Live-cell FRET imaging-based efficacy prediction of EGFR-TKIs in EGFR L858R-mutant NSCLC

Functional assays can assist individualised selection of treatment regimens by evaluating the ex vivo response of tumour cells to anticancer therapeutics. However, the two major strategies used to assess responses to targeted drugs, target-based profiling and phenotypic profiling, have inherent limitations. Target-based profiling does not directly capture downstream cell-fate outcomes, whereas phenotypic profiling provides limited mechanistic information on drug action.

Live-cell fluorescence resonance energy transfer (FRET) microscopy was used to integrate the interaction between EGFR^L858R and GRB2 with cellular phenotypic features, thereby predicting the efficacy of EGFR-tyrosine kinase inhibitors (TKIs) at the single-cell level.

The target-based drug score (T-score), derived from the interaction between EGFR^L858R and GRB2 in living cells co-expressing CFP-EGFR^L858R and YFP-GRB2, enabled evaluation of the on-target activity of EGFR-TKIs against EGFR^L858R. The phenotype drug score (P-score), constructed from cellular phenotype features, characterised the phenotypic responses induced by EGFR-TKIs. The TP-score, integrating the T- and P-scores, accurately predicted the efficacy of EGFR-TKIs. In living H3255 cells, the TP-score of osimertinib (OSI) was higher than that of gefitinib (GEF). Notably, in H3255 cell-derived xenograft (CDX) mouse models, OSI showed superior in vivo antitumour efficacy compared with GEF. Furthermore, live-cell TP scoring revealed that primary tumour cells from a patient with NSCLC did not respond to OSI, a finding consistent with the absence of actionable EGFR mutations in this patient.

This study demonstrates the clinical feasibility of predicting EGFR-TKI efficacy by integrating EGFR^L858R-GRB2 interactions with cellular phenotypic features in single living cells.