Real-world demographics, clinical characteristics, and testing and treatment patterns in patients with non-small-cell lung cancer and extensive-stage small-cell lung cancer in Belgium: the AIBED study

<h3>Background</h3> Real-world evidence is essential to understand the implementation of lung cancer treatments. This study describes the characteristics, testing patterns, and treatment of lung cancer patients in Belgium. <h3>Methods</h3> This multicenter study used routinely collected longitudinal electronic health record data from five Belgian hospitals (2019-2021). Natural language processing was used to process unstructured data, which was combined with structured data to generate an Observational Medical Outcomes Partnership (OMOP) common data model (OMOP-CDM) data warehouse. We generated a lung cancer cohort and subcohorts of special interest: stage IB-IIIB resected non-small-cell lung cancer (NSCLC), unresected stage III NSCLC, (EGFR-mutated) stage IV NSCLC, and extensive-stage small-cell lung cancer (ES-SCLC). <h3>Results</h3> We included 1952 patients, of whom 87% (1699) had NSCLC and 13% (253) SCLC (86.6% ES-SCLC). Stage IB-IIIB resected NSCLC included 17.3%, unresected stage III NSCLC 12.3%, and stage IV NSCLC 45.0% of NSCLC patients. Brain metastasis around diagnosis was reported in 16.4% of stage IV NSCLC (22.3% in EGFR mutated) and 15.1% of ES-SCLC patients. Across NSCLC subcohorts, 53.9% of patients had a programmed death-ligand 1 (PD-L1)- positive tumor (≥1%) and 12.7% an EGFR mutation-positive tumor. In stage IB-IIIB resected NSCLC patients, resection status was retrieved for 83.0% of patients, of whom 93.0% had complete resection. Among 327 stage III NSCLC patients, 63.9% (<i>n</i> = 209) were unresected, 36.7% (<i>n</i> = 120) had a PD-L1-positive tumor, 18.3% (<i>n</i> = 60) completed chemoradiotherapy, and 11.3% (<i>n</i> = 37) initiated durvalumab. <h3>Conclusions</h3> This is the first Belgian study to provide real-world insights into patient trajectories in lung cancer patients leveraging artificial intelligence technologies.