Finding cancer early changes the odds Pancreatic cancer, five years on: 3 in 100 alive if found late, 44 if found early.1
Scroll ↓
Tumours shed DNA into the blood Read those fragments and you read the tumour, without cutting into it.
Too little to find cancer early About 6,000 genome copies. An advanced tumour fills a share of them; an early one may leave none.2
What if we just had more material? The kidneys pass the blood's DNA into urine A whole cup, at home, with no needle.3
Urine gives 50× more DNA Each square is one blood draw.4
We break the extraction wall A whole cup goes in. Pure, concentrated cell-free DNA comes out, ready for any test.
More molecules Blood draw: ~6,000 copies Urine kit: up to 18,000 Hologenomix, one whole void: ~300,000 We pull the DNA from a whole morning void, collected at home.4
Blood expects 0.06 tumour copies. The cup, 3 Say 1 in 100,000 copies comes from the tumour, in urine as in blood.
A 6% chance becomes 89% or more 100% 0% 1 in 100k 1 in 1M rarer tumour DNA → 10× less in urine Hologenomix, 300 mL Blood Chance the sample holds a tumour copy. Dashed: if urine holds ten times less tumour DNA than blood.5
Shown in bladder cancer. Next: pancreatic cancer First pancreatic data: Q2 2027.
Better tests, bigger incentives 600,000 liquid-biopsy tests a quarter
6.7 → 13.2 months with the first RAS drug
$1,495 Medicare pays per blood screening test
Testing grows, drugs work, and payers pay.6 Why now →
We move fast A trial running, our device nearly finished, pancreatic data next.
In patients: more DNA than blood Blood draw: ~20 ng Urine kit: up to 60 ng Hologenomix (85 mL) a whole void (300 mL) 4,080 ng 3,600 ng 3,200 ng 824 ng 780 ng 500 ng Every bladder-cancer sample beat a blood draw, from 85 mL. Patent filed.7
Running clinical trial
PRECIS-BC: can a urine test spare the bladder? With Radboudumc, 60 patients, 2026 to 2028.
About the trial →
Team In Delft, on the Biotech Campus.
Eric van der Toorn Co-founder, CEO Bioinformatician and AI researcher
David Calderón Franco Co-founder, CTO A leading expert in cfDNA isolation in Europe
Thomas Abeel Co-founder, CSO. Associate professor Leading AI bioinformatician
Harshitha Hanji Business development Biotech partnerships and grant funding
Gonçalo Macedo Research scientist Quality and regulatory, ex-MSD Advisors Each has taken a life-science company to a sale.
Prof. Jack Schalken Urology professor. Co-founded NovioGendix, sold to MDxHealth
Joris Schuurmans CEO SkylineDx. Sold Cergentis in 2022. Before: Biocartis, MyCartis
Jan Groen, PhD Chaired Vesica Health to its 2026 sale. Ex-CEO MDxHealth Supported by Health Holland funds the PRECIS-BC trial (TKI-LSH SME Call 2025). Provincie Zuid-Holland: MIT feasibility study. Kansen voor West is co-funded by the European Union (EFRO).
Notes and sources Five-year survival for pancreatic cancer: 44% if localised, 3% if it has spread (SEER). US deaths 2026: American Cancer Society. ↩ Dying tumour cells release short fragments of DNA into the blood; reading them is a liquid biopsy. They thin out in five litres of blood, and a test reads one tube: a median 6,000 genome copies per blood sample in 194 patients (Zviran et al., Nature Medicine 2020). At one tumour copy in a million, the tube expects 0.006 copies of any one tumour mutation, so usually zero. “If no physical fragment that supports a targeted [mutation] is present in the sample, even ideal ultra-deep targeted sequencing will fail to discover a signal.” ↩ Short DNA fragments from the blood cross the kidney into the urine (transrenal DNA): first shown with donor and fetal DNA in urine (Botezatu et al., Clinical Chemistry 2000); urine holds small, 150–250 nucleotide DNA from the circulation (Su et al., Journal of Molecular Diagnostics 2004); methylation maps the tissues it comes from (Cheng et al., Clinical Biochemistry 2017). Review: trans-renal cell-free tumour DNA for urine liquid biopsy (Frontiers in Genetics 2022). ↩ Measured in healthy donors: 250–1,000 ng DNA per 100 mL of urine, rising in step with volume up to 160 mL. A morning void is ~300 mL (our device takes 100–500 mL). 3.3 pg per genome copy, so ~300,000 copies per void against a median 6,000 in a blood draw: about 50×. Counted from DNA mass; the usable human share is part of what we measure first. ↩ At 1 tumour copy in 100,000: blood 6%, a whole void 89% to ~100%. At 1 in a million: blood 0.6%, a whole void 20–59%. Urine also carries DNA from the kidneys and bladder, so its tumour share may be lower than blood's. At a tenth of blood's share (dashed), the cup still gives 20–59% against blood's 6% at 1 in 100,000. For pancreatic cancer, that share is the first thing we measure. Poisson sampling from the copy numbers in note 4. ↩ Liquid-biopsy testing: about 600,000 tests in Q3 2026 from the four largest US liquid-biopsy labs, up 70% on Q3 2025 (company reports). The first RAS drug for pancreatic cancer was approved by the FDA on 26 August 2026: median survival 6.7 → 13.2 months in previously treated patients whose cancer has spread (phase 3, NEJM 2026). Medicare pays $1,495 for an FDA-approved blood screening test for bowel cancer (CMS rate, 2026). ↩ Our device: six urine samples (85 mL) from four bladder-cancer patients gave 500–4,080 ng of DNA each, every one more than a blood draw (~20 ng) or a urine kit (up to 60 ng); our 2026 poster. Grey bars: the same yield per mL in a 300 mL void, not yet measured. ↩ 
