GI Oncology Daily Digest

KRAS-G12D is the single most common KRAS variant in pancreatic ductal adenocarcinoma — present in roughly 40% of tumors — and until recently it was considered undruggable: unlike KRAS-G12C, it lacks the reactive cysteine that covalent inhibitors exploit. HRS-4642 is a first-in-class intravenous, liposomal, noncovalent KRAS-G12D inhibitor, and this phase 1b/2 trial (NCT06520488) paired it with the standard gemcitabine plus nab-paclitaxel backbone in advanced KRAS-G12D-mutant disease. In the phase 2 expansion, the confirmed objective response rate was 63.3% (95% CI 43.9-80.1) among 30 treatment-naive patients, at a median follow-up of 12.3 months. For perspective, first-line gemcitabine/nab-paclitaxel alone has historically produced response rates in the low-to-mid 20% range, so a confirmed 63% — if it holds — would be a substantial jump. Safety was driven by the chemotherapy backbone: grade ≥3 treatment-related adverse events occurred in 90.3% of the 31 enrolled patients and were predominantly hematologic, with no treatment-related discontinuations or deaths. The caveats are real and worth stating plainly: this is a small, single-arm expansion cohort at mostly Chinese sites, survival data are immature, and response rate is a surrogate that has disappointed before in pancreatic cancer. But KRAS-G12D has been the field's most-wanted target for a decade, and this is among the first clinical signals that an intravenous G12D-selective agent can meaningfully move the needle in first-line metastatic disease. It warrants close follow-up and randomized confirmation.