Lung cancer burden peaks at 2.5 million new diagnoses and 1.8 million deaths per-year worldwide, with tobacco smoking as the main risk factor, together with exposures to radon gas, asbestos and other carcinogens, infections, living in heavily polluted areas, and unhealthy diet contributing to the huge incidence. The high prevalence of non-localized tumors leads to <32% surviving at 5 years. However, when diagnosed at an early stage, the survival ranges from 68% to 92%. Therefore, the strengthening of primary and secondary prevention strategy could definitely contribute to mitigation of lung cancer burden. Indeed, smoking cessation campaigns, asbestos ban, air pollutions reduction, and healthy lifestyles promotion are welcomed as effective primary prevention strategy, coupled with LD-CT screening program to promote early detection. Several biomarkers circulating in body fluids have been proposed as promising minimally-invasive tool to complement lung cancer screening program and improve nodules risk assessment; however, lack of validation in large multi-centric cohort, and a lower sensitivity in early stage tumors limit their transfer to the clinic. The most studied and attractive circulating biomarkers are microRNAs (miRNAs), with high stability in harsh storage conditions and resistance to circulating RNAses, and strongly associated with lung cancer diagnosis with altered levels in the blood and other body fluids. In this dissertation, several works related to the role of miRNAs as biomarkers for lung cancer prevention are presented, based on large study cohorts and robust analytical approaches. These studies are part of different translational projects of the Cancer Biomarkers Unit, Fondazione IRCCS Casa Sollievo della Sofferenza (San Giovanni Rotondo, Italy). The first work (chapter 4) highlights the connection between miRNAs-based molecular mechanisms and nutritional pattern, with implication on lung cancer primary prevention, and emphasis on the anti-inflammatory and antioxidant components of the Mediterranean Diet. Indeed, this work proposed a new cross-discussion on epidemiology, nutrition and omics sciences, with a deep description about miRNAs modulated by nutrients and food-derived miRNAs, as possible biomarkers for personalized nutrition for health optimization and disease prevention. The second work (chapter 5) is focused on the identification of useful clinical and molecular tools for the screening of population at high risk of developing lung cancer and other pre-malignant pulmonary diseases. Indeed, a large cohort of highly exposed to air pollution and asbestos has been studied, and a non-invasive screening protocol based on thoracic ultrasound has been effectively identified. This study will be completed with the profiling of plasma samples of enrolled individuals, to develop a miRNAs-based risk model from liquid biopsy, as a minimally-invasive tool to complement the ultrasound screening. The following work (chapter 6) presents a signature for the early detection of lung cancer, based on circulating miRNAs. This project is established in lung cancer screening framework, with the analysis of large multi-centric cohorts and application of machine learning techniques. The identified signature is very promising, given the high accuracy of distinguishing lung cancer from controls, independently from other well-known risk factors; moreover, the interesting diagnostic ability in stage I tumors and benign nodules makes this signature particularly attractive in lung cancer screening context. The last work presented (chapter 7) expands the application of the circulating miRNAs signature proposed in chapter 6. Indeed, the signature was successfully applied also in a clinical setting to ameliorate the malignancy risk-assessment of CT-detected indeterminate nodules, which usually present an unclear clinical management. This project is currently ongoing, and the promising results obtained should be further validated.
Il carico mondiale legato al tumore al polone ha raggiunto 2.5 milioni di nuove diagnosi e 1.8 milioni di morti per anno, con il fumo come principale fattore di rischio, insieme a radon, asbesto e altri carcinogeni, infezioni pregresse, vivere in aree molto inquinate e diete non corrette come altri fattori di rischio che contribuiscono all’enorme incidenza. L’alta prevalenza di tumori non localizzati conduce a una sopravvivenza <32% a 5 anni. Tuttavia, quando diagnosticato ad uno stadio precoce, la sopravvivenza varia tra il 68% e il 92%. Di conseguenza, il rafforzamento di strategie di prevenzione primaria, accoppiato con programmi di screening per promuovere la diagnosi precoce, potrebbero contribuire a ridurre il carico mondiale. Diversi biomarcatori circolanti in fluidi corporei sono stati proposti come strumenti poco invasivi per complementare i programmi di screening; tuttavia, l’assenza di validazione in grandi coorti multicentriche, e una bassa sensibilità nei tumori precoci limita la loro applicazione in clinica. I biomarcatori circolanti più interessanti e studiati sono i microRNA (miRNA), con alta stabilità e resistenza alle RNAsi circolanti, e fortemente associati alla diagnosi di tumore al polmone, con livelli alterati nel sangue e in altri fluidi corporei. In questa tesi sono presentati diversi lavori sul ruolo dei miRNA come biomarcatori per la prevenzione del tumore al polmone, basati su coorti ampie e approcci analitici robusti. Questi studi fanno parte di diversi progetti traslazionali della Cancer Biomarkers Unit, Fondazione IRCCS Casa Sollievo della Sofferenza (San Giovanni Rotondo, Italy). Il primo lavoro (capitolo 4) evidenzia la connessione tra meccanismi molecolari associati ai miRNA e modelli nutrizionali, con implicazioni sulla prevenzione primaria del tumore al polmone, ed enfasi sui componenti anti infiammatori e antiossidanti della Dieta Mediterranea. Questo lavoro propone una nuova discussione incrociata sull’epidemiologia, la nutrizione e le scienze omiche, con una descrizione dei miRNA modulati dai nutrienti e miRNA derivati dal cibo, come possibili biomarcatori per una nutrizione personalizzata per ottimizzazione della salute. Il secondo lavoro (capitolo 5) si concentra sull’identificazione di utili strumenti clinici e molecolari per lo screening di popolazioni ad alto rischio di sviluppare tumore al polmone e altre patologie polmonari pre-neoplastiche. E’ stata studiata un’ampia coorte di altamente esposti a inquinamento dell’aria e asbesto, ed è stato individuato un protocollo di screening basato sull’ecografia toracica. Questo studio sarà completato con l’analisi dei campioni di plasma degli arruolati, per sviluppare un modello di rischio basato su miRNA come strumento per complementare lo screening ecografico. Il lavoro seguente (capitolo 6) presenta una firma molecolare per la diagnosi precoce del tumore al polmone, basata su miRNA circolanti. Il progetto è sviluppato nel contesto dello screening per tumore al polmone, con analisi di ampie coorti multi-centriche e applicazione di tecniche di machine learning. La firma molecolare identificata è molto promettente, data la sua alta accuratezza nel distinguere i tumori al polmone dai controlli, in modo indipendente da altri fattori di rischio; l’interessante capacità diagnostica nei tumori di stadio I e nei noduli benigni rende questa firma molecolare particolarmente interessante nel contesto dello screening per tumore al polmone. L’ultimo lavoro presentato (capitolo 7) espande l’applicazione della firma molecolare proposta nel capitolo 6. Infatti, la firma molecolare è stata applicata con successo anche in un contesto clinico per migliorare la valutazione del rischio di malignità dei noduli indeterminati identificati alla TAC, che solitamente hanno un management clinico non chiaro. Questo progetto è attualmente attivo, e i promettenti risultati ottenuti dovranno essere ulteriormente validati.
A machine learning approach to discover new biomarkers for lung cancer early detection and risk assessment
DAMA, ELISA
2025
Abstract
Lung cancer burden peaks at 2.5 million new diagnoses and 1.8 million deaths per-year worldwide, with tobacco smoking as the main risk factor, together with exposures to radon gas, asbestos and other carcinogens, infections, living in heavily polluted areas, and unhealthy diet contributing to the huge incidence. The high prevalence of non-localized tumors leads to <32% surviving at 5 years. However, when diagnosed at an early stage, the survival ranges from 68% to 92%. Therefore, the strengthening of primary and secondary prevention strategy could definitely contribute to mitigation of lung cancer burden. Indeed, smoking cessation campaigns, asbestos ban, air pollutions reduction, and healthy lifestyles promotion are welcomed as effective primary prevention strategy, coupled with LD-CT screening program to promote early detection. Several biomarkers circulating in body fluids have been proposed as promising minimally-invasive tool to complement lung cancer screening program and improve nodules risk assessment; however, lack of validation in large multi-centric cohort, and a lower sensitivity in early stage tumors limit their transfer to the clinic. The most studied and attractive circulating biomarkers are microRNAs (miRNAs), with high stability in harsh storage conditions and resistance to circulating RNAses, and strongly associated with lung cancer diagnosis with altered levels in the blood and other body fluids. In this dissertation, several works related to the role of miRNAs as biomarkers for lung cancer prevention are presented, based on large study cohorts and robust analytical approaches. These studies are part of different translational projects of the Cancer Biomarkers Unit, Fondazione IRCCS Casa Sollievo della Sofferenza (San Giovanni Rotondo, Italy). The first work (chapter 4) highlights the connection between miRNAs-based molecular mechanisms and nutritional pattern, with implication on lung cancer primary prevention, and emphasis on the anti-inflammatory and antioxidant components of the Mediterranean Diet. Indeed, this work proposed a new cross-discussion on epidemiology, nutrition and omics sciences, with a deep description about miRNAs modulated by nutrients and food-derived miRNAs, as possible biomarkers for personalized nutrition for health optimization and disease prevention. The second work (chapter 5) is focused on the identification of useful clinical and molecular tools for the screening of population at high risk of developing lung cancer and other pre-malignant pulmonary diseases. Indeed, a large cohort of highly exposed to air pollution and asbestos has been studied, and a non-invasive screening protocol based on thoracic ultrasound has been effectively identified. This study will be completed with the profiling of plasma samples of enrolled individuals, to develop a miRNAs-based risk model from liquid biopsy, as a minimally-invasive tool to complement the ultrasound screening. The following work (chapter 6) presents a signature for the early detection of lung cancer, based on circulating miRNAs. This project is established in lung cancer screening framework, with the analysis of large multi-centric cohorts and application of machine learning techniques. The identified signature is very promising, given the high accuracy of distinguishing lung cancer from controls, independently from other well-known risk factors; moreover, the interesting diagnostic ability in stage I tumors and benign nodules makes this signature particularly attractive in lung cancer screening context. The last work presented (chapter 7) expands the application of the circulating miRNAs signature proposed in chapter 6. Indeed, the signature was successfully applied also in a clinical setting to ameliorate the malignancy risk-assessment of CT-detected indeterminate nodules, which usually present an unclear clinical management. This project is currently ongoing, and the promising results obtained should be further validated.File | Dimensione | Formato | |
---|---|---|---|
phd_unimib_070328.pdf
embargo fino al 27/01/2028
Dimensione
2.34 MB
Formato
Adobe PDF
|
2.34 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/190601
URN:NBN:IT:UNIMIB-190601