From Neuropsychiatric Symptomatology to Oncological Diagnosis: Progression of Grade 2 to Grade 4 Cerebral Astrocytoma over a Four-Year Period and the Potential of AI in Precision Medicine

Abstract

The progression of low-grade gliomas, particularly IDH1-mutant diffuse astrocytoma, remains a complex and clinically relevant topic, especially when early manifestation are predominantly neuropsychiatric. In certain clinical scenarios, young individuals presenting with severe depressive and anxiety symptoms, initially suggestive of primary psychiatric disorders, may be found to harbour underlying low-grade brain tumours upon neuroimaging. Left temporal IDH-1 mutant astrocytoma (WHO grade II) are among such pathologies that can initially mimic affective or anxiety disorders, delaying appropriate diagnosis and intervention. Longitudinal clinical observation over a four-year period have shown that these tumours may undergo malignant transformation into high-grade (WHO grade IV) IDH-mutant astrocytoma, characterised histologically by increased cellularity, mitotic activity, vascular proliferation, and necrosis. Molecular analysis typically reveals continued IDH1 mutation, ATRX loss and p53 overexpression, with proliferative indices increasing significantly at the point of progression. The neuropsychiatric trajectory may parallel this evolution, transitioning from affective symptoms to more profound cognitive and personality disturbances associated with frontal lobe dysfunction. Despite multimodal treatment strategies including surgery, radiotherapy and chemotherapy, prognosis tends to worsen significantly upon transformation, particularly when early symptoms are misattributed to primary psychiatric conditions. A synthesis of current literature supports the recognition of early neuropsychiatric symptoms, particularly those resistant to conventional treatment, as potential early indicators of intracranial neoplasms. Systematic neuroimaging in such cases could facilitate early detection. Additionally, recent advances in artificial intelligence (AI) show promise in enhancing diagnostic precision through imaging analysis and multimodal clinical data integration, thereby contributing to more personalised and timely therapeutic approaches in neuro-oncology.