2024 World Conference on Lung Cancer (WCLC) - Posters-P1.06A.05 Assessing Pathologic Response to Neoadjuvant Chemoimmunotherapy for NSCLC: Aligning Current Protocols with Artificial Intelligence

P1.06A.05 Assessing Pathologic Response to Neoadjuvant Chemoimmunotherapy for NSCLC: Aligning Current Protocols with Artificial Intelligence

Back to course

Pdf Summary

This study, conducted by researchers at Hospital Universitario 12 de Octubre in Madrid, Spain, evaluates the pathologic response to neoadjuvant chemoimmunotherapy in patients with early-stage, resected non-small cell lung carcinoma (NSCLC) using manual and artificial intelligence (AI) assessments. The study included 24 patients.<br /><br />Tumor characteristics such as tumor bed size and percentage of tumor cells were assessed both manually and through AI. Notably, there was a major pathologic response (MPR), meaning less than 10% tumor cells, in a majority of the cases (79.2% manually and 70.8% via AI). <br /><br />There was a moderate agreement between the manual and AI assessments of MPR, with a kappa statistic of 0.78, indicating substantial agreement. However, in two cases, the AI predicted a higher percentage of tumor cells compared to manual evaluations, highlighting discrepancies that need further exploration.<br /><br />The study also found a significant association between manual MPR and the AI-predicted estimation of the total number of tumor cells and tumor cell density, suggesting that AI might offer a robust quantification method. Yet, a non-significant trend was observed toward better outcomes in patients showing MPR, whether assessed manually or via AI.<br /><br />The researchers emphasize the potential of AI to enhance the accuracy of pathologic assessments and support clinical decision-making in NSCLC treatment. Despite its promise, the study underscores the need for long-term follow-up and further investigation into the benefits of AI in clinical settings to confirm its utility and effectiveness. Overall, this research aligns current pathological protocols with AI tools, aiming to improve the precision of therapeutic evaluations in lung cancer.

Asset Subtitle

Fernando Lopez-Rios

Meta Tag

Speaker Fernando Lopez-Rios

Topic Pathology & Biomarkers

Keywords

neoadjuvant chemoimmunotherapy

non-small cell lung carcinoma

pathologic response

artificial intelligence

manual assessment

major pathologic response

tumor cell density

AI quantification

clinical decision-making

lung cancer treatment