Characterization of TKI-induced drug-tolerant persister cells from patient-derived cell lines

Introduction: Targeted therapies provide substantial clinical benefit in oncogene-driven cancers, yet relapse remains inevitable. Increasing evidence suggests that a small population of drug-tolerant persister (DTP) cells survives initial therapy and ultimately gives rise to resistance. Understanding the vulnerabilities of these DTP cells is therefore essential for developing strategies that could prevent resistance before it emerges. Materials and Methods: Patients-derived cell lines (PDC) were established from NSCLC patients included in two prospective clinical trials: MATCH-R study (2015 to 2022) and the ongoing STING trial. The two main models used were MR57, harboring EML4-ALK fusion and ALK C1156Y/G1269A mutation and sensitive to the 3rd generation ALK inhibitor lorlatinib, and ST6566, carrying an EGFR L858R mutation and highly sensitive to osimertinib. Results and Discussion: DTP cells are characterized by marked phenotypic plasticity. In two PDC models, we observed the co-expression of epithelial and mesenchymal markers, supporting a hybrid EMT state. This plasticity is supported by BRD4, which regulates EMT-associated transcriptional programs, and by FGFR signaling, through pathway rewiring. Phospho-RTK and phosphokinase arrays revealed increased IGF1R and STAT3 phosphorylation level, consistent with adaptative signaling in DTP cells. In parallel, DTP cells displayed elevated gammaH2AX foci relative to treatment-naive cells, consistent with TKI-induced genomic instability that may induce a dependency on DNA-damage response pathways. In vitro models also enable to functionally assess these features, as well as other key DTP cells hallmarks, using long-term survival readouts, including crystal-violet clonogenic assays and IncuCyte live-cell imaging. These experiments revealed that targeting EMT-associated transcriptional control with the BRD4 inhibitor JQ1, inhibiting FGFR-driven signaling with erdafitinib, or blocking DNA-damage repair via the ATM inhibitor AZD-0156 each significantly delayed DTP regrowth. Together, these findings underline multiple actionable weaknesses emerging during the DTP state. Conclusion: Overall, our results demonstrate that DTP cells regrowth can be delayed in vitro , suggesting exploitable vulnerabilities. While further validation is required, they provide a rational for testing these strategies in vivo , with an ultimate goal of informing patient therapeutic approach.