Long-Term Follow-Up of 1036 Cerebral Aneurysms Treated by Bare Coils: A Multicentric Cohort Treated between 1998 and 2003

Retreatment

Late retreatment is an objective end point when comparing this series to other published reports. Rates of aneurysm recurrence after coiling have been provided in various single and multicenter studies and ranged from 4.5% to 33%.^{5–9,13,16,17,19,21,22} In our series, 62 recanalized aneurysms were treated several times, for a total of 72 new procedures. Most, 66%, were retreated during the first 18 months posttreatment. Considering all procedures, we found the rate of retreatment was 6% in this cohort of patients, with only 6.6% of the total treated population having been lost to follow-up. We reviewed 402 aneurysms at 5 years posttreatment, which represents 44% of the initial population of treated patients. It is very important to follow all patients regularly to better assess the time posttreatment at which an aneurysm is most susceptible to reopening. Most morphologic losses or recurrences are early, within the first year after treatment.¹⁶ In our series, >50% of the aneurysm retreatments were performed within the first year, and 65%, within the 18 first months.

In the publication of Campi et al,²¹ which compares frequency, timing, and consequences of aneurysm recurrence in the ISAT population, late retreatment was 6.9 times more likely after endovascular treatment. The mean interval to retreatment was 20.7 months, with a median of 13.7 months, which is similar to that in our series. We agree that after the first year, a proportion of aneurysm remnants remains stable and is unlikely to grow and rebleed.^16,21 We also found a higher frequency of recurrences in the posterior communicating artery and the anterior communicating artery aneurysms, which is similar to the findings of Campi et al.²¹ These aneurysms are more proximal to the supra-aortic vessel and, therefore, change with variation of vascular pressure.

Angiographic and MRA follow-up are necessary during the first 2 years to detect and retreat early recanalization.^9–11 After 5 years, we found de novo aneurysms in the same places as the initially treated aneurysms. Holmin et al,¹⁶ who published the results of a monocentric study with long-term clinical outcome data and a 4- to 5-year angiographic follow-up in 466 aneurysms, found 3 de novo aneurysms and performed 90 retreatments, representing 15.2% of all procedures. They found that a stable morphologic result during the first-year interval is a good predictor of a low risk of recurrence, and we agree with this hypothesis. Long-term follow-up is necessary to identify the presence of new aneurysms and late recurrences. There appears to be no safe time limit for follow-up patients treated endovascularly.²¹

With regard to new aneurysms, it is very difficult to determine if they are true de novo aneurysms or a late recurrence. How does one explain this growth in the same location but not observed until 6–7 years posttreatment? Is the de novo a real aneurysm or simply a late compaction of coils due to modification of pressure? Late recurrences do occur, and this may be due to initial bad visualization of the neck of the aneurysms or due to late modification of the circulation and pressure secondary to vessel occlusion. Sometimes, it is a new aneurysm near the location of the initially treated aneurysm with an intact wall between. The same observation was made in a recent surgical study.²³ Sprengers et al^11,20 found an incidence of de novo aneurysms in 1.5% of patients after 5 years. De novo aneurysms are extremely rare, but we should continue to follow all patients to proactively detect these malformations. It is certainly difficult to propose only observational follow-up without action, especially in patients with an initial history of subarachnoid hemorrhage. There is also a psychological aspect of follow-up to be considered, and both patients and their physicians value the importance of regular consultations during the first years after hemorrhage.^12,19,20,23

With regard to the device material used, these results are interesting because at the time of our study, only bare platinum coils were used without intracranial stents or modified coils. In 10% of all procedures, however, we did use the balloon remodeling technique. Stability of occlusion was obtained in 75% of all aneurysms, and spontaneous amelioration was identified in 11% of all aneurysms. With this classic material, we obtained very good occlusion results. More than 10 years after the introduction of endovascular treatment, safety and satisfactory occlusion with bare platinum coils have been established. It is, however, important to develop new technology to investigate whether outcomes can be improved, but bare coils are used as a reference to treat standard aneurysms. New technologies such as modified coils and/or intracranial stents could be necessary for the 10%–15% of aneurysms that recanalize easily and for very large aneurysms. Results of these new devices, however, are sometimes disappointing, and it remains to be proved that modified coils provide better results.^24–26

One of the limitations in our study, which began in 1998, is the absence of an independent core laboratory to analyze our results. However, in each of the 5 participating centers in our study, the physician who collected and reviewed all of the aneurysms was independent of the treating physician. We did lose access to some patients with time, which resulted in a loss of available follow-up in 6.6% of all patients treated due to patient death or other medical reasons, patients moving, or failure to respond to requests for clinical and imaging follow-up. In a multicentric study such as ours, with such a large patient population, it is difficult to achieve compliance of follow-up in 974 patients, despite considerable effort to re-contact missing patients. Finally, we believe that we had a good rate of follow-up, and our results were objective.

In comparison with other series, to our knowledge, there are few with an equivalent extent of long-term follow-up; however, some may not yet be published. Our results are comparable with the publication of Holmin et al.¹⁶ Patient selection was similar and their follow-up was comparable, yet with more retreatments overall. In our country, EVT has existed since 1994, yet at the beginning, there were only a few select centers that performed this treatment. In 1998, before the published results of ISAT, we continued to develop this new technique and primarily treated ruptured aneurysms. It was more difficult to treat unruptured aneurysms because neurosurgeons preferred to operate on these patients, and open surgery was still very much the accepted mode of treatment for this type of aneurysm and for middle cerebral artery aneurysms. In 2009, a far broader scope of aneurysms is endovascularly treated and only very difficult aneurysms are surgically treated. EVT is considered safe, and the development of other new devices is improving the feasibility of EVT. We believe that all physicians agree with the importance of long-term follow-up of their patients and the consideration of follow-up beyond 5 years is warranted.