Rasonque (daraxonrasib): the FDA has approved the first broad RAS(ON) inhibitor — what it changes and what it does not

On 26 August 2026 the FDA approved Rasonque (daraxonrasib) from Revolution Medicines — for adults with metastatic pancreatic adenocarcinoma who have received at least one prior line of systemic therapy or are not candidates for multi-agent chemotherapy. It is the first approved drug in the class of broad RAS(ON) inhibitors.

The framing first, because it governs everything else. This is late-stage oncology, not the slowing of ageing. Median overall survival rose from 6.7 to 13.2 months — a grim prognosis became less grim, and that is not a victory over cancer. The significance lies elsewhere, and it is real: RAS was considered an "undruggable" target for decades, and this is the first randomised evidence to the contrary.

What the trial showed

The registrational trial is RASolute 302 (NCT06625320), open-label, randomised, 500 patients, 1:1. The control arm was not a single regimen but the investigator's choice of four standard second-line options; in practice almost nine out of ten received gemcitabine plus nab-paclitaxel, or liposomal irinotecan with 5-FU. Crossover between arms was not permitted.

The trial had two primary populations: patients with a RAS G12 mutation (91.8% of them) and the overall population. News coverage usually quotes the overall-population figures — here are both.

Three clarifications that get lost in retellings:

• HR 0.40 refers to overall survival, and to nothing else. For progression-free survival the risk reduction is smaller: 0.45 and 0.49.
• The "30% versus 11%" figures circulating in the news are rounded. The publication reports 31.6 versus 11.2 in the overall population and 33.2 versus 11.8 in the primary one; the FDA summary gives a rounded 30% (95% CI 25–36) versus 11% (7–15). The difference is small, but it shows how a number reaches the reader.
• This is the first interim survival analysis. Median follow-up is 8.5 months, and the upper bound of the confidence interval for median survival on the drug has not yet been reached. The trial is ongoing; final data are not expected before 2027.

Published in the New England Journal of Medicine (2026;395:325–337, DOI 10.1056/NEJMoa2605555) simultaneously with the plenary presentation at ASCO 2026.

What it costs

Toxicity in this class is not fine print at the end — it is frequent, and the label itself names it:

Serious adverse events occurred in 30% of patients. 2.9% discontinued treatment entirely because of adverse events. Rash prophylaxis is part of the label. The label lists no contraindications; it does carry a warning about embryo-fetal toxicity.

What level of evidence this is

It is worth pausing here, because for our audience this matters more than the numbers themselves. A randomised phase III trial with a hard endpoint — overall survival — followed by regulatory approval sits at the very top of the evidence scale. This is a fundamentally different class of data from "extended life in mice", "improved markers in organoids" or "a correlation in an observational study". Death is an endpoint that cannot be measured in a convenient way and cannot be redefined after the fact.

And precisely for that reason it is worth naming the limitations the investigators named themselves.

What the criticism says — including from the authors

• The trial is open-label. Patients and physicians knew who was receiving what. In the control arm 15.1% received no chemotherapy dose at all — mostly having declined after learning their assignment; in the drug arm that figure was 2.8%. The analysis included all randomised patients, untreated ones among them. An independent review (Targeted Oncology, 26.08.2026) notes that such an imbalance in an open-label trial may inflate the observed benefit.
• This is second line, not first. The official ASCO discussant, Jennifer Knox, called the survival curve "absolutely beautiful" — and immediately pointed to the main practical flaw: a substantial share of patients with metastatic pancreatic cancer simply do not survive to a second line, or are too frail by the time they reach it. A drug available in the second line does not help those who never get there. First-line trials are running; there are no results yet.
• Resistance has already been described. In paired samples from 40 patients, acquired alterations were found in 45% — most often secondary KRAS Y64 mutations that disrupt drug binding.
• Small subgroups prove nothing. There were 41 patients without a G12 mutation; their results are descriptive, with wide intervals, and for progression-free survival the direction in that subgroup was not even in the drug's favour.
• None of the investigators calls this a cure. The lead investigator's phrasing: the next goal is "durable responses and ultimately cure" — that is, a goal, not an achieved result.

Why this matters beyond one tumour

RAS is the most common oncogene family in human cancer and the founding event in pancreatic cancer. For forty years it was considered undruggable: the protein has no convenient binding pocket, and its affinity for its natural ligand is so high that there is nothing to compete with it.

The first way around this came as KRAS G12C inhibitors, which act on the inactive state of the protein (RAS(OFF)) and only on one specific mutation. RAS(ON) is a different approach: the drug binds the active state and is not tied to a single amino-acid substitution. Hence the wording of the indication: it does not mention RAS mutations at all, and no companion diagnostic test is required.

The class is broader than one drug: the same company has three further RAS(ON) molecules in the clinic targeting individual mutations, and daraxonrasib itself is in phase III trials in non-small-cell lung cancer and in first-line pancreatic cancer. Which of these will hold up is an open question: in colorectal cancer the signal so far is weak.

Availability

The stated US wholesale price is $39,800 for a 30-day supply (Revolution Medicines Form 8-K, 26.08.2026). Outside the US the drug was not approved anywhere as of publication: the EMA is conducting a rolling review ahead of a full marketing-authorisation submission, and there is no decision. No registration records were found in Kazakhstan or the EAEU.

What follows from this

None of the above is a recommendation. We do not advise discussing any particular drug with a physician, and we do not undertake to judge applicability to anyone's case — that is a conversation between a patient and their oncologist, who has the medical history.

What is worth following next: the final survival analysis of RASolute 302, the first-line results, the lung-cancer trials, and whether the resistance mechanism turns out to be surmountable. No one can name timelines here, and we are not going to guess.

---

Sources. FDA approval announcement · FDA Oncology Center of Excellence · NEJM 2026;395:325–337 · ASCO 2026, LBA5 · ClinicalTrials.gov NCT06625320 · Targeted Oncology: critical review · The ASCO Post: review and resistance