Preoperative assessment in lung-limited metastatic patients with colorectal cancer: The Meta-Lung Score

3598 Background: Due to the inconsistent data on curative resection of isolated lung metastases (LM) from colorectal cancer (CRC) and the lack of specific recommendations, the therapeutic approaches for metastatic liver disease are mostly applied to LM treatment, despite differences in their biological behavior. Like Fong's criteria in CRC liver metastases, our study proposed a score associated with clinical outcomes in patients (pts) undergoing CRC lung metastasectomy, aiming to better pts' selection for surgery and the most appropriate therapeutic strategy. Methods: We retrospectively collected data from 260 pts (aged 18-85) who underwent CRC lung metastasectomy with curative intent from December 2002 to January 2022 at four Italian Centers: the Division of Thoracic Surgery at “A. Businco Cancer Center” in Cagliari, the Division of Thoracic Surgery at “Città della Salute e della Scienza” in Tourin, the Department of Thoracic Surgery at “IRCCS Azienda Ospedaliero-Universitaria” in Boulogne, and the Medical Oncology Unit of the University Hospital of Cagliari. Statistical analysis was performed with MedCalc (survival distribution: Kaplan-Meier; survival comparison: log-rank test; association between categorical variables: Fisher's exact test). Results: We analyzed the impact of different clinicopathological features on overall survival (OS). At the univariate analysis: higher baseline CEA levels (p = 0.0001), disease-free survival less than or equal to 12 months (m) (p = 0.0043), LM size larger than 2 cm (p = 0.0187), multiple resectable nodules (p = 0.0083), and positive nodal status of the primary tumor (p = 0.0011) were associated with poorer prognosis. In a Cox regression model, these five characteristics retained their independent role for OS (p < 0.0001) and were chosen as criteria to be assigned one point each for clinical risk score. The 5-year survival rate in pts with 0 points was 88%, while no pts with a 5-points score survived at 2 years. Based on the 0-2 versus 3-5 score range, we obtained a significant difference in median OS: 101.7 m (95%CI 64.6 - 101.7) vs. 39.5 m (95%CI 27.3 - 87.5) respectively (p < 0.0001) stratifying pts into good and poor prognosis. Conclusions: The Meta-Lung Score appeared to be an exciting prognostic tool in selecting CRC pts’ candidates for radical surgical treatment when LM were diagnosed. Whether early surgery should be considered in pts with a score 0-2, the same choice should be taken with caution in pts with a score 3-5, for whom early chemotherapy would allow better assessment of tumor biology and appropriate selection for surgery.[Table: see text]