UNCOVERING THE GENETIC LANDSCAPE OF LUNG ADENOCARCINOMA: A CLINICAL STUDY TO IDENTIFY BIOMARKERS FOR TUMOR SPREAD THROUGH AIR SPACES STATUS

Cancer is the second leading cause of mortality in the modern world with approximately 10 million deaths every year. From those, lung cancer is still the most frequently diagnosed cancer worldwide and remains the leading cause of cancer-related mortality globally, with five-year survival rates lingering below 20% for all stages combined. In particular, non-small cell lung cancer (NSCLC) is the most common type of lung cancer, with lung adenocarcinoma (LUAD) being the most prevalent. Despite advances in targeted therapies and immunotherapy, the prognosis of patients, especially those with advanced-stage NSCLC, is still dismal. Recently, the clinical and research community have widely discussed the importance of tumor spread through air spaces (STAS) concept as a pattern of invasion in pulmonary cancer. The STAS concept has been widely debatable among the clinician's community, specifically the pattern of invasion is described as unique to the lung, which increased criticism. On one hand, STAS represents air space invasion by the tumor cells. On the other hand, STAS was considered as an artifact induced by cutting through a tumor with a knife. STAS lesions are identified when tumor cells, which vary from single cells to micropapillary clusters, morphologically appear to be situated within air spaces and detached from alveolar walls. It is usually found in the first alveolar layer, close to the main tumor, which is confirmed by diagnostic H&E-stained slides and added in the pathologist's report. The presence of STAS correlates with multiple pathological and clinical features of aggressiveness in LUAD, including lymphatic and vascular invasion, high-grade morphologic histological patterns as well as intrathoracic recurrence. The type of surgical removal of STAS-positive tumors has impact on cancer recurrence, namely STAS was associated with a shorter 5-year recurrence free survival (RFS) in limited resection. It is hypothesized that STAS positivity may be a late event during LUAD evolution.Therefore, a more detailed analysis of the relationship between histological growth pattern, STAS and patient outcome, including the site of relapse is urgently needed. Interestingly, the early detection of LUAD by biomarker identification has been extensively studied in recent years. Several groups have identified the expression of microRNAs in the plasma of LUAD patients as potential noninvasive biomarkers for early detection of LUAD. Until now, there has been no published research on early biomarkers for STAS-positive LUAD detection, which is crucial for exploring miRNA signatures and their validation with tumor progression.

By harnessing the power of computer-assisted approaches fueled by artificial intelligence (AI), researchers can automate and expedite the analysis of complex datasets using predictive AI models, as we previously discussed. As AI continues to evolve, its integration into the field of cancer is creating opportunities to develop diagnostic and treatment protocols that can be finely tuned to each patient's condition, leading to more effective outcomes and minimized adverse effects. Recently, AI-based model ANORAK enabled the precision mapping growth patterns in LUAD and detected the degree of spatial heterogeneity. This study highlights the importance of AI in the heterogeneous landscape of LUAD to better target individual patients for adjuvant therapies. Until now, there has been a significant gap in current cancer research on STAS-positive LUAD behavior that promotes ineffective diagnostic tools and personalized treatment strategies that could dramatically jeopardize survival rates and quality of life of lung cancer patients. Indeed, early detection and patient-oriented therapies for STAS-positive tumours are not just about extending life expectancy, but also about enhancing the quality of life of patients. By using deep learning models, we intend to evaluate in a relatively objective and practical manner the validation of STAS presence on patient-derived LUAD samples by evaluating the extent of STAS according to how far the tumor cells had spread from the edge of the tumor; determining the number of cell clusters and its potential correlation with the epithelial to mesenchymal transition (EMT) phenomenon; and studying the correlation of these parameters with sequencing data and clinical patient data, including surgical approaches and survival rate.

In addition, the recreation of the three-dimensional (3D) organization is crucial to study the spatial distribution and structural interaction of the STAS positive tumor microenvironment. In the past few years, Tissue Engineering and Regenerative Medicine (TERM) drew heavily on an explosion of new knowledge that broadens the range of potential research strategies. In particular, in vitro 3D platforms for drug development and native pathophysiological mechanisms. The current developed lung cancer in vitro models have shown limited structural integrity, stability over the cell culture period, and limited recreation of the complex 3D tumor microenvironment. Recently, the fabrication of 3D models with patient-derived lung cancer organoids have shown to retain pivotal features of the original tumors, which functionally complement molecular and pathological tumor analysis. However, the current protocols to grow these organoids almost exclusively depend on culture within Matrigel-based systems.

Strikingly, matrigel systems have already shown limitations on defined culture conditions, introduction of animal components, enhance tumorigenicity when used in vivo, and results in heterogeneous organoids (i.e., shape, size, composition).

Our research aims to unravel the complexities of the STAS phenomenon in LUAD, a factor that has emerged as a critical player in tumor aggressiveness and patient prognosis. The identification of reliable biomarkers for STAS-positive LUAD has the potential to revolutionize treatment outcomes. By enabling earlier and more accurate diagnoses, our research could reduce the need for invasive diagnostic procedures, decrease time-to-treatment, and allow for more personalized and effective treatment plans.