A Grading System for Resected Invasive Squamous Cell Carcinoma of the Lung: A Multi-Institutional Study by the IASLC Pathology Committee

Abstract

Introduction: Tumor grading informs therapy and patient management across many organs, yet no consensus exists for grading invasive squamous cell carcinoma of the lung (LUSC). This study aimed to develop a globally applicable grading system using international cohorts. Methods: Histologic features, including tumor budding, smallest tumor nest size, nuclear size, and tumor spread through air spaces (STAS), were evaluated in two training sets comprising 262 and 427 LUSCs resected without neoadjuvant therapy from three institutions. Kaplan-Meier and Cox proportional hazards models were used to identify features associated with recurrence-free survival (RFS) and overall survival (OS). Features significant in both training sets were used to construct a grading system, which was then validated in a test set (n = 827, five institutions). Interobserver agreement was assessed among 10 pathologists on 25 cases. Results: Tumor budding (two-tier: cutoff at 10 buds per 0.785 mm2) was the only histologic feature significantly associated with both RFS and OS in multivariable analyses across both training sets. The proposed two-tier grading system-low-grade (0-9 buds), high-grade (>= 10 buds)- was validated in the test set, demonstrating median RFS of 4.8 versus 1.6 years for low-versus high-grade tumors in the entire cohort and 7.2 versus 3.4 years within stage I patients. Interobserver agreement was moderate (Fleiss' kappa = 0.524). Conclusions: The authors propose a simple, prognostically relevant grading system for resected invasive LUSC based on tumor budding. It is reproducible across international data sets and practical for routine pathology, offering a unified framework for clinical and research use. (c) 2025 International Association for the Study of Lung Cancer. Published by Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.