Competing interests br Ethics approval br Funding br Acknowl
Introduction An increased understanding of the molecular drivers of oncogenesis and resistance in non-small lung cancer (NSCLC) has led to a rapid expansion in the number of targeted therapies available to patients. Accordingly, molecular diagnostics in NSCLC have rapidly expanded, with guidelines now recommending analysis of multiple biomarkers including EGFR and BRAF gene mutations and rearrangements in ALK and ROS. The most common activating mutations in EGFR, del19 and L858R, are predictive of response to tyrosine kinase inhibitors (TKIs) in the advanced setting. In patients receiving first or second generation TKIs, the emergent EGFR mutation (EGFRm) T790M accounts for 60% of resistant cases. Osimertinib, a third generation TKI selectively inhibits both sensitizing EGFRm and T790M and has demonstrated improvement in progression free survival (PFS) compared to first generation TKIs in the first-line setting [1,2]. The optimal sequencing of EGFR-TKIs remains unclear. Resistance mechanisms to EGFR-TKIs are heterogeneous and multiple aberrations may be present simultaneously . Although tumour tissue genotyping remains the gold standard in detection of driver mutations, testing of circulating cell-free DNA (ctDNA) from liquid biopsies is growing and is particularly useful for patients with insufficient tumour tissue or where there may be substantial risk to the patient from invasive diagnostic procedures. Many next generation sequencing (NGS) panels are available that allow simultaneous detection of multiple genes and mutation types relevant in NSCLC from ctDNA, with one recently approved by the Food and Drug Administration (FDA) as a companion diagnostic in the United States (Oncomine™ Dx Target Test, ThermoFisher, Waltham, MA). While repeat tumour biopsies may help characterise the AUY922 (NVP-AUY922) of resistance in patients receiving targeted therapy, this is challenging in lung cancer patients who often are not well enough to undergo serial biopsies, nor does an isolated biopsy reflect the potential heterogeneity of resistance mechanisms at play. Monitoring levels of cell free circulating tumour (ct) DNA has been shown to correlate with disease response and progression with higher levels of ctDNA emerging at the time of radiographic progression [4,5]. We explored ctDNA levels in patients with stage IV EGFRm lung cancer in different phases of therapy, including pre-treatment, during EGFR-TKI treatment, upon progression and with chemotherapy.
Methods An ongoing prospective study at the Princess Margaret Cancer Centre is enrolling consenting patients with stage IV EGFRm NSCLC for longitudinal blood collection. Patients may enrol at any point in their treatment. In this study, we performed a cross-sectional analysis of 72 patient liquid biopsies using an NGS panel assessing regions of 11 genes (ALK, BRAF, EGFR, ERBB2, KRAS, MAP2K1, MET, NRAS, PIK3CA, ROS1, TP53) and >150 hotspots with a limit of detection (LOD) down to 0.1% (Oncomine™ Lung cfDNA Assay; Thermofisher). Conduct of this study was approved by the University Health Network Research Ethics Board.
Discussion As the treatment landscape in EGFRm NSCLC continues to evolve, resistance mechanisms are becoming better understood. More comprehensive analysis of ctDNA not only overcomes limitations of single site tissue biopsies but may also have an important role in longitudinal analyses, both in monitoring for clearance of primary EGFR mutations but also for the emergence of potential subclones and acquired mutations. While the main focus of research to date has been on the detection of T790M in ctDNA, our analyses show that when stable on treatment with an EGFR-TKI, the presence not only of T790M but other EGFR mutations, including co-occurring mutations, predicts a shorter progression free survival. This finding is supported by recent data from AURA3, suggesting that the persistence of an EGFRm results in an inferior response to osimertinib . Also, in the small group of patients receiving a TKI in our study the percentage variant allele fraction of the EGFRm also predicted time to confirmed progression. Others studies have shown that the % allelic fraction can impact treatment outcomes and survival  although this requires validation in well conducted large trials. Furthermore, it remains unknown as to whether the detection of resistance mechanisms prior to confirmatory RECIST progression or the rise in mutation allelic fraction should prompt a change in systemic treatment. The APPLE trial (NCT02856893) will evaluate sequencing of gefitinib and osimertinib and provides an opportunity to assess the impact of switching treatment based on the detection of T790M in ctDNA.