Verastem Bets the Frontline on a Chemotherapy-Free KRAS G12D Regimen

Verastem is taking an early KRAS G12D signal into first-line pancreatic cancer, including a chemotherapy-free arm, before it has an overall-survival result of its own. In an Oct 14, 2026 release, Verastem Oncology (Nasdaq: VSTM) reported initial efficacy and updated safety from TARGET-D 101, a Phase 1/2 trial of VS-7375, an oral KRAS G12D dual ON/OFF inhibitor, in advanced KRAS G12D-mutated solid tumors. Data cutoff was Oct 7, 2026; more than 200 patients were enrolled, and efficacy was assessed in modified intent-to-treat groups dosed on or before July 22.

At 900 mg once daily in metastatic pancreatic ductal adenocarcinoma, the confirmed objective response rate was 50% (4 of 8) in second line. Across second line and later (n=22, median follow-up 6.4 months, median two prior lines), the rate was 41% (9 of 22), including two complete responses (one confirmed, one unconfirmed with treatment ongoing). In third and fourth line it was 36% (5 of 14). Disease control was 86%, six-month landmark progression-free survival 67%, and six-month landmark overall survival 80%, with 16 of 22 patients still on treatment. Chief executive Daniel Paterson said no target lesions grew among all 22 patients at 900 mg. Those figures are still early and uncontrolled, and the second-line denominator is eight patients. Verastem still set 900 mg as the go-forward dose after a weaker 600 mg look (n=27).

Molecular clearance, then a frontline bet

Fifteen of 17 patients with elevated baseline CA19-9 at 900 mg saw the marker fall by more than half, and all 12 evaluable second-line and later pancreatic patients cut KRAS G12D variant allele frequency by at least 50% by cycle 3 day 1, with 11 of 12 reaching 90% or deeper and 7 of 12 (58%) clearing it. TARGET-D investigator Dr. Manuel Hidalgo of NYU Langone Health said the clinical activity is complemented by those deep variant-allele reductions and ctDNA clearance in more than half of evaluable patients, and that the tolerability profile provides a foundation for combination regimens and earlier settings.

The company said it had aligned with FDA on TARGET-D 301, a first-line metastatic pancreatic Phase 3 with three arms: a chemotherapy-free regimen of VS-7375 900 mg plus an anti-EGFR antibody; VS-7375 900 mg plus gemcitabine and nab-paclitaxel (three weeks on, one week off); and investigator choice of standard-of-care chemotherapy. Dual primary endpoints are progression-free survival and objective response rate; overall survival is secondary. The trial opens by year-end 2026. No FDA-approved therapy specifically targets KRAS G12D, the most prevalent KRAS mutation at 26% of all KRAS mutations, occurring most commonly in pancreatic (40%), colorectal (15%), and non-small cell lung (5%) cancers.

A rival already has overall survival

On Oct 15, 2026, EMA's CHMP recommended EU marketing authorization for Rasonque (daraxonrasib, Revolution Medicines), an oral RAS inhibitor for metastatic pancreatic adenocarcinoma after at least one prior treatment. In a Phase 3 open-label study of 500 patients, 92% with a confirmed RAS G12 mutation, median overall survival was 12.4 months versus 6.7 months with chemotherapy, and median progression-free survival was 7.2 versus 3.6 months. Verastem is asking first-line regulators to accept progression-free survival and response as co-primary endpoints where a competitor has shown an overall-survival benefit one line later in a much larger controlled trial.

Lung, colon, and a safety window

In second-line and later non-small cell lung cancer, 600 mg monotherapy produced a confirmed 43% response rate (10 of 23) and six-month landmark progression-free survival of 72%. A first-line triplet with carboplatin/pemetrexed plus pembrolizumab produced responses in 6 of 8 patients (75%), and all five evaluable second-line and later lung patients with detectable baseline ctDNA cleared it. In third-line and later colorectal cancer, 900 mg plus anti-EGFR produced a 33% response rate (2 of 6, one confirmed, one unconfirmed).

At a Sept 29, 2026 cutoff in 48 patients on 900 mg monotherapy, 2% discontinued for adverse events and dose intensity ran at 97%. Treatment-related events were mostly low-grade nausea, vomiting, and diarrhea; Grade 3 events hit 10% of patients, with only 2% any Grade 3 gastrointestinal event. Two rash cases stayed no greater than Grade 2, with no mucositis, stomatitis, cytopenias, or meaningful liver-function abnormalities. CHMP listed diarrhea, rash, stomatitis, nausea, and vomiting among daraxonrasib's most common side effects. The datasets are not head-to-head, but the absence of mucositis and stomatitis at 900 mg is the practical opening for EGFR combinations and a chemotherapy-free frontline arm.

Cash covers the start, not the readout

Three Phase 2 registration-directed trials are running, and FDA meetings on first-line colorectal and lung designs are planned before year-end. In its second-quarter 2026 update, Verastem said pro forma cash, expected AVMAPKI FAKZYNJA CO-PACK revenue, and a future Oberland facility tranche should extend its runway into the second half of 2027. That covers the planned year-end start of TARGET-D 301, not a full Phase 3 readout. The pancreatic 900 mg group is 22 patients, second-line confirmed responses rest on eight, and dual progression-free-survival-and-response primary endpoints may not satisfy reviewers who can point to a rival's controlled overall-survival result one line later. The activity looks real. The registrational leap is what remains unproven.