A first-in-human study of DYP688, an antibody drug conjugate delivering a direct Gq/11 inhibitor, in patients with metastatic uveal melanoma (MUM) and other GNAQ/11 mutant melanomas.

9509 Background: GNAQ/11 mutations occur in up to 95% of uveal melanomas (UM) and a subset of non-uveal melanomas. Cell surface PMEL17 (gp100) is highly and broadly expressed in melanoma (including UM). SDZ475 (FR900359) is a potent GNAQ/11 inhibitor, however in vivo toxicity has precluded clinical development. DYP688 is an antibody drug conjugate that binds to PMEL17 to deliver the payload SDZ475. Methods: This first-in-human, open-label, multicenter, single-arm study (NCT05415072) of DYP688 in patients (pts) with MUM and other GNAQ/11 mutant melanomas aimed to evaluate safety and tolerability, determine recommended dose(s) (RDs) of DYP688 (primary objective), and evaluate antitumor activity, pharmacokinetics (PK), and immunogenicity (secondary objectives). Here we present data from the ongoing Phase I dose-escalation. Results: As of 25 Oct 2024, 66 pts were treated with DYP688 at 4 (n=5), 8 (n=12), 12 (n=13), 16 (n=14), and 24 (n=11) mg/kg biweekly (Q2W) and at 12 (n=5) and 16 (n=6) mg/kg once weekly (QW) in 28-day cycles. Tumor types included MUM (n=60) and non-MUM (n=6). Of the 66 treated pts, 60 (90.9%) had prior antineoplastic therapy; 38 (57.6%) received ≥2 lines, and 22 (33.3%) received prior tebentafusp. The majority (n=57, 86.4%) of pts had liver metastases and elevated LDH (n=43, 65.25%) at baseline. Preliminary PK demonstrated a nearly dose-proportional exposure of total monoclonal antibody and active conjugated payload. Most treatment-related adverse events (TRAEs) were grade ≤2 with 4 grade 3 events: hypotension, hypercalcemia, anemia, and increased GGT. One dose limiting toxicity was reported (grade 3 hypotension at 24mg/kg Q2W). Most common TRAEs (all grades/doses, >15%) were hypercalcemia (22.7%), dry mouth (19.7%), fatigue (18.2%) and peripheral edema (16.7%). At data cutoff, 27 (40.9%) pts remained on study treatment and 39 (59.1%) pts had discontinued, mainly due to disease progression (PD) and none due to AEs. Of the 55 pts treated at doses ≥8 mg/kg Q2W who were eligible for RECIST v1.1 evaluation, confirmed objective responses were seen in 12 (21.8%) pts with 1/12 at 8 mg/kg (Q2W), 3/13 (1 complete response) at 12 mg/kg (Q2W), 5/8 at 16 mg/kg (Q2W), 2/5 at 12 mg/kg (QW) and 1/6 at 16 mg/kg (QW), with evidence of deepening response over time. Best response of PD was seen in 6/55 (10.9%) pts and stable disease in 35/55 (63.6%) pts. Median (range) duration of treatment by Kaplan Meier was 7.0 (<1 – 20.7) months. Analysis of mutational profiles from tissue and circulating tumor DNA is ongoing. Conclusions: DYP688 shows favorable safety and tolerability at all doses tested and promising preliminary clinical efficacy at doses ≥ 12mg/kg Q2W; the RDs for dose optimization are yet to be declared and dose exploration is ongoing. Clinical trial information: NCT05415072 .