Management of intracranial aneurysms that do not occlude on initial follow up after treatment with the Pipeline Embolization Device

The Pipeline embolization device (PED) for the treatment of intracranial aneurysms has been shown to have high aneurysmal occlusion rates and a low rate of complications.^1–5 Aneurysm occlusion rate after treatment with PED is reported to be as high as 95.2%, yet does not reach complete occlusion in all aneurysms. ^2,6 Several studies have evaluated predictors of failure in these aneurysms. Poor wall apposition, inadequate coverage of the aneurysm neck, older patient age, branch vessel arising from the aneurysm dome, fusiform morphology, fetal posterior communicating artery, and larger aneurysm size have all been shown to be associated with incomplete aneurysm occlusion after treatment with PED.^7–9 Although, the predictors of aneurysm incomplete occlusion has been widely studied, the fate of these aneurysms that do not go on to complete or near complete occlusion has not been. The clinician is left with decisions to continue observation, alter antiplatelet management, or place another flow diverting device. For aneurysms with incomplete occlusion after treatment with PED the risk of aneurysm rupture as well as further growth is not well reported in the literature. Therefore, the aim of this study was to analyze aneurysm occlusion status after longer term follow up with varied treatment strategies and determine the risk of aneurysmal rupture in those aneurysms with incomplete occlusion at initial follow up.

Methods or Case Description:

Study design

A prospectively collected databases of consecutive patients treated with flow diverter (FD) deployment from 2013 to 2019 at two academic US institutions was retrospectively analyzed. Following institution review board approval, the need for patient consent was waived due to the retrospective nature of the study. Aneurysms that had radiological follow-up with the first imaging performed within the first 18 months post-FD deployment were included in the analysis.

Variables and outcomes

Variables were retrospectively collected through medical charts and neuroimaging review at both institutions. These included patient demographics, aneurysm characteristics, procedural details, complications (immediate and delayed), functional outcomes (modified Rankin Scale [mRS]),  radiographic studies results and retreatments.  At both institutions occlusion status was classified as either completely occluded (100%), near completely occluded with neck remnant (90-100%), or incompletely occluded ( < 90%) at each follow-up. For final follow-up, we collected available DSA, computed scan angiography (CTA) or magnetic resonance angiography (MRA) data. 

Statistical Analysis

For the analysis, the observations were defined as patients, aneurysms, or procedures, as applicable. Categorical variables are reported as proportions. Continuous variables are reported as mean ± standard deviation (SD) or median [Interquartile range (IQR)] as appropriate according to the distribution of the data. In each group, categorical variables were compared using the Chi-square (χ2) test and continuous variables were compared using the Mann-Whitney U test. Follow-up analysis was done using time-to-event data (for which patients were censored at the time of last follow-up). Kaplan-Meier (KM) failure function estimator methodology was used for estimating the number of expected events (complete and near-complete aneurysm occlusion) during follow-up. Comparison between the KM failure function estimator for the indication for aneurysm treatment according to the 2018 FDA criteria was done with log-rank test. Finally other factors that could have impacted directly on complete occlusion rates were assessed. Statistical significance was set to a p-value of less than 0.05. All statistical analyses were performed using the Stata statistical software package 14.0 (StataCorp., College Station, TX).

Outcomes:

Baseline characteristics

A total of 134 aneurysms treated with 127 FD procedures in 125 patients were included in the analysis. These aneurysms were not completely occluded within the first 18 months after FD deployment. Baseline patient and procedural related characteristics are found in Tables 1 and 2.

For initial PED treatment, the majority of patients were female (80.0%), with a median age of 62 years old. Baseline mRS was ≤ 2 in 97.6% of cases. Platelet function testing was performed prior to the procedure in 78.7% of cases, from which 60.6% were clopidogrel responders. Patients who were clopidogrel non responders which switched to ticagrelor or prasugrel. The most common dual antiplatelet therapy was ASA 81mg and Clopidogrel 75mg daily (54.3%). Adjunctive coiling was used in 4.7% of cases, and most procedures (93.7%) used a single flow diverter.

For aneurysms that were non occluded at initial follow up, most aneurysms (85.1%) had a saccular morphology and were less than 7 mm (58.2%) in maximum diameter. Aneurysm median diameter was 5.6 mm. Aneurysms were mostly located in the anterior circulation (82.8%). The C6 internal carotid artery segment (49.3%) was the most common location. 6 % of aneurysms presented with rupture or with evidence of subarachnoid hemorrhage prior to the FD deployment . There was no evidence of intraprocedural or postprocedural aneurysm rupture in the analyzed sample. (Table 3).

Outcomes of aneurysms that did not occlude after initial follow-up

The median first radiographic follow-up occurred 9.8 months post-treatment, and all the aneurysms included in the analysis failed to occlude completely at this stage. The aneurysm status recorded was nearly completely occluded in 42.5% and incompletely occluded in 57.5% (Table 4).  On the last imaging follow-up (median of 28.2 months post-treatment), 17.9% (n=24) of the total aneurysm sample reached complete occlusion. The majority of cases that progressed to complete occlusion had a near-complete occlusion status in the first follow-up as shown in the KM failure function estimator (66.5% vs 42.6%; p < 0.001; Figure 1). The maximum complete occlusion rate was reached around 5 years post-treatment. A total of 29 (21.6%) aneurysms retreated, 28 patients were retreated with flow diversion and 1 patient was retreated with microsurgical clipping. Figure 2 shows the results after retreatment strategies for individual cases. We attempted to evaluate the effect of stopping ASA as a maneuver to aid aneurysm thrombosis and, ultimately, occlusion after the first follow-up. We identified only 13 cases in which this was performed, and so far, these aneurysms remain non-occluded after stopping all antiplatelets when it is considered safe. Eight patients (6.6%) had a poor functional outcome (mRS 3-5), and 4 patients (3.3%) died (Table 4). The breakdown of mortality included a patient who had a distal delayed ipsilateral intraparenchymal hemorrhage which may be related to the FD placement. The other three cases passed due to other reasons not related to the FD placement (large right MCA stroke contralateral to treated side, metastatic cancer, and pneumonia with sepsis).

Delayed complete occluders and factors assessed that can impact complete occlusion rates.

The median time to complete occlusion after treatment for delayed complete occluders was 35.6 months (rate 24.5 – 47.0).  These patients tended to have a younger age (57 vs 64 years old, p=0.05) and had a significantly higher rate of fusiform aneurysms compared to patients who had complete occlusion (29.2% vs 11.8%, p=0.03). There was no statistical differences in characteristics of smoking status, diabetes, the presence of an aneurysm daughter sac, maximal diameter, branch arising from the aneurysm (p=0.84, p=0.56, p=0.7, p=0.97, and  p=0.22, respectively).

Discussion:

The Pipeline for Uncoilable or Failed Aneurysms (PUFS) trial was the initial trial in the United States that allowed the PED to gain FDA approval in 2011. This prospective, multicenter, single-arm trial was initiated in 2008 and concluded in 2014. It included 109 large and giant (≥10 mm) wide-necked aneurysms of the ICA proximal to the posterior communicating segment. Results were made available to the public in three papers that described the early, 3-year, and 5-year follow-up results. ^1,2,10 Complete occlusion rates were 86.8% at 1 year, 93.4% at 3 years, and 95.2% at 5 years. Low thromboembolic serious complication (5.6%) and retreatment (5.7%) rates were evidenced in this study. Further real-world retrospective case series report occlusion rates at 6 and 12 months to be 52–93.3% and 86.8–95%, respectively. ^3,4,11–15 Although, the PED results in high aneurysmal occlusion rates, not all aneurysms occlude over time as evidenced by multiple prospective and retrospective studies. Treatment decision making of these incompletely occluded aneurysms after PED is varied with management including retreatment with additional flow diverter, surgical clipping of the residual aneurysm, antiplatelet regimen change, and continued observation with various angiographic follow up intervals. Salem et al. performed a multicenter, retrospective analysis of aneurysms treated with initial PED followed by retreatment with a second PED after at least 6 months in patients with incomplete occlusion.¹⁶ In their study, retreatment with a second PED showed complete occlusion in 46.2% and near-complete occlusion (90%–99%) in 20.4% of aneurysms.¹⁶ In our current study, 29 aneurysms were retreated with 28 aneurysms retreated with flow diversion and 1 aneurysm retreated with microsurgical clipping. Aneurysm retreatment with flow diversion resulted in 50% complete or near complete occlusion rates and retreatment with microsurgical clipping resulted in 100% complete occlusion rate. Thus further demonstrating retreatment of incompletely occluded aneurysms after PED as a safe and efficacious treatment strategy.

The PUFS trial and several other studies have shown that PED is an endoluminal aneurysm treatment device that only continues to improve over time in terms of aneurysmal occlusion rates. ^{2,3,5,17–19} However, imaging follow up after initial treatment with PED is varied, one study by Gupta et al. aimed to optimize angiographic follow up. ¹¹The authors analyzed 259 aneurysms treated with PED at least 1 follow-up imaging session to assess aneurysm occlusion status and proposed a follow up diagnostic cerebral angiogram at 12 months followed by MRA at 24 months in patients younger than 70 years and a single diagnostic cerebral angiogram at 12 months in patients older than 70 years, due to the trend toward decreased aneurysm occlusion over time in patients over 70 years.¹¹ In our current study, median last follow up was 28.2 months (IQR 13.8 - 44.3) with final complete or near complete occlusion rate 90.4%. As angiographic follow up data increases in longevity, aneurysm occlusion rates after treatment with PED will likely only increase over time, as multiple studies have shown increased aneurysm occlusion rates over longer term follow up. ^{2,3,5,17, 19}

Incompletely occluded aneurysms or residual aneurysms after endovascular treatment pose a potential risk of delayed aneurysmal rupture. The risk of rupture of incompletely occluded aneurysms after treatment with flow diversion has only rarely been reported in the literature, with one such study showing no increased risk of rupture.¹¹ Our study evaluated 606 aneurysms treated over 9 years at two large academic institutions with incomplete aneurysm occlusion on initial follow up in 134 aneurysms. On further long term follow up, none of these aneurysms ruptured even in those aneurysms with continued non occlusion.

Conclusion:

The pipeline embolization device is a safe and efficacious device, however there are still aneurysms that fail to completely occlude. In this study we analyzed treatment paradigms that lead to further increased aneurysmal occlusion as well as risk of rupture in incompletely occluded aneurysms. Retreatment with additional flow diversion as well as longer imaging follow up leads to higher aneurysm occlusion rates. Furthermore in our study, aneurysms with incomplete occlusion at initial follow up after treatment with PED did not rupture over further follow up.