基于一体化CT-linac的直肠癌全自动放疗工作流首次实现
单中心前瞻性试点研究纳入20例直肠癌患者,在集成CT-linac平台上完成从CT模拟、自动分割、自动计划到治疗实施的All-in-One全自动放疗工作流,无需患者重新摆位。工作流全部成功完成,平均总治疗时间37.5±6.9分钟,靶区覆盖和危及器官保护在多数病例中临床可接受,早期疗效和毒性结果支持进一步临床推广。
要点
- 20例直肠癌患者均成功完成全自动一体化放疗工作流。
- 平均总治疗时间37.5±6.9分钟,自动分割与计划占比最大。
- 3D gamma通过率全部超过95%,早期完全缓解率15.0%。
核心结论
- 该试点研究证实基于一体化CT-linac的直肠癌全自动放疗工作流可行、省时且临床可用。
- 良好的靶区适形性、稳定的计划质量和有前景的早期结果支持进一步临床实施。
研究设计与样本
- 单中心前瞻性试点研究,共纳入20例直肠癌患者。
- 患者在集成CT-linac系统上接受All-in-One放疗,流程包括CT模拟、人工智能自动分割、自动计划、在线验证和束流出束,全程无需重新摆位。
- 收集靶区和危及器官剂量学参数及gamma分析结果,采用Dice相似系数和平均表面距离评估自动分割精度,随访评估短期毒性、手术结局和肿瘤退缩分级。
主要发现
- AIO工作流在全部20例患者中成功完成,平均总治疗时间37.5±6.9分钟,自动分割和计划占工作流时间比例最大。
- AIO自动计划系统在多数病例中达到临床可接受的靶区覆盖和危及器官保护,表明队列内计划质量稳定可重复。
- GTV、CTV45和CTV50的平均Dice相似系数分别为0.83、0.95和0.89。
- Gamma分析显示3D通过率在所有患者中均超过95%,2D通过率在3%/3 mm标准下各机架角度范围为84.0%至100.0%。
- 血液学毒性最常见,白细胞减少发生率为55.0%;20例完成随访的患者中19例接受手术、1例达临床完全缓解,总体完全缓解率为15.0%,手术患者中68.4%达肿瘤退缩分级0-1。
关键数值与推荐
- GTV平均Dice相似系数0.83,CTV45为0.95,CTV50为0.89。
- 3D gamma通过率在所有患者中超过95%,2D gamma通过率在3%/3 mm标准下为84.0%至100.0%。
- 白细胞减少发生率55.0%,总体完全缓解率15.0%,手术患者中肿瘤退缩分级0-1者占68.4%。
临床怎么用
- 可在集成CT-linac平台上对直肠癌患者实施全自动一体化放疗,减少患者摆位和流程环节。
- 自动分割和自动计划可达到多数病例临床可接受的靶区覆盖和危及器官保护,但需关注其占用的工作流时间。
- 早期毒性以血液学毒性为主,白细胞减少发生率较高,治疗中需监测血常规。
局限性
- 单中心前瞻性试点研究,样本量为20例,结果需更大规模研究验证。
- 随访时间有限,短期毒性和手术结局为初步结果。
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
PURPOSE: This study aimed to prospectively evaluate, in a single-center pilot study, the feasibility, workflow efficiency, and dosimetric performance of an integrated All-in-One (AIO) radiation therapy solution for rectal cancer that combines computed tomography (CT) simulation, automatic segmentation, planning, and treatment delivery on a single CT-linac platform. METHODS AND MATERIALS: A total of 20 patients with rectal cancer underwent AIO radiation therapy on an integrated CT-linac system. The workflow comprised CT simulation, artificial intelligence-based autosegmentation, autoplanning, online verification, and beam delivery, all performed sequentially without patient repositioning. Dosimetric parameters for target volumes and organs at risk, together with gamma analysis results, were collected. Autosegmentation accuracy was assessed using Dice similarity coefficient and average surface distance. Short-term toxicities, surgical outcomes, and tumor regression grade were assessed at follow-up. RESULTS: The AIO workflow was successfully completed in all 20 patients. The mean total treatment time was 37.5 ± 6.9 minutes, with autosegmentation and planning accounting for the largest proportion of workflow time. The AIO autoplanning system achieved clinically acceptable target coverage and organs at risk sparing in most cases, indicating stable and reproducible plan quality across the cohort. The mean Dice similarity coefficients for the gross tumor volume, clinical target volume 45, and clinical target volume 50 were 0.83, 0.95, and 0.89, respectively. Gamma analysis demonstrated excellent delivery accuracy, with 3-dimensional pass rates exceeding 95% in all patients, and two-dimensional pass rates across gantry angles ranging from 84.0% to 100.0% under the 3%/3 mm criterion. Hematologic toxicity was the most common adverse event, with leukopenia observed in 55.0% of patients. Among 20 patients who completed follow-up, 19 underwent surgery and 1 achieved a clinical complete response. The overall complete response rate was 15.0%, and 68.4% of the surgical patients achieved tumor regression grade 0-1. CONCLUSIONS: This pilot study demonstrates the feasibility of a fully automated, time-efficient, and clinically applicable radiation therapy workflow for rectal cancer. Favorable target conformity, stable plan quality, and promising early outcomes support further clinical implementation.
引用
- Li X, Wang L, Guo R, et al. First Implementation of an All-in-One Solution With Fully Automatic Workflow for Rectal Cancer Based on an Integrated CT-linac. Pract Radiat Oncol. 2026. PMID: 42742501. doi:10.1016/j.prro.2026.07.009
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。