DeWalt Nicholas C, Yanagihara Theodore K, Gallitto Matthew, Kinslow Connor J, Chiuzan Codruta, Xu Yuanguang, Singh Gagandeep, Bruce Jeffrey N, Wang Tony J C, Grinband Jack
Department of Neurological Surgery, Columbia University Irving Medical Center, New York, New York.
Department of Radiation Oncology, Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, North Carolina.
Int J Radiat Oncol Biol Phys. 2025 Jul 1;122(3):561-571. doi: 10.1016/j.ijrobp.2025.03.037. Epub 2025 Mar 28.
Radiation therapy (RT) planning for high-grade glioma (HGG) typically relies on a single postoperative magnetic resonance image (MRI), assuming stable tumor and brain architecture throughout treatment. Though the gross tumor volume and surrounding tissues can shift after surgery and throughout RT, its rate of change, predictors of the movement magnitude, and impact on different dosing strategies are not well understood. This study uses meta-analysis and prospective MRI data to optimize MRI timing and identify patients who may benefit from adaptive RT (ART).
We performed a meta-analysis of 12 studies (405 patients) to quantify postresection brain morphology changes. Additionally, we prospectively collected MRIs from 17 newly diagnosed HGG patients who underwent resection and RT. Nonlinear image registration tracked voxel-wise movement from postsurgery through RT and follow-up. We analyzed changes in gross tumor volume, clinical target volume (CTV), and non-CTV influx volume (nCIV) across 8 CTV strategies. Exponential models predicted the magnitude and rate of postresection changes and optimal MRI timing.
Both meta-analysis and prospective data showed that morphological changes followed an exponential decay, with 80% of shifts occurring within ∼30 days of resection. The nCIV was strongly predicted by T2-weighted fluid-attenuated inversion recovery volume and CTV margin strategy (86% accuracy for T1-based and 80% for T2-weighted fluid-attenuated inversion recovery-based CTVs). To minimize nCIV in patients starting RT >3 weeks postsurgery, our model suggests acquiring an additional MRI before RT. Otherwise, RT planning should be adaptive.
Acquiring a delayed planning MRI or performing ART based on recent MRIs may optimize RT delivery for patients with larger tumors and specific CTV strategies. We propose a simple, clinically feasible algorithm for selecting patients and determining optimal MRI timing in prospective clinical trials of HGG ART.
高级别胶质瘤（HGG）的放射治疗（RT）计划通常依赖于术后单一的磁共振成像（MRI），假定在整个治疗过程中肿瘤和脑结构稳定。尽管手术切除后以及整个放疗过程中肿瘤总体积和周围组织会发生移位，但其变化速率、移位幅度的预测因素以及对不同剂量策略的影响尚不清楚。本研究采用荟萃分析和前瞻性MRI数据来优化MRI检查时间，并识别可能从自适应放疗（ART）中获益的患者。
我们对12项研究（405例患者）进行了荟萃分析，以量化切除术后脑形态学变化。此外，我们前瞻性地收集了17例新诊断的接受了切除和放疗的HGG患者的MRI数据。非线性图像配准追踪了从手术后到放疗及随访期间体素级别的移动。我们分析了8种临床靶体积（CTV）策略下肿瘤总体积、临床靶体积（CTV）和非CTV流入体积（nCIV）的变化。指数模型预测了切除术后变化的幅度和速率以及最佳MRI检查时间。
荟萃分析和前瞻性数据均显示，形态学变化呈指数衰减，约80%的移位发生在切除术后约30天内。T2加权液体衰减反转恢复序列体积和CTV边界策略能强烈预测nCIV（基于T1的CTV准确率为86%，基于T2加权液体衰减反转恢复序列的CTV准确率为80%）。为使术后超过3周开始放疗的患者的nCIV最小化，我们的模型建议在放疗前额外进行一次MRI检查。否则，放疗计划应采用自适应方式。
获取延迟的计划MRI或基于近期MRI进行ART可能会优化对较大肿瘤患者和特定CTV策略患者的放疗实施。我们提出了一种简单、临床可行的算法，用于在HGG ART的前瞻性临床试验中选择患者并确定最佳MRI检查时间。
