Zhang Ying, Chan Mark Ka Heng
Department of Medical Physics and Biomedical Engineering, University College London, London, UK.
Department of Radiation Oncology, University Nebraska Medical Center, Omaha, USA.
Med Phys. 2024 Dec;51(12):9207-9216. doi: 10.1002/mp.17412. Epub 2024 Sep 19.
In head and neck (H&N) cancer treatment, a conventional setup error (SE) of 3mm is often used in robust optimization (cRO3mm). However, cRO3mm may lead to excessive radiation doses to organs at risk (OARs) and does not purposefully compensate for interfractional anatomy variations.
This study introduces a method using predicted images from an anatomical model and a reduced 1mm SE uncertainty for robust optimization (aRO1mm), aiming to decrease the dose to OARs without affecting the coverage of the clinical target volume (CTV).
This retrospective study involved 10 nasopharynx radiotherapy patients. Validation CT scans (vCT) from treatment weeks 1 to 6 were analyzed. A predictive anatomical model, designed to capture the average anatomical changes over time, provided predicted CT images for weeks 1, 3, and 5. We compared three optimization scenarios: (1) aRO1mm, using three predicted images with 1mm setup shift and 3% range uncertainty, (2) cRO3mm, with a robust 3mm setup shift and 3% range uncertainty, and (3) cRO1mm, a robust 1mm setup shift and 3% range uncertainty. The accumulated dose to CTVs and serial organs was evaluated under these uncertainties, while parallel OARs were assessed using the accumulated nominal dose (without errors).
The accumulated volume receiving 94% of the prescribed dose (V94) for CTVs in cRO3mm exceeded 98%, meeting the clinical goal. For high-risk CTV, the minimum V94 was 96.44% in aRO1mm and 94.05% in cRO1mm. For low-risk CTV, these values were 97.68% in aRO1mm and 97.15% in cRO1mm. When comparing aRO1mm to cRO3mm on OARs, aRO1mm reduced normal tissue complication probability (NTCP) for grade 2 xerostomia and dysphagia by averages of 3.67% and 1.54%, respectively.
aRO1mm lowers the radiation dose to OARs compared to the traditional approach, while maintaining adequate dose coverage on the target area. This method offers an improved strategy for managing uncertainties in radiation therapy planning for H&N cancer, enhancing treatment effectiveness.
在头颈癌治疗中，稳健优化（cRO3mm）通常采用3毫米的传统摆位误差（SE）。然而，cRO3mm可能导致危及器官（OARs）接受过量辐射剂量，且未针对性地补偿分次间的解剖结构变化。
本研究介绍一种方法，利用解剖模型的预测图像和降低至1毫米的SE不确定性进行稳健优化（aRO1mm），旨在在不影响临床靶区（CTV）覆盖的情况下降低OARs的剂量。
这项回顾性研究纳入了10例鼻咽癌放疗患者。分析了治疗第1至6周的验证CT扫描（vCT）。一个旨在捕捉随时间变化的平均解剖结构变化的预测解剖模型，提供了第1、3和5周的预测CT图像。我们比较了三种优化方案：（1）aRO1mm，使用三张具有1毫米摆位偏移和3%范围不确定性的预测图像；（2）cRO3mm，具有稳健的3毫米摆位偏移和3%范围不确定性；（3）cRO1mm，具有稳健的1毫米摆位偏移和3%范围不确定性。在这些不确定性条件下评估CTV和系列器官的累积剂量，同时使用累积标称剂量（无误差）评估平行OARs。
cRO3mm中CTV接受94%处方剂量的累积体积（V94）超过98%，达到临床目标。对于高危CTV，aRO1mm中的最小V94为96.44%，cRO1mm中为94.05%。对于低危CTV，这些值在aRO1mm中为97.68%，在cRO1mm中为97.15%。在OARs方面，将aRO1mm与cRO3mm进行比较时，aRO1mm使2级口干和吞咽困难的正常组织并发症概率（NTCP）分别平均降低了3.67%和1.54%。
与传统方法相比，aRO1mm降低了OARs的辐射剂量，同时在靶区保持了足够的剂量覆盖。该方法为头颈癌放射治疗计划中的不确定性管理提供了一种改进策略，提高了治疗效果。
