Stretch-Resistant Coils for Intracranial Aneurysms: One Step Forward or Two Steps Back?

Endovascular treatment of cerebral aneurysms was boosted by the introduction of the Guglielmi detachable coil system (Boston Scientific, Natick, Massachusetts) in 1991. The concept of “detachability” made the selective placement of long coils into the lumen of an aneurysm much safer: Inserted coils could be repositioned when needed or even withdrawn and replaced by another coil. When the position of the coil in the aneurysm was satisfactory, the coil could be electrolytically detached. Occasionally, operators experienced unintended unraveling of the primary coil winding, on withdrawal or retrieval of the coil. This unraveling of the proximal part of the coil could occur when the distal part was stuck inside the mesh of previously inserted coils. An unraveled coil cannot be repositioned, and further withdrawal either leads to the removal of the remaining part of the coil or to coil fracture, resulting in thrombogenic coil material left in normal cerebropetal vessels.

To solve this technical problem of coil unraveling and fracture, manufacturers built in a filament centrally in the primary winding of the coil. The filament was made of nitinol, polyglycolic acid, or polypropylene and was attached to the proximal and distal ends of the coil. These coils were called stretch-resistant (SR) coils. In these SR coils, the force of withdrawal is transmitted by the inner filament and not by the winded coil wire itself. The concept of stretch resistance appealed to many operators and manufacturers, and all manufacturers currently have ranges of SR coils.

However, later it was suggested from clinical and experimental studies^1,2 that the SR filament had a negative influence on the physical properties of the coil, such as coil softness, shape memory, and flexibility. In the experimental study of Miyachi et al,² various types of SR coils caused hardening and straightening of the last few millimeters of the coil. The straightening phenomenon was due to relative SR line shortening and subsequent condensation of pitches of the first loops at the coil end. Coil tail flexibility was lost, and the SR coil for the last part behaved like a stiff wire. This straightening of the last few centimeters of the coil caused catheter kickback and thus progressive difficulty in inserting the final part of the coil. This technical issue was specific to SR coils and did not occur with standard coils.

When the last part of the coil is straightened and cannot be placed inside the aneurysm, the coil has to be withdrawn. With more manipulation, the risk of complications increases. In addition, the handling drawback of SR coils may also result in placement of fewer coils in comparison with standard coils and thus in lower packing attenuations and possibly more recurrences at follow-up. To test the hypothesis of lower packing attenuations obtained with SR coils by impaired handling, we compared the packing attenuations of 74 aneurysms treated with newly introduced SR Galaxy coils (Codman & Shurtleff, Raynham, Massachusetts) with those of 74 volume-matched aneurysms treated with standard Trufill/Orbit coils (Codman & Shurtleff) (Table). The recently introduced SR Galaxy coils only differ from the standard coils in the presence of the SR filament; all other properties are equal.³ The mean packing of aneurysms treated with standard coils was 29.3%, and the mean packing of aneurysms treated with new SR Galaxy coils was 25.7%. This difference of 3.6 percentage points was statistically significant (P = .0021).

The result of this comparison confirmed our personal and subjective experience in the handling properties of the 2 compared coil types. Standard coils produce less catheter kickback, are less stiff, and are easier to deliver. While oversizing of the first coil is mostly possible with standard coils, with the new SR Galaxy coils, undersizing is imperative to accommodate the first coil. The better handling properties of the standard coils, therefore, result in higher packing attenuations. In our view, the importance of packing attenuation is 3-fold: First, the relation of high packing attenuation and stable aneurysm occlusion at follow-up has been firmly established. Therefore, it is sensible to place as many coils as possible in a cerebral aneurysm. Second, packing attenuation is the only objective parameter available in comparing the handling performance of different types of coils. Finally, high packing attenuations reflect the ease of use and therefore safety: When coils can be easily and quickly placed inside an aneurysm, the procedure is effective and safe. On the other hand, when coils are difficult to place with repeated catheter kickbacks, the operator will likely, after a period of several futile attempts, withdraw the final coil resulting in lower packing attenuation and increased risk of complications due to microcatheter manipulations.

Although both experimental and clinical data indicate better handling and thus better obtained packing attenuations for the standard coils, most coils available on the market are stretch-resistant. Apparently, the fear of unintended stretching and unraveling of coils during withdrawal generally outweighs the impaired handling. In balancing the pros and cons of stretch resistance, one should know the frequency and impact of the stretching phenomenon. However, there are no data on the incidence of this technical problem and its clinical implications. In our experience, stretching and unraveling on withdrawal of the standard coils can be largely avoided. First, if friction is encountered during delivery of a coil, it is unsafe to try to force the coil into the aneurysm: The coil may be damaged or kinked. When withdrawal is then necessary, unraveling is likely to occur (forced in = stretched out). Second, if a coil has to be withdrawn, it is better to withdraw the microcatheter first for a few millimeters to align the coil with the catheter without kinking at the tip. Third, when an aneurysm is completely occluded, it is not necessary to try to force another (“last”) coil in, with subsequent risk of the need for retrieval.

With these technical precautions, unraveling of a coil during withdrawal will be rare. We believe that the drawback of possible coil stretching and unraveling in standard coils without stretch resistance is only a minor clinical issue that is outweighed by the shortcomings of the SR filament in terms of handling, safety, and obtained packing attenuation.

Standard coils are hardly available on the market any more. We plead for a renewed appreciation of the better physical properties of standard coils without SR filaments, so that operators can choose between standard or SR coils in every coil type.

• © 2014 by American Journal of Neuroradiology