Development of adverse event (AE) derived biomarkers for predicting clinical outcomes in lung cancer

PROJECT SUMMARY/ABSTRACT
Adverse event (AE) is a critical component of clinical trial for safety data to protect patients from unnecessary
risk and to develop safety profile of the drug for benefit-risk assessment. Meanwhile, various studies showed
association of AE with clinical outcomes in lung cancer, such as hypertension and immune-related adverse event
with improved survival outcomes and high neutrophil-to-lymphocyte ratio correlated with worse survival.
However, due to the complexity, AE data has been underutilized with sub-optimal AE reporting (e.g., the worst
grade method) to inefficiently assess a patient safety and efficacy profile. While existing approaches have tried
to improve characterization of AE profile, limitations have prevented their board application due to either limiting
to single AE analysis, lack of full utilization of AE parameters, difficulty of interpretation, or uncertainty of clinical
relevance. This highly translational study aims to change the paradigm of AE data usage from descriptive
summary into modern informative AE biomarkers to fulfill precision medicine. We hypothesize that AE-derived
biomarkers have clinically predictive values in lung cancer because of its ability to extensively assess grade,
treatment relatedness, occurrence, frequency, and duration. We also believe that comprehensive AE data
analysis by modern strategy from overall AEs, toxicity category AEs, to individual AEs, will facilitate development
of integrative AE biomarker signatures in predicting clinical outcomes.
Two innovative research Aims are proposed to achieve the overall study objective. Aim 1 is to evaluate AE
biomarkers for predictive values of clinical outcomes in lung cancer. Given the promising AE preliminary data,
we will first test our new framework for identifying AE’s association with treatment response and survival
outcomes in a unique set of new study cohorts in non-small cell lung cancer (NSCLC) treated with
immunotherapy (Aim 1.1). We will further evaluate AE biomarkers for clinical association in NSCLC treated with
chemotherapy or precision therapy (Aim 1.2). Aim 2 is to build a modern AE analysis tool to facilitate identification
of clinical-associated AE biomarkers and generation of automated informative AE report. We will develop new
clinically relevant AE analysis functions to enhance data analysis and AE report (Aim 2.1). We will then integrate
the developed AE analysis functions into a user-friendly tool for streamline data analysis and automated report
generation (Aim 2.2).
If successful, the AE-derived biomarkers could help clinicians early determine who should stay in the trial for
treatment benefit (e.g., treatment-related low-grade AEs) and who should be discontinued due to ineffective
treatment (e.g., higher-grade AEs). The informative AE analysis tool could be leveraged by clinical and research
scientists to discover novel AE biomarkers and facilitate vast meaningful AE research hypotheses.

Grant Number: 5R21CA286417-02
NIH Institute/Center: NIH
Principal Investigator: Dung-Tsa Chen