Pangea Biomed Takes EIC Accelerator Pitch to the Pathology Slide
A tumour biopsy is routinely cut into thin slices and stained with haematoxylin and eosin, the H&E slide a pathologist reads under the microscope. Reading the tumour's RNA takes a separate sequencing test. Pangea Biomed, a Tel Aviv company, says software can stand in for that test: one model estimates the tumour's gene expression from a scanned slide, and a second model turns that estimate into a ranking of the drugs the patient should respond to best, by the model's estimate.
The European Innovation Council selected the project, ENLIGHT-DP, described as "AI-based Precision Oncology Diagnostics from Routine Pathology", for blended finance in the list of selected companies from the May and July 2026 cut-offs, published with the EIC's announcement of 16 November 2026. The list states that selection "does not constitute a formal commitment for funding". This article sets out what the project promises and what can already be checked.
Company file
- Company
- Pangea Biomed, founded in 2018, per the company's press releases
- Headquarters
- 2A HaBarzel Street, Tel Aviv, Israel, per the company website
- Co-founders
- Eytan Ruppin, Ze'ev Ronai and Eyal Gottlieb, per GEN in April 2022
- Leadership
- Tuvik Beker, CEO; Ranit Aharonov, CTO
- Investors
- NFX and angel investor Danny Tocatly, per the company
- Products
- ENLIGHT (drug response from RNA data) and ENLIGHT-DP (the same from H&E slides)
- EIC project
- ENLIGHT-DP, May or July 2026 cut-off, blended finance; no CORDIS project yet
Two models, one slide
ENLIGHT came first. In a paper in Med, published online in December 2022 with Pangea's Gal Dinstag as first author, the company describes a method that predicts response to targeted drugs and immunotherapies from a tumour's RNA profile without being trained on any treatment-response data. Instead of learning from past patients on a given drug, it looks at the network of genes that interact with the drug's targets and asks whether the tumour's expression of those genes favours or opposes the treatment. Across 21 blinded clinical trial datasets the authors report an odds ratio of 2.59. The paper discloses that several authors are Pangea employees.
GEN reported in April 2022 that the method rests on nearly a decade of work by the co-founders, among them Eytan Ruppin, chief of the US National Cancer Institute's Cancer Data Science Lab. The same article records a limit acknowledged by the paper's co-authors: ENLIGHT has "very limited utility" for chemotherapy, because it works at the level of drug targets.
ENLIGHT-DP adds the slide. Its first stage, DeepPT, was developed by the Australian National University with the National Cancer Institute and Pangea, according to an ANU release of 3 July 2024, which says the model was trained on more than 5,500 patients across 16 cancer types. The joint paper appeared in Nature Cancer in September 2024 (volume 5, pages 1305 to 1317), with 31 authors including Pangea's CEO Tuvik Beker and CTO Ranit Aharonov. It reports that the combined method picked out true responders in five independent patient cohorts covering four treatments and six cancer types, with an overall odds ratio of 2.28. The ANU release puts the response rate among patients the model would have matched at 46.5%, against 33.3% without it.
The evidence
Matched patients responded more often in retrospective cohorts
Across five retrospective cohorts in the Nature Cancer study, the response rate among patients ENLIGHT-DP would have matched to their treatment was 46.5%, against an overall rate of 33.3%.
Australian National University release via ScienceDaily, 3 July 2024; Hoang et al., Nature Cancer, September 2024.
What the studies are, and are not
Both the Nature Cancer cohorts and the lung study below are retrospective: the model scored slides from patients whose outcomes were already known. HIT Consultant reported in July 2024 that further validation and testing were planned for regulatory approval. Sources reviewed on 17 November 2026 show no prospective trial results and no regulatory clearance for ENLIGHT-DP.
The lung cancer test
The most specific clinical figures come from a study in the Journal for ImmunoTherapy of Cancer, published on 11 January 2025, whose authors include Beker and Aharonov. The team scanned pre-treatment H&E slides from 50 patients with metastatic lung adenocarcinoma treated with first-line immune checkpoint inhibitors, 46 of them together with platinum chemotherapy and 4 without. Ranking patients by likelihood of response, ENLIGHT-DP reached an area under the curve of 0.69, against 0.52 for tumour mutational burden and 0.46 for PD-L1 expression.
The lung study
In 50 lung cancer patients, the slide score beat PD-L1 and TMB
Area under the curve for predicting response to first-line checkpoint inhibitors; 0.5 equals chance. The study was retrospective and single-cohort.
Arnon, Dinstag et al., Journal for ImmunoTherapy of Cancer, 11 January 2025.
At a cut-off fixed in advance, every patient ENLIGHT-DP flagged as a responder did respond (a positive predictive value of 100%), but the tool caught 44% of the responders. PD-L1 above 50% scored 65% and 38% on the same measures, and high tumour mutational burden 82% and 26%. ENLIGHT-DP was the only marker in the cohort significantly associated with progression-free survival, with a hazard ratio of 0.45 (95% confidence interval 0.2 to 0.99, p=0.048), an interval whose upper end sits just below 1, the value for no effect. The authors write that studies in other tumour types, in immunotherapy without chemotherapy and in bigger lung cancer cohorts are warranted.
What the company has done since 2022
Pangea came out of stealth in April 2022 with a $7 million seed round led by NFX, GEN reported. In a press release of 15 December 2022 the company announced $5 million more, led by Danny Tocatly with NFX, which it said brought the seed round to $12 million, to expand US operations, commercial partnerships and products. The same release says Pangea had begun two clinical trials and formed partnerships with several pharmaceutical companies during 2022; it names none of the companies or trials.
The money
$12M of seed funding, all announced in 2022
Both amounts are company announcements. Sources reviewed on 17 November 2026 show no later round; no EIC amount per company is published.
GEN, 25 April 2022; Pangea Biomed press release, 15 December 2022. Values in US$ millions.
2024 brought a series of named collaborations, each announced by the company. In August, Pangea, the Sharett Institute of Oncology at Hadassah University Hospital Ein-Kerem and Ultima Genomics announced a multi-year project to produce ENLIGHT reports for more than 2,000 patients a year from RNA data and whole-slide scans, as reported by HIT Consultant. In September the company said it had joined the second phase of the LiquidBX consortium, funded by the Israel Innovation Authority, which plans multi-omics data from more than 1,000 solid-tumour patients at Hadassah. In October it announced a pilot with the OHSU Knight Cancer Institute in Oregon to run ENLIGHT-DP on slides and RNA data from longitudinal biopsies, planned for about five months and a joint manuscript.
The record
Papers and partnerships through 2024, then the EIC in 2026
Most dated milestones fall between 2022 and 2024. The company's news page lists nothing after November 2024; the next public record is the EIC selection.
- 2018FoundedPangea Biomed founded in Tel Aviv to commercialise the co-founders' research
- Apr 2022Out of stealthSeed round led by NFX$7M
- Dec 2022Seed extensionLed by Danny Tocatly with NFX; Med paper on ENLIGHT published online$5M
- Jul 2024ENLIGHT-DP paperDeepPT and ENLIGHT results released with ANU and the US National Cancer Institute; printed in Nature Cancer in September
- Aug to Oct 2024CollaborationsHadassah and Ultima Genomics, the LiquidBX consortium, and a pilot with the OHSU Knight Cancer Institute
- Jan 2025Lung cancer study50-patient retrospective study in the Journal for ImmunoTherapy of Cancer
- 2026EIC applicationENLIGHT-DP submitted to the May or July 2026 EIC Accelerator cut-off
- Nov 2026SelectedOne of 35 companies on the EIC list for the May and July 2026 cut-offs, recommended for blended financeBlended finance
GEN, 25 April 2022; Pangea Biomed press releases 2022 to 2024; Med, Nature Cancer and Journal for ImmunoTherapy of Cancer; EIC, 16 November 2026.
Then the public record thins. The company's news page, read on 17 November 2026, lists as its newest item a Fast Company recognition dated 19 November 2024. Sources reviewed on 17 November 2026 show no funding round after December 2022, no named pharmaceutical customer and no published OHSU manuscript. Among the sources reviewed, the EIC selection is the most recent primary record of the company's activity; the EIC list does not say whether it applied to the May or the July 2026 cut-off.
The EIC selection
The EIC's announcement of 16 November 2026 covers both 2026 cut-offs together: 453 full applications, 88 interviews and 35 selected companies from 17 EU and associated countries, with €84 million in proposed grants and 32 companies to be assessed by the EIC Fund for equity of up to €10 million each. It gives no amount per company, and the grant agreements are expected within three months, so CORDIS lists no ENLIGHT-DP project yet. Pangea is eligible as an Israeli company because Israel signed its association agreement with Horizon Europe on 6 December 2021.
The competition
35 companies selected from 453 full applications
The figures cover the May and July 2026 cut-offs together; the EIC list does not say which cut-off each company applied to.
EIC, 16 November 2026.
The policy fit is direct. Europe's Beating Cancer Plan includes a flagship, "Cancer Diagnostic and Treatment for All", meant to widen access to diagnosis and treatment "through personalised medicine" and to foster next-generation sequencing. A test that works from "routine pathology", in the words of the project title, would be one route to that access where sequencing is not done; whether it works well enough in routine care is what the project has to show.
What can be checked now
- Peer-reviewed methodPublishedENLIGHT in Med; ENLIGHT-DP in Nature Cancer, with ANU and NCI co-authors.
- Clinical signal in a named indicationRetrospective50-patient lung adenocarcinoma study, January 2025; no prospective result found in sources reviewed on 17 November 2026.
- Regulatory approvalNot publicPlanned, per July 2024 coverage of the Nature Cancer study; no clearance found in sources reviewed on 17 November 2026.
- Paying customersNot namedPharmaceutical partnerships announced in 2022 without names; hospital and consortium collaborations announced in 2024.
- EIC grant agreementToo earlySelected on 16 November 2026; agreements expected within three months; no CORDIS record yet.
What the company says
ENLIGHT is, in its words, the world's most advanced multi-cancer, multi-therapy response predictor, and ENLIGHT-DP brings it to patients whose tumours were never sequenced, using only the slide that already exists.
What the record shows
Two peer-reviewed method papers and a 50-patient retrospective lung study, three named collaborations announced in 2024, $12 million of seed funding announced in 2022, and no posted company news since November 2024.
Verdict
Pangea Biomed brings more published science to the EIC than most companies at its stage: the method behind ENLIGHT-DP has been through peer review in Med and Nature Cancer, with government research institutes as co-authors. What sources reviewed on 17 November 2026 do not show is the step that turns a predictor into a diagnostic: published prospective results, a regulatory clearance and a named buyer. Its funding record stops at a 2022 seed round. The EIC project is the test of whether the slide-based method can make that step.
The reporting is based on publicly available sources, listed below.
Sources
- EIC, The European Innovation Council selects 35 deep tech companies for its Accelerator support, 16 November 2026
- EIC, list of selected companies, May and July 2026 cut-offs
- Pangea Biomed, company website and news page, read 17 November 2026
- Dinstag et al., Clinically oriented prediction of patient response to targeted and immunotherapies from the tumor transcriptome, Med, 2023
- Hoang et al., A deep-learning framework to predict cancer treatment response from histopathology images through imputed transcriptomics, Nature Cancer, September 2024
- Australian National University via ScienceDaily, release on DeepPT, 3 July 2024
- Arnon, Dinstag et al., Predictive value of ENLIGHT-DP in metastatic lung adenocarcinoma, Journal for ImmunoTherapy of Cancer, 11 January 2025
- GEN, Neighborhood Watch: Pangea pursues group approach to precision oncology, 25 April 2022
- Pangea Biomed press release via BioPharma Dive, $5M additional funding, 15 December 2022
- HIT Consultant, ENLIGHT-DP study in Nature Cancer, 3 July 2024
- HIT Consultant, Pangea Biomed partners with Hadassah and Ultima Genomics, 7 August 2024
- Pangea Biomed press release via WebWire, LiquidBX consortium second phase, 5 September 2024
- Pangea Biomed press release via WebWire, pilot with the OHSU Knight Cancer Institute, 24 October 2024
- European Commission, Israel joins Horizon Europe, 6 December 2021
- European Commission, Europe's Beating Cancer Plan flagship initiatives
Industry
HealthIn vitro diagnostics
- Focus
- Diagnostic software
- Companion diagnostics
- Technologies
- Artificial intelligence
- Computer vision
- Bioinformatics
- Keywords
- Oncology
- Precision oncology
- Digital pathology
- Transcriptomics
- Immuno-oncology
- Lung cancer