First-in-human study of MRG007 (ARR-217), a CDH17-directed exatecan ADC, in patients with locally advanced or metastatic solid tumors

MRG007

Patient / Population Intervention / Exposure Comparison Outcome

In patients with pretreated advanced colorectal cancer, the CDH17-directed exatecan ADC MRG007 produced an objective response rate of 44.8% at the preliminary recommended phase 2 dose, with a manageable safety profile.

N

123patients

Design

First-in-human phase 1 study in China and the US with dose escalation and backfill followed by dose optimisation and expansion cohorts, in patients with advanced or metastatic solid tumours after standard therapies (113 of 123 with colorectal cancer) receiving the CDH17-directed exatecan ADC MRG007 intravenously every 3 weeks (NCT07066657).

Endpoint

Safety in dose escalation. Objective response rate per RECIST v1.1 in expansion.

ResultIn efficacy-evaluable colorectal cancer patients at the preliminary recommended phase 2 dose, ORR was 44.8% (13/29; 95% CI 26.4%-64.3%), including 7 confirmed PRs and 4 PRs pending confirmation, and DCR was 93.1% (27/29). ORR was 50.0% (4/8) in RAS wild-type and 42.1% (8/19) in RAS-mutant colorectal cancer. Across the pooled higher dose levels, ORR was 34.2% (25/73; 95% CI 23.5%-46.3%). The most common TEAEs were cytopenias and GI toxicities, predominantly grade 1 or 2. No ILD was reported and 1/123 patients (0.8%) discontinued due to a TEAE.

Li J, Gong J, Shen L, et al. Presented at ESMO Congress 2026, abstract LBA34, 25 October. Not yet published.

Discussion & critique

This early single-arm phase 1 study shows encouraging response rates in pretreated colorectal cancer, with a low discontinuation rate and no ILD reported. The headline response rate comes from only 29 patients and includes partial responses still awaiting confirmation. The listed confirmed and pending responses also do not account for all 13 responders. The abstract reports no duration of response, progression-free survival or overall survival, and no grade 3 or higher adverse event rates. Randomised dose optimisation has only just begun, so the dose and the size of the benefit remain provisional.