News & Comment

Over the past decade, multiple lines of inquiry have converged on the appreciation that the nervous system plays crucial roles in cancer pathophysiology. Key conceptual advances will mark 2025 as a milestone year in which the ‘emerging’ cancer neuroscience field developed into an established field. Here, we discuss the latest developments.

Foundation models hold transformative promise for oncology, yet their clinical implementation remains limited, largely owing to their current model design as narrow specialists optimized for static tasks, whereas clinical oncology requires generalist systems capable of integrating multimodal data, capturing disease evolution over time and considering patient perspectives. Design along these requirements is essential to integrating foundation models as trusted partners in cancer care.

Cancer is causing a global health crisis, straining even wealthy nations with rising cases, costs and workforce limits. Europe lags in translating basic research discoveries into clinical applications. We must invest in prevention, fundamental research, clinical trials, biotech support, workforce and patient involvement to stay competitive.

Fostering a nimble, innovative and resilient biomedical workforce is crucial. Here, we consider how to accelerate the readiness of early career investigators to lead, inverting the traditional pyramid view of academia to place them at the top, with expansive possibilities for advancement.

Training the next generation of cancer researchers is essential for the cancer research enterprise. However, training programs and methods to evaluate their effectiveness vary greatly across the USA and other countries. Here we discuss strategies to enhance cancer education and processes by which training may be standardized.

Spatial-omics tools, together with deep learning models, are revealing interactions between tumors and their environment, including immune cells, and are aiding human-like performance in histopathological analysis, with the potential to improve diagnostic accuracy and treatment choices.

Deep learning models are advancing cancer research and oncology but require human engagement to perform complex multi-step workflows. Autonomous artificial intelligence agents, empowered by large language models, present a promising solution by enabling the planning, execution and optimization of multi-step reasoning in biomedical research.

Chimeric antigen receptor (CAR)-T cells have revolutionized the treatment of hematological malignancies; however long-term safety concerns, associated with infections and secondary malignancies, have been reported. Here we discuss the clinical challenges of CAR-T cell therapy in light of this emerging benefit–risk controversy.

Evidence-based reductions in cancer treatment that still preserve outcomes can result in an improved quality of life for patients and optimized healthcare resourcing. Using melanoma as an example, we define treatment deintensification, outlining barriers to its implementation, as well as existing guidance.

As quantum technology advances, it holds immense potential to accelerate oncology discovery through enhanced molecular modeling, genomic analysis, medical imaging, and quantum sensing.

Gold-standard cancer data management is pivotal to enable precision medicine for European citizens. Achieving this goal relies on key elements: adopting standardized data formats, ensuring robust data privacy, educating professionals about the infrastructure’s benefits and leveraging cutting-edge technologies to transform cancer care.

Drug regulatory agencies in the USA and Europe have mechanisms to provide patients faster access to novel treatments, expecting that follow-up trials will confirm clinically meaningful results. However, some early approvals are subsequently withdrawn. Here we discuss the insights gained from withdrawn accelerated approvals for oncologic agents in the past decade.

Real-world data (RWD) and real-world evidence (RWE) from heterogeneous data sources has the potential to transform oncology research, especially when coupled with artificial intelligence (AI). We discuss the issues involved in primary data capture and post-hoc AI analysis and propose using AI to support the capture of primary RWD.

Neoantigen immunogenicity prediction is a burgeoning field with vast potential; however, the shortage of high-quality data and biases in current datasets limit model generalizability. Here we discuss some of the pitfalls that may underly this limited performance and propose a path forward.

Owing to high response rates, the Food and Drug Administration has approved both gene- and immune-targeted drugs for tumor-agnostic, genomic biomarker-based indications, for lethal solid and blood cancers. We posit that current data support tissue-agnostic activity as a paradigm, rather than an exception to the rule.

Liquid biopsies of circulating tumor DNA offer a non-invasive tool with many potential applications in oncology, including early cancer detection, profiling, disease prognosis, prediction of therapy response and monitoring disease status. A growing body of literature and clinical trials support an increasingly valuable role for liquid biopsies in the care of patients with solid malignancies.

Recent progress indicates a considerably improved mechanistic understanding of CAR T cell biology and delivers important insights into why some patients achieve durable remissions and others do not. In addition, although most success has been achieved in the context of CAR T cells targeted to B cell tumor antigens, namely CD19 and BCMA, we are seeing promising clinical trial outcomes for solid tumor malignancies.

As guidelines, therapies and literature on cancer variants expand, the lack of consensus variant interpretations impedes clinical applications. CIViC is a public-domain, crowd-sourced and adaptable knowledgebase of evidence for the clinical interpretation of variants in cancer, designed to reduce barriers to knowledge sharing and alleviate the variant-interpretation bottleneck.

Rapid progress in the molecular characterization of cancer genomes has been enabled by technology and computational analysis, and large databases now exist. Novel cancer therapeutics have resulted that more precisely target the vulnerabilities revealed by genomic analysis. Emergent efforts that link the two, using machine learning approaches and circulating DNA from cancer cells, are furthering cancer diagnosis and precision medicine.

The COVID-19 pandemic has impacted cancer care globally, the consequences of which are still not well understood. Through the lens of the impact in India, we emphasize the importance of continuing cancer care even during extenuating public health circumstances, and of strengthening health systems as a global priority.