Ariana 6-Gene Signature Boosts NSCLC Survival Prediction
Overview
Non-Small Cell Lung Cancer (NSCLC) is the most common type of lung cancer accounting for over 85% of all cases. Yet, there is a high heterogeneity of survival for patients treated with Immunotherapy (IO) hence the importance of identifying biomarkers of survival to adapt the sequence of treatment lines. Taking advantage of transcriptomic data generated in a cohort of 81 NSCLC patients, KEM® identifies molecular biomarkers linked with NSCLC disease progression to predict survival in patients treated with immunotherapy.
Impact
• The xAI model developed by Ariana Pharma revealed a biologically relevant 6-gene survival signature, achieving high specificity and sensitivity (specificity greater> 0.90 and sensitivity of 0.84)
• The signature identified could lead to the development of a Companion Diagnostic aimed at improving therapeutic decisions for NSCLC patients.
Objectives
• Establish a robust molecular signature to accurately predict survival and enhance patient-specific treatment strategies and treatment sequences.
• Refine the previously developed CLB’s Hot Score by reducing the number of markers, maintaining the signature’s efficacy while simplifying its clinical application in routine settings
Method
KEM® explored all associations within the clinical data comprising 2559 gene expressions from 81 NSCLC patients, employing systematic extraction, identification, and interpretation of formal concepts to determine the most pertinent relations affecting survival.
Results
• KEM® significantly reduced the number of markers in the survival signature down to 6 from the 27 markers present in the Hot Score.
• The balanced accuracy (sensitivity + specificity) was increased by 30 points.
