Following different paths

Two general pathways of resistance to targeted therapies have been proposed: the selection of clones carrying a pre-existing resistance mutation, or the eventual de novo development of a resistance mutation in cells that survive initial therapy (for example, through epigenetic adaptation or microenvironmental stimuli). Although evidence has been reported to support the first model, direct evidence of the second is lacking. Hata, Niederst et al. present data suggesting that genetic resistance to the epidermal growth factor receptor (EGFR) inhibitor gefitinib can develop through both paths.

Credit: Philip Patenall/NPG

The authors previously developed gefitinib-resistant cell lines from EGFR-mutant non-small-cell lung cancer (NSCLC) PC9 cells, and observed differences in the time required for the acquisition of the EGFR^T790M gatekeeper mutation; for example, PC9-GR2 cells developed resistance in 6 weeks whereas PC9-GR3 cells required 24 weeks. To examine this further, they cultured >1,200 small pools of parental PC9 cells in the presence of gefitinib for 2 weeks. This resulted in two classes of surviving cells: rapidly growing 'early-resistant' colonies and small 'intermediate-resistant' colonies of drug-tolerant cells. Early-resistant colonies all carried the EGFR^T790M mutation; these were derived from rare pre-existing EGFR^T790M cells in parental PC9 cells that were selected by gefitinib treatment.

The drug-tolerant intermediate-resistant cells did not carry the EGFR^T790M mutation. However, further long-term (10–30 weeks) culture in gefitinib led to the development of 'late-resistant' cells, several clones of which had acquired EGFR^T790M, suggesting that this mutation can arise de novo following prolonged exposure to gefitinib. Other resistant clones remained EGFR^T790M-negative, and the authors identified various mutations in other oncogenes (for example, KRAS, BRAF and RET) that might account for this resistance. Further experiments using PC9 cell subclones derived from single cells (to eliminate any pre-existing EGFR^T790M cells) confirmed that resistant cells carrying EGFR^T790M can arise from drug-tolerant cells that do not initially carry this mutation. Similar results were observed in another patient-derived NSCLC cell line that did not carry an initial EGFR^T790M mutation. Although these results seemed surprising, given the timescales and initial cell population sizes, mathematical modelling predicted that the EGFR^T790M mutation could indeed arise over a time course of several months.

Late-resistant cells had transcriptional profiles similar to those of intermediate-resistant drug-tolerant cells. Furthermore, intermediate- and late-resistant cells had decreased sensitivity to apoptosis induced by the third-generation EGFR inhibitor WZ4002 compared with early-resistant cells, suggesting that cells in which resistance arises later are less reliant on EGFR signalling for survival. The authors then examined several other EGFR^T790M-positive NSCLC cell lines derived from patients who had developed EGFR inhibitor resistance; these had differing sensitivities to WZ4002 in vitro and in xenograft models, suggesting that these different paths of resistance might be clinically relevant. Interestingly, the cells that were least sensitive to WZ4002 were derived from patients who had a long duration of response to initial EGFR inhibitor therapy, indicating that perhaps these cells arose from drug-tolerant cells in vivo. In addition, the combination of WZ4002 and navitoclax, which inhibits anti-apoptotic BCL-2 family proteins, could induce apoptosis of late-resistant cells and regression of xenograft tumours derived from these cells, suggesting a possible therapeutic strategy.

mutations that promote drug resistance can be both pre-existing and acquired de novo

Overall, these data support the idea that mutations that promote drug resistance can be both pre-existing and acquired de novo. Furthermore, as genetic resistance might arise from small populations of drug-tolerant cells in patients, this argues that it is important to develop therapeutic strategies to eliminate these cells before resistance develops.