A New Aneurysm Occlusion Classification after the Impact of Flow Modification

SUMMARY: A new classification is proposed for cerebral aneurysms treated with any endovascular technique, for example, coiling with or without adjunctive devices, flow diversion, intrasaccular flow modifiers, or any combination of the above. Raymond-Roy Occlusion Classification is expanded with novel subgroups such as class 1 represents complete occlusion and is subdivided if a branch is integrated to, or originated from, the aneurysm sac; class 2 represents neck filling; class 3 represents incomplete occlusion with aneurysm filling as in the previous classification; and class 4 describes the immediate postoperative status after extra- or intrasaccular flow modification treatment. A new concept, “stable remodeling,” is included as class 5, which represents filling in the neck region that stays unchanged or reduced, as shown with at least 2 consecutive control angiographies, at least 6 months apart, for not <1 year, or the remodeled appearance of a dilated and/or tortuous vessel in continuation with the parent artery without sac filling.

A lthough endovascular cerebral aneurysm treatment has already been established, 1,2 there is still controversy in regard to the possibility of not always providing complete occlusion or showing recanalization. However, it has also been shown that only a small group of incomplete occlusions or aneurysm remnants have clinical relevance. 3,4 However, the Cerebral Aneurysm Rerupture After Treatment study 5 reported that the degree of aneurysm occlusion after treatment was strongly associated with risk of rerupture. Aneurysm re-treatment may or may not carry a higher risk than the stable incomplete occlusion or recanalization. 6,7 Unfortunately, the relevant information mostly originates from coiled aneurysms and, therefore, may not be generalized to the entire population of cerebral aneurysms, particularly those aneurysms treated with new devices that modify flow from inside or outside of the aneurysm sac.
Flow modifiers (FMs) (dedicated extrasaccular flow diverters, multiple stent-in-stent applications, and intrasaccular flow disrupters) have been introduced as a new concept for treatment of cerebral aneurysms; FMs cause, over time, curative reconstruction of the aneurysm neck. The reconstruction after extrasaccular FMs (ie, flow diverters) starts immediately after the construct is in place and then evolves over a period of weeks to months. 8 Intrasaccular FMs (ie, flow disrupters) may result in instantaneous occlusion after placement or may require some time for final effect. 9 The delay in aneurysm occlusion with FMs occurs consistently if the flow through the sac continues due to flow demand through a branch coming off or integrated into the sac. Hence, in most cases, the initial result immediately after treatment does not represent the ultimate goal, and reconstruction proceeds after surgery over time. 8,9 Overall, it is evident that evolution of aneurysm management necessitates development of unified terminology to describe both initial and delayed procedural efficacies, regardless of treatment technique.
In the literature, there have been several classifications published to describe the appearance of aneurysm and/or remnant filling. The Raymond-Roy Occlusion Classification, also known as the Montreal scale 10 has been the most widely used, and it classifies the results after aneurysm coiling, which can be applied immediately after the treatment as well as during the followup. 10,11 In the literature, there exist some articles about the implications on future management of aneurysm and/or neck remnants that state class 3 aneurysm remnants, according to Raymond-Roy Occlusion Classification, are more likely to be retreated than class 2 neck remnants, which are most often followed up. 12,13 The introduction of extrasaccular flow diverters created a need for different classifications, not only to describe the initial results but also to anticipate the outcome, including the risk of infrequent but severe complications of postoperative rupture. [14][15][16] However, these classifications are exclusively for the extrasaccular FMs, and, yet, none has gained common acceptance.
Subsequently, with the use of intrasaccular FMs, the control angiographic findings become even more controversial. Lubicz et al 17 described 4 patterns in the follow-up of intrasaccular FM treatment, namely, complete occlusion, filling of the proximal recess of the device, neck remnant, and aneurysm remnant. Based on their follow-up experience, the investigators suggested a different definition, that is, "adequate occlusion," which includes the first 3 patterns.
So far in the literature, none of the classifications address the entire spectrum of current endovascular aneurysm treatment, irrespective to treatment technique. To solve this problem of ambiguity and to exclude the need for the use of different classifications for different modalities, we propose a new classification system, by expanding on the widely used Raymond-Roy Occlusion Classification scale, that describes results with current and evolving treatment techniques.

MATERIALS AND METHODS
The proposed classification is presented in the Table, and the schematic drawing (Fig 1) highlights the newly added class 1 A, B, and C as well as class 5 in different case settings. Class 1 calls for complete occlusion of an aneurysm when a branch is not directly involved, and the subgroups of class 1-A, B, C-are to be used when a branch is originating directly from a sidewall aneurysm (Figs 2 and 3) or filling from a bifurcation aneurysm (Fig 4). To assign class 5 (Figs 4B and 5-7; On-line Figs 1 and 2), at least 2 angiographic controls, at least 6 months apart, and expanding for a period of not Ͻ1 year are required to demonstrate the stability. DSA is the criterion standard at the moment, although CTA may serve the purpose for the aneurysms treated with FMs. However, with the evolving technology, noninvasive angiographic imaging may replace this in the future. The remodeling concept 18 was first defined in a more restricted concept that referred to 1 as "infundibulum like" enlargement of the branch coming off the aneurysm sac after the shrinkage of the aneurysm after flow diverter treatment ( Fig 5) and 2, as a tortuous course of the branch (which is coming off the aneurysm sac) at its proximal segment after flow diverter treatment of the sac and the sac not filling (Figs 4 and  7). 19 These 2 situations, in which the branch that originates from Classification of angiographic results after endovascular treatment with any technique Classification Class 1: Complete occlusion of the aneurysm sac. When there is a branch integrated with the aneurysm sac, ie, coming off the aneurysm, at any point of the sac, further analysis is carried out with subgroups 1A: Complete occlusion with the full patency of the integrated branch 1B: Complete occlusion with the branch reduced in caliber 1C: Complete occlusion with no antegrade filling of the branch Class 2: Neck filling Class 3: Incomplete occlusion with aneurysm filling Class 4: Aneurysm filling. This class is reserved for an immediate postoperative result based on end-of-treatment DSA; after extraand/or intrasaccular flow modification treatment 4A: With contrast stagnation-contrast stagnation is referred to when there happens to be any change in the duration of the contrast stay within the aneurysm sac after treatment 4B: Without contrast stagnation Class 5: Stable remodeling with flow modification. Filling in the neck region, which stays unchanged or reduced; to be included in this group, there have to be at least 2 consecutive control angiographies, by definition, at least 6 months apart, and expanding for a period of not Ͻ1 year; exceptionally, 1 control angiography could be sufficient for definition of class 5, only in selected cases of contrast filling the branch coming off the sac, with an appearance of a different vessel course than the original, eg, tortuous or dilated, given that it is in continuation with the parent artery with no sac filling In the other situations, remodeling is called for after confirmation of the stability (as described above) when 3, the bifurcation appears enlarged after the intrasaccular FM obliterates the sac, which stays stable in the controls (Fig 6), and 4, when there is an unchanged focal bulging at the neck region (On-line Fig 2). Therefore, in such cases, the first control result would be classified as class 2 (neck remnant) by definition; then, after confirmation of stability, it can be changed to class 5 (Fig 6 and On-line Fig 2).

DISCUSSION
The healing process or so-called reconstruction of the vessel wall at the aneurysm neck after FM and/or flow diverter treatment is different from that after endosaccular coiling, 18,20-22 and aneurysms may behave differently when treated with flow modification treatments versus endosaccular coiling. 23,24 Therefore, the previous classifications, particularly Raymond-Roy Occlusion Classification, described for coiled aneurysms could not address the results of flow modification treatments. Several new classifications 14,17 have been introduced. In these previous classifications for extrasaccular flow diverter treatment, flow stagnation is defined as a determining feature. 14-16 However, the flow stagnation does not necessarily have a direct implication regarding the future or final treatment result. That is, flow stagnation within the aneurysm may or may not result in total occlusion of the aneurysm sac. On the contrary, an aneurysm that shows no or little, if any, postoperative contrast stagnation may end up in complete occlusion in the follow-up. Yet, the presumed importance of the contrast stagnation is whether or not it has any predictive value in regard to postoperative rupture of the index (treated) aneurysm. This is a controversial issue, with no proven data to date. Our classification system does not emphasize the degree of contrast filling (eg, the pattern, timing) immediately after the FM treatment, not only for these reasons but also because it may be relatively subjective or may vary in regard to technical parameters, such as contrast injection power, acquisition parameters, and so forth. Contrast stagnation is referred to in the immediate postoperative DSA findings (class 4) after FM treatment to differentiate that grade from the aneurysm filling after aneurysm packing (class 3).
In some grading scales previously described exclusively for use after flow diverter treatment, 15,16 parent artery size is taken into   FIG 2. Class 1A. A and B, Preoperative images show the ICA aneurysm in which the anterior choroidal artery (arrow) is originating from the aneurysm at the neck. C, Six-month control angiography after single Pipeline device (Covidien, Irvine, California) placement demonstrates total occlusion of the aneurysm with the anterior choroidal artery preserved (arrow). Reprinted from Saatci et al. 18   FIG 3. Class 1C. A, Preoperative angiography shows right vertebral artery aneurysm with the posterior inferior cerebellar artery originating from the sac. B, Single Pipeline device was placed, and a 6-month control angiography demonstrates complete occlusion of the aneurysm sac, along with the posterior inferior cerebellar artery. The patient was asymptomatic.
consideration. Parent artery occlusion may occur, at least in some cases, related to the inefficiency of antiaggregating medication, but, the rate of in-stent stenosis is reported to be not any higher than that of conventional stents 18,22 and in-stent stenosis is not a part of this classification.
The patency of the branch that originates from, or is integrated to, the aneurysm is assessed in this classification, which has not been included in previous classifications. The flow demand through the relevant branch is important in its patency after the treatment. [18][19][20] Although the aneurysm is getting occluded, the branch originating from the aneurysm may remain the same, reduce in caliber, or may not be filling antegrade. This classification can apply to the results of flow diverter treatment not only in sidewall aneurysms but also in bifurcation aneurysms 19 ; the series of exclusive MCA bifurcation aneurysms treated with flow diverter placement demonstrated 84% of class 1 occlusions with a variety of occlusion patterns, including branch occlusion in 12%. Another 8% showed class 5 occlusion. In addition, transition between classes is also possible (Fig 4) in the group of aneurysms with integrated branches, whether located at the side wall or the bifurcation.
This classification is also well grounded for the intrasaccular FM treatments in that it takes not only the involving branches into consideration but also the interval change in regard to stability. With these features, this classification is differentiated from the classification for intrasaccular FMs previously described. 17 Another advantage of this proposed classification is that it can be applied by single angiographic images, in several planes, and does not necessitate seeing the entire series of the angiographic run to evaluate the flow pattern within the aneurysm. Therefore, objective findings on the angiographic image are enough for classification, with no further subjective interpretation. However, rotational angiography may be a requisite, particularly in the definition of class 5 in some cases.

CONCLUSIONS
With preventing rupture as the ultimate goal of cerebral aneurysm treatment, if the treatment serves this purpose and the aneurysm is secured, then the treatment can be called "successful" and the angiographic appearance of the vascular reconstruction can be considered irrelevant. Long-term stability of remodeling is to be determined and not free of concern until that time. How- B, A 3D image from a 6-month control angiography after treatment with single Pipeline device shows remodeling of the flow with a tortuous appearance of the branch proximally at its direct continuation with the parent artery, and no sac filling. This appearance is referred to as class 5. C, An 18-month control angiography shows complete occlusion of the aneurysm with the originating branch in reduced caliber, that is, class 1B. Reprinted from Yavuz et al. 19 FIG 5. Class 5. A, Right internal carotid angiogram shows a posterior communicating artery aneurysm (the ipsilateral P1 is aplastic, not shown). B, A 2-year angiography after a single Pipeline device placement shows that the aneurysm sac is not filling and the origin of the posterior communicating artery is remodeled. Reprinted from Saatci et al. 18 ever, in complex aneurysms where "perfect" anatomic results may not be possible or at least have increased risks, then, "remodeling" may be acceptable, given that it is stable. This classification not only addresses the new concept of remodeling that emerged after the use of extra-and intrasaccular flow modifications but also covers the entire spectrum of result possibilities with any current technique.
We acknowledge that this classification should be validated in regard to the interobserver agreement to test its reproducibility. In addition, we also plan to extract a random sample of credible size among our entire aneurysm population to review the stability in long-term follow-up (eg, 5 years) to investigate the outcome of "designated" classes. This classification may then be ready to establish a clinical algorithm for follow-up, including the risk of bleeding after endovascular treatment.