The biology
Resistance is not one mutation.
Tumours that start with the same EGFR mutation can stop responding to osimertinib for different reasons[1]. A genetic profile shows what changed. A functional assay on living cells shows how they now respond. Cytogent’s work starts from that gap.
Escape routes
Three broad ways a tumour escapes.
- 01
On-target
The drug’s target changes. New mutations in EGFR itself, such as C797S, can stop osimertinib from binding.[1]
- 02
Bypass
Another growth signal takes over, for example through MET or HER2 amplification, so blocking EGFR is no longer enough.[1]
- 03
Histologic or cell-state shift
The tumour changes its type, for example towards small-cell histology, and no longer depends on EGFR.[1]
Why the lab still decides
A profile is a map. An assay is a test drive.
Sequencing catalogues mutations, but it does not by itself tell you how adapted living cells will respond to two drugs together. Patient-derived models such as organoids can keep features of the tumour they came from and can be used for drug testing[4], and open catalogues help researchers find them[5].
These models are scarce, so each experiment has to count. Choosing the experiments expected to teach the most is an active-learning problem[6], and the readouts are only useful if the assays are well designed and quality-checked[7].
Starting focus
Why start after osimertinib.
Osimertinib is a standard treatment option for advanced EGFR-mutant non-small cell lung cancer[2], and resistance to it is a well-described, varied problem[1]. That makes it a sharp place to test whether better experiment choice changes decisions.
Our inaugural programme blueprint is combination experiment selection in this setting.
What a pilot measures
Three signals, three different questions.
| Signal | Compares | What it can tell you |
|---|---|---|
| Combination benefit | The pair against each drug alone, in the same model | Whether two drugs together do more than either alone in that model. Not what happens in a patient. |
| Resistance-associated differential activity | The resistant model against its matched parental model | Whether a combination hits what resistance changed. Only as good as the match between the two models. |
| Clinical therapeutic window | Needs in vivo pharmacology and toxicology | Whether an effective exposure could be tolerated. A parental cancer line is not a healthy-tissue safety proxy. |
Unfamiliar terms? See the glossary.
Sources
These primary sources describe the scientific field, not Cytogent product performance.
- [1]Leonetti A, Sharma S, Minari R, Perego P, Giovannetti E, Tiseo M. Resistance mechanisms to osimertinib in EGFR-mutated non-small cell lung cancer. British Journal of Cancer. 2019;121(9):725–737. (review) doi.org/10.1038/s41416-019-0573-8
- [2]PDQ Adult Treatment Editorial Board. Non-Small Cell Lung Cancer Treatment (PDQ®) – Health Professional Version. National Cancer Institute. (reference) www.cancer.gov/types/lung/hp/non-small-cell-lung-treatment-pdq
- [3]National Cancer Institute. Drug resistance. NCI Dictionary of Cancer Terms. (reference) www.cancer.gov/publications/dictionaries/cancer-terms/def/drug-resistance
- [4]Kim M, Mun H, Sung CO, et al. Patient-derived lung cancer organoids as in vitro cancer models for therapeutic screening. Nature Communications. 2019;10:3991. (primary study) doi.org/10.1038/s41467-019-11867-6
- [5]Perova Z, Martinez M, Mandloi T, et al. PDCM Finder: an open global research platform for patient-derived cancer models. Nucleic Acids Research. 2023;51(D1):D1360–D1366. (The resource now operates as CancerModels.org.) (database) doi.org/10.1093/nar/gkac1021
- [6]Reker D, Schneider G. Active-learning strategies in computer-assisted drug discovery. Drug Discovery Today. 2015;20(4):458–465. (review) doi.org/10.1016/j.drudis.2014.12.004
- [7]Markossian S, Grossman A, et al., editors. Assay Guidance Manual. Bethesda (MD): Eli Lilly & Company and the National Center for Advancing Translational Sciences; 2004–. (guideline) www.ncbi.nlm.nih.gov/books/NBK53196/
