Search

e21557 Background: Just like classical EGFR mutations, EGFR uncommon mutations non-small cell lung cancer (NSCLC) will still have acquired resistant problem to icotinib. The mechanism of icotinib resistance in such uncommon EGFR mutant patients has also risen to be a difficult question in lung cancer research. In order to explore the resistance mechanism in EGFR uncommon mutant NSCLC patients treated by icotinib, it is necessary to first identify the resistant gene profiles and clinic-pathologic characteristics of those patients. As far as we know that there is no large cohort of EGFR uncommon mutant NSCLC study in evaluating the efficacy and resistant genomic profiling of icotinib. Methods: We screened 3279 patients with NSCLC for EGFR uncommon mutations. Among them, 106 patients received icotinib treatment, and a total of 69 patients with stage IIIb-IV EGFR uncommon mutations NSCLC were undergoing tumor biopsies or blood withdrawing by the time of primary or acquiring to icotinib, in including formalin-fixed paraffin-embedded (FFPE) samples, serum samples and serous effusions. We used targeted next-generation sequencing to detect genes status of patients. Results: Among 69 patients treated with icotinib, 69.57% (48/69) developed acquired resistance, and 30.43% (21/69) had primary resistance. Using the specimens at the baseline, there were 39(81.25%) patients with EGFR T790M (including 7 patients with EGFR T790M, 32 patients with EGFR T790M plus EGFR amplification), 3(36.25%) patients with EGFR amplification, 1(2.08%) patient with BRAF mutation, 1(2.08%) patient with PIK3CA mutation, 1(2.08%) patient with CTNNB1 mutation, 1(2.08%) patient with ALK fusion, 1(2.08%) patient with ROS1 fusion, and 1(2.08%) patient with unknown status. Conclusions: EGFR T790M, EGFR amplification, BRAF mutation, PI3K-AKT-mTOR signaling pathway (PIK3CA mutations), CTNNB1 mutation, ALK fusion or ROS1 fusion might contribute to molecular mechanisms of acquired resistance to icotinib in EGFR uncommon mutations NSCLC. Our study uncovered EGFR uncommon mutational profiles of NSCLC patients with icotinib resistance with potential therapeutic implications.

e14572 Background: Mammography screening for breast cancer results in large number of impalpable lesions without clear determination of the malignancy. Analysis of breast cancer related gene mutations in blood circulating tumor DNA (ctDNA) may provide clarification. This analysis aims to provide insights into the feasibility of the approach. Methods: The clinical trial was conducted at top tier teaching hospitals in China to recruit patients with breast diseases for surgery. Eligible patients were consented and the breast lesions were pathologically diagnosed. Peripheral blood was collected prior to surgical resection. For breast cancer patients, samples of resected tissue were also collected. The samples were analyzed using our proprietary NGS technique called systematic error correction sequencing (Sec-Seq) (detailed in Abstract ##e23057, ASCO 2018). Results: In total, 69 patients with breast lesions (57 malignant and 12 benign) were included in this analysis. Tumor gDNA and plasma ctDNA were analyzed by deep NGS sequencing using a panel of 62 breast cancer-associated genes. The average sequencing depth is 35000. After deduplication, the average number of unique reads is 1500. Detection limit for mutant allele frequency was set at 0.2% for ctDNA and 1% for tumor tissue. For ctDNA mutation detection, 2 out of 12 patients with benign diseases were found with mutations while 10 out of 55 breast cancer patients had no mutations, resulting in an overall sensitivity of 82% and specificity of 83%. By cancer stage, the two Stage 0 (carcinoma in situ) patients had no mutation, and the range of mutations detected is between 53% to 75% from Stage I to III. The tumor tissue samples have higher rate of mutations (only 2 cancer patients, 1 Stage 0 and 1 Stage 2, had no mutations). 15% patients have at least one common mutation detected in both the tumor tissue and ctDNA, and 27% patients have mutations in the same genes in the two matching samples. The concordance increases as the clinical stage advances. The most commonly mutated genes are previously reported breast cancer drivers of PIK3CA (79% of tumor and 18% of ctDNA samples), TP53 (56% and 39%), and BRCA1 (6% and 15%). Conclusions: In this hypothesis generating analysis, we showed the feasibility of plasma ctDNA sequencing for gene mutation detection in early stage breast cancer and differentiation from the benign breast diseases. Although with limited number of samples, the data encourage further improvement of the gene panel and the validation of ctDNA assay as a non-invasive approach to the cancer screening. Clinical trial information: ChiCTR1800017345.