314P Natural language processing-based model to predict radiation pneumonitis in patients with locally advanced non-small cell lung cancer undergoing chemoradiotherapy

Background: Stage III NSCLC represents a heterogeneous disease.Platinum-based chemotherapy remains the corner stone.Accurate staging is essential; however, any substantial discordance between clinical and pathological staging may affect treatment allocation and outcome.This study uses real-world data from the Baden-Wrttemberg Cancer Registry (BWCR) to characterize staging accuracy in operated patients and outcomes of major curative-intent treatments in stage III NSCLC.Methods: We identified 10,700 adults with non-metastatic NSCLC diagnosed 2017-2024 who received treatment in curative intent.Two cohorts were analyzed: 1) Primary surgery with pathological stage III (pIII, n = 1010), 2) Definitive chemoradiotherapy (RCT) durvalumab (Durva) with clinical stage III (cIII, n = 604).Only treatments according to the German S3 guidelines were included.Endpoints were overall survival (OS) and time to progression. Results:Staging accuracy was evaluated in all patients with initial surgery (n=3981).Clinical-pathological staging discordance here was substantial, with agreement only in 42.0%.Upstaging was more common (31.8%) than down-staging (26.2%).In group I, adjuvant platinum therapy (n=537) improved OS versus surgery alone (n=473; HR 0.52, 95%CI 0.4-0.7).Cisplatin (n=340) and carboplatin (n=197) showed comparable OS (HR 1.0).Median time to progression was 452 days for both platinum groups versus 329 days for surgery alone (p = 0.034).In group II, Durva maintenance conferred the highest benefit (RCT+Durva vs. RCT; HR 0.55, 95%CI 0.4-0.7).Here cisplatin was more effective in non-squamous tumors, while in squamous tumors both platinums performed comparable.Median time to progression was longest for cisplatin-RCT + Durva (339 days), followed by cisplatin-RCT (283), carboplatin-RCT + Durva (280) and carboplatin-RCT (218).Conclusions: Real-world evidence from BWCR interestingly reveals considerable staging discrepancy in non-metastatic NSCLC.Adjuvant platinum therapy confers benefit to pIII-patients irrespective of platinum regimen, while in cIII-disease, Durva maintenance and histology-specific platinum selection significantly influence outcomes.

Authors

Institutions

Publication Details

Journal
ESMO Open
Published
2026-03-17
DOI
https://doi.org/10.1016/j.esmoop.2026.106624
Primary Topic
Radiomics and Machine Learning in Medical Imaging
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

314P Natural language processing-based model to predict radiation pneumonitis in patients with locally advanced non-small cell lung cancer undergoing chemoradiotherapy

Nicole Tignor, Sooyun Caroline Tavolacci, Tielman Van Vleck, David Mendoza et al.
ESMO Open
Radiomics and Machine Learning in Medical Imaging
article

314P Natural language processing-based model to predict radiation pneumonitis in patients with locally advanced non-small cell lung cancer undergoing chemoradiotherapy

Nicole Tignor, Sooyun Caroline Tavolacci, Tielman Van Vleck, David Mendoza, Robert Samstein, J. Wisnivesky, P. Wang, C. Tackaberry, D. Simmons, J. Bloom, R. Veluswamy, I. Cotarla
article en

Abstract

Background: Stage III NSCLC represents a heterogeneous disease.Platinum-based chemotherapy remains the corner stone.Accurate staging is essential; however, any substantial discordance between clinical and pathological staging may affect treatment allocation and outcome.This study uses real-world data from the Baden-Wrttemberg Cancer Registry (BWCR) to characterize staging accuracy in operated patients and outcomes of major curative-intent treatments in stage III NSCLC.Methods: We identified 10,700 adults with non-metastatic NSCLC diagnosed 2017-2024 who received treatment in curative intent.Two cohorts were analyzed: 1) Primary surgery with pathological stage III (pIII, n = 1010), 2) Definitive chemoradiotherapy (RCT) durvalumab (Durva) with clinical stage III (cIII, n = 604).Only treatments according to the German S3 guidelines were included.Endpoints were overall survival (OS) and time to progression. Results:Staging accuracy was evaluated in all patients with initial surgery (n=3981).Clinical-pathological staging discordance here was substantial, with agreement only in 42.0%.Upstaging was more common (31.8%) than down-staging (26.2%).In group I, adjuvant platinum therapy (n=537) improved OS versus surgery alone (n=473; HR 0.52, 95%CI 0.4-0.7).Cisplatin (n=340) and carboplatin (n=197) showed comparable OS (HR 1.0).Median time to progression was 452 days for both platinum groups versus 329 days for surgery alone (p = 0.034).In group II, Durva maintenance conferred the highest benefit (RCT+Durva vs. RCT; HR 0.55, 95%CI 0.4-0.7).Here cisplatin was more effective in non-squamous tumors, while in squamous tumors both platinums performed comparable.Median time to progression was longest for cisplatin-RCT + Durva (339 days), followed by cisplatin-RCT (283), carboplatin-RCT + Durva (280) and carboplatin-RCT (218).Conclusions: Real-world evidence from BWCR interestingly reveals considerable staging discrepancy in non-metastatic NSCLC.Adjuvant platinum therapy confers benefit to pIII-patients irrespective of platinum regimen, while in cIII-disease, Durva maintenance and histology-specific platinum selection significantly influence outcomes.

ESMO OpenVol. 11
Bexley College (GB), NYU Langone Health (US), AstraZeneca (Australia) (AU), AstraZeneca (Finland) (FI), Icahn School of Medicine at Mount Sinai (US)
Quality Education
Openalex Percentile: Top 19%
Radiomics and Machine Learning in Medical Imaging
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.