Edge Restenosis After Implantation of High Activity 32P Radioactive β-Emitting Stents
To the Editor:
We compliment Albiero et al1 on their innovative attempts to reduce the incidence of edge restenosis in association with radioactive stent implantation. Their lack of success, however, is simply explained by the basic physics of a radioactive line source (RLS). Along the long axis of a RLS, the prescribed isodose extends beyond both source ends for a variable distance. However, what is of real interest is the length of the prescription isodose (in relation to the length of the RLS) at the prescribed distance from the RLS. This is the length that is really available to us for treating the “injured segment” of the artery. This length is always shorter than the RLS because, as we move away from the RLS (in a radial direction), the isodoses “bow inward” (ie, they curve toward the RLS as one approaches the ends). Thus, if the RLS length was selected to be exactly equal to the length of the vessel that we wanted to treat to a given dose (as is the case with the radioactive stent), the ends of that vessel length would always be underdosed. This can be understood if we consider the following: at the treatment distance (ie, at the level of the prescription isodose, say 2 mm), points in the central portion of the arterial segment receive radiation not only from the source in direct line with that point, but also from the 2 to 3 sources on either side of that source. It is easy to appreciate that as one moves to the end of the “injured arterial segment” (at the prescription distance), the terminal points can receive radiation only from the source directly in line with that point, which translates to a lower dose compared with the central points and a consequent curving inward of the isodose lines.2
If we examine their data (Figure 3) closely, we find that the neointima begins to increase just within the ends of the stent, and this increase extends for the first millimeter or so on either side. This would be in keeping with an underdosing of the 2 ends. However, the predominance of negative remodeling that they observed beyond the first millimeter on either side of the stent probably reflects a restenotic process occurring in the nonstented segment of the artery, which has sustained an occult injury, either due to the predilatation balloon or due to the extension of medial fractures (which may have originated in the stented segment). The very effective inhibition of neointimal proliferation within the stent results in the minimal luminal diameter shifting to these edges, with the appearance of edge restenosis.
An understanding of these concepts also makes it clear why the approach of the “cold-ends stent” is probably not logical. This model would be associated with arterial wall “injury” at either end of the stent, with no radiation at the ends to counteract the neointimal proliferation. However, the “hot-ends stent” approach merits further study, if engineering considerations allow us to produce such a stent.
- Copyright © 2001 by American Heart Association
We appreciate the comments of Professors Parikh and Noti regarding our study.R1
They suggest that edge restenosis can be explained simply by the basic physics of a radioactive stent, which delivers a lower dose of radiation at the ends compared with the central part of the stent. This is the result of the shorter length of the prescription isodose compared with the length of the radioactive stent. In our first articleR2 on radioactive stents with an initial activity up to 12 μCi, we explained the occurrence of edge restenosis in a similar way, as “the result of a low dose of radiation at the stent edges, due to a sharp decline of dose rate within millimeters from the stent margins.”R2 This conclusion was based on the dosimetry of a 32P stent, as reported by Janicki et al.R3 The 2D plot of the dose rate along the stent obtained using this model shows that the nonuniformity of dosing, reflective of the stent geometry, decreases at distances 1 to 2 mm from the stent surface. At these distances, the dose delivered is more uniform (isodose) along the length of the stent. However, if we look carefully at the plot, the isodose line starts to decrease ≈1 mm before the end of the stent. Therefore, in the case of a radioactive stent, the length of the prescription isodose is always shorter than the stent length.
Professors Parikh and Noti interestingly comment on the pattern of neointima formation along the stent length, observing that the neointima begins to increase within the ends of the stent, probably as a result of underdosing the 2 ends.
The rationale for using a “cold-ends stent” is to prevent negative remodeling at the edges. This concept is supported by the results of our study,R1 in which increasing the initial stent activity level and limiting the balloon-induced injury outside the stent resulted in a reduction of edge restenosis related to plaque growth but not of that related to negative remodeling. In addition, despite the technical complexity of creating a radioactive stent with a higher activity at the ends, this task has been successfully accomplished. In Europe (Milan and Rotterdam), 56 lesions have already been treated by implanting a “hot-ends stent.” We will know shortly if this approach proves effective in diminishing edge restenosis.
Albiero R, Nishida T, Adamian M, et al. Edge restenosis after implantation of high activity 32P radioactive β-emitting stents. Circulation. 2000;101:2454–2457.
Albiero R, Adamian M, Kobayashi N, et al. Short- and intermediate-term results of 32P radioactive β-emitting stent implantation in patients with coronary artery disease: The Milan Dose-Response Study. Circulation. 2000;101:18–26.