QUANTITATIVE DETECTION OF CLINICALLY ACTIONABLE MUTATIONS IN LUNG CANCER

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Existing chemotherapy improves survival by only 4%-15% in most patients with non-small cell lung cancer (NSCLC). There is a tremendous need to improve the outcomes in these patients given the limited success of current therapies. To date, comprehensive genomic studies have shown that lung tumors are remarkably diverse between patients. On the other hand, the recent use of agents that target specific molecules in lung cancer and the realization that the genetic makeup of a lung tumor can predict how patients will respond to each of these drugs has been considered one of the most dramatic advances in lung cancer treatment over the past three decades. The hypothesis of this proposal is that, using 'next-generation' sequencing technology, a focused and sensitive analysis of mutations in genes which are already known targets of currently available therapeutic agents will identify NSCLC patients who would be ideal candidates for novel clinical trials. To test this hypothesis, we will use a capture-based, 'next-generation' sequencing assay to perform 'deep' sequencing of the exons and surrounding sequence of approximately 48 genes that are the known targets of currently available agents. The assay will be performed on DNA from routinely fixed surgical pathology specimens to demonstrate practical clinical utility. Based on a carefully selected, highly annotated cohort of 400, early stage NSCLC patients, tumor DNA sequence data generated from this study will allow us to address several practical questions important for translating this approach into a clinical diagnostic tool. First, based on a vastly increased levelof sensitivity (1,500-fold sequencing coverage), we will determine the overall frequency of genomic variants (single nucleotide substitutions, small insertions/deletions, and amplifications) in early stage NSCLC. Second, we will quantitatively assess the variability of mutated allele frequency ('mutation burden') between patients and seek correlations with other clinical or pathological parameters such as clinical relapse or tumor histological features. Finally, we will examine potential correlations between specific genomic variants and traditional clinical and pathological parameters used for classification. Using both a novel annotated bio specimen resource and an innovative technical approach, this work will test the clinical relevance of performing prospective, quantitative mutation profiling on NSCLC patients for 'personalized' treatment selection.