Identification of non-coding RNAs-based molecular signature for early detection of Lung Cancer

• Grant number: 23/03918-0
• Support Opportunities: Scholarships in Brazil - Doctorate (Direct)
• Start date: October 01, 2023
• End date: July 31, 2026
• Field of knowledge: Biological Sciences - Genetics - Human and Medical Genetics
• Principal Investigator: Letícia Ferro Leal
• Grantee: Giovanna Maria Stanfoca Casagrande
• Host Institution: Hospital do Câncer de Barretos. Fundação Pio XII (FP). Barretos , SP, Brazil

Abstract

Lung cancer is the most incident tumor worldwide and the most lethal. Screening programs have been implemented worldwide in an attempt to reduce lung cancer mortality rates. Currently, strategies based on the analysis of body fluids (liquid biopsy) are being increasingly used, however, the application of this type of analysis in screening programs remains incipient. The detection of free circulating non-coding RNAs (ncRNAs) in plasma, serum, urine and saliva samples shows promise due to the stability of these molecules in different body fluids. Objective: To identify a molecular signature based on miRNA/circRNA for screening and early detection of lung cancer in samples obtained through minimally invasive procedures. Materials and methods: Plasma samples from a group of control individuals (n=60) and a group of cases (n=60) diagnosed with non-small cell lung cancer (NSCLC), both attended by the screening program will be evaluated. lung cancer in a mobile unit at the Hospital de Câncer de Barretos or at the fixed unit at the Hospital de Câncer de Barretos. In addition, a cohort from the Quirón Dexeus Hospital (Barcelona - Spain) will be analyzed, with 86 cancer samples (NSCLC) and 100 healthy individuals and/or with benign nodules, and a validation cohort with 53 cancer samples and 124 samples from individuals with benign nodules, all from the Barretos Cancer Hospital. All plasma samples will be subjected to RNA extraction followed by analysis of the expression profile of ncRNAs (miRNA and circRNA), using the NanoString platform. The ncRNA counts will be normalized and filtered by R Studio software, using p<0.05 and fold-change<1.3 as criteria. The development of the molecular signature of ncRNAs will be performed using machine learning (ML) employing Recursive Feature Elimination and the accuracy of the signatures will be determined by the Area Under the ROC Curve (> 0.70). Partial results: As a result between the comparison of plasma samples (Brazil) from individuals with cancer and controls, seven miRNAs were differentially expressed. Among these seven miRNAs, three miRNAs showed high accuracy. As a result of comparing a subset of sputum samples from individuals with cancer and controls, three miRNAs were differentially expressed. For a subset of plasma samples from the independent cohorts (Spain), six circRNAs were differentially expressed. Next steps: i) Continue the analysis of differential expression of miRNAs in sputum samples; ii) Start analysis of differential expression of circRNA in sputum; iii) Develop signature ncRNAs (circRNA + miRNA) for early diagnosis of lung cancer using plasma samples and/or sputum samples; iv) Validate the signature obtained in samples from independent cohorts. (AU)