Hybrid Breath Analysis/Computer-Assisted Image Processing System for Early Assessment of Lung Nodule Malignancy
Abstract
The objective of this grant is to develop a novel clinical diagnostic system that integrates data from a single low-dose computer tomography (LDCT) scan and a single breath test to significantly improve the accuracy (>95%) and speed of early lung cancer diagnosis. Lung cancer is the leading cause of cancer death and the second most common cancer among both men and women in the United States. Military personnel and Veterans have a 23%-45% higher prevalence of lung cancer and associated mortality compared to the civilian population due to exposure to carcinogens (depleted uranium used in weapons and armor shielding, beryllium, fuel exhaust, and other battlefield emissions) and a higher incidence of smoking. The 5-year survival rate for lung cancer is very low (17.7%), as only 32% of patients diagnosed with lung cancer are at an early stage (Stage I or II). Patients diagnosed with smaller, early-stage tumors have a significantly higher survival rate (55.2%) than patients with larger tumors. However, early detection of lung cancer relies on imaging techniques that identify small nodules (stage IA), only 6% of which are cancerous (94% false positive rate). Consequently, patients must be followed for 2 years to estimate the growth rate of lung nodules to diagnose lung cancer. This approach requires multiple CT scans over a 2-year period and results in significantly higher diagnostic costs (over $10,000/patient), radiation exposure, and a significant delay in final diagnosis of cancer. Further, patients with non-cancerous, benign nodules (false positive cases) are also required to undergo further interventions, including bronchoscopy or percutaneous biopsy or invasive surgical procedures for the correct diagnosis. Importantly, the prolonged time period (up to 2 years) required to diagnose lung cancer using serial CT scans delays cancer treatment, increases diagnostic costs, and reduces the overall survival rate of military personnel, Veterans, and civilians. The prolonged follow-up period may also reduce patient compliance and impact the deployment of Soldiers with suspected nodules. To overcome these limitations, we are developing a novel clinical diagnostic system that integrates data from a single LDCT scan for computed tomography markers and a single breath test for cancer biomarkers to significantly improve the accuracy and speed of early lung cancer diagnosis for cavity, solid, and ground glass nodules. The accuracy of lung cancer detection using current approaches is approximately around 60-80%, which is not adequate for cancer diagnosis (requires greater than 95% accuracy). Our group has developed and patented a novel, non-invasive breath test (similar to a breathalyzer) for lung cancer that is simple and only requires the patient to exhale into a non-reactive bag. The collected breath sample can be analyzed for cancer biomarkers (volatile organic compounds [VOCs], i.e., carbonyl compounds) that are elevated in the exhaled breath of patients with lung cancer. We have identified 27 VOCs and tomography nodule markers based on a single LDCT scan (shape, appearance, and size) that are indicative of lung cancer. Our approach combines both breath biomarker and imaging data to enable a rapid, and highly accurate diagnosis of early-stage lung cancer (our feasibility study from 47 patients demonstrated greater than 97% accuracy) for all nodule types. The primary contribution of this study is the development of an accurate, non-invasive early-stage lung cancer diagnostic system that would: (1) eliminate the need for repetitive CT/LDCT scans and radiation exposure, (2) eliminate the need for biopsies or other invasive surgical approaches for diagnosis, and (3) improve the speed of cancer diagnosis to within a few days as opposed to up to 2 years. This project is relevant to military Service members, Veterans, and their families because it will (1) significantly reduce the costs for diagnosis of lung cancer (over a
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Nov 19, 2019
- Source ID
- W81XWH1910799
Entities
People
- Ayman El-Baz
Organizations
- United States Army
- University of Louisville