LONDON, UK, 13 October 2026 – Concr, the techbio company using Bayesian AI for precision oncology, today announced new results showing that its molecular digital twins accurately predict first-line chemotherapy response for patients with metastatic pancreatic ductal adenocarcinoma (PDAC).
In the PASS-01 trial (NCT04469556), a randomised phase II study of modified FOLFIRINOX (mFFX) versus gemcitabine/nab-paclitaxel (GnP) in 160 patients with metastatic PDAC, Concr's FarrSight® platform predicted each patient's chance of disease control on each regimen. The predictions were made blind to outcomes and scored independently by the PASS-01 Challenge Team at University Health Network (UHN), Toronto.
Concr's analysis formed part of the PASS-01 Challenge, an arm's-length evaluation of treatment-selection algorithms led by UHN. FarrSight® was the first external model to be evaluated.
Concr received each patient's genomic, transcriptomic and routine clinical data. Outcomes stayed with the Challenge Team, and the model was not tuned to the trial dataset. Of the 160 patients, 115 were evaluable.
Across those patients, FarrSight® stratified patients who achieved disease control from those who did not, and its predicted probabilities were in line with observed outcomes:
AUC of 74% (p<0.001) for predicting disease control across both arms, and 78% in patients on mFFX, the more intensive regimen, where knowing who is likely to respond matters most.
Patients FarrSight® predicted would achieve disease control did so in 76% of cases, versus 48% of those predicted not to, making a clinically meaningful stratification.
Predicted disease control rate reasonably matched those observed (expected calibration error 9.8%), the property that allows predictions to inform simulated control arms and trial design.
Because patients were randomised, some received the regimen FarrSight® favoured and some did not. Those who did had a 26% lower hazard of death (HR 0.74), with a C-for-benefit of 0.62, a measure of how well a model identifies who benefits more from one regimen than the other: an early treatment-selection signal that a cohort of this size is underpowered to confirm statistically.
Dr Matthew Griffiths, CTO and co-founder of Concr, said: "The ultimate test of any predictive model is its ability to work on blinded data. We are gratified to see FarrSight® maintain its consistent performance and pass this type of test without tuning to the trial dataset.”
Pancreatic cancer has a five-year survival of around 13%. Patients starting treatment face a choice between two standard chemotherapy regimens that perform similarly on average, and no biomarker has yet been validated in a randomised trial to say which will work for whom.
Prof. Andrew Biankin, Regius Professor of Surgery at the University of Glasgow and a key opinion leader in pancreatic cancer, commented: “For many patients with advanced pancreatic cancer, the first treatment is the only one they are likely to receive, so choosing it well matters enormously. In this disease, tumours often don't shrink much even when a treatment is working, which makes disease control a crucial signal, capturing benefit that response rates might miss and is more closely linked to how long patients live.”
“Results like these are a step in the right direction for how we develop drugs today, and how we will treat cancer patients in the future. Most approved treatments deliver only incremental benefit, largely because only a fraction of patients actually benefit. The rest bear the side effects without the gain, and health systems bear the cost. Predicting how an individual patient will respond, before treatment starts, would go a long way to addressing both."
Instead of learning from the trial cohort, FarrSight® draws on proprietary Bayesian priors learned from 25 billion data points spanning 190,005 patients, 17,633 compounds, 890 cell lines and 335 patient-derived xenograft models. It builds a digital twin of each patient's cancer from their molecular biology and predicts how that patient will respond to each available treatment through in silico n-of-1 studies.
PASS-01 builds on Concr's peer-reviewed COMPASS study, published in Frontiers in Artificial Intelligence in August 2026, in which FarrSight®-derived biomarkers outperformed conventional stratification in the basal-like subtype. COMPASS used observational data; PASS-01 takes the same approach to a randomised cohort with externally sequestered outcomes. A full manuscript validating the digital twin approach across more than 5,000 patients is under editorial review.
Dr Griffiths added: “In this study, FarrSight® picked out which patients were likely to respond, and the probabilities it gave were aligned to what actually happened. That's the part sponsors care about, because it's what lets them use digital twins to simulate a control arm in a hard-to-recruit indication, or check whether an early signal will hold up before committing capital to a larger trial."
Biopharma and diagnostics partners already trust FarrSight® to prioritise indications, explore new drug combinations, build synthetic control arms and identify likely responders. Partners defining patient-selection strategies or trial designs before committing to pivotal studies are invited to get in touch.
-Ends-
About Concr
Concr is a global, UK-headquartered techbio company applying astrophysics-derived Bayesian AI to simulate human cancer biology and drive Model-Informed Drug Development (MIDD). Uniquely engineered to overcome the limitations of traditional "big data" AI by generating predictive patient-level digital twins from sparse and fragmented multi-modal data, Concr delivers precision from early biomarker discovery through to simulated clinical trials and synthetic control arms. Validated in partnership with industry leaders like Roche and Debiopharm and premier clinical centres including The Royal Marsden and NHS Christie, Concr provides drug developers with the biological fidelity needed to de-risk R&D and accelerate effective cancer therapies to market.
For Media Enquiries
Emre Misirlioglu | Concr | emre@concr.co
Acknowledgements and disclosures
Concr acknowledges University Health Network (UHN) and the PASS-01 Challenge team at the Schwartz Reisman HPB Research Innovation Centre for providing the clinical data and independent evaluation, and thanks the patients and investigators who took part in the PASS-01 trial.
This study was conducted using data provided by the Ontario Institute for Cancer Research which is funded by the Government of Ontario. The views expressed in the publication are those of the author and do not necessarily reflect those of the Ontario Institute for Cancer Research or the Government of Ontario.
Concr products and services are for research use only. They are intended to support research and strategic decision-making in therapeutic and diagnostic development. They are not a medical device and are not intended to diagnose, treat, cure, or prevent any disease.
Prof. Andrew Biankin is a Scientific Strategy Advisor at Concr Ltd.
Back