Abstract
522
Introduction: Advanced non-small cell lung cancer (NSCLC) can be treated with epidermal growth factor tyrosine kinase inhibitors (EGFR TKI). Treatment response of patients treated with EGFR TKI is greater when an activating EGFR mutation is found in tumor DNA (collected through histological biopsy) when compared to wild type (WT) EGFR1-2. Activating EGFR mutations can be subdivided in common (e.g. exon 19 deletions, EGFR com+ ) and uncommon mutations (e.g. exon 18 G719X point mutation, EGFR uncom+)3. Most EGFR TKI have shown efficacy in EGFR com+. In contrast to first generation EGFR TKI, afatinib, a second-generation, irreversible EGFR TKI, has also shown efficacy in EGFR uncom+4-5. Identifying the patients who will benefit from EGFR TKI therapy is a challenge: 1. representative tumor biopsies might be impossible to obtain due to tumor localization. 2. there may be intra- and interlesional differences in EGFR expression in tumor tissue. These challenges may be identified by performing positron emission tomography (PET) scans using radiolabeled afatinib ([18F]afatinib). This study aimed to 1: assess the value of PET and [18F]afatinib in identifying patients with EGFR mutations by establishing the optimal pharmacokinetic model for quantification of uptake and 2: assess differences in tracer uptake between WT, EGFR com+, and EGFR uncom+.
Methods: 8 advanced stage NSCLC patients were included in this study: 3 WT, 3 EGFR com+, 2 EGFR uncom+. EGFR mutational status was determined by immunohistochemical analysis of tumor DNA. In order to assess tumor perfusion and to determine whether [18F]afatinib uptake is perfusion-dependent 20-minute, dynamic 370 ± 37 MBq [15O]H2O PET scans were performed in all patients prior to [18F]afatinib scanning. Immediately after injecting 342 ± 24 MBq[18F]afatinib, 60-to-90-minute dynamic PET scans were performed. Single-tissue (1T2K), 2-tissue-irreversible (2T3K) and -reversible (2T4K) plasma and image-derived input function (IDIF) input models were used to analyze [18F]afatinib data.
Results: Out of 8 patients, 19 tumors were analyzed (9 WT, 5 EGFR com+, 5 EGFR uncom+). Analysis of [15O]H2O data showed adequate perfusion and showed [18F]afatinib uptake to be perfusion independent. Tumor kinetics were best described using the irreversible, 2T3K model which produced the best fits to the data for both metabolite-corrected plasma input and IDIF (Akaike 37% and 42% respectively). This was consistent throughout all 3 groups. Mean Ki values of the 2T3K model using plasma and IDIF input are shown in table 1. In both models difference in Ki was significant between WT and EGFR com+ (plasma: p= 0.004, IDIF: p=0.002), between WT and EGFR uncom+ (plasma: p=0.003, IDIF: p=0.0009) and between WT and all EGFR mutations combined (plasma: p=0.0012, IDIF: p=0.005).
Conclusions: The preferred pharmacokinetic model for quantitative assessment of [18F]afatinib tumor uptake was the irreversible, 2T3K model for both plasma and IDIF input. Using this model, [18F]afatinib uptake in tumors has great potential to differentiate between patient groups harboring EGFR WT, EGFR com+, and EGFR uncom+ in order to select the right NSCLC patient for treatment with afatinib.