The Rise of Endovascular Technique

While there were earlier efforts to introduce materials into aneurysms to induce thrombosis,^6,7 the use of entirely endovascular techniques to treat intracranial aneurysms began with initial attempts in the 1960s. These treatments included endovascular techniques to occlude parent vessels and the use of detachable balloons, cyanoacrylate, and “pushable” metallic coils to directly occlude aneurysms while preserving the parent artery.^8–14 With progressive improvement in angiographic and computer capabilities, the ground work was established for the development of modern methods of endovascular therapies.

The Guglielmi detachable coil was developed in the 1980s with the first use in a human brain aneurysm occurring in 1990.¹⁵ The crucial breakthrough that allowed this technology to flourish was the “detachable” component.¹⁶ Before this development, platinum coils could be pushed through microcatheters into intracranial aneurysms, but there was no practical way to retrieve or reposition the coils if the initial placement was not satisfactory. With the development of a reliable detachment system, the coils could be positioned, and if needed, easily retrieved or repositioned as desired, to be detached as a permanent implant only when the positioning was deemed acceptable. The clinical trial that resulted in approval from the U.S. Food and Drug Administration (FDA) ran through the early 1990s with FDA approval in 1995. In the years that have followed, increased experience and many incremental technological changes have resulted in further improvements. Multiple coil systems now exist with a wide variety of available coil characteristics and detachment systems. Although there are many coil systems, these recent developments are the direct descendent of the original Guglielmi detachable coils.

As indicated in FDA “instructions for use” package inserts, coils are approved for intracranial aneurysms that “because of their morphology, their location, or the patient’s general medical condition—are considered by the treating neurosurgical team to be:

With the early experience, however, it was soon apparent that many of the aneurysms that were difficult to treat by surgical clipping were also more difficult than other aneurysms to treat with detachable coils. This problem was most readily apparent for aneurysms that were wide necked and/or large.^17–19 Over time, this limitation shifted the dynamic away from viewing detachable coils as a complementary technology fulfilling the unmet need of treating “unclippable aneurysms” to a situation where endovascular techniques were increasingly perceived to be, and used as, a competitive or even replacement technology for surgical clipping. Posterior circulation aneurysms were particularly singled out as aneurysms to be preferentially treated by endovascular means.^17,20,21

Throughout the 1990s, the majority of literature published regarding endovascular therapy (EVT) for intracranial aneurysms consisted of single-center, retrospective, self-reported series.^22–25 End points varied widely, as would be expected with a developing treatment modality, and lack of a common reporting language made comparisons difficult. As noted previously for reports regarding surgical clipping, the question of publication bias and generalizability of results loomed large. While there was early evidence suggesting that endovascular treatment was protective against subsequent rehemorrhage,²⁶ the recognition that many aneurysms were not completely occluded left concerns regarding the potential risks of delayed recurrence, the need for retreatment, or worse, the possibility of new subarachnoid hemorrhage despite endovascular treatment.^27–29

This essential phase, however, laid the groundwork for more sophisticated future analyses. In the course of reporting their case series of unruptured intracranial aneurysms, Roy and coworkers described a simple grading system for classifying the degree of aneurysm occlusion achieved.³⁰ They went on to apply this scale to their overall series of aneurysms, noting that recurrences often occurred more than 1 year after the initial treatment, emphasizing the need for diligent long term follow-up.²⁹

During this period of the late 1990s and early 2000s, as experience and confidence with the use of endovascular techniques were growing, there were at the same time increasingly pointed questions being asked about open surgical treatment of intracranial aneurysms. In particular, the International Study of Unruptured Intracranial Aneurysms (ISUIA) raised questions about both the natural history of unruptured intracranial aneurysms and the safety of surgical clipping.³¹ In the initial report from this group, published in 1998, the authors suggested that the risk of rupture in untreated aneurysms was much lower than previously thought, and for this reason the study attracted considerable attention and called into question the value of prophylactic surgical clipping of unruptured intracranial aneurysms. However, this study was significantly biased because the aneurysms included in the study were primarily lesions that clinicians felt were better managed by conservative means. This flaw was particularly striking in the initially retrospective study.

The other aspect of the trial, however, was that it captured in a prospective fashion data from multiple centers regarding the morbidity and mortality of aneurysm treatment. Combined surgical morbidity and mortality at 1 year ranged from 13.1% to 15.7%, depending on group assignments. Age was an independent predictor of outcome. These results revealed higher rates of treatment morbidity and mortality than had been previously reported. The authors suggested that this higher rate of clipping-related morbidity may have been in part because of the use of more sensitive assessments of cognitive outcomes than typically revealed in self-reported series.

In 2003 the ISUIA group published an update of their study and noted a substantial increase in the observed rate of aneurysm rupture, with the updated rates being more consistent with those seen in previously published, well-accepted natural history studies.^32,33 The treatment-related morbidity and mortality as evaluated 1 year after treatment remained high, particularly from surgical clipping: 12.6% in patients that had not previously suffered subarachnoid hemorrhage. The morbidity and mortality for endovascular therapy in patients that had not previously experienced subarachnoid hemorrhage was lower at 9.8%, but the patients treated by the two modalities were not matched cohorts. Indeed, the patients treated endovascularly were older, had larger aneurysms, and more posterior circulation aneurysms.

Other evidence was accumulating during this time period similarly suggesting that surgical clipping morbidity and mortality were both higher than generally reported and that outcomes were more likely to be favorable in centers with larger case volumes where endovascular therapy was also available. Johnston and coworkers examined data from university hospitals in the United States looking at discharge disposition of patients having unruptured aneurysms treated either by clipping or endovascular techniques.³⁴ They found that 18.5% of surgically clipped patients died or had been discharged other than to their home but that this undesirable outcome was seen in only 10.6% of the patients treated by endovascular means. Once again, however, it is important to note that these were not matched cohorts. Johnston then looked at a statewide database for the state of California and found similar disparities between the two modalities, again with endovascular therapy having fewer poor outcomes (25% versus 10%) and with fewer in-house deaths (3.5% versus 0.5%).³⁵ A similar review of outcomes for New York State was published in 2003 by Berman and associates. Again in this analysis the hospital volume and propensity of a hospital to use endovascular therapy were both associated independently with better patient outcomes.³⁶ Barker and coworkers and Hoh and associates analyzed nationwide Medicare data regarding surgical clipping and endovascular therapy, respectively, for unruptured aneurysms. These Medicare databases covered the years 1996-2000 and similarly showed better outcomes in larger volume centers and overall fewer poor outcomes in patients treated with endovascular methods.^37,38 It is important to reiterate that these studies were not by any means comparing matched patients and that in no way did they suggest that one modality of treatment was better than the other for any individual patient. What was important about these studies was that they helped to quantify treatment-related morbidity and mortality. In addition, they drew attention to the possibility that a comprehensive approach, which included endovascular techniques, may result in overall better outcomes.