Biomarker Development Laboratory

Project summary (BDL) Recently, the use of culture independent techniques to characterize the microbiome has led to identification of microbial signatures in the systemic circulation associated with lung cancer diagnosis and prognosis. Consultants in this proposal have described how shotgun metagenomics, which identifies microbial DNA, can be used to identify microbial signatures in plasma predictive of different cancers, including lung cancer. Our preliminary data from our NYU EDRN archives show that metagenomic signatures can be predictive of early- stage non-small cell lung cancer (NSCLC) compared to non-NSCLC. In this cohort, we have also identified microbial and host transcriptomic signatures present in the lower airways associated with prognosis (recurrence). These data support that microbial and host genomic signatures could be used to develop novel biomarkers in early stages of this disease. Omic approaches can explore these signatures in an unbiased fashion, allowing for identification of best performing features for predicting, in this case, NSCLC diagnosis and prognosis. In addition, evaluation of the metabolomic environment can further uncover other potential biomarkers as it relates to the metabolism of microbes and host. Therefore, the goal of this proposal is to utilize our NSCLC archives to evaluate microbial metagenomic and host transcriptomic features paired with metabolomic approaches using blood samples to develop novel biomarker signatures that predict early-stage NSCLC disease (Aim 1). We will then evaluate the metagenome, metabolome and host transcriptomic data from lower airway samples from patients with early-stage NSCLC to identify features predictive of lung cancer recurrence (Aim 2). Finally, using an integrated multi-omic approach, we will optimize the selection of best performing features in Aim 3. The cohort selected for these investigations will be divided in Discovery and Validation. Successful biomarkers will then undergo external validation. The data generated here will serve as the foundation of an agnostic approach to identify highly predictive biomarkers that will feed the development and validation for targeted approaches under the Biomarkers Reference Laboratory (BRL).