Introduction: Accurate staging of non-small cell lung cancer (NSCLC) is crucial to determine treatment and prognosis.
However, the clinical stage, based on imaging elements and small biopsies, is not always consistent with the pathologic stage provided by the surgical specimen, which may have implications in treatment and prognosis.

Aims: Investigate the agreement between the T clinical (cT) and pathologic stage (pT) for NSCLC, and evaluate factors associated with discordance.

Methods: Retrospective study, including all NSCLC patients, followed in our Thoracic Tumors Unit, who underwent surgical treatment, between 2012 and 2021. Demographic data, histology, clinical staging techniques, cT stage, pT stage, extent of resection and time between staging and surgery were collected from patient?s clinical files. TNM classification 8th edition of Lung Cancer was used. Agreement between clinical and pathologic stage was calculated by Cohen?s k.

Results: We included 157 patients (table 1). cT and pT were different in 54 (34.4%) patients: 30 (19.3%) were upstaged and 24 (15.5%) downstaged (table 2), ?=0.566. Table 3 describes the reasons for disagreement. Agreement was not associated with age, gender, histology, staging method or time between staging and surgery.

Conclusion: Although all patients performed an abdominal and chest CT scan for clinical staging, about a third of patients the pT disagreed with the cT and in most cases there was an upstaging. This may have implications in the decision of treatment, extent of resection, in adjuvant treatment and prognosis. This study points to the need to improve clinical staging, possibly through artificial intelligence.