Author + information
- Received December 18, 1998
- Revision received October 7, 1999
- Accepted January 14, 2000
- Published online May 1, 2000.
- Joachim Schofer, MDa,*,
- Michael Schlüter, PhDa,
- Thomas Rau, MDa,
- Falk Hammer, MDa,
- Natalie Haaga and
- Detlef G Mathey, MD, FACCa
- ↵*Reprint requests and correspondence: Dr. Joachim Schofer, Othmarscher Kirchenweg 168, 22763 Hamburg, Germany
This retrospective study was designed to determine the six-month angiographic outcome after stenting of native coronary arteries in insulin-treated (ITDM) and non-ITDM patients with diabetes mellitus (DM) and compare the results with those in non-DM patients.
The influence of the treatment modality for DM on restenosis in patients undergoing coronary artery stenting has not been elucidated sufficiently.
A total of 1,439 (70%) of 2,061 patients underwent repeated angiography within six months of coronary stenting. The ITDM and non-ITDM (oral hypoglycemic drugs or diet) were documented in 48 (3.3%) and 177 patients (12.3%), respectively, leaving 1,214 non-DM patients.
Baseline reference vessel diameter tended to be smaller in ITDM patients (mean, 2.73 mm) than in non-DM and non-ITDM patients (2.88 mm and 2.85 mm, respectively). However, percent diameter stenosis was not different. The median number of stents deployed was 1; median stent length was 15 mm. Statistically significant differences were present after stenting for the means of minimal lumen diameter (MLD) and acute gain between ITDM patients (MLD: 2.67 mm, acute gain: 1.98 mm) and non-DM patients (MLD: 2.81 mm, acute gain: 2.16 mm). At follow-up, percent diameter stenosis, late lumen loss and loss index were significantly higher in both non-ITDM lesions (42%, 1.14 mm and 0.56, respectively) and ITDM lesions (48%, 1.26 mm and 0.65, respectively) than in non-DM lesions (35%, 0.96 mm and 0.45, respectively). The corresponding differences between non-ITDM and ITDM lesions did not reach statistical significance. Restenosis rates in non-DM, non-ITDM and ITDM lesions were 23.8%, 32.8% (p = 0.013 vs. non-DM) and 39.6% (p = 0.02 vs. non-DM, p = 0.477 vs. non-ITDM), respectively.
This study showed that compared with stenting in non-DM patients, stenting of native coronary arteries in DM patients is associated with significantly increased lumen renarrowing, regardless of the treatment modality for DM.
Diabetes mellitus (DM) is a risk factor for a variety of vascular disease states that are ultimately the primary causes of morbidity and mortality in DM patients. The increased propensity for atherosclerosis predisposes DM patients to stroke as well as acute and chronic coronary syndromes. Balloon angioplasty to treat coronary artery disease in patients with DM has been associated with an increased rate of acute complications and a lower long-term success rate compared with non-DM patients (1–4). The randomized Bypass Angioplasty Revascularization Investigation trial comparing clinical outcomes after balloon angioplasty versus coronary artery bypass grafting in patients with multivessel disease revealed that DM patients randomized to the angioplasty arm exhibited a higher coronary morbidity and mortality than non-DM patients (2). It has therefore been suggested that multivessel balloon angioplasty should be abandoned in DM patients (5,6).
To date, conflicting data have been published regarding the long-term outcomes after coronary stent placement in DM patients. While Van Belle et al. (7) found no difference in the restenosis rates between DM and non-DM patients, Kastrati et al. (8) reported a significantly higher late lumen loss and an increased risk for restenosis in the former group of patients. In a recent study distinguishing insulin-treated (ITDM) from non–insulin-treated (non-ITDM) diabetic patients, Abizaid et al. (9) observed an increased target lesion revascularization (TLR) rate versus non-DM patients only in ITDM but not in non-ITDM patients. Repeated angiography to determine restenosis rates had not been performed in that study. It was therefore the purpose of the present study to assess angiographic restenosis after coronary stenting in both ITDM and non-ITDM patients and compare the results in either group with those in non-DM patients.
From July 1, 1994, until January 1, 1997, 2,061 consecutive patients, among them 313 with documented DM, underwent stent implantation in native coronary artery lesions at our institution. Angiographic follow-up within six months was obtained from 1,439 patients (70%) who comprised the study group. Diabetes mellitus was present in 225 (15%) of the 1,439 patients; 48 DM patients were treated with insulin, whereas the 177 non-ITDM patients either took hypoglycemic drugs (n = 100) or adhered to a strict diet such that their fasting blood glucose level was maintained at <140 mg/dl. Pertinent patient characteristics for the three patient subgroups are given in Table 1. Compared with non-DM patients, DM patients were more often female, were slightly older and more often hypertensive. The prevalence of single-vessel versus multivessel coronary artery disease was equally distributed in the three patient groups.
The patients were informed about the coronary interventional procedure and gave their written consent. Stent implantation was performed routinely after recanalization of a total occlusion and in nonoccluded lesions if the following criteria were met: 1) a suboptimal result of balloon angioplasty (residual stenosis >20% or major dissection) or 2) a bail-out situation. During the procedure, a total of 10,000 to 15,000 U of heparin were given as a bolus injection. Balloon angioplasty utilized the monorail system after passage of the lesions with a guide wire. High-pressure stent implantation using noncompliant balloons was angiographically guided. A visually estimated residual diameter stenosis of <10% was aimed at. Patients were discharged from the hospital on a combination regimen of 100 mg/d of acetylsalicylic acid and 500 mg/d of ticlopidine for four to 12 weeks; they were scheduled for a repeat angiography after six months.
Quantitative coronary angiography
Angiographic variables obtained for every patient included the lesion length (length of vessel segment visually stenotic by at least 50%), the reference vessel diameters proximal and distal to the lesion and the minimal lumen diameter (MLD). Measurements were taken in identical, “worst view” projections using a hand-held digital caliper on optically magnified images. The diameter of the guiding catheter served as angiographic reference. Proximal and distal reference vessel diameters were averaged to obtain a mean reference vessel diameter.
The primary end point of the study was restenosis at follow-up, defined as a diameter stenosis of ≥50%.
Definitions of derived angiographic variables
Definitions are as follows: acute gain = MLD poststenting minus MLD at baseline; relative gain = acute gain divided by reference vessel diameter; late lumen loss = MLD poststenting minus MLD at follow-up; and loss index = late lumen loss divided by acute gain.
Definition of myocardial infarction
Myocardial infarction was classified as the development of new pathologic Q waves in two or more contiguous leads and/or elevation of postprocedure creatine kinase levels to more than two times normal.
In patients with multiple coronary lesions, only a single lesion, chosen at random, was entered into the analysis.
Continuous variables are presented as mean ± 1 SD, where appropriate. In case of a non-Gaussian distribution, median and range are given. Group differences among normally distributed continuous variables were assessed utilizing analysis of variance with Bonferroni’s correction; not normally distributed continuous variables were tested using the Kruskal-Wallis and Mann-Whitney U tests. Nominal variables were analyzed with the chi-square test. Statistical significance was assumed at p values <0.05; whenever three between-group comparisons were made, statistical significance was assumed at a p value <0.05/3 = 0.0167 (Bonferroni’s correction).
Following the definition of acute procedural success as a <10% residual stenosis without major adverse cardiac events during hospitalization and/or subacute stent thrombosis, the procedure was successful in 964 non-DM patients (79%), 138 non-ITDM patients (78%) and 41 ITDM patients (85%) (p = 0.526). ⇓Myocardial infarctions during hospitalization occurred in 13 patients (9 non-DM and 4 non-ITDM patients, p = 0.109); in 2 of the 9 non-DM patients, subacute stent thrombosis was observed.
Target vessel distribution was statistically not different in the three patient groups; on average, the left anterior descending coronary artery was affected in 44% of lesions, the right coronary artery in 34% and the left circumflex coronary artery in 21% (Table 2).
Stenting was performed as a bail-out measure in a total of 10 patients, 9 of whom were non-DM patients and 1 who took hypoglycemic drugs (p = 0.828). Adjunctive debulking devices such as the rotablator (n = 48) or the laser catheter (n = 15) were used in 63 patients, 47 of whom were non-DM patients, 14 belonged to the non-ITDM group (10 oral, 4 diet), and 2 were treated with insulin (p = 0.049). The median number of stents deployed per lesion was 1 in all groups, ranging from 0.5 (i.e., a “half” stent of ≤9 mm in length) to 5.0, and the median stented vessel segment length was 15 mm. The median smallest stent diameter used was 3.0 mm in all groups; smallest stent diameter ranged from 2.0 to 4.5 mm in the non-DM group, from 2.5 to 4.0 mm in the non-ITDM group and from 2.5 to 3.5 mm in the ITDM group (p = 0.087). No difference between patient groups was present in the distribution of stent types (Table 2).
The median lengths of non-DM lesions, non-ITDM lesions and ITDM lesions were 8.5 mm, 8.0 mm and 7.1 mm, respectively (Table 2); these differences were statistically not significant. At baseline, the mean reference vessel diameter in ITDM lesions (2.73 mm) was found to be less than in both non-DM lesions (2.88 mm, p = 0.021) and non-ITDM lesions (2.85 mm, p = 0.109). The mean MLD in ITDM lesions (0.69 mm) was slightly increased, though not significantly so, over that in non-ITDM (0.66 mm) and non-DM lesions (0.65 mm). Mean percent diameter stenosis was not different in the three groups.
Mean MLD after stenting in ITDM lesions, but not in non-ITDM lesions, was significantly less than in non-DM lesions (corresponding to a significantly lower acute gain); the means of relative gain and diameter stenosis were not different between groups.
At follow-up, the means of diameter stenosis (35%), late lumen loss (0.96 mm) and loss index (0.45) were all significantly less in non-DM lesions compared with both non-ITDM lesions (41% [p = 0.0006], 1.12 mm [p = 0.009] and 0.54 [p = 0.002], respectively) and ITDM lesions (47% [p = 0.0007], 1.25 mm [p = 0.008] and 0.64 [p = 0.0002], respectively) (Fig. 1 and 2)⇓ . ⇓ Consequently, the restenosis rates were increased in DM lesions, but the difference versus non-DM lesions was statistically significant only for non-ITDM lesions (Fig. 3). Differences between non-ITDM and ITDM lesions did not reach statistical significance for any angiographic variable obtained at follow-up.
The prevalences of TLR in the three patient groups were 17.4% (non-DM), 22.0% (non-ITDM) and 29.2% (ITDM), with an overall p value of 0.047 but no statistical significance between groups.
This study sought to assess the long-term angiographic outcome after stent placement in native coronary arteries in diabetic versus nondiabetic patients. The former were distinguished into two groups with regard to the treatment modality for DM, that is, patients were either treated with insulin or not. Reference vessel diameter was smallest in the ITDM group of patients, yet target vessel distribution, lesion length and percent diameter stenosis were statistically not different in the three groups. Lesions were supplied with a median of one 15-mm stent, resulting in identical mean residual stenoses after the intervention but a significantly lower MLD (vs. non-DM lesions) in ITDM lesions. At follow-up, mean diameter stenosis, late lumen loss and loss index were significantly increased in both ITDM and non-ITDM lesions compared with non-DM lesions; mean values for these variables were highest in the ITDM subgroup, but statistical significance versus the non-ITDM subgroup was not reached for any parameter. Consequently, the restenosis rate in non-ITDM lesions (32.8%) as well as in ITDM lesions (39.6%) was higher than in non-DM lesions (23.8%), yet statistically significantly so only for non-ITDM lesions.
Our findings support recent studies reporting an adverse outcome after stent placement in native coronary arteries for DM compared with non-DM patients (8,10). Lau et al. (10) studied 42 DM patients and an equal number of non-DM patients matched for stent design, reference vessel and stent diameter, target vessel treated and residual diameter stenosis after stenting; the restenosis rate in DM patients was 40.5%, significantly higher than in non-DM patients (16.7%). This study is in concordance with a recent analysis of predictive factors of restenosis after coronary stent placement in 1,084 patients which revealed that the presence of DM increased the risk for restenosis by a factor of 1.86 (8). In neither study was a distinction made between DM patients treated with insulin and those either taking oral hypoglycemic drugs or adhering to a diet.
This distinction has been made by Abizaid et al. (9) who investigated the clinical outcomes after coronary stenting in 954 consecutive patients. There were 151 non-ITDM patients and 97 ITDM patients. Target lesion revascularization was significantly increased in the latter patients (28% vs. 16.3% in non-DM patients), whereas no statistically significant difference versus non-DM patients was observed in this variable for non-ITDM patients (17.6%). However, the rate of angiographic follow-up in the different subgroups has not been reported. To our knowledge, no study to date has addressed the angiographic outcome after stenting in DM patients differentiated by treatment modality. In our study, the rates of angiographic follow-up in the three subgroups were not different. Late lumen loss as a measure of neointimal hyperplasia was significantly increased in non-ITDM and ITDM lesions (by 15% and 27%, respectively) compared with non-DM lesions. Although late lumen loss was higher in ITDM lesions than in non-ITDM lesions, this difference was statistically not significant. Target lesion revascularization based on the findings at follow-up angiography was increased in both non-ITDM and ITDM patients compared with non-DM patients, but statistical significance was not reached. Thus, the bottom line of our study is that patients with DM, regardless of the treatment modality for DM, exhibit a markedly higher propensity for neointimal hyperplasia after stenting of native coronary arteries than do patients without DM.
With the lack of elastic recoil and constrictive vascular remodeling in stented coronary lesions, neointimal hyperplasia secondary to smooth muscle proliferation constitutes the major mechanism responsible for lumen renarrowing after coronary stenting (11,12). Several in vitro and animal studies have supported the hypothesis that the mechanisms underlying smooth muscle cell proliferation are affected by the presence of DM (13). Serum from DM patients enhances smooth muscle cell proliferation to a higher degree than does serum of non-DM patients, indicating a difference in the concentration of growth promoting factors (14). An increased glucose concentration has been shown to result in an increased expression of the basic fibroblast growth factor, the transforming growth factor-alpha and the receptor for platelet-derived growth factor-beta in smooth muscle cells (15–17). High insulin levels stimulate the expression of endothelin-1 in endothelial cells and potentiate the effect of the insulin-like growth factor-1 on smooth muscle cells (18–20). Finally, in DM patients, local thrombus formation after stenting may be enhanced as a consequence of increased platelet reactivity, an impaired thrombolytic system and higher blood levels of clotting factors (21–24). Another factor predisposing to the high restenosis rate in ITDM lesions may be the significantly smaller poststenting MLD (25).
This was a retrospective study with all the inherent shortcomings of such a study design. The small number of ITDM patients may have given rise to type II statistical errors relative to the comparisons with non-DM patients. If the absolute difference in restenosis rates between non-ITDM and ITDM patients of 6.8% (corresponding to a 21% relative increase) was to be accepted or rejected with a statistical power of 80% and a type I error of 5%, approximately 785 patients would be needed in each group. The impact of metabolic parameters such as insulin and C-peptide levels as well as the concentrations of glucose and advanced glycolization end products has not been assessed.
This study showed that, compared with stenting in nondiabetic patients, stenting of native coronary arteries in diabetic patients is associated with an increased propensity for neointimal hyperplasia, regardless of the treatment modality for diabetes. In particular, this finding holds true for diabetic patients who are not treated with insulin.
- diabetes mellitus
- insulin-treated diabetes mellitus
- minimal lumen diameter
- target lesion revascularization
- Received December 18, 1998.
- Revision received October 7, 1999.
- Accepted January 14, 2000.
- American College of Cardiology
- Stein B.,
- Weintraub W.S.,
- Gebhart S.P.,
- et al.
- ↵(1997) Influence of diabetes on 5-year mortality and morbidity in a randomized trial comparing CABG and PTCA in patients with multivessel disease. The Bypass Angioplasty Revascularization Investigation. Circulation 96:1761–1769.
- Kip K.E.,
- Faxon D.P.,
- Detre K.M.,
- Yeh W.,
- Kelsey S.F.,
- Currier J.W.
- Weintraub W.S.,
- Stein B.,
- Kosinski A.,
- et al.
- O’Neill W.W.
- Ferguson J.J.
- Van Belle E.,
- Bauters C.,
- Hubert E.,
- et al.
- Kastrati A.,
- Schömig A.,
- Elezi S.,
- et al.
- Abizaid A.,
- Kornowski R.,
- Mintz G.S.,
- et al.
- Kornowski R.,
- Mintz G.S.,
- Kent K.M.,
- et al.
- Dussaillant G.R.,
- Mintz G.S.,
- Pichard A.D.,
- et al.
- Aronson D.,
- Bloomgarden Z.,
- Rayfield E.J.
- McClain D.A.,
- Paterson A.J.,
- Roos M.D.,
- Wei X.,
- Kudlow J.E.
- Inaba T.,
- Ishibashi S.,
- Gotoda T.,
- et al.
- Oliver F.J.,
- de la Rubia G.,
- Feener E.P.,
- et al.
- Banskota N.K.,
- Taub R.,
- Zellner K.,
- King G.L.
- McGill J.B.,
- Schneider D.J.,
- Arfken C.L.,
- Lucore C.L.,
- Sobel B.E.
- Kuntz R.E.,
- Gibson M.,
- Nobuyoshi M.,
- Baim D.S.