Object Micro-arteriovenous malformations (mAVMs) represent about 8%–10% of surgically treated brain AVMs, and may be an unrecognized etiology of intracerebral hemorrhages of unknown origin. The aim of our study was to assess the diagnostic value of intra-arterial cone-beam CT angiography (IA-CBCTA) relative to digital subtraction angiography (DSA) in the diagnosis, anatomic identification and localization of intracranial mAVMs. Additionally, we assessed the clinical presentation, radiological features and management of mAVMs

Material and methods We performed a retrospective review of all diagnosed mAVMs (defined as AVMs with <1 cm nidus) from our institution’s neurointerventional database from 2010–2016. Two blinded interventional neuroradiology observers scored DSA and IA-CBCTA images of mAVMs based on a qualitative scale from 0–2 (2: excellent/good visibility; 1: Poor/relevant visibility with restrictions; 0: non-diagnostic) for the following parameters: detection of arterial feeders, venous drainers and subtle nidal components. For evaluation of IA-CBCTA treatment planning efficacy, two vascular neurosurgeons reported their presumed treatment strategy at the end of DSA and IA-CBCTA of the last follow up imaging study before the definitive management, and scored the adjunctive value of IA-CBCTA according to a scale of 0–2 (2: altered treatment plan, 1: more confident treatment plan, or 0: no value). Inter-observer agreements between the two readers for qualitative image criteria and the two neurosurgeons for IA-CBCTA treatment efficacy were assessed using Kendall’s τ coefficient. Wilcoxon’s test was used to compare scores of image quality parameters, as appropriate.

Results Ten patients (6F:4M; mean age 43±19; range 10–69 years) harbored mAVMs with 9/10 patients (90%) presenting with neurological deficits and 8/10 patients presenting with intracerebral hemorrhages (80%) (Mean volume 20.3 mL, range 0.5–55.5 mL). Following image analysis, both observers assigned significantly higher scores to IA-CBCTA for overall diagnostic value (both observers: p<0.05, observer 1: p=0.004, observer 2: p=0.016). Although no significant differences were seen between DSA and IA-CBCTA scores (low relative IA-CBCTA efficacy value) when evaluating venous drainage (observer 1: p=0.157, observer 2: p=0.564), both observers assigned significantly higher scores to IA-CBCTA (high relative IA-CBCTA efficacy value) when evaluating arterial feeders (observer 1: p=0.02, observer 2: p=0.03) and the micro-nidus (observer1: p=0.006, observer 2: p=0.009). Only 1/10 patients were definitively diagnosed harboring a micro-nidus by DSA alone by both observers. There were inter-observer differences in DSA grading of the micro-nidus resulting in fair agreement (τ=0.47), but perfect agreement with IA-CBCTA (both observers graded 2 for all the micro-nidi). Both neurosurgeons agreed that integrating the IA-CBCTA data in the neuro-navigation system would allow more confident localization and microsurgical resection/stereotactic radiotherapy. The agreement between the two neurosurgeons for the treatment planning efficacy of IA-CBCTA was good (τ=0.73).

Conclusions Micro-AVMs are often occult neurovascular lesions that are diagnosed after the presentation of an intracranial hemorrhage, and/or neurological deficits. Intra-arterial CBCTA improves the diagnostic identification and anatomic delineation of mAVMs, with the potential to improve microsurgical or stereotactic radiation treatment planning.

Disclosures A. Al-Smadi: None. A. Elmokadem: None. A. Honarmand: None. A. Shaibani: None. M. Hurley: None. M. Potts: None. B. Jahromi: None. S. Anasari: None.