Cancer therapy was revolutionized in the last two decades by molecularly targeted therapies. However, in many cases, the efficient and manageable treatment declines within months or few years due to repopulation of the tumor with resistant cells, which leads to progression and metastasis – the two major factors of cancer mortality. Cell line model treatments can show individual cell changes following treatment, but found targets usually are failing to be translatable to clinics. Patient-derived tumor xenograft (PDTX) models are resembling heterogeneity, structure, plasticity of the tumors more than cell lines. Therefore, we wanted to see if resistance evolution in PDTX models assimilate to those in cell lines, or can reveal novel pathways of interest.

We performed long-term treatment on melanoma (MM), renal cell carcinoma (RCC), and non-small cell lung cancer (NSCLC) with the respectively approved targeted drugs: vemurafenib, sunitinib and erlotinib. NOD-SCID mice used have limited timespan for the experiment, thus tumors were serially transplanted and treated again. This fashion enabled us to collect flash-frozen samples from each generation of mice. The samples were either analyzed using mRNA-seq to show genetic background and changes in the landscape during treatment, or possible suspects were further confirmed by rt-pcr, western blot or immunohistochemistry.

In melanoma, BRAF V600E mutation specific inhibitor usually is reported to enhance MAPK or AKT pathway activities, as well as loss of the mutation, or enhancement of CRAF. Intriguely, none of these changes were recorded in PDTX resistance model, whilst upregulation of multidrug resistance determinant ABCB1, immune checkpoint stimulator CD27, or interferon inducible protein IFI27. In our RCC models, two different resistance strategies were identified, one described by the overexpression of MAPK pathway, the other by silencing cell cycle and apoptosis, generating the tumors from a slower, mesenchymal, motile cell phenotype. In NSCLC, the resistance mutation of EGFR was identified in one case, while altered MAPK and PI3K signaling in case of the other model.

Our PDTX resistance experiments show clear evidence that previously published resistance mechanism share role with unknown, patient-specific resistance strategies. With an increasing number of such model experiments, typical background landscapes might be identified to propose and intervene probable drug tolerance in solid tumors.