C77-43 Dlco as a Functional Predictor of Oncogene Addiction in Non-Small Cell Lung Cancer

Rationale Actionable driver mutations are detected in up to 65% of patients with advanced non-small cell lung cancer (NSCLC), making oncogene addiction a central determinant for targeted treatment. Although demographic and clinical features such as smoking history, sex, age, and imaging characteristics have been linked to oncogene status, no validated pre-test clinical model currently exists. Identifying practical and inexpensive predictors could streamline diagnostic workflows and shorten the time to appropriate therapy. This study investigated whether specific clinical and functional respiratory parameters—particularly the diffusing capacity for carbon monoxide (DLCO)—could help predict oncogene addiction in NSCLC prior to molecular testing. Methods We retrospectively analyzed consecutive patients with newly diagnosed NSCLC evaluated at the Pulmonology Unit of Spedali Civili Hospital (Brescia, Italy) between December 2022 and May 2024. Data on demographics, smoking history, comorbidities, and pulmonary function tests were collected. Oncogene addiction was defined by the presence of actionable driver mutations (EGFR, ALK, ROS1, and others). A total of 375 patients were included. DLCO values were compared between oncogene-addicted and non-oncogene-addicted groups. Diagnostic accuracy was assessed using receiver operating characteristic (ROC) analysis, and sensitivity, specificity, and predictive values were calculated. Results Among 375 patients, 71% carried actionable oncogenic alterations. This group included a higher proportion of women and never-smokers compared with the non-oncogene-addicted population. Mean DLCO values were significantly lower in the non-oncogene-addicted group. A DLCO threshold <80% predicted the absence of oncogene addiction with high sensitivity (92.9%) but limited specificity (35.7%). The negative predictive value was 88.2%, indicating that preserved DLCO was strongly associated with a higher likelihood of oncogene addiction. These findings suggest that DLCO, a routinely available and noninvasive parameter, may serve as a simple functional marker for pre-test patient stratification. Conclusion DLCO measurement may contribute to early identification of NSCLC patients more likely to harbor oncogenic driver mutations, particularly in resource-limited contexts or when molecular testing results are delayed. Although not diagnostic, the high sensitivity of DLCO supports its role as a potential “rule-out” tool for non-oncogene-addicted disease. Prospective validation studies are warranted to confirm these findings and define the integration of DLCO-based assessment into clinical decision-making algorithms.