In the present study, WEB embolization of SSAs was feasible in 94%, with 19% experiencing deployment problems and 4% requiring stent support. The overall complication rate was 6%, 2% of which were symptomatic. The adequate occlusion rate increased from 90% at 6 months to 96% at long-term follow-up, and the recanalization rate was 9%. From a technical and procedural point of view endovascular treatment of SSAs still remains a highly challenging and cumbersome approach. To our knowledge, this is the first study focused on WEB embolization of SSAs using the 2 mm height variants.

There are few clinical studies of the WEB 17 in unselected cohorts of small aneurysms. The technical success rates and the need for additional stents were 98% and 10% in the study by Maurer et al., 95% and 9% in the study by Pagano et al. and 100% and 4% in the prospective CLEVER study (CLinical EValuation of WEB 17 device in intracranial aneuRysms), respectively [7,8,9]. Hence, the feasibility rates in these studies were slightly better than in the present study (94%), while a similar subset of patients required an additional stent (4%).

However, it should be noted that the present study focused on SSAs characterized by a small aneurysm height (3.4 mm) and a low aspect ratio (1.3). In the previous WEB 17 studies, these parameters were larger, with a height of 4.5 mm in the study by Maurer et al. and a maximum diameter of 5.1 mm in the CLEVER study [7, 9].

Due to the flat shape of the aneurysm, WEB protrusion into the parent artery was the most common technical problem, occurring in 12 cases (12%). This rate is higher than in the WEBCAST/WEBCAST-2/French Observatory benchmark studies where protrusion was observed in only 4% of cases [10]. Despite the shallow aneurysm morphology, the principle of WEB oversizing was consistently applied, as represented by a WEB/dome ratio of 1.2. In the univariate analysis, WEB protrusion was associated with larger dome width and smaller WEB/dome ratio. This association seems unlogical and cannot be explained by the authors, as WEB protrusion would be expected with smaller aneurysm heights and more pronounced oversizing (WEB/dome ratio).

Several strategies to address WEB protrusion have been identified in this study. Post-deployment repositioning of the WEB without changing to a different WEB size is straightforward if the WEB was deployed too low in the aneurysm and there is free space in the upper aneurysm dome. Changing to a smaller WEB size was also efficient when the originally selected WEB size proved to be too large, even when selected according to the manufacturer’s sizing chart. In the present study, selecting a smaller size resolved WEB protrusion in 3 cases. An established method to counteract WEB protrusion is additional stent implantation [11]. However, this should only be used as a salvage option, as stenting nullifies the advantages of the WEB as an intrasaccular device because it requires antiplatelet medication and is associated with an increased risk of procedural events [11]. The same applies to a switch to another treatment modality, particularly if the aneurysm requires stenting or clipping.

Overall, 9% of aneurysms could not be treated with WEB only as originally planned, with either stenting or clipping required in these cases. Although the rate of additional stent implantation did not exceed that of other WEB studies, we recommend that SSAs be treated with dual antiplatelet therapy to allow for additional stent implantation or conversion to another stent-assisted modality, if required [10].

The clinically relevant issues of WEB protrusion are the risk of apposition thrombus formation at the protruding WEB portion with occlusion or distal embolization and narrowing of the parent artery [6]. In the present study, procedural thromboembolic complications occurred in 4%, of which 3% were due to WEB protrusion, but none were symptomatic. The association between WEB protrusion and thromboembolic complications has been shown in the univariate analysis. Previous WEB 17 studies reported comparable rates of thromboembolic events with 4% (1% symptomatic) in the study by Maurer et al., 7% in the study by Pagano et al., and 6% (2% symptomatic) in the CLEVER study [7,8,9].

For endovascular coiling, treatment of small aneurysms is associated with an increased procedural risk of aneurysm perforation. In this context, the procedural rupture rate was 8% for small aneurysms ≤ 3 mm versus 4% for larger aneurysms in the study by van Rooij et al.[12] Similarly, in WEB procedures, there may be an increased risk of aneurysm perforation with the tip of the folded WEB device due to the shallow anatomy. To avoid this complication, the WEB must be opened very proximal, approximately at the level of the neck, and after partial unfolding, pushed distally into the aneurysm. Iatrogenic aneurysm perforation with the WEB tip occurred in 1 case in the present series. This complication could be managed by leaving the deployed WEB into the aneurysm to provide stasis and additional temporary balloon occlusion of the parent artery. Another hemorrhagic complication occurred during probing the aneurysm with the microwire. For comparison, hemorrhagic events with WEB 17 occurred in 0.8% in the study by Maurer et al., 0.9% in the study by Pagano et al., and 0.6% in the CLEVER study [7,8,9].

After endovascular treatment, small aneurysms usually have better occlusion results than large aneurysms, and large aneurysm size is a risk factor for recanalization. In this context, van Rooij et al. reported lower retreatment rates after coiling of small aneurysms (5%) compared to large aneurysms (10%).[12] A wider neck and larger aneurysm size also correlate with incomplete occlusion after WEB treatment, as larger aneurysms require more time for complete intrasaccular thrombosis to occur [13, 14]. In this context, studies using the WEB 17 for small aneurysms < 7 mm report higher occlusion rates than WEB studies on larger sizes. Complete and adequate occlusion rates were 71% and 93% in the WEB 17 study by Maurer et al. and 68% and 86% in the WEB 17 study by Pagano et al.[8, 9] In the present study of SSA, the 12-month angiographic results showed complete occlusion in 76% and adequate occlusion in 88%, confirming the high efficacy of WEB embolization for small aneurysms. Notably, there were no major recanalizations leading to aneurysm remnants beyond the 6-month follow-up. In contrast, complete and adequate occlusion rates of 54% and 85% were reported in the WEB-IT study, which included aneurysms with a mean diameter of 6.4 mm treated by diverging WEB types [15].

For selected aneurysms, other endovascular techniques might a viable approach for small aneurysms. For coiling, van Rooij et al. reported feasibility in 196 aneurysms ≤ 3 mm with a procedural morbidity rate of 2% and a high 6-month adequate occlusion rate of 94%.[12] Gao et al. reported complete and adequate occlusion in 84% and 93% after coiling and 89% and 100% after SAC with a low-profile visualized intraluminal support (LVIS) device (Microvention) and low morbidity in both groups [16]. Zheng et al. reported 93% complete occlusion and 8% complication rate for SAC with various stents [17]. Flow diversion is also a viable option for small sidewall aneurysms, even though there are no specific series for this aneurysm subset.

However, the studies cited seem to focus on sidewall aneurysms. In particular, coiling of bifurcation aneurysms is often not feasible due to a wide neck, and stent-assisted coiling or flow diversion may be complicated by thromboembolic events [3]. In contrast, the proportion of bifurcation aneurysms in the present study was 65% and almost all aneurysms were wide-necked, which represents a challenging morphology for conventional endovascular treatment. Therefore, some of the aneurysms presented in this series might have been treated by clipping if the WEB had not been available. However, the favorable results of the present study support the WEB as a promising endovascular option for SSAs, including wide-necked bifurcation aneurysms.

This study has several limitations. The results are based on a retrospective analysis with a moderate number of patients. Short- and medium-term follow-up is incomplete, and long-term angiographic follow-up information is lacking due to current unavailability. There is no control group, which limits generalizability. Finally, the lack of core laboratory evaluation raises the possibility of confounding the reported angiographic and clinical results.