Pancreatic cancer remains a highly lethal cancer where therapeutic response is limited by both intrinsic and acquired chemoresistance. Understanding resistance mechanisms may therefore lead to improved therapeutic strategies. We have recently defined specific molecular subgroups of pancreatic cancer associated with pre-clinical and clinical response to select tailored treatment strategies [1-3]. Using robust patient-derived xenografts (PDXs) of pancreatic cancer, we have generated novel in vivo models for the study of intrinsic and acquired chemoresistance mechanisms to clinically-used agents, gemcitabine, mitomycin C and cisplatin.
Whole-genome sequencing and microarray profiling of gemcitabine-resistant tumours revealed complex but potentially targetable resistance mechanisms, including increased DNA repair through activation of PARP1, MCM genes and RRM1, and changes within the tumour microenvironment. Importantly, acquired resistance to gemcitabine was effectively reversed by a novel PARP inhibitor, rucaparib, indicating that combination therapy involving this low toxicity agent may be useful in treating gemcitabine-resistant tumours defined by high genomic instability. Similarly, modulation of key components of the tumour microenvironment with fasudil, as recently achieved [2], provided another effective way of reversing gemcitabine resistance.
Our findings demonstrate the promise of PDX models for the study of in vivo mechanisms of chemotherapy resistance and efficacy testing of novel agents for the treatment of human pancreatic cancer.