P1.01-035 A Next Generation Sequencing and Characteristics Based Model for Predict Clinical Benefits of Advanced NSCLC Patients

Y. Zhang,N. Yang,J. Ye,L. Zhang,X. Mao,H. Zhang,N. Zhang
DOI: https://doi.org/10.1016/j.jtho.2017.09.689
IF: 20.121
2017-01-01
Journal of Thoracic Oncology
Abstract:The development of targeted therapies has revolutionized the treatment of non-small cell lung cancer. Interrogating the status of driver mutations has become routine practice. In this study, we applied next-generation sequencing to investigate the association between molecular signature and clinical benefit. We performed capture-based targeted ultra-deep sequencing on 204 samples obtained from NSCLC patients at a single center, including 93 FFPE, 70 fresh tissue and 41 plasma samples. One hundred and twenty two samples were subjected to a panel consisting of 8 driver genes; 50 samples were subjected to a 56-gene panel. The remaining 32 samples were subject to a 168 gene panel. In 159 TKI-naïve patients, driver mutation was identified in 95.2% of (79/83) patients using the 8-gene panel; among them, 65.1% (54/83) carried EGFR mutations. Larger panels identified mutations in 68.1% of patients; 21% carried mutations other than driver mutations. Treatment-naïve patients were primarily subject to the 8-gene panel; in contrast, patients progressed on chemotherapy were subject to larger panels. Seventy-two percent of patients (80/111) undergone matched targeted therapy (MTT) according to sequencing results had a significantly longer PFS than 29 patients who chose chemotherapies despite the fact of harboring driver (p=4.58x10-4 HR=0.342, 95% CI: 0.158, 0.74). Next, we investigated whether the number of EGFR mutations a patient carries and the presence of concurrence EGFR amplification have an effect on PFS. Our data revealed that both parameters are not associated with PFS. Among 46 patients receiving chemotherapy, patients with KRAS mutations were associated with a shorter PFS, 133 days vs 207 days (p= 0.073, HR=2.06 95% CI; 0.791, 5.36). For TKI-naïve patients, primary tumor tissue was collected from 86 patients and tumor tissue from metastatic lymph nodes was collected from 35 patients. Interestingly, we observed that lymph node samples had a higher maximum mutation allelic fraction (MAF) than primary lung tumor samples in patients with distance metastasis, especially with visceral metastasis (p=0.0986); such trend was not observed in patients without distant metastasis. We also analyzed samples obtained after TKI-treatment. Among 36 TKI-treated patients, patients with visceral metastasis were more likely to harbor TP53 mutations (p=0.04), which were primarily missense mutations not loss of function mutations, primarily seen in tumorigenesis. TP53 missense mutations can potentially promote distant visceral metastasis after the development of resistance to TKIs. Our study highlighted the utility of sequencing-based screening technologies and characteristics in providing treatment guidance.
What problem does this paper attempt to address?