Where are you stuck?
Lack of a good starting point for a challenging target
We pick the right entry point with you. Diversity screening against 400,000+ compounds, fragment-based discovery where the pocket is shallow, virtual screening to narrow the field first, or display technologies when the answer is a peptide or an antibody.
Works in 2D, fails in vivo
Reevaluate it in organoids, then orthotopic models. Find where the activity is lost, before chemistry commits again.
Unsure which modality fits
Small molecule, peptide, antibody or ADC. Same target, same hands, comparable data.
The model you need is hard to build
Isogenic lines, patient-derived cultures and orthotopic models require careful engineering and validation. We edit them here, then confirm the model represents the tissue before a program depends on it.
It works, but you can’t say why
Target engagement, pathway effects, and where in the tumor the drug actually reaches. Mechanism evidence is what turns a compound into a translational plan with defined biomarkers and a defensible dose.
Which modality fits the target?
Click a hexagon
Each one opens the science behind it.
Small molecules
Enzymatic, allosteric, covalent and protein–protein interaction inhibitors, carried from hit to optimized lead
Proximity inducers
Leveraging advanced proximity-induction platforms for the discovery of targeted protein degradation and stabilization therapeutics, including PROTACs, molecular glues, and emerging modalities
Peptides
Non-natural amino acids, backbone edits and conformational constraint, moving exposure from hours to weeks
ADCs & PDCs
Payload conjugation, DAR analytics, aggregation, cytotoxicity, bystander effect, in vivo stability and efficacy
Antibodies
scFv ID & maturation, binding & internalization, direct signaling blockade, immune effector recruitment, checkpoint inhibition
Cell therapies
scFv discovery & maturation, CAR construct design & engineering, CAR-T production, in vitro and in vivo functional validation
How complex does the model need to be?
Complexity buys physiological relevance but costs throughput. Each model is deployed according to the biological question and the throughput the analysis requires.
Subcutaneous and orthotopic xenografts, syngeneic models, intracranial engraftment.
Complex 3D Patient Organoids (C3PO) platform
A patient-derived tumor line and selected tumor-microenviroment (TME) lines, each carrying its own fluorescent label, are seeded together and grown as spheroids. Compounds are then tested on that co-culture and imaged channel by channel, so the effect on the tumour cells and on the microenvironment cells is quantified in the same well.
Seeding and spheroid formation
Mixing of different labelled cell populations at seeding, left to form spheroids by day 6.
TUMOR CELL LINE
TME CELL LINE 1
TME CELL LINE 2
Mixing of different labelled cell populations
Day 0 · seeding
Day 6 · spheroid
Spheroid, merged channels
Compound testing and high content analysis
Drug treatment across a dose-response, high content multi-channel imaging, then image analysis and quantification.
GFP
tdTomato
RFP
Merge
Segmentation
Orthogonal views
GFP intensity (×10⁹) by day
CELL VIABILITY (%) VS CONCENTRATION (µM)Four patient-derived glioma models, imaged channel by channel
Click a model to view individual channels.
From challenging targets to preclinical candidates
Two different modalities and targets, each advanced to a preclinical candidate.
An orally dosed SHP2 inhibitor that reaches the brain, without trading away potency.
Fifteen SHP2 inhibitors have entered clinical trials. None has reached the market, and none is in active trials for brain cancers.
Structure-guided optimization of I-0436650 delivered I-1000233: a potent, orally bioavailable allosteric SHP2 inhibitor that evades efflux and crosses the blood-brain barrier.
In orthotopic glioblastoma xenograft models, oral administration of I-1000233 produced marked tumor growth inhibition, robust pERK suppression, and significantly extended survival.
I-1000233 treatment of patient-derived GSCs and TCGA patient data integration revealed selective sensitivity in the GBM proneural subtype: a biomarker signature for targeted patient selection.
A screening hit with the allosteric site confirmed by co-crystal, and an SPR assay in place to rank everything that came after it.
Scaffold hopping and a modified thiol fragment cleared the hERG liability without giving up potency.
The preclinical candidate: 4 nM enzymatic, clean safety panels, and oral exposure in every species tested.
A brain-permeable follow-up compound, aimed at brain tumors and metastases.
Discovered, optimized and characterized at IRBM.
See the latest publication on this program →
Discover, conjugate, de-risk: a complete CRLF2 ADC development pipeline
CRLF2 overexpression defines a high-risk subset of Ph-like B-ALL. There is no approved targeted therapy for it.
Phage display against recombinant protein and cells identified AM E3, a fully human α-CRLF2 mAb with single-digit nanomolar affinity for the target and a rapid internalization rate.
Upon conjugation with the payload tesirine (SG3249), a clinically validated cross-linking agent, AM E3-SG3249 demonstrated a potent anti-tumor effect on several CRLF2-overexpressing cell lines and orthotopic xenograft models.
An ADC surrogate against murine CRLF2, which has limited homology with the human counterpart, proved to be efficacious and well-tolerated in preclinical mouse models.
CAR-T Platform
From scFv discovery to functional validation of CAR-T cell therapy.
& engineering
production of CAR-T cells
Who we work with
We partner with world-renowned organizations devoted to cancer patient care, including MD Anderson Cancer Center. The NIH selected us for the NCI Experimental Therapeutics (NExT) Program’s Chemical Biology Consortium, addressing unmet needs in therapeutic oncology.
IRBM chemistry contributed to the approved drug.
Where should we start?
One screen. One model. One mechanism of action. Or the entire research program, with one team accountable for it.
Frequently Asked Questions
Which oncology models does IRBM run in-house?
Cell models spanning different levels of complexity: 2D stable and primary cells, 3D spheroids, patient-derived organoids including C3PO brain tumor organoids, co-cultures with immune or stromal compartments, and iPS-derived cells. In vivo work covers subcutaneous and orthotopic xenografts, syngeneic models and intracranial engraftment.
Which therapeutic modalities does IRBM support in oncology?
Small molecules including allosteric, covalent and protein–protein interaction inhibitors; proximity inducers such as PROTACs and molecular glues; therapeutic peptides; ADCs and PDCs; antibodies; and the discovery and functional validation stages of cell therapy programs. Because the campaigns run in the same building, a target can be tested across more than one modality on comparable data.
How does IRBM approach brain-penetrant drug discovery?
Design criteria are fixed before chemistry begins: a CNS MPO score of 4 or above, efflux ratio below 2 against both MDR1 and BCRP, and in vivo free brain-to-plasma partition above 0.3. Compounds are assessed against an iPSC-derived blood–brain barrier model built in-house, then confirmed in vivo. Applied to I-0436650, this converted a peripherally restricted compound into I-1000233, with a rodent Kp,uu of 1.0 and potency held at 17 nM on p-ERK.
Where is the work carried out, and under what accreditations?
All oncology work is carried out at IRBM’s single research site in Rome, Italy. In vivo studies run under AAALAC International accreditation, with study protocols reviewed by the internal Animal Welfare Body / OPBA and authorized under Italian and EU legislation. The site holds ISO 9001 certification, and IRBM is a member of the NCI Experimental Therapeutics (NExT) Program’s Chemical Biology Consortium.
Start a conversation.
Great science begins here.
