For about four months, it looked like one of medicine's longest-running nightmares was finally ending. A Spanish oncologist in his late seventies, working out of a research center in Madrid, published data showing complete, durable elimination of pancreatic tumors in mice — no resistance, no relapse, minimal side effects. The scientific world held its breath. Patient advocacy groups wept. Donations poured in. Then, on April 27, 2026, the journal that published the study took it back.
The retraction of Mariano Barbacid's triple-combination therapy paper from PNAS did not happen because the experiments were wrong. It happened because three of the paper's authors — including Barbacid himself — held undisclosed financial interests in a startup company built around the very drugs they were testing. That distinction matters enormously, and almost every headline about this story has either buried it or missed it entirely.
This article gives you the full picture: what the research actually showed, why it was retracted, what the retraction does and does not mean for the science, and what people with a stake in pancreatic cancer treatment — patients, families, researchers, investors — should take away from the wreckage of what briefly looked like a miracle.
Table of Contents
- The Disease That Made This Research Feel Urgent
- What Barbacid's Team Actually Found
- The Three Drugs and the Logic Behind the Combination
- Why PNAS Retracted the Paper — The Exact Reason
- Does the Retraction Invalidate the Science?
- The Vega Oncotargets Question
- What This Means for Pancreatic Cancer Research Going Forward
- Who Should Be Paying Attention
- Verdict: Hope, Caution, and the Longer View
- Frequently Asked Questions
The Disease That Made This Research Feel Urgent
Pancreatic ductal adenocarcinoma — PDAC, the form that accounts for more than 90% of pancreatic cancer cases — is the kind of diagnosis that changes a room. It hides until it has spread. It resists the standard chemotherapy regimens that work reasonably well elsewhere. And it kills fast: the global five-year survival rate sits at roughly 10 to 13%, dropping to around 3% once the cancer has metastasized. In Spain alone, more than 10,300 people receive this diagnosis every year. Fewer than one in ten will be alive five years later.
The reason it is so hard to treat comes down to molecular stubbornness. Over 90% of PDAC cases carry mutations in the KRAS gene — a discovery Barbacid himself made in 1982, which is part of what makes his career arc so striking. KRAS drives uncontrolled cell division, and for decades it was considered essentially undruggable. When the first targeted KRAS inhibitors finally reached patients around 2021, they worked — briefly. Then the tumors adapted, activated backup signaling pathways, and kept growing. That resistance pattern is what Barbacid's triple-combination approach was specifically designed to break.
What Barbacid's Team Actually Found
The study, published in PNAS in December 2025 with Vasiliki Liaki and Sara Barrambana as first authors and Carmen Guerra as co-lead, tested a three-drug combination across multiple mouse models of PDAC. The models included orthotopic tumors, genetically engineered mice, and — most significantly — patient-derived xenografts, meaning tumors taken from actual human patients and implanted into mice.
The results, as reported before the retraction, were unlike anything previously published in this disease context. Tumors regressed completely and durably. In the xenograft models, 16 of 18 mice showed full remission with no evidence of relapse for over 200 days post-treatment. The animals tolerated the therapy well — stable weight, no significant changes in blood counts, no visible organ damage.
"This triple combination is well tolerated in mice. These studies open the road to design novel combination therapies that may improve the survival of PDAC patients." — Liaki et al., PNAS, December 2025 (subsequently retracted)
Barbacid was careful in public statements. He told reporters and CNIO's own press office that the team was not ready for human clinical trials — that further optimization and safety work remained necessary. That caution was genuine and appropriate. What was not disclosed at the time was something else entirely.
The Three Drugs and the Logic Behind the Combination
Understanding why this combination was scientifically interesting — regardless of what happened afterward — requires a brief look at what each component targets.
Daraxonrasib (RMC-6236): Blocking the Primary Engine
Daraxonrasib is a RAS(ON) inhibitor — part of a newer class of drugs that target mutant KRAS proteins directly while leaving normal KRAS largely alone. Unlike earlier KRAS inhibitors that could only block a specific mutation variant, daraxonrasib works against a broader range of KRAS mutations. It has already entered human clinical trials for pancreatic and lung cancer, including a randomized Phase 3 trial comparing it against standard chemotherapy in metastatic pancreatic cancer patients. The drug's single-agent results have been described as improvements over historical data, but resistance still develops.
Afatinib: Closing the Escape Route
When KRAS is suppressed, tumors frequently activate alternative growth signals through the EGFR and HER2 receptors — essentially flipping on a backup switch. Afatinib is an FDA-approved irreversible inhibitor of both receptors, already used in lung cancer treatment. Its role in Barbacid's combination was to slam that escape route shut before the tumor could use it.
SD36: Dismantling the Resistance Shield
SD36 is a PROTAC — a newer class of molecules that work not by blocking a protein but by tagging it for destruction by the cell's own internal disposal machinery. Its target is STAT3, a protein that cancer cells use to survive hostile conditions, evade immune surveillance, and resist therapeutic pressure. By degrading STAT3 entirely rather than merely inhibiting it, SD36 removes a layer of protection tumors rely on when other pathways are under attack.
The combined logic was elegant: instead of attacking one lock on a door the tumor has many ways to open, the combination changes all three locks simultaneously. The tumor, in theory, has no molecular path left to exploit. In the mouse models — before the retraction — that theory appeared to hold.
Why PNAS Retracted the Paper — The Exact Reason
This is the part that deserves precision, because the nuance is real and it matters.
The retraction, formally published on April 27, 2026, was not triggered by failed experiments, data manipulation, or errors in methodology. PNAS explicitly stated that the decision was based on a procedural breach: Barbacid, Liaki, and Guerra held financial interests in a company called Vega Oncotargets — a biotech startup created specifically around the commercial development of the therapy described in the paper — and those interests were not disclosed at submission.
Under PNAS editorial policy, papers contributed by members of the National Academy of Sciences (a process that gives those members significant influence over peer review selection) must follow stricter disclosure rules when financial conflicts exist. Specifically, papers with competing financial interests must be submitted as "Direct Submissions," which routes them through an independent editorial process. Because the conflict was not declared, the paper went through the wrong channel, bypassing the safeguards designed for exactly this situation.
The journal did not issue a correction or add a disclosure note. It retracted the paper entirely. That is the most drastic option available, and PNAS chose it.
Barbacid's Response
Barbacid publicly maintained that the data remain sound and described the retraction as an "administrative problem." He indicated his intention to resubmit the paper through the correct channel with full disclosures. Critics in the scientific community were less dismissive — several researchers noted publicly that undisclosed financial ties to a company built around research findings are not a minor administrative issue, particularly in a field where patients are watching every development with desperate attention.
Does the Retraction Invalidate the Science?
This is genuinely complicated, and anyone who gives you a simple yes or no is oversimplifying.
A retraction based on conflict-of-interest non-disclosure does not mean the experiments did not happen, the results were fabricated, or the conclusions were wrong. Multiple independent scientific sources — including analysts covering the case after the retraction — have pointed out that the laboratory findings themselves have not been formally challenged. The CNIO mouse data, as far as outside reviewers can assess, appear to reflect real experimental results.
What the retraction does mean is that the paper's peer review process was compromised by a procedural failure that existed specifically to catch financial conflicts. Peer review, however imperfect, is the mechanism science uses to catch problems before findings enter the public record. When that mechanism is bypassed — intentionally or not — the community loses confidence in the result, not necessarily because the result is wrong, but because the system of verification failed.
There is also a subtler problem. When researchers hold equity in a company commercializing the drugs they are testing, the pressure to frame results favorably — in which experiments to run, which data to highlight, how to interpret ambiguous findings — is real and documented across decades of pharmaceutical research. That pressure does not require conscious dishonesty. It operates at the level of experimental design choices that look entirely reasonable to the person making them.
The practical answer for anyone following this story: the underlying science deserves replication by independent groups with no financial stake in the outcome. Until that happens, the results sit in an uncomfortable middle ground — neither confirmed nor refuted, but no longer carrying the validation stamp that PNAS publication was supposed to provide.
The Vega Oncotargets Question
The existence of Vega Oncotargets — a company apparently created to develop the therapy commercially — raises questions that go beyond this single paper. Building a startup around research findings before those findings are publicly validated is not inherently unethical; it is, in fact, a standard feature of how academic science gets translated into actual treatments. The ethical requirement is transparency: readers, peer reviewers, and journal editors need to know about those interests when they are evaluating the research.
What complicated this case further was its public footprint. The CRIS Cancer Foundation, a Spanish nonprofit focused on cancer research funding, had organized fundraising efforts connected to the clinical development pathway following the paper's initial publication. After the retraction, CRIS announced it was reviewing how those funds would be allocated. The story had already moved from the laboratory to the fundraising world before the procedural failure came to light.
What This Means for Pancreatic Cancer Research Going Forward
Set the retraction aside for a moment and look at the broader scientific landscape. The Barbacid paper — regardless of its current status — sits within a genuine paradigm shift in how oncologists think about KRAS-driven cancers.
Daraxonrasib itself is still in active human clinical trials. The Phase 3 RASOLUTE-302 trial is comparing it against standard chemotherapy in metastatic pancreatic cancer patients and was ongoing at the time of writing. Results from that trial will carry independent weight, untouched by the Barbacid paper's fate. If daraxonrasib shows meaningful survival benefits at the single-agent level, the scientific logic for exploring triple combinations with it becomes even more compelling.
The broader principle Barbacid's team was testing — that blocking KRAS, EGFR, and STAT3 simultaneously can prevent the resistance that kills single-agent approaches — remains scientifically plausible and worth investigating. Combination approaches have already extended survival meaningfully in melanoma (BRAF plus MEK inhibitors), lung cancer (targeted therapy plus immunotherapy), and HER2-positive breast cancer. The same logic applied to PDAC is reasonable. The question is whether independent researchers, without financial ties to any specific compound, will pursue it rigorously.
KRAS mutations are not limited to pancreatic cancer. Roughly 30% of lung cancers and about 40% of colorectal cancers carry them. Whatever combination strategy ultimately proves effective against PDAC could in principle be adapted for those diseases too. That is the scale of what is scientifically at stake — which is precisely why getting the evidence right, through transparent processes, matters so much.
Who Should Be Paying Attention
- Pancreatic cancer patients and families: The retraction does not close a door — it leaves one open, but without confirmation. Enrollment in clinical trials testing daraxonrasib and related agents remains the most direct way to access emerging therapies. Discuss options with an oncologist who specializes in PDAC and check active trials at ClinicalTrials.gov.
- Oncology researchers: The experimental models described in this paper, and the scientific rationale for the three-target approach, represent a legitimate research direction that now needs independent replication — without the commercial entanglement that compromised the original publication.
- Science journalists and editors: This case is a reminder that "published in PNAS" is not a final verdict, and that conflict-of-interest disclosures exist for reasons that become visible only when they are violated. The media's initial coverage — which treated preliminary mouse data as a near-cure — deserves its own honest postmortem.
- Biotech investors and pharmaceutical companies: Vega Oncotargets and the drug combination it was built around remain in an uncertain position. The clinical logic for the approach has not been disproved, but the foundational published evidence supporting it has been formally withdrawn and awaits resubmission.
- Science policy advocates: This is a case study in how academic conflict-of-interest rules interact with the growing pressure on researchers to commercialize their work. The rules exist; the failure was in following them.
Verdict: Hope, Caution, and the Longer View
Mariano Barbacid spent four decades mapping the molecular landscape of KRAS-driven cancer. The discovery that opened this story — his 1982 identification of the first human oncogene — genuinely changed how science understands cancer. Nothing about the retraction erases that contribution or makes the question he was pursuing less important.
But the retraction is not nothing, either. Pancreatic cancer patients and their families followed this story with the kind of attention that comes from having no good options. The procedural failure that led to the retraction — financial interests in a commercial vehicle, undisclosed to the journal that validated the work — was not an abstract administrative matter to people who reorganized their hope around the results.
The science may ultimately be confirmed. Barbacid has said he will resubmit with proper disclosures. Independent groups may replicate the findings. The clinical logic of the triple combination is coherent and worth pursuing. None of that is foreclosed. What is foreclosed, for now, is treating the original results as validated — because the validation process broke down at a critical step.
Pancreatic cancer will yield eventually. The molecular understanding of it has never been deeper, the tools available to test combinations have never been more sophisticated, and the first KRAS inhibitors to reach patients at all arrived only in 2021, after half a century of nothing. Progress is real. It just needs to be built on foundations that can hold the weight of what is being promised.
Frequently Asked Questions
Was the Barbacid pancreatic cancer paper retracted because the data was fraudulent?
No. PNAS explicitly stated that the retraction was due to an undisclosed conflict of interest — three authors held financial stakes in Vega Oncotargets, a company created around the research, and this was not disclosed at submission. The journal has not alleged that the experimental data was fabricated or manipulated. That said, the peer review process was compromised procedurally, and the findings now require independent replication before they carry scientific consensus weight.
What is Vega Oncotargets and why does it matter?
Vega Oncotargets is a biotechnology startup created to develop the triple-combination therapy described in the retracted paper commercially. Barbacid and two co-authors held financial interests in the company without disclosing this to PNAS at the time of submission. This matters because financial ties to a company that profits from positive research results create pressures — not necessarily conscious ones — that peer review is specifically designed to account for through disclosure and independent editorial oversight.
Can pancreatic cancer patients access the triple-combination therapy now?
No. The therapy was tested only in mouse models and was never in human clinical trials. Even before the retraction, Barbacid himself stated that the team was not ready to begin human trials. The retraction does not affect the current availability of the component drugs individually — daraxonrasib, for instance, is being tested in separate ongoing trials — but the specific triple combination has no active clinical development pathway currently announced.
What is daraxonrasib and is it still being tested?
Daraxonrasib (RMC-6236) is a RAS(ON) inhibitor that targets mutant KRAS proteins across a broad range of mutation types. It is currently in Phase 3 human clinical trials for metastatic pancreatic cancer, including the RASOLUTE-302 trial comparing it against standard chemotherapy. Its clinical development continues independently of the retracted Barbacid paper.
Does the retraction mean the triple-combination approach is scientifically dead?
Not at all. The scientific rationale — blocking KRAS, EGFR, and STAT3 simultaneously to prevent resistance — remains a plausible and testable hypothesis. The retraction removes the published validation of one set of experiments, not the underlying logic of the approach. Independent researchers can and likely will test similar combinations, and Barbacid has indicated plans to resubmit the paper through the correct editorial channel with full disclosures.
Why did PNAS retract the paper instead of just adding a disclosure note?
PNAS policy distinguishes between papers submitted by Academy members with competing interests and those without. When competing interests exist, the submission must go through a "Direct Submission" process with independent editorial oversight. Because that process was bypassed entirely, the journal determined that a correction or note was insufficient — the editorial integrity of the submission process had been fundamentally compromised, requiring full retraction.
How common are retractions in high-profile cancer research?
More common than public awareness suggests. Retraction Watch, which tracks withdrawn scientific papers, has documented thousands of retractions across medical research, with conflict-of-interest failures representing a meaningful share of procedural retractions. High-profile cancer research is particularly vulnerable because the stakes — commercial, reputational, and emotional — are high for all involved. Most retractions do not involve fraud; many involve exactly the kind of procedural and disclosure failures seen here.
What should I do if I was following this story hoping for a new treatment?
Stay connected to clinical trial registries — particularly ClinicalTrials.gov and the EU Clinical Trials Register — where trials testing KRAS inhibitors and combination regimens for pancreatic cancer are listed in real time. Organizations like the Pancreatic Cancer Action Network and the CNIO maintain updated information on research developments. Discuss any emerging trial with a specialist in pancreatic oncology who can assess eligibility and fit.
Sources: PNAS (Proceedings of the National Academy of Sciences), CNIO (Centro Nacional de Investigaciones Oncológicas), EurekAlert, Technology Networks, SciTechDaily, Ara.cat, Demócrata, GlobalPortalNews, Wikipedia. Pricing and specifications reflect the latest available data at time of writing. Always verify current details with official sources.