Holland Martin, Krishnan Chetana, Sotoudeh Houman, Nabors Louis, Fiveash John, Riley Kristen, Huang Wei, Barboriak Daniel, Kim Harrison
Department of Interdisciplinary Engineering, University of Alabama at Birmingham, Birmingham, AL, USA.
Department of Biomedical Engineering, University of Alabama at Birmingham, Birmingham, AL, USA.
Quant Imaging Med Surg. 2025 May 1;15(5):4321-4332. doi: 10.21037/qims-2024-2566. Epub 2025 Apr 28.
Currently, no definitive method reliably differentiates pseudoprogression from true progression. Misclassification can either halt effective therapy or prolong ineffective treatment. We hypothesized that the diagnostic accuracy could be improved using quantitative dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) after error correction via point-of-care portable perfusion phantoms (P4s). This study aimed to develop a P4 for quantitative DCE-MRI of the brain and enhance accuracy in distinguishing between pseudo and true glioblastoma progression.
Twelve patients with potential glioblastoma progression after adjuvant chemoradiation therapy were recruited. Each subject underwent two DCE-MRI exams within a week using a single 3T MRI scanner. Quantitative DCE-MRI parameters were retrieved based on the extended Tofts model (ETM), Tofts model (TM), and shutter speed model (SSM) before and after P4-based error correction. The consistency of the pharmacokinetic (PK) parameter measurements was evaluated based on the within-subject coefficient of variation (wCV) before and after P4-based error correction. Glioblastoma progression status was determined using the Response Assessment in Neuro-Oncology (RANO) criteria about five months after DCE-MRI exams.
Among the participants, five had true progression, and seven had pseudoprogression. The wCVs of the measurement based on TM, ETM, and SSM were 22%, 22%, and 24%, respectively, before error correction but improved to 7%, 6%, and 8%, respectively, after correction. Similarly, their accuracies in differentiating between pseudo and true progression were 0.88 regardless of the PK models before error correction. However, those after error correction were improved to 100% in TM (or ETM) and 96% in SSM.
Following P4-based error correction, a quantitative DCE-MRI parameter, , demonstrated 100% accuracy in discriminating between pseudo and true progression when TM or ETM were employed.
目前，尚无可靠的确定性方法能可靠地区分假性进展与真性进展。错误分类可能会导致有效治疗中断或无效治疗延长。我们假设，通过即时检验便携式灌注体模（P4）进行误差校正后，使用定量动态对比增强磁共振成像（DCE-MRI）可提高诊断准确性。本研究旨在开发一种用于脑部定量DCE-MRI的P4，并提高区分胶质母细胞瘤假性进展和真性进展的准确性。
招募了12例辅助放化疗后疑似胶质母细胞瘤进展的患者。每位受试者在一周内使用一台3T MRI扫描仪进行两次DCE-MRI检查。在基于P4的误差校正前后，根据扩展Tofts模型（ETM）、Tofts模型（TM）和快门速度模型（SSM）获取定量DCE-MRI参数。基于P4的误差校正前后，根据受试者内变异系数（wCV）评估药代动力学（PK）参数测量的一致性。在DCE-MRI检查后约五个月，使用神经肿瘤学疗效评估（RANO）标准确定胶质母细胞瘤的进展状态。
参与者中，5例为真性进展，7例为假性进展。校正前，基于TM、ETM和SSM测量的wCV分别为22%、22%和24%，校正后分别提高到7%、6%和8%。同样，在校正前，无论采用哪种PK模型，它们区分假性进展和真性进展的准确率均为0.88。然而，校正后，TM（或ETM）的准确率提高到100%，SSM的准确率提高到96%。
基于P4的误差校正后，当采用TM或ETM时，定量DCE-MRI参数在区分假性进展和真性进展方面显示出100%的准确率。
