数字病理生物标志物指导前列腺癌放疗决策:系统综述与专家共识
本文通过系统综述和改良德尔菲法专家调查,总结数字病理联合人工智能(DP-AI)在前列腺癌放疗中的作用。
要点
- DP-AI对无远处转移生存和总生存的预后预测优于传统风险模型
- DP-AI可辅助原发放疗中ADT时长的治疗决策
- 专家支持DP-AI在局限、术后和转移场景的潜力,但需前瞻性验证
核心结论
- DP-AI工具在前列腺癌放疗中通过指导患者分层和优化原发放疗ADT时长,展现出较强预后和预测潜力,但广泛临床采用前需前瞻性研究和现代诊疗队列验证。
研究设计与样本
- 系统综述评估DP-AI用于接受放疗的前列腺癌患者的证据,并采用改良德尔菲法进行两轮专家调查,以识别未来研究方向。
主要发现
- DP-AI对无远处转移生存和总生存等终点显示出较强预后和预测性能,优于传统风险模型并辅助ADT时长等治疗决策。
- 调查确认DP扫描仪采用日益增多,但报销机制和可用性存在区域差异,专家认可DP-AI在局限、术后和转移场景的潜力。
关键数值与推荐
- 专家调查中≥75%一致性的问题被认为相关并纳入定性分析。
临床怎么用
- DP-AI可辅助前列腺癌放疗患者的分层和原发放疗中ADT时长优化,但需在前瞻性研究和现代诊疗队列验证后广泛采用。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
Current risk-stratification systems for prostate cancer (PCa) do not sufficiently reflect the disease heterogeneity, and digital pathology (DP) combined with artificial intelligence (AI) tools (DP-AI) may offer a solution to this challenge. The aim of this work is to summarize the role of DP-AI for PCa patients treated with radiotherapy (RT), and to point out future areas of research. We conducted (1) a systematic review on the evidence of DP-AI for patients treated with RT and (2) a survey of experts using a modified Delphi method, addressing the current role of DP-AI in clinical and research practice to identify relevant fields of future development. Eleven studies investigated DP-AI in PCa RT, with most using the multimodal AI (MMAI) classifier and four ongoing studies are currently prospectively testing the DP-AI performance. DP-AI showed strong prognostic and predictive performance for endpoints like distant metastasis free survival and overall survival, outperforming traditional risk models and assisting treatment decisions such as androgen deprivation therapy (ADT) duration. Fifty-one and 35 experts responded to round 1 and round 2 of the survey respectively. Questions with ≥75 % agreement were considered relevant and included in the qualitative analysis. Survey results confirmed growing adoption of DP scanners, although regional differences in re-imbursement mechanisms and availability persist, with experts endorsing DP-AI's potential across localized, postoperative, and metastatic settings, though further prospective validation is needed. DP-AI tools show strong prognostic and predictive potential in various PCa by guiding patient stratification and optimising ADT duration in primary RT. Prospective studies and validation in cohorts using modern diagnostic and treatment methods are needed before broad clinical adoption.
引用
- Zamboglou C, Doncker W, Christoforou AT, et al. oDigital pathology biomarkers for guiding radiotherapy-based treatment concepts in prostate cancer - a systematic review and expert consensus. Radiother Oncol. 2025. PMID: 40645505. doi:10.1016/j.radonc.2025.111039
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。