The classification of cancer traditionally has relied on organ of origin and histological characteristics observable through microscopy. This approach has been foundational to obtaining an understanding of cancer biology and treatment for more than a century, shaping how to diagnose, research, and treat various forms of this disease. As knowledge of cancer biology has advanced, however, it has become increasingly clear that the traditional classification system is at odds with the principles of modern precision oncology, according to an article (Khozin, S.) published in the September 2024 issue of the journal Clinical and Translational Science. The essential tenet of precision medicine postulates that administering the right therapy to the right patient at the right time and dose should theoretically result in 100% response to treatment. The goal is not reached because current disease classification methods frequently fail in matching the right therapy to the right patient. Recent developments in molecular profiling and advanced imaging techniques have revealed a more complex picture of cancer biology. It is understood now that tumors arising in different organs can share key molecular and morphological features, while tumors in the same organ can be strikingly heterogeneous at multiple levels of analysis.
This heterogeneity has an impact on clinical trials, drug development, and patient care. Recent advances in artificial intelligence (AI), particularly machine learning and deep learning, offer promising avenues for reclassifying cancers through comprehensive integration of molecular, histopathological, imaging, and clinical characteristics. AI-driven approaches have the potential to reveal novel cancer subtypes, identify new prognostic variables, and guide more precise treatment strategies for improving patient outcomes. In recent years, artificial intelligence, particularly machine learning and deep learning, has emerged as an important tool for addressing these challenges. Advanced computational methods can analyze vast amounts of multimodal data, identifying patterns and relationships that may not be apparent through traditional disease classification methods. By embracing the full complexity of tumor biology through AI-enabled multimodal analyses, it is possible to move beyond the historical constraints of organ-based and histological classifications and towards a new era that truly delivers on the promise of personalized medicine for patients with cancer.