Predictive scoring of high‐grade histology among early‐stage lung cancer patients: The MOSS score

• Published in: Thoracic cancer 2023-07, Vol.14 (19), p.1865-1873
• Main Authors: Woo, Wongi, Cha, Yoon Jin, Park, Chul Hwan, Moon, Duk Hwan, Lee, Sungsoo
• Format: Article
• Language: English
• Subjects: adenocarcinoma; Adenocarcinoma - pathology; Biopsy; Carcinoma, Non-Small-Cell Lung - pathology; Chronic obstructive pulmonary disease; Datasets; Histology; Humans; International Association for the Study of Lung Cancer; Lung cancer; Lung Neoplasms - pathology; Lung Neoplasms - surgery; Male; Medical prognosis; Medical screening; Neoplasm Staging; non‐small cell lung cancer; Original; Patients; prediction model; preoperative factors; Prognosis; Regression analysis; Retrospective Studies; Surgery; Tomography; Tumors; Variables
• ISSN: 1759-7706; 1759-7714
• DOI: 10.1111/1759-7714.14932

Background Poor prognosis associated with adenocarcinoma of International Association for the Study of Lung Cancer (IASLC) grade 3 has been recognized. In this study we aimed to develop a scoring system for predicting IASLC grade 3 based before surgery. Methods Two retrospective datasets with significant heterogeneity were used to develop and evaluate a scoring system. The development set was comprised of patients with pathological stage I nonmucinous adenocarcinoma and they were randomly divided into training (n = 375) and validation (n = 125) datasets. Using multivariate logistic regression, a scoring system was developed and internally validated. Later, this new score was further tested in the testing set which was comprised of patients with clinical stage 0–I non‐small cell lung cancer (NSCLC) (n = 281). Results Four factors that were related to IASLC grade 3 were used to develop the new scoring system the MOSS score; male (M, point 1), overweight (O, point 1), size>10 mm (S, point 1), and solid lesions (S, point 3). Predictability of IASLC grade 3 increased from 0.4% to 75.2% with scores from 0 to 6. The area under the curve (AUC) of the MOSS was 0.889 and 0.765 for the training and validation datasets, respectively. The MOSS score exhibited similar predictability in the testing set (AUC: 0.820). Conclusion The MOSS score, which combines preoperative variables, can be used to identify high‐risk early‐stage NSCLC patients with aggressive histological features. It can support clinicians in determining a treatment plan and surgical extent. Further refinement of this scoring system with prospective validation is needed. The MOSS score is composed of four preoperative factors; male sex (M, point 1), overweight (O, point 1), tumor size>10 mm (S, point 1), and solid lesions (S, point 3). Predictability of IASLC grade 3 increased from 0.4% to 75.2% with scores from 0 to 6. This new score can be used to identify high‐risk early‐stage NSCLC patients with aggressive histological features.