Individual assessment of rupture risk of intracranial aneurysm (IA) is still challenging, and increased knowledge of predictors for IA rupture is needed. Aneurysm wall enhancement (AWE) in post-contrast black-blood MRI, reflecting the IA wall inflammation/angiogenesis, was found to have higher prevalence in ruptured IAs (1, 2). Recently, the permeability of IA wall quantified by dynamic contrast-enhanced MRI (DCE-MRI) has been proposed to predict IA progression (3). However, the relationship of IA wall permeability with other risk factors and its value in IA rupture risk assessment are unknown. Thus, the purpose of this study is to investigate the relationship of IA wall permeability with established clinical/imaging risk factors and explore its potential in IA rupture prediction.

MR imaging: After institutional ethics approval and obtained informed consent, 32 patients with unruptured saccular IAs diagnosed by DSA were recruited. The MR imaging was performed on a 3.0T Philips scanner with a 32-channel head coil. The MRI protocol included: TOF, 3D pre- and post-contrast black blood T1W-VISTA (4) for IA wall imaging (voxel size=0.6mm isotropic), pre-contrast T1 mapping, and 3D DCE-MRI. DCE-MRI was performed for 6mins with 8.8s interval for 10 slices. A bolus of 0.1mmol/kg Gd-DTPA (Magnevist; Bayer Healthcare) was injected at the third dynamic. Other imaging parameters were: FOV=160x160mm²; spatial resolution=0.8x0.8mm²; TR/TE=3.9/2ms; slice thickness=4mm (interpolated to 2mm).

Image analysis: An experienced neuroradiologist evaluated the IA size, location, multiplicity and blebs by the DSA. And the AWE was evaluated in post-contrast VISTA images compared with the pre-contrast. For DCE-MRI, the extended Kety/Tofts model (5) was used to generate transfer constant (K^trans), which reflects IA wall permeability. ROIs were placed by another neuroradiologist blinded to patient information and AWE, on the region with the highest K^trans values adjacent to the aneurysm wall for IA wall permeability quantification and a reference region near a normal artery. Then the mean K^trans was calculated for each ROI after excluding blood contaminations (3).

Statistical analysis: IA wall permeability was compared with the reference by paired t-test. The relationship of IA wall permeability with IA size, AWE, location, multiplicity, blebs, age, gender, hypertension, and smoking were analyzed by Pearson correlation or independent t-test. The agreement between the locations of high IA wall permeability and the AWE region was reported. For the part of the population with follow-up after MR imaging, independent t-test was used to compare IA size and K^trans between the ruptured and unruptured patients.

Three patients were excluded for DCE analysis for the cavernous sinus corruption after contrast injection. Of the remaining 29 patients (16-74 years; 22 females), IA wall permeability were larger than the reference (0.0428±0.0380min^-1 vs. 0.0033±0.0017min^-1, P<0.001). As shown in Fig 1a, K^trans was weakly correlated with IA size (P<0.01). For the relationship between IA wall permeability and AWE, although higher K^trans (P=0.036) were found in patients with AWE (n=19) compared with patients without AWE (n=10), a considerable portion of patients (9 cases, 47%) with AWE has relatively low K^trans (arrow in Fig 1b). Moreover, of the 12 patients with partial AWE, the locations of high IA wall permeability and AWE mismatched in 4 cases (33.3%). There were no significant associations found between K^trans and other clinical/imaging markers (P>0.05). In this population, 9 patients received conservative treatments and were followed for 6-10 months, of which 2 patients died from aneurysmal subarachnoid hemorrhage. Fig. 3 shows the K^trans maps and the post-contrast VISTA images of the 2 ruptured cases. One ruptured patient had AWE (Fig. 2b), while the other had no obvious AWE (Fig. 2c). The IA size was also not significantly different between the ruptured and unruptured patients (P=0.607, Fig. 3a). But, the K^trans of the 2 ruptured patients were significantly larger than the 7 unruptured cases (P<0.01, Fig. 3b).In this study, K^trans near the IA wall was found to be higher than normal arteries, suggesting the permeability of IA wall was elevated. Among all the tested clinical/imaging markers, IA wall permeability was weakly correlated with IA size, and was associated with AWE. However, there were patients whose AWE and high IA wall permeability disagreed in both K^trans values and locations. The weak correlation with IA size and disagreement with AWE indicate that IA wall permeability may provide additional information. More importantly, rather than IA size and AWE, IA wall permeability was significantly different between the ruptured and unruptured IAs in follow-up patients, suggesting that aneurysm wall permeability may be an independent risk factor for IA rupture risk prediction.