Author + information
- Received May 7, 2003
- Revision received January 2, 2004
- Accepted February 10, 2004
- Published online June 16, 2004.
- Hans-Peter Bestehorn, MD*,* (, )
- Franz-Josef Neumann, MD*,
- Heinz Joachim Büttner, MD*,
- Peter Betz, MD, PhD*,
- Peter Stürzenhofecker, MD†,
- Eberhard von Hodenberg, MD‡,
- Antoine Verdun, MD§,
- Laszlo Levai, MD§,
- Jean Pierre Monassier, MD∥ and
- Helmut Roskamm, MD*
- ↵*Reprint requests and correspondence:
Dr. Hans-Peter Bestehorn, Herz-Zentrum Bad Krozingen, Südring 15, 79189 Bad Krozingen, Germany.
Objectives We investigated the effect of oral verapamil on clinical outcome and angiographic restenosis after percutaneous coronary intervention (PCI).
Background Thus far, there is no established systemic pharmacologic approach for the prevention of restenosis after PCIs. Five small studies reported encouraging results for calcium channel blockers.
Methods Our randomized double-blind trial included 700 consecutive patients with successful PCI of a native coronary artery. Patients received the calcium channel blocker verapamil, 240 mg twice daily for six months, or placebo. Primary clinical end point was the composite rate of death, myocardial infarction, and target vessel revascularization (TVR) during one-year follow-up; the angiographic end point was late lumen loss at the six-month follow-up angiography.
Results We obtained complete clinical follow-up in 95% of the patients, and scheduled angiography was performed in 94%. The proportion of patients treated with stents was 83%. The primary clinical end point was reached in 67 (19.3%) patients on verapamil and in 103 (29.3%) patients on placebo (relative risk [RR] 0.66 [95% confidence interval (CI) 0.48 to 0.89]; p = 0.002). This difference between the groups was driven by TVR (17.5% with verapamil vs. 26.2% with placebo; RR 0.67 [95% CI 0.49 to 0.93]; p = 0.006). Late lumen loss was 0.74 ± 0.70 mm with verapamil and 0.81 ± 0.75 mm with placebo (p = 0.11). Compared with placebo, verapamil reduced the rate of restenosis ≥75% (7.8% vs. 13.7%; RR 0.57 [95% CI 0.35 to 0.92]; p = 0.014).
Conclusions Verapamil compared with placebo improves long-term clinical outcome after PCI of native coronary arteries by reducing the need for TVR. This was caused by a reduction in the rate of high-grade restenosis.
The long-term benefit from percutaneous coronary interventions (PCI) is limited by restenosis (1). Despite intensive research in recent years, it is still controversial whether any systemic drug therapy can prevent this adverse outcome. To this end, various drugs with promising efficacy in animal experiments have been tested in patients. Yet, most of the studies failed to show an unequivocal beneficial effect of systemic drug therapy on restenosis. There is reasonable skepticism whether systemic administration can achieve the local drug levels needed to suppress restenosis (2–4).
Calcium channel blockers are among the few drugs with promising results in the prevention of restenosis. The putative antirestenotic properties of calcium channel blockers have been attributed to concentration-dependent inhibition of smooth muscle cell transformation and proliferation after stimulation with platelet-derived growth factor (5–7). Meta-analysis (8)of five small studies (9–13)yielded a significant 30% relative reduction in the risk of restenosis by calcium channel blockers as compared with placebo. This finding, however, has not been tested in an adequately powered trial. Therefore, we conducted the multicenter, randomized, double-blind, placebo-controlled, Verapamil Slow-Release for Prevention of cardiovascular Events After Angioplasty (VESPA) trial to assess the effect of the calcium channel blocker verapamil on clinical outcome and angiographic restenosis after PCI.
Patients selection and stent placement
Patients of age ≥35 and ≤80 years with successful PCI of a native coronary artery were eligible for the study. Successful intervention was defined by residual stenosis <30% on visual estimation and—in case of stenting—desired position of stent. We excluded patients with restenotic lesions, occlusions, lesions in bypass grafts, and left main location as well as patients with unstable angina, acute myocardial infarction (MI), ad-hoc and multistage PCI, insulin-dependent diabetes, renal insufficiency, sick sinus syndrome, atrioventricular node block, congestive heart failure and/or left ventricular ejection fraction <40%, severe concomitant diseases, contraindications to verapamil, and patients with inability to provide informed consent for participation. The study was approved by the institutional ethical review board of each hospital, and all patients gave written informed consent before inclusion in the study.
Although the use of stents was encouraged, the percutaneous treatment modality was left to the operator's discretion. All patients received aspirin 100 mg, once daily, indefinitely and, in case of stenting, ticlopidine 250 mg, twice daily, or clopidogrel 75 mg, once daily, for four weeks. Thienopyridines were started immediately after the intervention.
We designed our study as a prospective, multicenter, randomized, double-blind, placebo-controlled trial. Within 30 min after PCI, we assigned eligible patients to receive either verapamil, 240 mg slow-release tablets, or identical-appearing placebo twice daily. At each participating center, allocation to the study treatment was based on computer-generated random numbers that were used for double-blind labeling. In patients undergoing multivessel PCI (n = 83), the most clinically relevant lesion was defined as the study lesion at the time of randomization. The study drug was started immediately after PCI and was continued until three days before six-month follow-up angiography.
After PCI, we determined plasma concentrations of creatine kinase and its MB isoenzyme systematically for 48 h. We scheduled outpatient visits at one and three months after randomization for clinical follow-up including 12-lead electrocardiogram recordings. Patients returned to the hospital for routine angiographic restudy and clinical evaluation including bicycle ergometry at six months. Follow-up angiography was performed earlier if the patient had recurrent symptoms or signs of ischemia. Patients who had undergone angiography at <4 months after recruitment without meeting the criteria for a clinical end point were encouraged to undergo repeat angiography at six months. At each follow-up visit, we assessed compliance by pill counts.
Angiographic images were stored on compact discs (one center) or grabbed from cine films (other centers) and analyzed before the study was unblinded. Quantitative analysis was performed as described previously (14–16). Using the same two orthogonal views throughout the study, we obtained minimal luminal diameter (MLD), reference diameter, percent diameter stenosis, and the diameter of the maximally inflated balloon from the analysis system (Cardiovascular Angiographic Analysis System, Department of Medical Informatics, University of Limburg, Limburg, the Netherlands).
Acute gain was calculated as the difference between post-stenting and predilation MLD, late loss as the difference between post-stenting MLD and MLD at follow-up, net gain as the difference between MLD at follow-up, and predilation MLD and loss index was calculated as the ratio of late loss to acute gain. All measurements were performed by the same blinded operator; intraobserver variability was 0.10 mm for MLD and 2.58% for diameter stenosis.
Study end points
Our clinical primary end point was the combined incidence of death, MI, and target vessel revascularization (TVR). Myocardial infarction was defined as the presence of new Q waves (≥40 ms) in two or more contiguous electrocardiographic leads or an elevation of creatine kinase or its MB isoenzyme to at least 3 × the upper limit of normal (80 U/l, 10 U/l, respectively) in two samples during hospitalization or to 2 × the upper limit of normal after discharge. We defined TVR as coronary artery bypass surgery or repeat percutaneous angioplasty involving the treated vessel and performed for symptoms or signs of ischemia in the presence of angiographic restenosis (≥50%). Target vessel revascularization by coronary artery bypass surgery, which was indicated at the time of follow-up angiography, was counted as an event, even when performed during subsequent hospital admission. As our primary angiographic end point, we assessed late loss. We also analyzed other angiographic indexes of restenosis, including loss index, percent diameter stenosis, and the incidence of severe restenosis, defined as diameter stenosis ≥75%.
Sample size estimation and statistical analysis
For calculation of sample size, we assumed a 25% incidence of our primary clinical end point (17). We designed the study to have a power of 80% to detect a 40% reduction in our primary clinical end point by verapamil as compared with placebo with a two-sided alpha value of 0.025. According to these assumptions, 325 patients were required in each treatment arm. To account for losses to follow-up, we intended to include 700 patients. Assuming a normal distribution with a standard deviation of 0.5 mm for late loss (17)and allowing for missing angiograms in 100 patients, this sample size gave us an 80% power to detect a 0.13 mm difference in late loss with a two-sided alpha value of 0.025.
All analyses were performed according to the intention-to-treat principle. Data are presented as mean ± SD or as counts or proportions. We assessed differences between the groups with use of a two-sided chi-square test for categorical variables. For continuous variables we used the unpaired ttest. A value of p < 0.05 in the two-tailed test was considered to statistically significant. The impact of baseline characteristics and other pertinent covariables on the primary clinical end point was adjusted using multivariate logistic regression. To account for the double primary end point, we adjusted the level of significance for our primary end points to p = 0.025. For all statistical analyses, we used SAS 6.12 software package (SAS Institute, Cary, North Carolina).
Study cohort and follow-up
The trial population is shown in Figure 1. The study enrolled 700 patients; 348 were assigned to verapamil and 352 to placebo. A coronary stent was placed in 581 patients (83%). The study groups were homogeneous with respect to baseline demographic, clinical, and angiographic characteristics (Tables 1 to 3). ⇓⇓⇓Thirty-seven patients were lost to clinical follow-up, and five additional patients refused second angiography. Thirteen patients of the verapamil group and 33 patients of the placebo group underwent follow-up angiography prematurely.
Of the patients on verapamil, 19.5% (68 of 348) discontinued their study medication because of constipation (n = 23), second- or a third-degree heart block (n = 8), or other cardiovascular side effects (n = 37). In the placebo group, the discontinuation rate was 15.6% (55 of 352). In addition, 11.8% (n = 41) of the verapamil group and 10.8% (n = 38) of the placebo group had a pill count that differed by more than 30% from the expected pill count.
The primary clinical end point was reached in 19.3% (67 of 348) of the verapamil group and in 29.3% (103 of 352) of the placebo group (Table 4). Thus, adverse cardiac events were significantly (p = 0.002) fewer in the verapamil group than in the placebo group with a relative risk (RR) reduction by verapamil of 34% (Table 4). The difference in the incidence of our primary end point was driven by a reduction in TVRs (Table 4). There were no appreciable differences in the incidences of death or MI. The risk reduction for the TVR by verapamil as compared with placebo was similar in patients receiving a stent and in patients treated with plain percutaneous transluminal coronary angioplasty (PTCA): (RR, 0.74 [95% confidence interval (CI) 0.52 to 1.05]; p = 0.05 vs. RR 0.50 [95% CI 0.21 to 1.15]; p = 0.06). In a multivariable logistic regression model that took into account baseline variables listed in Tables 1 and 3, the adjusted odds ratio was 0.52 (95% CI 0.32 to 0.84; p = 0.007) for the clinical primary end point comparing both treatment strategies.
Angiographic indexes of restenosis
Mean late loss was 0.74 ± 0.70 mm in the verapamil group and 0.81 ± 0.75 mm in the placebo group (Table 5); thus, our primary angiographic end point did not reach statistical significance (p = 0.11). The cumulative distribution of percent diameter stenosis at follow-up (Fig. 2) shows a separation of the curves in the region of more severe stenoses. Accordingly, we found a significant reduction in the rate of high-grade restenosis (≥75%) by verapamil as compared with placebo, but only a trend towards a reduction in restenosis rate according to the 50% criterion (Table 5). Other indices of restenosis confirmed the trend towards attenuated restenosis by verapamil, but did not reach statistical significance either.
The differences in late loss between the verapamil and placebo groups were similar after stenting (0.82 ± 0.69 mm vs. 0.90 ± 0.72 mm; p = 0.20) and after plain balloon angioplasty (0.36 ± 0.64 mm vs. 0.41 ± 0.72 mm; p = 0.67), as were the risk reductions for high-grade restenosis (RR 0.56 [95% CI 0.33 to 0.97]; p = 0.03 vs. RR 0.58 [95% CI 0.12 to 1.75]; p = 0.30).
Our randomized, placebo-controlled multicenter trial investigated the effect of verapamil on the clinical and angiographic outcome after PCI. Verapamil, administered for six months after PCI, reduced the incidence of major adverse cardiac events, our primary clinical end point, by reducing the need for repeat TVR. Consistent with this outcome, we found a significant reduction in the incidence of high-grade restenosis, although our primary angiographic end point, late loss, failed to show significant differences. Previous trials on calcium channel blockers were flawed by small sizes and methodological issues, such as subjective evaluation of restenosis. To our knowledge, our trial is the first adequately powered study addressing the effect of verapamil on restenosis with the use of quantitative coronary angiography. Although the impact of verapamil on restenosis was considerably smaller than anticipated based on the earlier studies, the principle effect is confirmed. With the weight of the present study, currently available evidence suggests that calcium antagonists are capable of interfering with mechanisms involved in restenosis formation. Mechanisms that could be targeted by calcium antagonists include smooth muscle cell transformation, and migration and proliferation, as well as elaboration of extracellular matrix proteins (18–20).
Our analyses, specifically the cumulative distributions of percent diameter stenosis, suggest a predominant effect of verapamil in the prevention of excessive restenosis formation. Previous studies revealed a bimodal distribution of restenosis formation after plain PTCA and after stenting (21). This bimodal distribution delineates two populations, which develop distinctively different degrees of lumen renarrowing. From our findings, it is conceivable that verapamil acts predominantly on the population with the strongest propensity to renarrowing. In this population, the antiproliferative properties of verapamil that have been described in experimental studies may become particularly effective.
Of a large number of patients screened, only 6% were eventually included in the study. This was largely due to the strict inclusion criteria avoiding high-risk patients such as those with acute coronary syndromes and patients who were unlikely to complete follow-up (Fig. 1). Notably, only 2.1% were excluded because of contraindications to verapamil. On the other hand, the dosage of verapamil had to be reduced or withdrawn in about one-third of the patients.
The possibility has to be considered that the reduction in TVR was caused by the antianginal properties of verapamil. To reduce this effect, the study protocol mandated bicycle ergometry before follow-up angiography and after discontinuation of study medication for three days. Nevertheless, the role of antianginal properties of verapamil cannot be completely ruled out. Although the majority of our patients were treated with stents, there was some admixture from patients who underwent plain PTCA. The mechanisms of restenosis differ substantially between the two treatment modalities. Neointima formation accounts for about 90% of the lumen loss after stenting (22,23). After plain PTCA, neointima formation is less pronounced, and about two-thirds of lumen loss are caused by early elastic recoil and late vessel shrinkage. We cannot assume a uniform action of verapamil on each of these mechanisms. The low number of patients treated with plain PTCA prevented the detection of differences between the two percutaneous treatment modalities, but may have contributed to scatter.
Although the effect of verapamil on restenosis formation was statistically detectable, its extent was limited and has to be weighed against the disadvantages of withholding beta-blockers. In the meantime, stents releasing antiproliferative agents have proven, by far, to be more powerful tools in the prevention of restenosis (24). However, these devices are costly, and there is concern that the impairment of vascular healing processes after intervention might pose a substantial risk of late thrombotic events (25). Therefore, there is continued interest in the search for alternative approaches (26). To this end, the results of our study demonstrate the potential that resides in systemic administration of antiproliferative agents.
☆ Supported, in part, by a grant from Abbott GmbH & Co., KG Ludwigshafen, Germany.
- coronary artery bypass grafting
- coronary artery disease
- confidence interval
- myocardial infarction
- minimal lumen diameter
- percutaneous coronary intervention
- percutaneous transluminal coronary angioplasty
- relative risk
- target vessel revascularization
- Verapamil Slow-Release for Prevention of Cardiovascular Events After Angioplasty trial
- Received May 7, 2003.
- Revision received January 2, 2004.
- Accepted February 10, 2004.
- American College of Cardiology Foundation
- Von Hodenberg E.,
- Tost B.,
- Scheffold T.
- Block L.H.,
- Emmons L.R.,
- Vogt E.,
- et al.
- Withworth H.B.,
- Roubin G.S.,
- Hollman J.,
- et al.
- Unverdorben M.,
- Kunkel B.,
- Leucht M.,
- et al.
- Hoberg E.,
- Dietz R.,
- Frees U.,
- et al.
- Bestehorn H.P.,
- Rensing U.F.,
- Roskamm H.,
- et al.
- Frey A.W.,
- Hodgson J.M.,
- Muller C.,
- et al.
- Reiber J.H.,
- Serruys P.W.,
- Kooijman C.J.
- Karsch K.R.,
- Preisack M.B.,
- Baildon R.,
- et al.
- Hoberg E.
- Pepine C.J.
- Schomig A.,
- Kastrati A.,
- Elezi S.,
- et al.
- Fattori R.,
- Piva T.
- Liistro F.,
- Colombo A.
- Farb A.,
- John M.,
- Acampado E.,
- et al.