|
|
||||||||||
|
J Am Coll Cardiol, 2006; 47:2172-2179, doi:10.1016/j.jacc.2006.01.063
(Published online 12 May 2006). © 2006 by the American College of Cardiology Foundation |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||




,
,
,||
* San Carlos University Hospital, Madrid, Spain
University of Florida Cardiovascular Imaging Core Laboratories, Jacksonville, Florida
"Prínceps dEspanya" University Hospital, Barcelona, Spain
"Do Meixoeiro" University Hospital, Vigo, Spain
|| Clínico University Hospital, Valladolid, Spain
Manuscript received October 31, 2005; revised manuscript received December 29, 2005, accepted January 16, 2006.
* Reprint requests and correspondence: Dr. Manel Sabaté, Hospital Clínico San Carlos, Plaza Cristo Rey s/n, 28040 Madrid, Spain. (Email: manelsabate1{at}telefonica.net).
| Abstract |
|---|
|
|
|---|
BACKGROUND: The Diabetes and Sirolimus-Eluting Stent (DIABETES) trial is a prospective, multicenter, randomized, controlled trial aimed at demonstrating the efficacy of the SES compared with BMS in diabetic patients.
METHODS: Serial intravascular ultrasound analyses were performed in 140 lesions (SES = 75; BMS = 65) immediately after stent implantation and at nine-month follow-up. Vessel, luminal, and stent mean areas and volumes were evaluated at both edges and within the stented segment. Qualitative assessment of residual dissections and stent apposition were also performed.
RESULTS: Baseline clinical and angiographic characteristics were similar between groups. At 9 months, in-stent neointimal hyperplasia (NIH) mean area and volume were significantly reduced in the SES group (median NIH area 0.01 mm2 [0.0 to 0.1] vs. 2.0 mm2 [1.0 to 2.9] and median NIH volume 0.11 mm3 [0 to 2.1] vs. 35.3 mm3 [16.6 to 62.6]; both p < 0.0001). In the SES group, stent edges evidenced significant increase in lumen dimensions mainly due to significant increase in vessel volume, whereas those of the BMS group presented vessel shrinkage leading to significant lumen reduction. Late acquired incomplete stent apposition was observed in 11 lesions (14.7%) in the SES group and 0 in the BMS group (p = 0.001). At one year, no stent thromboses occurred in malapposed stents.
CONCLUSIONS: The SES implantation effectively inhibits NIH in diabetic patients. The antirestenotic effect of SES is also appreciated at the stent edges. Late acquired stent malapposition is a frequent phenomenon in diabetic patients treated with SES.
| |||||||||
The aim of the present study was to evaluate the vascular effects of SES compared with BMS by means of serial three-dimensional volumetric IVUS analyses as well as to report the one-year clinical results in patients enrolled in the DIABETES trial.
| Methods |
|---|
|
|
|---|
Volumetric IVUS. The IVUS studies were performed with a 40-MHz mechanical system (Atlantis; Boston Scientific, Natick, Massachusetts). Images were acquired during a motorized pullback of the IVUS probe at a constant speed of 0.5 mm/s, after intracoronary administration of 0.1 to 0.2 mg nitroglycerin, after the procedure, and at nine-month follow-up. All IVUS procedures were recorded on super-VHS videotapes. Qualitative and quantitative volumetric IVUS analyses were performed by an independent core laboratory (University of Florida Cardiovascular Imaging Core Laboratories, Jacksonville, Florida) using a methodology previously reported (12). The analysts were blinded to treatment allocation. The stented coronary segments and both edges (5 mm adjacent to the distal and proximal stent borders) were selected for quantitative and qualitative analysis. Quantitative three-dimensional IVUS analysis was performed using a dedicated quantitative intravascular ultrasound analysis system (QIVA, Pie Medical Imaging, Maastricht, the Netherlands) (12). Lumen, stent, and external elastic membrane (EEM) areas were measured along the entire stented and edge segments. Volumes were determined from a summation of measured cross-sectional areas in all frames of the pullback region based on Simpsons rule (12).
In-stent NIH volume was calculated as stent volume minus luminal volume at nine-month follow-up. Percent NIH volume was defined as the ratio of NIH volume to the stent volume multiplied by 100. Mean in-stent NIH area was calculated as in-stent NIH volume divided by stent length. For totally occluded vessels at follow-up, which were not associated with stent thrombosis, it was estimated that the entire length of the stent was filled with NIH (13).
Incomplete stent apposition was defined as
1 strut clearly separated from the vessel wall, with evidence of blood speckling behind the stent struts without overlapping side branches (14). To define ISA, consensus between two independent analysts blinded to treatment allocation was required. On the basis of serial intravascular analyses ISA was classified into three categories: 1) resolved: ISA present at baseline but no longer at follow-up; 2) persistent: ISA present at baseline and at nine-month follow-up; and 3) late acquired: ISA not present at baseline but present at nine-month follow-up. In the segment with ISA, the lumen contour was delineated outside the stent contour. Lumen area outside of the stent was also measured (ISA area). Further, vessel, stent, and lumen mean areas and volumes were measured in both the ISA and the non-ISA segments within the stent.
Clinical follow-up. After discharge, clinical follow-up was obtained at 1, 9, and at 12 months. At 12 months, clopidogrel was withdrawn in all patients who did not receive additional revascularization during follow-up.
Statistical analysis. The sample size for this prespecified IVUS substudy was calculated for a difference of 30% in the mean in-stent NIH area between the groups (estimated for a mean in-stent NIH area of 2.2 mm2 in the BMS arm vs. 1.5 mm2 in the SES arm), with a standard deviation of 1.5, alpha error of 0.05, and beta error of 0.20 (13). Thus, a minimum of 105 lesions were prespecified for enrollment to allow for 25% attrition due to noncompliance with invasive follow-up procedures, IVUS technical failures, and suboptimal IVUS quality.
Data are presented as frequencies, mean ± 1 SD, or median (interquartile range). Following evaluation of normal distribution (Kolmogorov-Smirnov test), comparisons between BMS and SES stents were performed with two-tailed, unpaired t tests for continuous parameters, paired t tests for change from post-procedure to follow-up, or median comparisons when applicable using the median test. Categoric variables were compared by means of the chi-square test or Fisher exact test when at least 25% of values showed an expected cell frequency below 5. Significance was set at alpha of 0.05.
| Results |
|---|
|
|
|---|
Baseline demographic and angiographic characteristics of this subset of patients were comparable between SES and BMS groups (Tables 1 and 2).
This cohort of patients assessed by IVUS presented similar baseline clinical characteristics to those not evaluated by IVUS. However, from the angiographic point of view, lesions evaluated by IVUS evidenced larger vessel size and more often were located in the right coronary artery both in the SES and in the BMS group. Besides, in the BMS group, IVUS-assessed stents were less often implanted in chronic total occlusions than in those without IVUS assessment. In terms of outcomes, stents studied by IVUS presented less frequent restenosis at follow-up. The length of restenotic segment was similar between lesions assessed or not by IVUS (18.7 ± 10 mm vs. 18.2 ± 11 mm, respectively, in the BMS group; p = 0.69; 11.8 ± 7 mm vs. 12.0 ± 8 mm, respectively, in the SES group; p = 0.76). Occlusive restenoses were not interrogated by IVUS (one in the SES group and six in the BMS group). Finally, two stents in the SES group and one in the BMS group with edge effect were not studied by IVUS.
|
|
Stented segment IVUS analysis. Post-procedure IVUS parameters were similar between groups (Table 3). There was an increase in EEM area from post-procedure to follow-up in the SES group (15.8 ± 4.8 mm2 to 16.7 ± 4.3 mm2 from post-procedure to follow-up; p = 0.04). In the BMS group, mean lumen area was significantly reduced from post-procedure to follow-up (8.0 ± 2.5 mm2 to 5.3 ± 2.5 mm2; p < 0.001). The NIH median area and volume were significantly reduced in the SES group as compared to the BMS group (median NIH area 0.01 mm2 [0.0 to 0.1] vs. 2.0 mm2 [1.0 to 2.9]; median NIH volume 0.11 mm3 [0.0 to 2.1] vs. 35.3 mm3 [16.6 to 62.6]; both p < 0.0001). Figure 1 depicts the non-gaussian distribution of main IVUS volumetric outcomes between groups, with the marked difference observed between mean and median values especially in the SES group.
|
|
|
|
|
Overall, stents with persistent malapposition (n = 10) did not evidence significant changes in EEM or lumen mean areas between post-procedure and follow-up (lumen area 8.9 ± 2.7 mm2 vs. 8.8 ± 3.7 mm2; p = 0.9; EEM area 18.4 ± 5.9 mm2 vs. 18.7 ± 5.8 mm2; p = 0.8). Finally, stents with resolved ISA (n = 2) presented an increase in mean plaque area (delta change 5.9 mm2) and virtually, no modification in mean EEM area (delta change 0.6 mm2). As a result, mean lumen area was reduced at follow-up (delta change 4.3 mm2).
None of the segments assessed by IVUS evidenced coronary aneurysm. Resolved and persistent rates of edge tears were comparable between groups (SES vs. BMS: 3.9% vs. 6.5% [p = 0.7] and 2.6% vs. 0% [p = 0.5], respectively).
Clinical outcomes at one year. One-year clinical follow-up of the entire population included in the DIABETES trial was available in all patients (100%). Overall, the incidence of major adverse cardiac events was significantly lower in the SES group (Table 5).
|
| Discussion |
|---|
|
|
|---|
The NIH was similarly inhibited in both insulin and noninsulin-dependent diabetics treated with SES. The relative reduction in NIH volume was higher in the insulin-dependent group, which was likely owing to a higher proliferative response observed in the BMS group. The subanalyses of the Sirolimus-Eluting Stent in Coronary Lesions (SIRIUS) and SES-SMART trials (7,16) suggested a reduced antiproliferative effect of SES in insulin-dependent diabetics, mainly related to restenosis at the stent edges (peri-stent). Such a hypothesis was not confirmed in the overall DIABETES trial and in the present IVUS substudy. The higher degree of NIH suppression demonstrated in insulin-dependent lesions may indicate that in the absence of peri-stent restenosis (7,16), rapamycin may be even more efficacious in this subgroup which usually presents an exacerbated proliferative response to stent implantation (3).
Opposite patterns of changes in EEM were observed at the edge segments in vessels treated with SES versus BMS (Fig. 2). As a result, lumen was preserved in the edge segments of patients treated with SES. Vessel shrinkage or lack of positive remodeling has been previously observed at the edges after BMS implantation (17,18). Whether the present IVUS findings are associated with a beneficial vascular effect of sirolimus at unstented adjacent coronary segments remains to be demonstrated. Interestingly, both groups in the present study showed minimal plaque proliferation at the edges, which may be related to stent deployment techniques used in this trial. In the DIABETES trial, judicious stent deployments with special attention to cover the entire diseased segment and avoid trauma outside the stented zone, namely, geographical miss, were performed (19,20).
The incidence of late acquired ISA was markedly increased in the SES group. This phenomenon was associated with local vessel enlargement at ISA sites. Recently, in the SIRIUS trial (21), late acquired ISA was documented in 8.7% of stents evaluated with IVUS and it was mainly located at the middle portion of the stent. Conversely, in our trial late acquired ISA was mainly located at stent edges. The positive remodeling appeared to be the main cause of ISA, which was extended beyond the boundaries of the stent, leading to a beneficial vascular effect at the edges (Fig. 2). On the other hand, trials by the use of paclitaxel-eluting stent demonstrated similar rates of ISA as compared with BMS (22). Late acquired ISA may be related to localized hypersensitivity response to the polymer that remains in situ after complete drug release (23). Eventually, diabetes mellitus as a proinflammatory disease (24) may exhibit an enhanced inflammatory reaction after drug-eluting stent implantation. However, at present no data exist regarding a higher incidence of late acquired ISA after SES implantation in diabetics than in nondiabetics. Other plausible mechanistic explanations for the ISA phenomenon include wash-out of plaque or thrombus entrapped behind the stent struts after the procedure, tissue necrosis, and positive vessel remodeling associated with the lack of NIH (10). Although no thrombotic event was associated with ISA even after withdrawal of clopidogrel, a word of caution must be given to keep these patients under clinical control given the possibility of the occurrence of very late stent thrombosis (25).
Study limitations. The main limitation of this study is the potential selection bias inherent to any IVUS evaluation. Therefore, this data can not be extrapolated to the diabetic population with vessels not amenable for IVUS assessment. However, in this DIABETES IVUS substudy >70% of patients could be assessed by IVUS and no baseline clinical or angiographic differences were observed between SES and BMS groups. Also, IVUS assessment was performed in only three out of eight SES presenting with restenosis (Table 2). Therefore, we cannot extrapolate these findings to the entire cohort of restenotic SES. This is also an inherent limitation of any IVUS study, especially when one tries to avoid the undesirable risk of inducing a nonclinically driven target lesion revascularization.
| Footnotes |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
E. Camenzind, P. G. Steg, and W. Wijns Response to Letter Regarding Article, "Stent Thrombosis Late After Implantation of First-Generation Drug-Eluting Stents: A Cause for Concern" Circulation, October 16, 2007; 116(16): e390 - e390. [Full Text] [PDF] |
||||
![]() |
P. Jimenez-Quevedo, M. Sabate, D. J. Angiolillo, F. Alfonso, R. Hernandez-Antolin, M. SanMartin, J. A. Gomez-Hospital, C. Banuelos, J. Escaned, R. Moreno, et al. Long-term clinical benefit of sirolimus-eluting stent implantation in diabetic patients with de novo coronary stenoses: long-term results of the DIABETES trial Eur. Heart J., August 2, 2007; 28(16): 1946 - 1952. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. R. Holmes Jr Coming close and then pulling away Eur. Heart J., June 1, 2007; 28(11): 1275 - 1276. [Full Text] [PDF] |
||||
![]() |
E. Camenzind, P. G. Steg, and W. Wijns A Cause for Concern Circulation, March 20, 2007; 115(11): 1440 - 1455. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |