Cancer is one of the leading causes of death worldwide, and its incidence is steadily increasing. Although years of research have been conducted on cancer treatment, clinical treatment options for cancers are still limited. Animal cancer models have been widely used for studies of cancer therapeutics, but these models have been associated with many concerns, including inaccuracy in the representation of human cancers, high cost and ethical issues. Therefore, in vitro human cancer models are being developed quickly to fulfill the increasing demand for more relevant models in order to get a better knowledge of human cancers and to find novel treatments. This review summarizes the development of in vitro human cancer models for biomedical applications. We first review the latest development in the field by detailing various types of in vitro human cancer models, including transwell-based models, tumor spheroids, microfluidic tumor-microvascular systems and scaffold-based models. The advantages and limitations of each model, as well as their biomedical applications, are summarized, including therapeutic development, assessment of tumor cell migration, metastasis and invasion and discovery of key cancer markers. Finally, the existing challenges and future perspectives are briefly discussed.

Brain tumors provide considerable diagnostic and treatment challenges due to their intricate nature and the hazards linked to direct tissue biopsies. Owing to the restricted sensitivity and specificity of conventional procedures, new techniques like liquid biopsy have garnered attention. Circulating tumor DNA (ctDNA), present in physiological fluids such as cerebrospinal fluid (CSF) and plasma, has emerged as a viable instrument for non-invasive tumor characterization. Hence, advancements in next-generation sequencing (NGS) and digital PCR have enhanced the sensitivity of ctDNA detection, rendering it a feasible method for monitoring tumor dynamics and evaluating therapy responses. Research indicates that ctDNA strongly correlates with tumor heterogeneity, providing a superior alternative to single-site tissue biopsies. CSF, due to its proximity to the brain, offers elevated amounts of ctDNA for examination relative to plasma, particularly in central nervous system (CNS) cancers. Research indicates that ctDNA can detect actionable mutations, forecast little residual illness, and enable real-time monitoring of disease development and resistance. Notwithstanding these advantages, difficulties, including poor ctDNA yield and heterogeneity in detection methodologies persist. This review examines the clinical efficacy of ctDNA in brain tumor diagnosis, emphasizes technical developments in ctDNA analysis, and stresses the necessity for standardized methods. Comprehending the capabilities and constraints of ctDNA can facilitate the development of more accurate, individualized therapy approaches in neuro-oncology.