. Its predominantly recurrent population is an important feature of the study, although relatively few patients had received prior immune checkpoint inhibition.
The September 2026 session highlighted several aspects of the trial that are particularly relevant to contemporary practice. PIVOTAL RFS was measured from randomization rather than surgery, whereas the later EFS analysis broadened the definition of failure during the preoperative interval. The randomized DMFS result is also important for an injectable therapy because it provides evidence of disease control beyond the directly treated field.
The context for intratumoral therapy has changed further with the approval of RP1 plus nivolumab after PD-1 progression in unresectable melanoma. PIVOTAL and IGNYTE evaluate different treatment strategies in different disease settings, but both reinforce the need to define where locally administered therapies fit within modern melanoma treatment. For daromun, the most informative next data will come from patients and treatment sequences that reflect current practice.
Materials and Methods
Bibliography
Appendix
Share Your Thoughts
Join the conversation
What did this article make you think about?
The Society of Cutaneous Oncology welcomes thoughtful reactions, questions, critiques, and ideas from readers. Share a brief note with the SoCO and Journal of Cutaneous Oncology team.