文献速递消化道肿瘤放疗放疗物理与技术

胰腺SBRT分次内误差与分次间形变对上消化道危及器官剂量的影响及可预测性

Intra-fraction errors and inter-fraction deformation during pancreas SBRT: Analysis and predictability of dose variations for upper gastrointestinal organs.
PubMed 文献 · Radiother Oncol · 2026-04-03

研究纳入31例胰腺癌患者,在CyberKnife上行5次SBRT,处方40 Gy/50 Gy至PTV/GTV,计划优化要求十二指肠、胃和肠道V35Gy<0.5 cm3。评估分次内呼吸追踪误差和分次间OAR形变对OAR剂量的影响,并探讨靶区-OAR邻近度(EIV)的预测价值。结果显示追踪误差影响较小,但EIV大者需最小化误差;OAR形变影响较大,45%患者仍满足约束;EIV阈值为3.5 cm3时预测准确率77%。

要点

  • CyberKnife追踪误差对OAR剂量影响小,但EIV大者需最小化误差。
  • 分次间OAR形变影响更大,45%患者仍满足V35Gy<1 cm3约束。
  • EIV阈值3.5 cm3预测OAR约束违反的准确率为77%。

核心结论

  • CyberKnife追踪误差对OAR剂量影响较小,但对EIV值大的患者必须最小化追踪误差。
  • OAR分次间形变对剂量影响更大,但45%患者仍满足约束。
  • EIV是有前景的预测参数,阈值3.5 cm3预测准确率77%。

研究设计与样本

  • 回顾性分析31例胰腺癌患者,CyberKnife 5次SBRT,处方40 Gy/50 Gy至PTV/GTV。
  • 计划优化要求十二指肠、胃和肠道V35Gy<0.5 cm3。
  • 利用系统日志文件计算分次内追踪误差,模拟扰动计划评估OAR剂量。
  • 基于分次前CT评估分次间OAR形变,重新勾画并计算剂量。
  • 扰动计划和形变解剖的OAR约束以V35Gy<1 cm3为满足标准。

主要发现

  • EIV值范围为0.54至28.11 cm3。
  • 追踪误差标准不确定度范围为0.6至6.4 mm。
  • 仅1例EIV最大(28.11 cm3)且追踪不确定度>5 mm的患者在扰动计划中OAR约束违反(V35Gy=1.2 cm3)。
  • 14/31例患者在形变OAR上重算剂量仍满足约束;其余16例V35Gy最高达8.0 cm3(中位2.0 cm3)。

关键数值与推荐

  • 计划优化约束:十二指肠、胃和肠道V35Gy<0.5 cm3。
  • 扰动计划和形变解剖的OAR约束满足标准:V35Gy<1 cm3。
  • EIV阈值3.5 cm3预测准确率77%。

临床怎么用

  • 对EIV值大的患者应最小化CyberKnife追踪误差以保护OAR。
  • 分次间OAR形变可能导致剂量超限,需关注形变较大的患者。
  • EIV可作为预测OAR剂量变化的参数,辅助个体化计划评估。

局限性

  • 样本量较小(31例),需更大样本验证。
  • 回顾性研究,形变评估基于分次前CT,可能未完全反映实时变化。
  • EIV预测准确率77%,仍需进一步优化阈值和验证。

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

PURPOSES: /Objectives. Effects on OAR doses of intra-fraction respiratory tracking errors and inter-fraction deformation were evaluated for pancreas SBRT and predictability in terms of target-to -OAR proximity investigated. MATERIALS: /Methods. 31 pancreatic cancer patients, treated at CyberKnife (CK) in 5 fractions prescribing 40 Gy/50 Gy to PTV/GTV, were considered. Plan optimization required for duodenum, stomach and bowel V35Gy < 0.5 cm3. Target-to-OAR proximity was quantified by the expansion-intersection volume (EIV), computed on each planning CT as the intersection between the PTV expanded by 5 mm and duodenum, stomach and bowel. Treatments were delivered using CK real-time respiratory tracking. Tracking errors occurring during each fraction were computed using the system log-files. To simulate their effects on OAR doses, a perturbed plan was created modifying each beam coordinates by the corresponding errors. Inter-fraction OARs deformation was retrospectively evaluated on pre-fraction CT scans. Deformed OARs were contoured and planned doses re-calculated on the deformed anatomy. OAR constraints of perturbed plan and deformed anatomy were considered met at V35Gy < 1 cm3. EIV thresholds were defined and their predictive accuracy evaluated. RESULTS: EIV values ranged from 0.54 to 28.11 cm3. Tracking errors standard uncertainties ranged from 0.6 to 6.4 mm. For plans perturbed by tracking errors, OAR constraints were violated (V35Gy = 1.2 cm3) in a single patient showing the largest EIV value (28.11 cm3) and large tracking uncertainties (>5 mm). Doses recalculated on deformed OARs still fulfilled constraints for 14/31 patients. For the remaining 16 cases, V35Gy values up to 8.0 cm3 (median 2.0 cm3) were observed. An EIV threshold of 3.5 cm3 showed 77% prediction accuracy. CONCLUSIONS: CK tracking errors have small effects on OAR doses, but they must be minimized for those patients with large EIV values. OAR deformation has larger effects, but constraints were still met in 45% of patients. EIV is a promising predictive parameter.

来源:Radiother Oncol,查看原文。版权归原作者所有。

引用

  1. Loi M, Doro R, Zani M, et al. Intra-fraction errors and inter-fraction deformation during pancreas SBRT: Analysis and predictability of dose variations for upper gastrointestinal organs. Radiother Oncol. 2026. PMID: 41936944. doi:10.1016/j.radonc.2026.111525

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。