Advances in Lung Cancer Diagnosis: From Biopsy-Based Pathology to Deep Learning Approaches

The histopathological examination of biopsy tissue by a pathologist remains the gold standard for the
diagnosis of lung cancer, forming the cornerstone of clinical decision-making. The 2015 World Health Organization
(WHO) Classification of Lung Tumors established a globally standardized framework for diagnosing lung
malignancies, integrating histological subtyping with molecular characteristics to guide personalized treatment
strategies. This classification refined the categorization of lung adenocarcinomas—introducing precursor lesions such
as atypical adenomatous hyperplasia (AAH), adenocarcinoma in situ (AIS), and minimally invasive adenocarcinoma
(MIA)—and emphasized invasive patterns with prognostic significance. Additionally, it reinforced the importance of
distinguishing between major tumor types, including adenocarcinoma, squamous cell carcinoma, and neuroendocrine
tumors, using both morphological and immunohistochemical markers. With the growing reliance on small biopsy and
cytology samples, especially in advanced-stage disease, the accuracy of diagnosis now depends heavily on the
integration of limited histological material with ancillary testing such as immunohistochemistry and molecular profiling
(e.g., EGFR, ALK, ROS1). These advancements have transformed pathology from a purely diagnostic tool to a critical
component in the era of targeted therapy and precision oncology. Collectively, these developments underscore the
indispensable role of tissue-based diagnosis in the modern management of lung cancer.