$RVMD: they made RAS druggable, and pancreatic cancer was only the proof
$RVMD · Revolution Medicines Inc · by Thomas Bouman ·
covered since
bullishRating
$220.00Price target · 12M
$188.85Written at
4Notes
Daraxonrasib doubled overall survival in second-line metastatic pancreatic cancer, a disease where nothing has worked for decades, and the NDA was accepted on July 22. RAS drives about a quarter of all human cancers. Pancreatic is the proof of concept, not the market.
The Thesis (The Story)
The bet
Revolution Medicines was founded in 2014 on a bet that RAS could be drugged. RAS is a gene mutated in roughly a quarter of all human cancers, and for forty years it was considered essentially impossible to target with a drug. Earlier attempts tried to hit RAS in its "off" state. RevMed's approach instead goes after the "on," active form of the protein, a real mechanistic departure rather than a tweak on what came before.
The pancreatic cancer result
That bet paid off in one of the hardest cancers there is. In the Phase 3 RASolute 302 trial, 500 patients with previously treated metastatic pancreatic cancer received either a once-daily pill, daraxonrasib, or one of four standard chemotherapy regimens chosen by their doctor. Median overall survival was 13.2 months versus 6.7 months on chemo, cutting the risk of death by 60% (a hazard ratio of 0.40, highly statistically significant). The benefit held across both the specific RAS-mutation subgroup and the full trial population. The trial hit every primary and key secondary goal, patients also reported delayed worsening of pain and quality of life, and the results were presented at ASCO's top plenary session with simultaneous publication in the New England Journal of Medicine, the highest-visibility venue in oncology.
To put that in context: metastatic pancreatic cancer has a five-year survival rate around 3%. Roughly 60,000 Americans are diagnosed each year, and about 50,000 die from it. Doubling survival at this stage isn't an incremental improvement, it's a genuinely rare result in this disease.
What's behind it
Daraxonrasib is already running in four separate Phase 3 trials covering first-line pancreatic cancer and RAS-mutant lung cancer, with colorectal cancer in the pipeline behind it. Beyond that lead drug, RevMed has built a full family of related drugs targeting other specific RAS mutations: elironrasib, zoldonrasib, RMC-5127, plus two more candidates still earlier in development. Combination data on zoldonrasib presented in July showed response rates up to 82% in pancreatic cancer, and a new trial (RASolute 305) began dosing patients with a different RAS mutation in first-line pancreatic cancer in June.
🐂 Bull Case (What has to go right?)
A clear, unambiguous Phase 3 win in a cancer with essentially no other effective treatment options, backed by FDA designations (Breakthrough Therapy, Orphan Drug) that speed up review
Pancreatic cancer is just the first win. Lung and colorectal cancer are far bigger markets, and daraxonrasib is already in Phase 3 testing for RAS-mutant lung cancer
Real pipeline depth: multiple mutation-specific drugs behind the lead asset, with early data suggesting they can work together rather than compete for the same patients
First-mover advantage in a newly proven drug class, RevMed has shown this approach to RAS works, and currently leads the field
One of the most strategically valuable independent cancer-drug companies in the sector, the kind of asset every large pharma with an oncology business needs some exposure to
🐻 Bear Case (What are the risks?)
Valuation is the single biggest risk here. A roughly $39B market cap with zero product revenue yet already prices in a lot of what still has to go right
Patients in the pivotal trial have only been followed for a median of 8.5 months so far. Survival results can shift as more time passes, and the eventual drug label will reflect what regulators conclude from mature data, not just the early headline numbers
No commercial infrastructure yet. Launching and selling an oral cancer drug to doctors and hospitals is a whole capability the company still has to build from scratch
Heavy cash burn. Running four separate Phase 3 trials for one drug, plus a broad early pipeline behind it, requires sustained and significant spending
Competition is coming. Lilly, Amgen, Erasca and others are working on their own RAS-targeting drugs, and being first doesn't guarantee staying the best option
Recent insider selling and a planned leadership change in R&D, announced in June, aren't disqualifying on their own, but both are worth watching
Conclusion
This is a platform bet, not a single-drug trade on pancreatic cancer. The RASolute 302 result did two things at once: it gave the company a clear path to its first approval in a disease with essentially no competition, and it proved the underlying science works, which matters for every other RAS-driven cancer behind it in the pipeline.
What matters from here is the transition from "proof it works" to "a real franchise", getting this first drug approved and launched, then finding out whether the same approach delivers in lung and colorectal cancer, where far more patients are affected.
The risk worth watching is the price. At this valuation, the market has already agreed the science works, so returns from here depend on the rest of the pipeline delivering too, not just on this one drug reaching patients.
Update notes & developments
$RVMD: Q2 2026 earnings, EAP uptake accelerates, strong early NSCLC combo data
Development · · written at $197.30
Q2 earnings show accelerating EAP penetration ahead of the daraxonrasib PDAC approval decision, plus encouraging early combination data for elironrasib and zoldonrasib in first-line NSCLC. Cash position remains solid at $3.9B.
PDAC: EAP momentum ahead of approval
The daraxonrasib NDA acceptance was covered in the prior note. The new data point this quarter is Expanded Access Program uptake: over 2,000 patients have received daraxonrasib through the EAP since it opened in May, across nearly all 50 states.
That is a meaningful share of the annual 2L+ metastatic PDAC population in the US, achieved in roughly a quarter of EAP activity. It functions as a pre-approval read on physician and patient demand, though the exact penetration rate depends on which population estimate you use and shouldn't be taken as precise.
NSCLC: early combination data
Two readouts this quarter, both small and early but directionally strong.
Median follow-up only 3.4 months, materially shorter than the elironrasib cohort
Both support recently initiated or planned Phase 3 studies (RASolve 307 for elironrasib, RASolve 308 for zoldonrasib). Small single-arm cohorts with short follow-up, so treat as hypothesis-generating rather than confirmed signal, particularly the zoldonrasib data given the shorter follow-up window.
Financials
Cash, equivalents and marketable securities: $3.9B, plus up to $1.5B more available under the Royalty Pharma arrangement subject to milestones
Net loss widened to $644.4M from $247.8M a year ago, driven by commercial readiness spend and a non-cash EQRx warrant fair value charge
FY2026 GAAP opex guidance updated to $2.1-2.2B
Takeaway
Nothing here changes the thesis. RASolute 302 OS benefit remains the core of the position, and this quarter is confirmation of execution around it rather than new efficacy evidence. The EAP data is the most useful new signal, since it's a real-world demand proxy ahead of an approval decision. NSCLC combo data adds optionality but needs more mature follow-up before it moves the needle on valuation.
$RVMD: the NDA is in, and the voucher is the part worth watching
Update note · · written at $189.81 · confirmed
The FDA accepted the daraxonrasib NDA on July 22. Acceptance was expected, but daraxonrasib carries a Commissioner's National Priority Voucher, and the last drug through that pathway was approved 50 days after filing. I expect approval far sooner than the market is pricing.
What happened
On July 22 the FDA accepted Revolution Medicines' NDA for daraxonrasib in previously treated metastatic pancreatic cancer. The filing rests on RASolute 302, where daraxonrasib roughly doubled median overall survival against chemotherapy (13.2 versus 6.7 months, HR 0.40, p<0.0001) and met all primary and key secondary endpoints.
Acceptance means the package is complete enough to review, not that it works. For a filing built on an ASCO plenary dataset published in the NEJM, it was close to a formality.
The voucher is the actual news
Daraxonrasib holds Breakthrough Therapy and Orphan Drug designation, and it was selected for the FDA Commissioner's National Priority Voucher pilot. The NDA went in as a rolling submission under that pathway.
CNPV is not Priority Review. It replaces the standard 10 to 12 month process with an intensive senior-level review that eliminates idle time through continuous sponsor-agency contact. The FDA's stated target is a decision within two months.
The precedent is concrete: on April 1 the FDA approved Lilly's orforglipron under CNPV, 50 days after filing and 294 days ahead of its PDUFA date, which the agency called the fastest new molecular entity approval since 2002.
Europe runs in parallel rather than behind. The EMA began a phased review on July 7, with orphan designation and high-priority status under the Cancer Medicines Pathfinder project.
🐂 What goes right
Approval within weeks rather than quarters, pulling first revenue forward and compressing the pre-commercial burn.
The clinical case is as clean as oncology gets: the comparator buys patients under seven months in a disease with roughly 3% five-year survival. Payer and prescriber resistance should be minimal.
The EMA phased review keeps ex-US timing close behind the US rather than the usual year-plus lag.
A fast first approval validates the RAS(ON) platform commercially ahead of the NSCLC and first-line PDAC readouts.
🐻 What would break it
CNPV is fast but not automatic. Several voucher holders have been delayed on safety or efficacy questions, and even orforglipron had its date pushed back before approval.
Median follow-up in RASolute 302 was 8.5 months. Label language around the RAS G12 versus intent-to-treat populations is where the FDA has room to be narrower than the headline.
Approval starts the commercial problem rather than ending it, and a faster approval means less time to build a launch organisation.
At a ~$39B market cap, an in-line approval may not move the stock much. The upside case rests on the pipeline behind daraxonrasib.
Conclusion & My Position
My angle: I expect this to be approved quickly, and quite possibly very quickly. The CNPV precedent points at weeks rather than quarters, sell-side has floated an approval by the end of August, and the underlying case is unusually easy for the FDA to sign off on: an oral drug that doubles survival where the standard of care buys six months.
I'll go further than that. I think daraxonrasib has a real chance of beating orforglipron's 50 days and becoming the fastest new molecular entity approval on record under this programme. The unmet need is more severe, the comparator is weaker, the effect size is larger, and pancreatic cancer is exactly the kind of indication the voucher pathway was created to prioritise. An obesity pill cleared in 50 days. This is a disease with a 3% five-year survival rate.
The acceptance itself was expected and isn't the event. What matters is the pace of what follows.
If it lands the way I think it will, the story changes shape fast. RevMed becomes a commercial company before most of the market has repriced it, and attention shifts to whether the same mechanism delivers in lung and colorectal, where the patient numbers are far larger.
The risk I accept is that this is a prediction about FDA behaviour, not about data. I could be right on the drug and wrong on the timing, and a fast approval is partly priced already. The real work starts the day it lands.
Position at publication: long RVMD, no change planned.
$RVMD: RASolute 302 lands within weeks, and the setup says it works
Update note · · written at $99.15 · confirmed
RASolute 302 topline is due in the first half and ASCO is eight weeks out. Merck offered $28–32B for this company in January and management turned it down rather than sell ahead of the readout. At $99 the stock sits well below that, with a setup that points to data that works.
What's coming
Revolution Medicines guided RASolute 302 topline for the first half of 2026. It is an event-driven, overall-survival-triggered readout in second-line metastatic pancreatic cancer, and ASCO runs May 29 to June 2. Those two facts sit close enough together that sequencing matters: a readout in the next several weeks gives the company time to take full data to Chicago, and a plenary slot at ASCO would be a signal in itself, since plenary sessions are reserved for results the society considers practice-changing.
This note is a call on that readout, written before it lands.
Merck already told us what it's worth
In January the Financial Times reported Merck in talks to acquire Revolution Medicines at $28B to $32B. AbbVie was separately reported to be circling, though it publicly denied being in discussions. Talks collapsed in late January over price, and the stock fell roughly 20% on the news.
Both halves of that story are informative.
A buyer with due diligence access, sitting on the full Phase 1/2 dataset and the trial design, was willing to pay a 25% to 40% premium to a ~$23B market cap for an asset whose pivotal trial had not read out. Merck does not make that offer if it thinks RASolute 302 is a coin flip. It is facing the Keytruda patent cliff and needs oncology revenue, which explains motivation, but it does not explain price.
The refusal is the stronger signal. Management, who know this trial better than anyone, chose to carry readout risk rather than lock in $30B. That is either conviction or arrogance, and given the alternative was the second-largest biotech acquisition in history, I read it as the former.
At $99 the stock now sits at roughly $20B, below both the Merck range and the level it traded at during those talks. You are buying beneath what a strategic buyer with diligence access thought it was worth before the data.
Why I think it works
The Phase 1/2 signal was real. Second-line RAS-mutant PDAC showed a 29% confirmed objective response rate, and first-line reached 47%. Single-arm oncology data against historical controls deserves scepticism as a rule, but the bar here is unusually well characterised: second-line metastatic PDAC on chemotherapy runs to roughly six months of median survival, and it has barely moved in two decades.
The trial design favours the drug. RASolute 302 enrolled across a wide range of RAS variants including G12D, G12V and G12R, plus patients with no identified RAS mutation. Roughly 90% of pancreatic cancers carry a KRAS mutation, so a multi-selective RAS(ON) inhibitor addresses nearly the whole disease rather than a sliver of it.
The Street already believes. Leerink puts probability of success at 97%. Consensus peak sales estimates for daraxonrasib in first and second-line metastatic PDAC cluster around $5B to $7.6B, and consensus price targets sit near $122 to $129 against a $99 share price.
Regulatory positioning is already in place. Daraxonrasib won an FDA Commissioner's National Priority Voucher in October 2025, which can compress review from 10 to 12 months down to as little as one or two. That only matters if the data works, but it means the path from readout to approval is far shorter than usual.
Company behaviour is consistent with confidence. Four Phase 3 trials for daraxonrasib are already running, including first-line PDAC and RAS-mutant NSCLC. You do not build a programme of that scale while waiting to find out whether your lead indication reads out positive.
What good looks like
A hazard ratio in the 0.6 to 0.7 range with two or three months of median OS benefit would be a clear win and enough for approval. Anything below 0.5, with median OS pushing past a year, would be a different category of result entirely: no Phase 3 in this disease has produced that in any line of therapy, and it would reframe daraxonrasib from a good drug to a landscape-changing one.
I think the second outcome is more likely than the market is pricing.
🐻 What would break it
PFS hits, OS misses. The most plausible failure mode. Crossover or post-progression therapy could produce significant PFS with an OS curve that does not separate cleanly. The FDA has signalled it wants OS to carry the weight, so this would be a genuine problem rather than a technicality.
Historical controls flatter single-arm data. Phase 1/2 populations are self-selected for fitness, and randomisation removes that advantage. Effect sizes routinely shrink.
The Merck read-through cuts both ways. A buyer walking away can be read as diligence that found something, not just a price disagreement. I think price is the better explanation, but I cannot prove it.
Timing slips. Event-driven means the readout arrives when enough deaths accrue. A slip past ASCO delays the narrative even if the data is fine.
Valuation asymmetry. At ~$20B on zero revenue, a merely good result may not move the stock much, while a miss is a long way down.
Tolerability. Rash is an on-target effect of RAS/MAPK inhibition, and high discontinuation would complicate the commercial case even alongside a survival win.
Conclusion & My Position
My angle: this is the readout that determines whether RAS(ON) inhibition is a real drug class or an expensive scientific idea. RAS is mutated in about a quarter of all human cancers and spent forty years labelled undruggable. RevMed built a company on cracking it, and pancreatic cancer is where the proof arrives first.
I expect the data to be strong, and possibly considerably better than the Street's models allow for. The Phase 1/2 response rates, the near-universal KRAS mutation rate in this disease, the weakness of the comparator, the voucher already in hand, and the fact that both Merck and this management team behaved like people who expect it to work, all point the same way.
If I'm right, the second-line indication is the smallest part of the story. First-line PDAC, NSCLC and colorectal are all far larger, and a positive readout de-risks a mutant-selective portfolio the market is currently valuing at close to nothing. It also puts Merck, or whoever else was circling, back at the table against a company with no reason to accept $30B any more.
If I'm wrong, particularly on OS, this is a long way down from here.
Position at publication: long RVMD, no change planned into the readout.