Sep 29

SS 22 - Radiation and Cancer Physics 2: Imaging Biomarkers for Response Monitoring

229 - Dynamic Imaging in AVM Follow-Up: Evaluating 4D MRI for Post-Radiosurgical Obliteration

10:45am - 10:55am PT

Room 22/23

Presenter(s)

Vangipuram Shankar, MD, MBBS - Apollo Proton Cancer Centre, Chennai , Tamil Nadu

V. Shankar¹, K. P. Srinivasan Paramasivam², S. Ghosh³, S. Cholayil¹, R. Adhityan⁴, A. S. Uday Krishna⁵, R. Jalali⁵, and D. Shyam⁶; ¹Apollo Cancer Centers, Chennai, India, ²Endovascular Neurosurgery, Apollo Hospitals, Greams Road, Chennai, India, ³Dept. of Neurosurgery, Apollo Proton Cancer Center, Chennai, India, ⁴Department of Radiology, Apollo Proton Cancer Center, Chennai, Chennai, India, ⁵Apollo Proton Cancer Centre, Chennai, India, ⁶Apollo Speciality Hospitals, Madurai, India

Purpose/Objective(s): Post-radiosurgery assessment of arteriovenous malformations (AVMs) traditionally relies on digital subtraction angiography (DSA). Conventional MRI/MRA during follow up imaging, however, lacks real-time flow evaluation, limiting differentiation between true obliteration and slow-flow residuals.

Four-dimensional magnetic resonance imaging (4D MRI) addresses this gap by providing time-resolved vascular imaging, capturing dynamic arterial filling, nidus perfusion, and venous drainage. This study evaluates 4D MRI’s diagnostic accuracy, predictive value, and clinical utility in AVM Surveillance post radiosurgery, aiming to refine imaging protocols and optimize clinical decision making.

Materials/Methods: In this prospective study, 45 AVM patients (mean age: 34.2 ± 9.6 years) underwent dose-staged stereotactic radiosurgery (35–40 Gy in 5 fractions) using a frameless robotic radiosurgery system.

A multi-modality imaging protocol integrated 3D rotational angiography (3D-RA; biplane DSA) for vascular mapping, contrast-enhanced CT for anatomical precision, and MRI (3D T1 post-contrast, T2, time-of-flight MRA) for nidus delineation. All datasets were co-registered within the frameless robotic radiosurgery planning system for inverse dose optimization, ensuring maximal target coverage and organ-at-risk sparing.

Patients underwent 4D MRI at 6, 12, 24,36 and 48 months, with obliteration graded via the Angiographic Obliteration Scale. DSA, done after 4D MRI documented obliteration, served as the reference standard.

Statistical analysis included sensitivity, specificity, predictive values, Cohen’s kappa for intermodality agreement, and Wilcoxon tests for volume reduction (p<0.05 significant).

Results: At median follow-up of 32 months (range: 6–48), 4D MRI identified complete obliteration (Grade 0) in 20 patients (44.4%), Grade 1 (minimal residual flow) in 12 (26.7%), Grade 2 (partial nidus persistence) in 8 (17.8%), and Grade 3 (significant residual AVM) in 5 (11.1%). Relative to DSA, 4D MRI demonstrated 88.9% sensitivity and 93.1% specificity, with positive and negative predictive values of 91.1% and 85.8%, respectively (Cohen’s ?=0.8). Delayed obliteration occurred in 41.9% (5/12) of Grade 1 cases. Persistent shunting on DSA confirmed 84.5% (11/13) of Grade 2/3 cases, prompting retreatment. Significant volume reduction observed across all grades (p=0.002).

Conclusion: 4D MRI emerges as a non-invasive, high-accuracy modality for post-radiosurgical AVM monitoring, enabling dynamic flow assessment with DSA reliance for nidus obliteration confirmation. Its capability to distinguish obliteration grades supports risk-adapted follow-up, reserving DSA for suspected residual shunting (Grades 2–3). Routine integration of 4D MRI into clinical workflows may enhance patient safety, minimize invasive procedures, and guide timely intervention decisions. Standardization of 4D MRI protocols is recommended to optimize long-term AVM management.