Effect of COPD diagnosis on TP53-mutated colony size distribution in airway epithelium of subjects with or without lung cancer

Background: We previously identified a biomarker for the early detection of lung cancer (LC) based on the prevalence of TP53-mutated colonies in airway epithelium, and determined that this biomarker performs well in subjects with or without COPD. To better understand the biological basis for this biomarker, we compared TP53-mutated colony size distributions in subjects with or without COPD and/or LC. Methods: Airway epithelial cell specimens were collected from subjects undergoing bronchoscopy under an IRB-approved research protocol. DNA was extracted from each specimen and ultra-sensitive NGS was used to measure TP53 mutations, with variant allele fraction (VAF) as low as 0.01%. TP53-mutated colony size was estimated assuming cells were heterozygous for the mutation. Specimens from 59 subjects were assessed in four cohorts: lung cancer with COPD [LC-CO) (N=14), non-lung cancer with COPD (NLC-CO) (N=7), lung cancer without COPD (LC-NCO) (N=16) and non-lung cancer without COPD (NLC-NCO) (N=22). Colony size distributions were assessed by piecemeal slope analysis and Kaplan-Meier analysis. Results: A total of 192 unique TP53 mutations were identified among the 59 subjects, distributed among the cohorts as follows: LC-CO, 72; NLC-C0, 19; LC-NCO, 87; and NLC-NCO, 14. The mutation VAF values ranged from 0.01% (0.0001) to 6% (0.06), corresponding to a TP53-mutated colony size range of 200 to 30,000 cells. For each cohort, colony prevalence was inversely proportional to colony size, and there was an inflection point (IP1) at 400-600 cells/colony. Prior to IP1 there was little change in colony number relative to size; after IP1 the decline had a steep slope. The IP1 value was smaller for CO subjects (400 cells) than for NCO subjects (600 cells). There was a second inflection point (IP2) at about 1,800 cells, after which there were few surviving colonies for NLC subjects, but many surviving colonies for LC subjects. According to Kaplan-Meier analysis, among LC-CO subjects the higher prevalence of colonies >1,800 cells in LC was key to accuracy of the TP53 biomarker. In contrast for LC-NCO subjects there was a higher prevalence of colonies across the range of sizes. Conclusions: Interactions between TP53 mutated colonies and the microenvironment are different in CO vs non-CO subjects. This difference may be partly due to the increased adaptive immune inflammation characteristic of CO lung tissue.