DRES-14. PROTEIN AGGREGATE FORMATION PREDICTS CLINICAL RESPONSES TO EGFR TKIs

Pim French,Ya Gao,Maurice de Wit,Darlene Mercieca,Iris de Heer,Martin van Royen,Joachim Aerts,Peter Sillevis Smitt
DOI: https://doi.org/10.1093/neuonc/noy148.321
2018-01-01
Neuro-Oncology
Abstract:EGFR tyrosine kinase inhibitors (TKIs) have thus-far not shown clinical benefit for glioma patients, which is in stark contrast to the benefit observed for patients with EGFR-mutated pulmonary adenocarcinomas (PA). We performed functional analysis on various EGFR-mutation constructs to better understand the differential sensitivity to EGFR TKIs. Response to EGFR TKIs was monitored using high-throughput automated imaging analysis, western blot, immunofluorescence, RT-QPCR, and immunoprecipitation-mass spectrometry (IP-MS). Addition of TKIs to cells ectopically expressing EGFR mutation constructs resulted in a rapid and strong formation of EGFR-protein aggregates. However, aggregates were found only cells expressing mutations common to PAs (n=3 mutations tested) but not in mutations common to gliomas (n=4) or those harboring secondary resistance mutations (n=1). Moreover, aggregates were only observed in TKIs with proven clinical benefit (erlotinib, gefitinib, dacomitinib or osimertinib) and not with a drug with no such benefit (lapatinib). We find a high concordance between the IC50 for viability of 13 cell lines harboring endogenous EGFR mutations (n=12 different mutations) and the IC50 of mutations to form protein aggregates. Patients harboring mutations that are sensitive to aggregate formation (IC50 < 500 nM) had significantly longer time to progression (median survival 7.5 vs 13 months, HR 0.25, P=0.012). These data demonstrate that formation of aggregates predicts response to EGFR TKIs in patients. The aggregates are formed, predominantly in the nucleus, first by a dephosphorylation of EGFR after which specific mutations and drugs both affect the (inactive) conformation state of EGFR. Such a state renders the protein more prone to aggregate formation and subsequent complete inactivation. Protein aggregation predicts response to EGFR TKIs. Since these aggregates inactivate all functions of the protein, the absence of aggregate formation with glioma specific mutations in EGFR explain why these tumors are insensitive to EGFR TKIs.
What problem does this paper attempt to address?