WCLC 2025 - Posters & ePosters-EP.07.03 Lung Cancer: Artificial Intelligence, Biometrics and Modeling of Alive Supersystems for Best Management

Pdf Summary

This study by Oleg Kshivets analyzes survival outcomes and prognostic factors in 786 patients with non-small cell lung cancer (NSCLC) who underwent radical surgery (lobectomies, pneumonectomies, and combined resections) at Bagrationovsk Hospital, Russia, from 1985 to 2025. The patient cohort (mean age ~58) had tumors averaging 4.1 cm, with extensive pathological, clinical, and biochemical data collected. Using complex system analysis methods—including neural networks, Cox regression, Monte Carlo simulation, and bootstrap techniques—the research aimed to identify key factors influencing 5-year survival (5YS) and longer-term outcomes.<br /><br />Key findings include a high cumulative 5YS of 73.4%, 10YS of 65.2%, and 20YS of 42.5% after radical surgery. 5YS was superior after bi-/lobectomies (78.2%) versus pneumonectomies (63.5%). Adjuvant chemoimmunoradiotherapy significantly improved survival for patients with nodal involvement (N1-2), raising 5YS from 34.8% to 65.6%. Cox modeling showed that survival correlated with tumor stage transitions (early vs. invasive cancer and N0 vs. N1-2), cell ratios (e.g., cancer cells to various blood cell types), hemostasis markers, biochemical factors (bilirubin), patient age, and surgery type. Neural networks achieved 100% accuracy in predicting 5YS, highlighting phase transitions and cell ratio factors as top prognostic variables.<br /><br />The study emphasizes the value of combining artificial intelligence, biometrics, synergetics, and modeling to capture the complex "alive supersystem" dynamics of lung cancer progression. Recommendations for optimal management include early detection through screening, availability of skilled thoracic surgeons to perform complex radical resections with thorough lymphadenectomy, use of precise predictive models for prognosis, and application of adjuvant chemoimmunoradiotherapy in high-risk patients.<br /><br />In summary, this comprehensive integrative approach provides an advanced predictive framework for improving long-term outcomes in NSCLC through individualized treatment planning and highlights critical biological and clinical determinants of survival post-surgery.

Asset Subtitle

Oleg Kshivets

Meta Tag

Speaker Oleg Kshivets

Topic Early-Stage Non-small Cell Lung Cancer

Keywords

non-small cell lung cancer

NSCLC

radical surgery

lobectomy

pneumonectomy

5-year survival

chemoimmunoradiotherapy

prognostic factors

neural networks

lung cancer prognosis