Usefulness of Non–Contrast-Enhanced MR Angiography Using a Silent Scan for Follow-Up after Y-Configuration Stent-Assisted Coil Embolization for Basilar Tip Aneurysms

Discussion

The Y-configuration stent-assisted coil embolization technique has been used for wide-neck bifurcation aneurysms such as those of the tip of the BA. In the present study, we evaluated the usefulness of silent MRA compared with that of 3D TOF-MRA for follow-up after Y-configuration stent-assisted coil embolization for basilar tip aneurysms. It has been reported that 3D TOF-MRA and CE-MRA have been used after single stent–assisted coil embolization.^{12,13,15⇓–17} Irie et al²⁰ reported follow-up after stent-assisted coil embolization by using silent MRA, and Cho et al¹² reported that the degree of agreement and correlation between 3D TOF-MRA and 3D rotational angiography was high. With respect to the status of neck remnants, the sensitivity was 80% (4/5 cases). It is thought that magnetic susceptibility artifacts and radiofrequency shielding were decreased because of the very small voxel size (reconstructed voxel size, 0.24/0.24/0.7 mm). Cho et al¹³ reported comparisons between the 3D TOF-MRA and conventional angiography, showing that the MIP plus source images had almost perfect agreement (κ = 0.892), with better agreement than with the MIP images alone (κ = 0.598). It is difficult, through the evaluation of source images, to assess the residual flow that is protruding in the direction of the body axis after stent-assisted coil embolization. To confirm the residual flow, MPR with an accurate image angle should be obtained. Irie et al²⁰ showed the usefulness of the silent MRA for stent-assisted coil embolization rather than a comparison between silent MRA and 3D TOF-MRA. Moreover, the visualization of neck remnants was good.

CE-MRA has a better depiction of the flow in a stent than 3D TOF-MRA, and it is reported that CE-MRA has a visualization ability similar to DSA.¹⁵ Previous reports have examined 3D TOF-MRA and CE-MRA for stent-assisted coil embolization by using 1 stent; however, a Y-configuration stent-assisted coil embolization case was not examined. Moreover, there have been no studies on the use of MRA for follow-up after Y-configuration stent-assisted coil embolization. The Y stent has a strong influence on magnetic susceptibility artifacts and radiofrequency shielding effects, particularly at the cross point of the 2 stents. It is thought that the depiction of the flow in Y stents is difficult in 3D TOF-MRA.

On the other hand, CE-MRA has certain problems such as nephrogenic systemic fibrosis, gadolinium accumulation, and anaphylactic shock, and it is not necessarily suitable for repeated examination. Therefore, we examined the usefulness of silent MRA in the present study. Silent MRA is a non–contrast-enhanced MRA technique as is 3D TOF-MRA. In the present study, the visualization ability of silent MRA was higher than that of 3D TOF-MRA for Y-configuration stent-assisted coil embolization. The reason silent MRA is superior in the depiction of the flow in a stent is the data acquisition by UTE. The TE of silent MRA was 0.016 ms and that of 3D TOF-MRA was 2.9 ms. Gönner et al²¹ reported that 3D TOF-MRA using UTE reduced the artifacts after coil or stent placement; however, a TE of 2.4 ms was used in that study. In addition, Yamada et al²² reported that a TE of 1.54–1.60 ms was used in their 3D TOF-MRA study. In silent MRA, UTE can minimize the phase dispersion of the labeled blood flow signal in the voxel and decrease magnetic susceptibility artifacts, and accordingly, the artifacts from stents or coils are decreased.

The Y stent deploys 2 stents from the BA to the bilateral PCA. Moreover, the PCA is positioned almost perpendicular to the BA by stent deployment. Therefore, the PCA has weak inflow effect parallel to a section direction. The visualization ability of the PCA decreased because 3D TOF-MRA was used inflow effect. The limitation of 3D TOF-MRA is spin dephasing by slow flow, turbulent flow, and horizontal directional flow. Moreover, the visualization ability of the BA trunk is worse because of the susceptibility artifacts with 2 stents.

On the other hand, the visualization ability of the BA and PCA is good in silent MRA. The arterial spin-labeling technique is used as a preparation pulse for the visualization of the blood flow. Therefore, visualization of the blood flow does not depend on the flow direction, such as with 3D TOF-MRA in the silent MRA.

In the series, we experienced an interesting case (Fig 1) that used both a Neuroform and Enterprise stent. The Neuroform stent was deployed from the BA trunk to the right PCA, and the Enterprise stent was deployed from the BA trunk to the left PCA. The stent sizes were 3.0 × 20 mm and 4.5 × 22 mm, respectively. In this particular case, the depiction of the distal stent edge of the right PCA showed strong signal loss in both MRAs. Neuroform and Enterprise stents are nitinol stents.^12,16,17,23 The Neuroform stent has an open-cell design (ie, stent-strut is not interconnected). On the other hand, the Enterprise stent has a closed-cell design (ie, stent-strut is interconnected). Open cell–design stents have lower susceptibility artifacts than closed-cell designs.^16,17,23 Therefore, the Neuroform stent has a better visualization ability compared with the Enterprise stent.

The marker of Neuroform comprises stainless steel and platinum.^15,23 On the other hand, the marker of Enterprise comprises tantalum.^12,16 Agid et al¹⁵ reported a “marker band effect” that appeared in small arteries with a diameter of 2 mm or less and markers of different alloy than the rest of the stent (usually made of platinum), which creates a stronger local susceptibility artifact.¹⁵ The markers of Neuroform stents are made of platinum, which creates a strong magnetic susceptibility artifact.¹⁵

In our study, this effect appeared in the right PCA, where the Neuroform stent (marker made of stainless steel and platinum) with a 3-mm diameter (the smallest diameter in our 7 cases) was deployed. This phenomenon corresponded with that in the report by Agid et al¹⁵; it is thought that a similar phenomenon is caused in the silent MRA.

Case number 3 (Fig 3) also used both a Neuroform and Enterprise stent. The left PCA segment was depicted more clearly than the right PCA segment. The Enterprise stent was deployed in the right PCA. Choi et al²³ reported that the open-cell design and thinner strut of the Neuroform stent provide better image quality than the Enterprise stent when using 3D TOF-MRA. In the present study, silent MRA more clearly depicted the Neuroform stent than the Enterprise stent, similar to the previous study. The cell design and thickness of the stent strut are important factors in both silent MRA and 3D TOF-MRA.

The depiction of the neck remnants in silent MRA was similar to that of DSA in the present study. Cho et al¹³ and Agid et al¹⁵ reported good depiction of the neck remnants in 3D TOF-MRA and CE-MRA in a case of single stent–assisted coil embolization. In the present study, 3D TOF-MRA had poor image quality compared with that reported in previous studies.^12,13 With respect to Y-configuration stent-assisted coil embolization, the aneurysm is sandwiched between 2 stents.

Therefore, the influence of 2 stents on the susceptibility artifact was stronger. On the other hand, silent MRA uses UTE and the arterial spin-labeling technique, decreasing susceptibility artifacts and depicting the neck remnants in comparison with 3D TOF-MRA.

As a similar sequence to silent MRA, Iryo et al²⁴ reported the contrast inherent inflow-enhanced multiphase angiography, or CINEMA, sequence. The CINEMA sequence is a non–contrast-enhanced MRA using arterial spin-labeling, such as silent MRA. This is a 4D MRA sequence. They evaluated patients with AVF. It is thought that the reduction of the susceptibility artifacts can possibly decrease if we shorten TE by this sequence. However, TE of silent MRA is 0.016 ms. TE of CINEMA cannot use UTE, such as silent MRA. They reported that a TE of 2.2 ms was used. Therefore, it is thought that silent MRA has a higher visualization ability of the artery treated with intracranial stenting compared with the CINEMA sequence. On the other hand, CINEMA has high temporal resolution; as a result, it is useful for a patient needing temporal resolution, such as one with shunt lesions.

On the other hand, Koktzoglou et al²⁵ reported the pointwise encoding time reduction with radial acquisition, or PETRA, sequence. PETRA uses the UTE sequence.²⁶ They evaluated carotid stenosis by using a vascular phantom. They reported that arterial spin-labeling MRA could improve the display of hemodynamically significant carotid arterial stenosis. Silent MRA and PETRA MRA are arterial spin-labeling with a UTE MRA sequence. If we use the PETRA MRA sequence, a result similar to that of our examination may be provided.

A major limitation of the present study is the small sample size. The examination of more cases and a variety of stents, lengths, and diameters is required. In the future, we should evaluate stent-specific visualization. There is a report that CE-MRA is useful,¹⁵ and comparative evaluations with CE-MRA are also required. It is important to elucidate whether silent MRA is not inferior to CE-MRA; however, patients in the current study did not undergo CE-MRA. In the current study, silent MRA picked up all neck remnants, with 1 false-positive test result. Thus, we postulate that silent MRA is noninferior to CE-MRA, which needs to be confirmed in future studies.