Emiltatug ledadotin (Emi-Le)

All clinical data for emiltatug ledadotin in ACC come from interim analyses of the dose escalation and backfill cohorts of NCT05377996, a first-in-human, open-label, single-group phase 1/2 study. There is no randomized or concurrent comparator, responses were assessed by investigators, and independent central review has not been reported. Results are available only as a congress abstract, a sponsor press release, conference news reports, and earlier company disclosures; no peer-reviewed full publication was identified. Participants with ACC received 57.4 to 89 mg/m2 every 3 or 4 weeks, which differs from the single regimen selected for phase 2 (80 mg/m2 every 4 weeks, capped at 160 mg).

The two ASCO 2026 sources report different data sets. The abstract (data cut-off October 1, 2025) reports 221 participants dosed across tumor types, 35 with ACC (median age 58 years, 57% female, median 1 prior line of therapy), and 25 evaluable participants with ACC, in whom the objective response rate was 40% (10 of 25; 9 confirmed and 1 unconfirmed), the disease control rate was 76% (19 of 25), and median progression-free and overall survival had not been reached. The oral presentation on June 1, 2026 reports 180 participants in dose escalation and backfill and 48 with ACC; in 45 evaluable participants with ACC, the objective response rate was 35.6% (95% CI 21.9% to 51.2%) with 1 complete response, the disease control rate was 82.2%, and the median duration of response was 7.4 months. The data cut-off of the oral presentation is not stated in the accessible sources, and the lower total count (180 compared with 221) with a higher ACC count (48 compared with 35) is not explained. At the earlier March 8, 2025 data cut-off, the objective response rate was 56% (5 of 9 evaluable participants with ACC-1).

The analysis closest to the target population is post hoc. A stricter definition of aggressive ACC required a histopathologic phenotype (solid or basaloid histology, or high-grade transformation) and at least one clinical risk factor. Of 48 participants with ACC, 13 were excluded (10 because histology was unavailable), leaving 35, of whom 32 were evaluable. In these 32 participants, the objective response rate was 46.9% (95% CI 29.1% to 65.3%) with 1 complete response, the disease control rate was 81.3% (95% CI 63.6% to 92.8%), the median time to response was 2.7 months, the median duration of response was 6.4 months (95% CI 3.1 to 9.5), and the median progression-free survival was 7.8 months (95% CI 4.7 to 11.6). A patient-directed news report described progression-free survival “approaching eight to nine months”. Sources disagree on response confirmation: the investigator was quoted as saying that “all responses have now been confirmed”, whereas another conference report described the 35.6% and 46.9% rates as combined confirmed and ongoing unconfirmed responses. B7-H4 expression (median tumor proportion score 77.5% in 40 evaluated participants) did not clearly predict benefit, and responses occurred with expression as low as 2%. Overall survival, health-related quality of life, and response by dose level have not been reported.

The safety data also differ by source. In the abstract safety set (221 participants), the most common treatment-related adverse events were proteinuria (49.8%), transient AST increase (49.3%), and fatigue (41.2%); grade 3 AST increase and grade 3 proteinuria each occurred in 17.6%, 3.6% discontinued treatment because of treatment-related adverse events, and no treatment-related deaths occurred. In the oral presentation (180 participants), the corresponding rates were AST increase 56%, proteinuria 54%, fatigue 42%, and nausea 33%, with grade 3 AST increase in 20%, grade 3 proteinuria in 23%, no grade 4 or 5 events, and discontinuation in 3.9%. Any-grade treatment-related adverse events occurred in 88.9% of all participants and 100% of participants with ACC, and grade 3 events in 51.1% and 79.2%, respectively. Among the 48 participants with ACC, treatment-related adverse events led to discontinuation in 1 (2.1%), dose reduction in 45.8%, and dose delay in 56.3%. Proteinuria, characterized on post hoc biopsy as glomerular podocytopathy, was the leading cause of dose modification; after a protocol amendment that added monitoring and renal-protective medications, proteinuria-related dose reductions decreased from 38.9% to 20% and delays from 50.0% to 6.7% among participants with ACC. AST elevations resolved to baseline or grade 1 by the next dose, with no hepatic failure or Hy's law cases. Neutropenia (2.2%), keratitis (3.9%), and neuropathy (2.8%) were infrequent, and no alopecia or interstitial lung disease was reported. Earlier in dose escalation, responses in several participants at high doses were not confirmed after proteinuria-related dose delays, and 115 mg/m2 was considered a non-tolerated dose.

Historical outcomes in recurrent or metastatic ACC provide the only context for these results. A meta-analysis of 17 VEGFR inhibitor studies (560 participants) reported a pooled objective response rate of 6%. Individual studies reported 15.6% (5 of 32 evaluable participants) and 11.5% (3 of 26) with lenvatinib and 21.7% (10 of 46) in a real-world lenvatinib cohort of 51 participants; 9.1% (3 of 33) with axitinib in ACC and 8% (2 of 26) with axitinib in a single-center trial in salivary gland carcinomas that did not meet its primary endpoint; 18% with axitinib plus avelumab (28 evaluable participants); 15.3% by investigator and 9.7% by blinded independent review with rivoceranib (72 evaluable participants); and 11% (2 of 19 evaluable participants) with sorafenib. In the randomized trial of axitinib versus observation (60 participants allocated, 54 evaluable), the objective response rate was 0.0% in both groups while the 6-month progression-free survival rate was 73.0% versus 23.0%, the disease control rate was 100.0% versus 51.9%, and median progression-free survival was 10.8 versus 2.8 months (hazard ratio 0.25); median overall survival was not reached with axitinib and was 27.2 months with observation (P = 0.226). Median progression-free survival in the other comparator studies ranged from 5.5 months (axitinib in salivary gland carcinomas) to 17.5 months (the 32-participant lenvatinib phase II trial), and reported median overall survival was 16.1 months in the real-world lenvatinib cohort, 19.6 months with sorafenib, 26.2 months with axitinib in salivary gland carcinomas, and 27 months in the 26-participant lenvatinib phase II trial. With axitinib plus avelumab, median progression-free survival in ACC-I was 1.8 months. Higher response rates were reported with apatinib in a single-arm trial in China (46.2% in 65 evaluable participants) and with CAP chemotherapy (25%). Toxicity in these comparator studies was frequent: grade 3 adverse events occurred in approximately half of participants in the 26-participant lenvatinib phase II trial and in 13 of 23 participants (57%) receiving sorafenib; dose reductions were required in 24 of the 28 participants enrolled in that lenvatinib trial, in 74% of participants receiving sorafenib, and in 32 of 51 participants (63.6%) in the real-world lenvatinib cohort; and median dose intensity with axitinib in salivary gland carcinomas was 59%. None of these studies restricted enrollment to aggressive ACC, and apart from the subtype analysis of the axitinib plus avelumab trial none reported efficacy by histologic or molecular subtype; the randomized axitinib trial states that it did not capture detailed information on pathologic subtype. Response assessment methods also differed, so comparisons with the phase 1 results are indirect.

EMBLEM-1, the phase 2 part of NCT05377996, enrolls participants with aggressive ACC and measurable disease per RECIST v1.1 who meet both the histopathologic and clinical criteria, and administers 80 mg/m2 (capped at 160 mg) every 4 weeks. The primary endpoint is investigator-assessed objective response rate, with duration of response, disease control rate, progression-free survival, overall survival, and safety as secondary endpoints. EMBLEM-1 has reported no results, its planned sample size and statistical hypotheses are not public, and the registry lists an estimated enrollment of 360 participants for the entire study and an estimated primary completion date of February 2027.