This is the second installment of our ASCO 2026 in Review series. Of the “five turning points” introduced in the Prologue (Vol.1), this piece confronts the most symbolic—and the one declared “impossible” for longest—of them all: targeting RAS. At the Plenary Session of the 2026 ASCO Annual Meeting (Chicago, May 29–June 2), RASolute-302 showed that a pan-RAS inhibitor improved overall survival (OS) over standard chemotherapy in metastatic pancreatic ductal adenocarcinoma (PDAC), one of the deadliest cancers of all. Long called “the king of oncogenes” yet resistant to drug development for four decades, RAS has finally begun to fall at its hardest wall—pancreatic cancer. Rather than cataloging trial numbers, this piece reads the lineage of “undruggable” KRAS, the ongoing KRAS G12D inhibitor race, and the “five arrows to come” through an original lens.
What “the King of Oncogenes” Refused for Forty Years
RAS (KRAS/NRAS/HRAS) is the most frequently mutated family of oncogenes in human cancer. Roughly 90% of pancreatic cancers, 40% of colorectal cancers, and 30% of lung adenocarcinomas are driven by aberrant RAS signaling. And yet RAS was for decades the very synonym for “undruggable.” The reason is simple, and formidable. The surface of the RAS protein is smooth and lacks a deep pocket—there is no foothold for a small molecule to “catch onto.” Worse, RAS clings to GTP inside the cell with extraordinarily high affinity, leaving almost no opening for an inhibitor to wedge in. Ever since RAS was identified as a human oncogene in 1982, drug-discovery teams worldwide tried, and were turned away.
What makes RAS so vexing is also that it sits near the very root of cancer. On the inner face of the cell membrane, RAS receives growth signals from outside and relays the “proliferate” command downstream through the RAF–MEK–ERK and PI3K pathways—a molecular switch. In normal cells this switch toggles tightly between a GTP-bound “ON” and a GDP-bound “OFF,” but mutations such as G12D, G12C, and G12V lock it in the ON position, so the proliferation signal keeps ringing without any external instruction. Because it sits so close to the top of the pathway, shutting RAS off would be enormously effective—yet its smooth surface left no means to shut it off. This was RAS’s paradox: “immense if you could hit it, but no one could.”
The tide turned in the 2010s. Researchers noticed that one particular mutant, KRAS G12C, carries a cysteine residue—a “reactive handle”—and designed small molecules to covalently bind there. These became sotorasib (Lumakras, Amgen) and adagrasib (Krazati, Mirati→BMS), granted FDA accelerated approval in May 2021 and December 2022 respectively—the first clinical breach of a 40-year “impossibility.” Both later expanded into colorectal cancer in combination with anti-EGFR antibodies (such as panitumumab), establishing in the clinic that “RAS is druggable.” But the victory carried a major caveat. While G12C is relatively common in lung adenocarcinoma, the mutations that dominate pancreatic cancer are G12D and G12V—and these lack a convenient reactive cysteine. The first-generation drugs simply could not reach them. Moreover, even the leading G12C agents helped only about 30% of patients, half of whom progressed within roughly six months—limits on both efficacy and resistance.
In other words, the 2021–2022 milestone amounted to “hitting the one RAS variant that happened to have a handle, in the one tissue—lung—where it happened to work.” The true wall—G12D, which rules pancreatic cancer, and “pan-RAS,” which suppresses all RAS regardless of variant—still stood tall. What made ASCO 2026 historic was that it made visible the moment when multiple arrows began reaching that inner keep at once.
The meeting’s overarching theme was “The Science and Practice of Translation”—translating discoveries in basic science into real patient benefit across barriers of geography and language. The RAS story is the purest example of that theme. The 40-year textbook dogma that “a smooth-surfaced protein cannot be drugged” was overturned one step at a time by accumulated chemistry, structural biology, and drug-discovery technology, until it finally began reaching the clinic in the deadliest of cancers. This piece is an attempt to map, as a chart of trends and lineage, just how far that “translation” has come.
RASolute-302: Proof at the “Final Wall” of Pancreatic Cancer
Presented at the meeting’s Plenary Session (its top-tier slot, LBA5), RASolute-302 was a Phase 3 trial comparing single-agent pan-RAS inhibitor daraxonrasib (RMC-6236) against standard chemotherapy in previously treated metastatic PDAC. Reported median OS was 13.2 months vs. 6.7 months, an approximately 60% reduction in the risk of death, with significant improvements also in progression-free survival (PFS) and objective response rate (ORR). In pancreatic cancer—a field that has repelled every new drug for two decades—the fact that a single-agent oral targeted molecule produced this margin carries real weight.
Daraxonrasib is a RAS(ON) inhibitor developed by Revolution Medicines, and its mechanism departs from conventional thinking. Whereas many early RAS drugs aimed at the “OFF” (inactive) state, daraxonrasib forms a “tri-complex” that glues RAS to cyclophilin A, an abundant intracellular chaperone, directly neutralizing RAS in its signaling “ON” (active) state. Crucially, it is a “multi-selective (pan-RAS)” design that targets multiple mutant RAS variants simultaneously, not just one. In tumors like pancreatic cancer, where diverse KRAS mutations coexist, the ability to hit RAS regardless of variant is tailor-made for the hardest-to-treat disease.
Why did the shift to “targeting the active state” work? Conventional G12C drugs were designed to catch RAS in the brief instant it returned to the OFF state and bind covalently. But many variants, including G12D, spend most of their time in the ON state, so an OFF-targeting strategy cannot capture them. Daraxonrasib inverts the logic, physically disabling the very ON-state RAS that is firing signals by gluing it to the abundant cyclophilin A. It turns the destination into which the target flees—the ON state—into the trap itself. Because of this mechanism, it can suppress multiple variants without depending on a specific handle. For a tumor like pancreatic cancer, where KRAS mutations can be heterogeneous even within a single patient, this “variant-agnostic” property is a decisive advantage.
The FDA has reportedly granted daraxonrasib both Orphan Drug Designation in pancreatic cancer and Breakthrough Therapy Designation in previously treated metastatic pancreatic cancer with KRAS G12 mutations. The regulator’s early judgment that the drug “offers substantial improvement over existing therapy” underscores RASolute-302’s clinical impact. The very fact that “an effective targeted drug” now exists in pancreatic cancer is one of the largest structural changes of the past decade. Until now, pancreatic drug therapy revolved around multi-agent chemotherapy such as FOLFIRINOX and gemcitabine plus nab-paclitaxel, with targeted agents playing essentially no role outside a few rare genetic alterations. A drug that directly reaches KRAS—the most common and worst-prognosis driver mutation—could become the starting point for rewriting the very structure of pancreatic cancer care.
G12D, the Inner Keep—The Five Arrows Now in Flight
If RASolute-302 is the standard-bearer for “pan-RAS,” the other main battlefield is direct conquest of the single mutation KRAS G12D. G12D is the most frequent KRAS mutation across pancreatic, colorectal, and lung cancers, and the value of being able to hit it is immense. In the meeting’s Expert Commentary (Day 2), the G12D inhibitor cohort was summarized as showing “promising efficacy with no observed dose-limiting toxicity, with particular promise in PDAC and NSCLC.” Strikingly, against the same goal of “hitting G12D,” developers are converging with entirely different molecular strategies.
| Drug (code) | Developer | Mechanism / design | Main target cancers |
|---|---|---|---|
| daraxonrasib (RMC-6236) | Revolution Medicines | RAS(ON) pan-RAS inhibition (tri-complex) | PDAC, others |
| zoldonrasib (RMC-9805) | Revolution Medicines | RAS(ON) G12D-selective, covalent | NSCLC, PDAC |
| setidegrasib (ASP3082) | Astellas | KRAS G12D targeted protein degradation | PDAC (Phase 3 ongoing) |
| INCB161734 | Incyte | G12D ON/OFF dual-state selective inhibition | PDAC, others |
| HRS-4642 | Hengrui | Non-covalent, long-acting G12D inhibition | Solid tumors, PDAC |
What this table shows is a remarkably rare situation in drug-discovery strategy: against a single target, G12D, three fundamentally different approaches are running in parallel—inhibition (stopping the protein’s function with a drug), degradation (destroying the protein itself), and covalent binding to the ON state. Notably, setidegrasib (ASP3082) is not an inhibitor but a targeted protein degradation agent that physically removes G12D, and Astellas has reportedly begun a Phase 3 trial in first-line PDAC. Zoldonrasib, a RAS(ON) G12D-selective agent, has reportedly shown an ORR of about 52% in NSCLC and earned Breakthrough Therapy Designation. The fact that multiple strong candidates with distinct mechanisms have lined up against one mutation is itself proof that the industry now regards this target as “clinically conquerable.”
Why do so many mechanisms converge on a single target? Because each strategy has its own weaknesses and strengths. Small-molecule inhibition is easy to deliver orally and has a track record, but its dependence on binding the active site makes resistance mutations more likely. Degradation removes the target protein wholesale, so it may work where inhibition falters, and it promises a “catalytic” mode of action that drives repeated degradation with minimal drug. ON/OFF dual-state designs reduce escape regardless of which state a variant favors. Each company is competing over “which weakness to exploit in order to surpass the predecessors’ limits—narrow indication, early resistance.” This very diversity is the clearest evidence that the assault on G12D has begun in earnest.
One more development not to miss is the idea of attacking the RAS pathway “from the side.” Onvansertib (Cardiff Oncology) is a PLK1 (polo-like kinase 1) inhibitor that strikes not RAS itself but a cell-division checkpoint on which RAS-mutant cells tend to depend. At the meeting, CRDF-004 (Phase 2, Abstract 3510), adding onvansertib to standard chemotherapy plus bevacizumab, was studied in first-line RAS-mutant metastatic colorectal cancer (mCRC). Unlike pancreatic cancer, colorectal cancer is not dominated by G12D alone but carries a mix of KRAS mutations—G12V, G13D, G12C, and more—so the value of a variant-agnostic “flank attack” is relatively higher. Where RAS cannot be hit directly, or where direct inhibition alone breeds resistance, a combination that exploits a synthetic-lethal vulnerability could be a powerful “second arrow” of a different lineage from pan-RAS and G12D drugs. Pairing direct inhibitors with pathway-dependency agents like PLK1 inhibitors is one of the next frontiers of RAS drug development.
Reading the Lineage—A Three-Stage Rocket: “G12C → G12D → Pan-RAS”
The key to grasping the 2026 RAS map in a single frame is to overlay, in chronological order, the three-stage evolution this target has traveled.
- Stage 1 (2021–2022): Breaching G12C. Sotorasib and adagrasib covalently struck the one variant with a reactive cysteine “handle,” proving for the first time that “RAS is druggable.” But the scope was limited, and resistance emerged early.
- Stage 2 (2023–2026): Storming the inner keep, G12D. Against G12D—the most frequent mutation, ruling pancreatic and colorectal cancer—diverse molecular strategies (inhibition, degradation, ON-state binding) converged. Multiple routes opened even to a “handle-less” mutation.
- Stage 3 (2026–): Pan-RAS, the integrated answer. Daraxonrasib, which suppresses active RAS in bulk regardless of variant, showed an OS benefit in the deadliest disease, pancreatic cancer. This is a shift from conquering individual mutations one by one to “suppressing the RAS signal itself as a whole surface.”
This three-stage rocket represents not mere accumulation of drugs but a deepening of drug-discovery thinking. From “find the weakness (cysteine) and strike a point,” through “attack the major mutations one surface at a time,” to “control the RAS pathway itself in an integrated way.” Each stage was born trying to overcome the prior stage’s limits—narrow indication, early resistance, mutational diversity. ASCO 2026 may be remembered as the first year all three coexisted within a single meeting.
What is intriguing is that the three stages are “layering,” not “replacing.” The arrival of pan-RAS drugs does not render G12C or G12D-selective agents obsolete. Rather, the repertoire thickens—these agents will be used selectively, or sequentially over time, according to a patient’s mutation profile, cancer type, and prior treatment. A “layered strategy” becomes realistic: use a pan-RAS drug first line, then switch to a variant-specific agent or a pathway-inhibitor combination once resistance emerges. RAS drug development has left the stage of seeking a single decisive blow and now stands at the threshold of maturity—the question of how to combine multiple weapons.
The Competitive Landscape—Revolution Medicines’ Lead and the Pursuing Giants
Surveying the current RAS race, Revolution Medicines is clearly out front on both wheels—pan-RAS (daraxonrasib) and G12D-selective (zoldonrasib). The company’s strength is that, anchored on a single platform—RAS(ON), which traps active RAS in a tri-complex with cyclophilin A—it covers multiple mutational territories from one base. Producing a positive Phase 3 result first in pancreatic cancer, the largest and hardest market, could prove a decisive competitive milestone.
Pursuing it are giants and emerging players wielding different weapons. Astellas has stepped into Phase 3 for G12D with its own “degradation” route, setidegrasib (ASP3082); Incyte enters with INCB161734, which captures both ON and OFF states; and Hengrui joins with the non-covalent, long-acting HRS-4642. The rise of China-originated biotech is conspicuous in RAS as well, making G12D drug development a global, multipolar contest rather than a Western monopoly. On top of that, “flank attacks” via PLK1 inhibition, such as onvansertib (Cardiff Oncology), aim at a distinctive niche as combination partners for direct inhibitors.
Viewed economically, the scale of this contest stands out. RAS-mutant cancers are broadly distributed across high-incidence major tumor types—pancreatic, colorectal, lung—and effective targeted therapy was essentially absent until now. RAS drug development is therefore a field where, given the large medical void, the winner could capture an enormous clinical and commercial impact. That is precisely why large, emerging, and China-originated biotechs are all pouring in resources, and candidates with distinct mechanisms are crowding into the clinic at once. With the first decisive blow—a positive Phase 3 in pancreatic cancer—now landed, the contest has shifted a gear from a “proof-of-concept phase” to a “phase of fighting over market and positioning.”
The focus of competition will now move from “survival extension as monotherapy” toward “optimal combinations and sequencing,” “control of resistance,” and “cross-variant coverage.” Will pan-RAS drugs become the first-line foundation? Will G12D-specific agents carve out precision-medicine niches by variant? Or will triple combinations of both with chemotherapy and immunotherapy become standard? This struggle for leadership will redraw the RAS drug-development map over the next several years. The keys will be how fast each company can deploy its platform across multiple cancer types and combinations, and how well it has prepared its next move (next-generation molecules or combination regimens) for when resistance surfaces.
What Proof in Pancreatic Cancer Means—and the “Five Arrows to Come”
That RASolute-302 demonstrated an OS benefit in pancreatic cancer means more than simply adding one new drug. Pancreatic cancer is often already advanced at diagnosis, offers few effective options beyond chemotherapy, and has long had the lowest five-year survival of any cancer. The fact that “a targeted drug works” in that disease shows that targeting RAS has crossed from “proof of concept” into “clinical reality.” So what comes next? Let us lay out the “five arrows to come” visible along the trajectory of the trend.
- ① Advance into first-line and standardized combinations: from single-agent later-line use to first-line combinations with chemotherapy (gemcitabine/nab-paclitaxel, etc.). A movement toward embedding RAS drugs in the opening move against pancreatic cancer.
- ② Precise niche-carving by variant-specific drugs: an era in which agents optimized per mutation—G12D, G12V, and so on—are used selectively alongside companion diagnostics.
- ③ Getting ahead of resistance mechanisms: against the acquired resistance that direct inhibition inevitably breeds, designs that combine up/downstream agents (SHP2, SOS1, MEK) or evade resistance via pan-RAS coverage.
- ④ Connection with immunotherapy: exploration of IO combinations, informed by evidence that RAS inhibition may turn the tumor microenvironment into a more “immune-permissive” state.
- ⑤ A second current—targeted protein degradation: the possibility that “destroying” RAS protein via degraders like setidegrasib, rather than inhibiting it, offers an alternative answer to the problems of resistance and selectivity.
Among these five arrows, the intersection of ① and ③ deserves particular attention. To advance RAS drugs from later-line into first-line necessarily presumes combination with chemotherapy, and at that moment the patterns of resistance and the toxicity profile look different from monotherapy. The reports that Revolution Medicines is concurrently advancing Phase 3 first-line combinations (with gemcitabine/nab-paclitaxel) and resected (post-surgical) pancreatic cancer reflect exactly this drive to claim the “advance” by the shortest path. Once efficacy is established in first-line, RAS drugs would be built into treatment design from the moment of initial pancreatic cancer diagnosis, changing the very standard of a field that had only chemotherapy to offer.
What these five arrows share is a shift in thinking—from “hit RAS once and you’re done” to “control the RAS pathway in three dimensions: time, combination, and modality.” ASCO 2026 was the year that shift sounded its starting gun on the most symbolic battlefield of all, pancreatic cancer. The granular deep dives into RAS targeting—detailed trial data and the molecular biology of resistance—we will revisit in separate pieces of this series. Here, we want to inscribe the large directional truth that “the wall has begun to crumble.” What comes next is a larger avalanche, pushing this collapse into lung and colorectal cancer, and into first-line therapy.
My Thoughts and Future Outlook
“KRAS cannot be drugged”—many of us who studied medicine were taught this for a long time. That is exactly why this news should reach readers outside the field. Pancreatic cancer is a cancer often spoken of as “diagnosis equals despair.” In that disease, an oral targeted pill nearly doubled survival time. This is not a wish that “a drug might work someday”; it is a reality demonstrated in Phase 3. A target that resisted for four decades has finally begun to crumble at its hardest wall—let us, first, simply take in that historic moment.
At the same time, as an expert one must temper the elation a notch. A median OS of 13.2 months is a genuine advance, but the absolute prognosis of pancreatic cancer remains harsh, and “cure” is far off. Acquired resistance inevitably stands in the way of direct RAS inhibition; as we learned with G12C, the pattern in which a fraction of patients progress within about six months is likely to repeat even with pan-RAS drugs. The real contest lies not in single-agent survival extension but in the design of “how to use it”—combination, sequencing, and resistance avoidance.
What is missing is long-term follow-up OS data, an understanding of where the ceiling lies across variants, and the optimal combinations. Even so, 2026—when multiple mechanisms (inhibition, degradation, pan-RAS) crowded onto the same target and results began to emerge in the deadliest cancer—will be remembered as a “turning-point year” for RAS drug development. What will be asked next is how to extend this victory to other RAS-driven cancers beyond the pancreas, and into first-line therapy.
This article is an independent summary and analysis by Morningglorysciences based on publicly available information from the 2026 ASCO Annual Meeting and related presentations. Always consult primary publications, the latest guidelines, and official prescribing information when making treatment decisions.
ASCO 2026 in Review — Series Index
This article is part of the six-part series “ASCO 2026 in Review.” For the full overview, start with Vol.1 Prologue (hub article). Each part stands alone but is designed to complement the others.
- Vol.1 Prologue: The Modality Revolution [Hub]
- Vol.2 RAS: The Undruggable Falls (this article)
- Vol.3 ADC: Crossing Tumor Barriers
- Vol.4 Bispecifics: PD-1 x VEGF
- Vol.5 GLP-1: Metabolic Oncology
- Vol.6 Finale: Equity, Liquid Biopsy & AI
Comments