Back

Leveraging death of drug-sensitive cancer cells to promote immune-mediated bystander killing of subclones of drug-resistant tumor cells

Tomaschko, M.; Ng, K.; Moore, C.; Pillsbury, C. E.; Rana, S.; Campbell, J.; Mukherjee, S.; Mikolajczak, A.; Anastasiou, P.; de Castro, A.; Alonso de la Vega, A.; de Carne Trecesson, S.; Goehring, N.; Molina-Arcas, M.; Downward, J.

2025-10-28 cancer biology
10.1101/2025.10.27.684825 bioRxiv
Show abstract

The response of lung cancer patients to drugs targeting the G12C mutant form of KRAS is limited by the development of resistance through multiple mechanisms. In order to achieve lasting benefit with these therapies, effective strategies for tackling the evolution of drug resistance are required. We have developed a preclinical model system to mimic the development of resistance to KRAS G12Ci inhibitors (G12Ci) such as adagrasib and RMC-4998. Treatment of tumours containing a minor subpopulation of resistant cancer cells with G12Ci leads to their rapid outgrowth to replace the drug-sensitive cells within a few weeks. However, when combined with therapies that, at least in part, target the immune response, such as SHP2 inhibitors or PD-1 blockade, drug-resistant cells can be eliminated, even by drug combinations that do not impact their growth in the absence of drug-sensitive cells. This bystander killing of drug-resistant cells when drug-sensitive cells are targeted is dependent on an intact adaptive immune system. Mechanistically, these combination therapies lead to profound remodelling of the tumour immune microenvironment, with influx of T cells recognising a tumour associated antigen shared between drug-resistant and drug-sensitive cancer cells. Promotion of immune-mediated bystander killing of drug-resistant cells may provide a paradigm for tackling the problem of drug resistance in cancer more broadly.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.