Article Text
Abstract
Introduction An imaging approach identifying patients who benefit from treatment of unruptured intracranial aneurysms (UIA) is desired. We investigate a potential biomarker of UIA instability, myeloperoxidase (MPO), in human aneurysm tissue and in parallel develop a MPO molecular imaging approach in experimental models.
Materials and Methods We harvested 20 aneurysms from 17 patients during surgical clipping. Angiograms were evaluated for: 1. maximum diameter, 2. blebs, 3. surface architecture (berry/irregular), and 4. single-lobe or multilobular. The tissue samples were stained against human MPO.
MRI sequence insensitive to complex intra-aneurysmal flow was optimised in vitro. A silicone vascular replica of a rabbit elastase aneurysm was built1 and embedded in coconut oil. Flow was generated by a pulsatile pump simulating the rabbit aortic waveform1 using a blood analogue matching T1- and viscosity of blood. At 3T, we acquired motion-sensitised driven-equilibrium (MSDE) (TE=10ms, variable TR and FA) and optimised the flow velocity encoded gradient echo imaging parameter (VENC, 1–8cm/s), while using spectral pre-saturation inversion recovery (SPIR) fat suppression.
Saccular aneurysm model in white New Zealand rabbits (n=8) was created2. The animals were imaged using the MRI protocol optimised in-vitro. Naïve aneurysms were imaged before and 3 hours after injecting MPO-specific contrast agent. Animals returned to the surgical suite 1-week later for lipopolysaccharide (LPS)-induced inflammation of the aneurysm wall3. Two-days after the LPS administration, the MRI study was repeated before and after MPO-specific contrast agent infusion. The animals were euthanised and the aneurysms explanted for histology.
Results Ten human aneurysms were positive for MPO. All ruptured aneurysms (n=3) were positive for MPO. UIAs described as irregular/complex had a positive trend for MPO infiltration (p=0.087). Aneurysms were more likely to be positive for MPO in patients who had a family history of subarachnoid haemorrhage (p<0.05). Aneurysms with MPO were 39% larger. All UIAs that were negative for MPO were described as berry aneurysms.
In the phantom experiment we confirmed that the MSDE sequence with VENC of 1 cm/s and SPIR eliminated the signal from blood flow and adjacent fat, respectively, yet provided sufficient contrast to image a representative amount of the MPO-contrast.
The optimised MSDE sequence was used in the rabbit aneurysm model. Significant motion artifact required respiratory-triggering. Consequently, T1-weighting was lost. An inversion pulse was added to the sequence as to gain T1-sensitivity by inversion recovery (IR, IR delay optimised to 800ms).
As compared to the naïve aneurysm, there was a 40-fold increase in the SNR change from pre to post-contrast MSDE imaging in the inflamed aneurysm model (p<0.0001). Histologically, LPS-induced inflammation demonstrated a large infiltration of MPO within the aneurysm wall.
Conclusion Human aneurysms with associated risk factors for rupture or that have ruptured contain MPO within the aneurysm wall; suggesting that MPO could be a valuable biomarker for assessment of aneurysm propensity for rupture. A diagnostic MR imaging protocol has been optimised in vitro and applied for detection of an MPO-specific contrast agent in an animal model of aneurysms.
Disclosures M. Gounis: 1; C; NIH. I. van der Bom: None. A. Wakhloo: None. S. Zheng: None. J. Weaver: None. A. Puri: None. A. Kuhn: None. A. Bogdanov: 1; C; NIH.
References
-
Seong, et al. Biorheology 2005;42:345–361
-
Cloft, et al. Radiology 1999;213:223–228
-
DeLeo, et al. Radiology 2009;252:696–703