Phase 1A/B study of AB248, a CD8+ selective IL-2 mutein fusion protein, alone or in combination with pembrolizumab, in patients with advanced solid tumor malignancies.

2616 Background: High-dose IL-2 demonstrates modest activity in melanoma and RCC but its use is limited by severe toxicity. AB248 is a novel IL-2 fusion protein of an attenuated IL-2 mutein linked to an antibody targeting CD8β that features >500-fold selectivity for CD8+ T cells and shows strong anti-tumor activity both alone and with anti-PD1 in preclinical models. In addition to CD8+ T cell expansion and activation, the drug avoids NK cell toxicity and Treg-mediated immunosuppression. Methods: NCT05653882 is a phase 1A/B study investigating the safety, pharmacokinetics (PK), pharmacodynamics (PD) and anti-tumor activity of AB248 alone or with pembrolizumab in locally advanced/metastatic solid tumor malignancies, including melanoma, having previously progressed through PD-1/PD-L1 checkpoint blockade. The study’s primary objective was to assess the safety and tolerability of AB248 alone and in combination with pembrolizumab. Results: As of December 1, 2025, 61 patients (pts.) were treated with monotherapy (MT) at 6 dose levels across 3 schedules (Q2W at 0.02mg/kg-0.75mg/kg; QW at 0.15mg/kg; Q3W at 0.3mg/kg). 68 additional pts. were treated with combo therapy (CT) at Q3W 0.15-0.3mg/kg and step-up dosing. The most common TEAEs (> 95% G1-2) among all pts. were fatigue (50%), rash (49%), nausea (43%), chills (33%), pyrexia (32%), vomiting (32%) and diarrhea (30%). Maximum tolerated doses (MTDs) were 0.5 mg/kg for MT Q2W and 0.15 mg/kg for CT Q3W. An MTD was not reached for CT step-up dosing. 41 cutaneous (cut) and mucosal (muc ) melanoma pts. were evaluable for response across MT and CT dose cohorts, with 95% of pts. having received prior IO doublet therapy (anti-PD1 + anti-CTLA or anti-LAG3) and 83% having received ≥ 2 anti-PD1 regimens. Among evaluable cut melanoma pts., 2 confirmed PRs (18%) were observed among 11 pts. enrolled at higher MT dose levels (net tumor reductions of 78.8% and 87.2%) and 2 confirmed PRs (12%) were observed among 16 CT pts. (37.5% and 57.5%). Among 11 muc melanoma pts., 3 confirmed PRs (27%) were observed in CT dosing. One additional muc melanoma pt. received 0.5 mg/kg MT and remained on study with SD > 15 mos. AB248 exhibited dose-proportional PK that is comparable between MT and CT dosing. Peripheral PD data demonstrated robust and preferential CD8+ T cell expansion, with >20X expansion in MT (0.3 mg/kg and above) and >11X expansion in CT dosing (0.15 mg/kg and above). Further data on safety, PK, PD, and preliminary anti-tumor activity will be presented. Conclusions: AB248 demonstrates robust anti-tumor activity across heavily IO-pretreated patients with metastatic melanoma (incl. mucosal melanoma). Combined with an acceptable safety profile, favorable PK and differentiated PD, the drug’s anti-cancer activity merits further investigation. Clinical trial information: NCT05653882 .