For decades, RAS was oncology’s quintessential “undruggable” target. Drugs such as sotorasib and adagrasib began changing that, before Revolution Medicines pushed further with daraxonrasib, a RAS(ON) inhibitor targeting active mutant and wild-type RAS. The implications are particularly significant in pancreatic cancer, where more than 90% of tumors harbor RAS mutations. At this year’s ASCO meeting, landmark Phase III daraxonrasib results in metastatic disease drew a rare 42-second standing ovation—a striking response in a cancer where therapeutic progress has historically come in frustratingly small increments.

But even a groundbreaking cancer drug does not bat a thousand. Daraxonrasib does not work for every patient, and even among those whose cancers initially respond, many tumors eventually begin growing again. The latter group is the focus of a new study published in Nature Medicine showing that cancers that daraxonrasib initially knocked down but that subsequently figured out how to get back up. The researchers looked at the progression of events within pancreatic cancer from the start of RAS inhibition until its escape, which could reveal what happens next in the RAS revolution.

When pancreatic cancer activates RAS

Researchers analyzed paired circulating tumor DNA (ctDNA) samples collected before treatment and at progression from 44 patients with RAS-mutant metastatic PDAC treated with daraxonrasib. These were not primary nonresponders. Every patient had initially experienced a complete response, partial response, or stable disease lasting more than three months before eventually progressing. Mutant KRAS in their blood also declined early during treatment, consistent with daraxonrasib initially suppressing their cancers.

When researchers examined what had changed by the time those cancers progressed, a striking pattern emerged. Twenty-six of the 44 patients—59%—had acquired genomic alterations involving the RAS signaling network. Most notably, 16 patients, or 36%, acquired amplification of mutant KRAS. Others developed alterations involving RAF, receptor tyrosine kinases (RTKs), or PI3K signaling. In effect, many tumors responded to RAS inhibition by finding ways to turn RAS signaling back up.

KRAS amplification was particularly revealing. Rather than acquiring a new RAS driver, tumors made additional copies of the mutant KRAS they already carried. In laboratory models, increasing mutant KRAS expression increased downstream ERK signaling and progressively reduced sensitivity to daraxonrasib.

That represents a notable difference from some earlier KRAS inhibitors. Drugs such as sotorasib and adagrasib target KRAS G12C in its inactive, GDP-bound—or RAS(OFF)—state, and secondary mutations that interfere with drug binding are among the resistance mechanisms that have emerged. No secondary KRAS mutations were detected in the 44 pancreatic cancer patients in the new study.

Instead, resistance largely converged on a broader principle: restore enough RAS pathway activity to overcome the drug. Paradoxically, this may strengthen the rationale for targeting RAS rather than weaken it. These cancers appear so dependent on the pathway that, under therapeutic pressure, many evolve specifically to reactivate it.

Preventing RAS inhibition escape

That observation also provided researchers with a roadmap for potential combination therapies. In some models, daraxonrasib treatment increased HER2 expression on the tumor cell surface. Researchers exploited that vulnerability by combining the RAS inhibitor with the HER2-directed antibody-drug conjugate trastuzumab deruxtecan (T-DXd). The combination produced sustained tumor regressions across several pancreatic cancer models, including those with acquired daraxonrasib resistance. Combining daraxonrasib with amivantamab, a bispecific antibody targeting EGFR and MET, similarly improved the depth and duration of responses in an RTK-driven model.

But perhaps the most intriguing strategy was simply to hit RAS harder. Researchers combined daraxonrasib’s broad RAS(ON) inhibition with zoldonrasib, Revolution Medicines’ mutant-selective RAS(ON) inhibitor targeting KRAS G12D—one of the most common KRAS mutations in pancreatic cancer. The logic is essentially pharmacological double coverage: daraxonrasib suppresses active mutant and wild-type RAS broadly, while zoldonrasib concentrates additional pressure on the principal oncogenic driver.

In treatment-naive mouse models, that doublet produced deeper RAS pathway suppression and substantially delayed progression compared with either drug alone. More importantly, the combination prevented resistance from emerging during the experiments. It also improved tumor control in models that had already developed resistance to RAS(ON) monotherapy. The experiments remain preclinical, but they illustrate where the broader RAS field may be heading.

The first generation of direct KRAS inhibitors established that the supposedly undruggable protein could, in fact, be drugged. Mutation-selective RAS(ON) inhibitors such as zoldonrasib are expanding the number of individual RAS mutations that can be attacked, while multi-selective agents such as daraxonrasib are attempting to suppress a much broader spectrum of RAS-driven cancers.

Now those approaches may begin converging—through combinations of broad- and mutation-selective RAS inhibitors, or RAS drugs paired with therapies aimed at the escape routes tumors reveal under treatment.

RAS inhibitor innovation’s next chapter

Important questions remain. The researchers found candidate genomic explanations for resistance in roughly 60% of patients, leaving about 40% unexplained. Nongenomic mechanisms, tumor heterogeneity, and alterations beyond those captured by the ctDNA assay could all contribute, and larger cohorts with paired tumor biopsies will be needed to complete the resistance map.

But the study also reveals something fundamental about pancreatic cancer biology. When daraxonrasib successfully suppresses RAS, many tumors do not simply abandon the pathway and adopt an entirely different means of survival. Instead, they amplify mutant KRAS, activate upstream receptors, or otherwise restore signaling through the same oncogenic circuitry. In that sense, resistance may be further evidence of just how dependent these cancers remain on RAS.

That changes the challenge facing the field. For decades, researchers asked whether RAS could be drugged at all. The success of daraxonrasib and the rapidly expanding class of direct RAS inhibitors has increasingly answered that question. Now researchers are asking how deeply, broadly, and durably RAS must be suppressed—and whether combinations can close off the escape routes before tumors exploit them.

The Nature Medicine study suggests that the next leap may come not from moving beyond RAS, but from doubling down on it: combining broad and mutation-selective RAS inhibitors, pairing them with therapies that target compensatory pathways, and eventually using the molecular fingerprints of resistance to determine which strategy a particular tumor needs.

Daraxonrasib helped crack one of cancer biology’s most notoriously difficult targets. The next phase of the RAS revolution will be figuring out how to keep it cracked.