- Design
- phase 1-2 multicentre dose-escalation and dose-expansion study, single group, primary endpoint safety
- Population
- 136 patients with previously treated advanced RAS-mutant non-small-cell lung cancer treated at 300 mg or less, data cutoff 21 July 2025
- Primary outcome
- safety; objective response by RECIST 1.1 as a secondary endpoint
- Effect
- objective response 31% at 120 mg or less, 34% at 160-220 mg, 37% at 300 mg; grade 3 or higher adverse events in 54%, four grade 5 events
RAS mutations drive around 30% of non-small-cell lung cancer, and until recently only the KRAS G12C subset had a targeted option. Daraxonrasib is an oral inhibitor of the active, GTP-bound form of multiple RAS isoforms, mutant and wild-type, which in principle covers far more of that group.
This was a phase 1-2 dose-escalation and dose-expansion study, not a randomised comparison, and its primary endpoint was safety. It enrolled patients with previously treated advanced RAS-mutant NSCLC and gave daraxonrasib at 10 to 400 mg once daily in 21-day cycles. By the July 2025 data cutoff, 136 patients treated at 300 mg or less had been assessed.
Objective response occurred in 31% at doses of 120 mg or less, 34% at 160 to 220 mg and 37% at 300 mg. Toxicity was substantial: adverse events of any grade in 99%, with rash, diarrhoea, nausea, vomiting and mucositis each in at least 30%, and grade 3 or higher events in 54%, including pneumonia in 10%, diarrhoea in 9%, rash in 8% and anaemia in 5%. Four grade 5 events occurred.
A response rate above 30% in previously treated RAS-mutant lung cancer is genuinely notable, and the dose-response pattern is consistent. What this study cannot say is anything about survival, since there is no comparator and response rate is a surrogate that has repeatedly failed to predict benefit in oncology. Set that against grade 3 or higher toxicity in more than half of patients and four deaths, and the honest summary is a drug worth taking into randomised trials rather than one whose place is established.
- Test for the full RAS mutation panel in advanced NSCLC, not KRAS G12C alone
- Read a 31-37% response rate as tumour shrinkage, not as evidence that patients live longer
- Expect dermatological and gastrointestinal toxicity to dominate: rash, diarrhoea, nausea, vomiting and mucositis each affected at least 30%
- Note grade 3 or higher events in 54% and four grade 5 events among 136 patients treated
- Refer suitable patients to trials rather than seeking the drug outside them; randomised evidence does not yet exist
The statistics, in plain English
This is a single-arm study, so the 31-37% response rate has nothing to be compared against; historical response rates in previously treated NSCLC are low, which is why the figure attracts attention, but historical comparison is not evidence. Objective response measures tumour shrinkage on imaging at a point in time and says nothing about how long patients live or feel well, and oncology has many examples of high response rates that did not translate. The apparent rise in response with dose is consistent but comes from non-randomised dose groups of modest size, so it suggests rather than establishes a dose-response relationship.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for top clinical updates, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free