Author + information
- Received November 29, 2004
- Revision received May 27, 2005
- Accepted July 11, 2005
- Published online January 3, 2006.
- Mandeep Singh, MD⁎,
- Brent A. Williams, MS†,
- Bernard J. Gersh, MB, ChB, DPhil⁎,
- Robyn L. McClelland, PhD†,
- Kalon K.L. Ho, MD, MSc‡,
- James T. Willerson, MD§,
- William F. Penny, MD∥,
- Donald E. Cutlip, MD‡ and
- David R. Holmes Jr, MD⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. David R. Holmes, Jr., Division of Cardiovascular Diseases, Department of Internal Medicine, Mayo Clinic, 200 2nd Street SW, Rochester, Minnesota 55905.
There are global differences in the revascularization strategies in the management of patients presenting with stable or unstable angina. Not only are invasive procedures, such as cardiac catheterization, percutaneous coronary interventions (PCIs), and coronary bypass graft surgery used more commonly in the U.S. than in other countries, but the timing of such interventions differs, with PCI performed earlier during the course of an acute coronary syndrome in patients treated in the U.S. (1–5). A more aggressive revascularization approach in most recent trials in patients with acute coronary syndromes is known to reduce major adverse cardiovascular outcome, but its effect on restenosis and target vessel revascularization (TVR) is largely undefined (6,7). An earlier analysis from the Platelet glycoprotein IIb/IIIa in Unstable angina: Receptor Suppression Using Integrilin Therapy (PURSUIT) trial suggested higher revascularization rates with earlier performance of angioplasty in a setting of acute coronary syndrome (5). However, the PURSUIT trial antedates current angioplasty results.
The Prevention of Restenosis With Tranilast and its Outcomes (PRESTO) trial was designed to evaluate the effects of tranilast, an oral anti-inflammatory agent, on major adverse cardiovascular events and angiographic and intravascular (i.e., intravascular ultrasound) end points (8). This trial not only provided important information on patients treated in the U.S. but also enrolled patients in non-U.S. sites, providing an ideal substrate for comparison of the angiographic restenosis and ischemia-driven TVR rates among patients treated in the U.S. and other countries. With this background, we sought to determine differences in these outcomes between patients treated at U.S. and non-U.S. sites.
The PRESTO trial has been described previously (8). In brief, it was a double blind, placebo-controlled, parallel group study of patients after PCI. The primary end point was the first occurrence of major adverse cardiovascular events within nine months, defined as death, myocardial infarction (MI), and/or ischemia-driven TVR. No differences in outcomes were found in patients treated with the study drug versus placebo. Furthermore, the PRESTO trial included an angiographic substudy in which enrollees were required to undergo follow-up angiography at nine months (or sooner if clinically warranted). Restenosis was defined as ≥50% stenosis in the treated segment at follow-up, or at least 50% loss of the original gain in the minimal luminal diameter. Ischemia-driven TVR was defined as intervention for chest pain or a positive test for ischemia (exercise stress test, stress echocardiogram, 24-h Holter monitor, resting electrocardiogram evidence of ST-segment depression or elevation in >1 lead, or radionuclide study showing reversible defect). Target vessel failure (TVF) was defined as a composite of death, MI, or ischemia-driven TVR. The type of intervention performed was at the investigator’s discretion, with the exclusion of intracoronary radiation. All patients gave informed consent to take part in the trial. The institutional review board of each hospital approved the study protocol.
The primary clinical outcome in this study was ischemia-driven TVR within nine months after PCI, and the primary angiographic outcome was restenosis at nine-month follow-up as defined previously. The rates of TVR are reported across geographic location, non-U.S. versus U.S. A logistic regression model was developed to estimate the association between geographic location and TVR. An odds ratio and corresponding 95% confidence interval are reported comparing U.S. with non–U.S-treated patients. An adjusted odds ratio and 95% confidence interval for U.S. versus non-U.S. patients was estimated after including variables in the logistic regression model that have known or assumed association with TVR. These variables previously were identified by our group as lesion length, vessel size, American College of Cardiology/American Heart Association (ACC/AHA) lesion type, previous PCI, diabetes, smoking status, current unstable angina, and gender (9). Vessel size was not available in the clinical data; therefore, it was not included in any of the models. Similar strategies were used for analyzing death, MI, TVF, and angiographic restenosis. Angiographic restenosis was classified on a per-person basis as any lesion with restenosis versus none. The number of lesions per person was included in the models for angiographic restenosis.
Baseline clinical and lesion characteristics are reported separately for U.S. and non-U.S. sites. Continuous variables are reported as mean ± one standard deviation, with differences between groups tested using ttests; categorical variables are reported as counts and percentages, with differences between groups tested using chi-square tests. Nonparametric tests were incorporated as necessary.
Four hundred thirty-two centers (224 in the U.S., 166 in Western Europe, 27 in Canada, 7 in Eastern Europe, and 8 in Australia/South Africa) enrolled 11,484 patients between April 1999 and July 2000. Among these patients, 5,026 (44%) were treated in the U.S. and 6,458 (56%) in other countries. The angiographic substudy of the PRESTO trial originally enrolled 2,823 patients, with 2,096 (74%) providing angiographic data at nine-month follow-up. A significantly larger percentage of U.S. patients did not provide angiographic follow-up at nine months (31% vs. 22%, p < 0.01).
There were several significant differences in baseline characteristics between patients treated in the U.S. and those treated elsewhere (Table 1).Patients treated in the U.S. were more likely to be female, diabetic, not currently smoking, to have unstable angina, and to have undergone a previous PCI. Lesions tended to be longer in U.S.-treated patients, but were less likely to be ACC/AHA class C (Table 2).U.S. patients also had shorter time to PCI after hospital admission. Rates of alcohol consumption and prevalence of previous MI were, however, were greater in patients treated outside the U.S.
Table 2highlights the differences in the lesion characteristics as assessed by quantitative coronary analysis in patients treated inside and outside the U.S. Among the pre-PCI characteristics, lesions in patients treated at non-U.S. sites were more likely to be complex and diffuse. In contrast, the pre-PCI stenosis was worse in lesions of patients treated in the U.S., and the pre-PCI reference vessel diameter was smaller in the U.S.-treated lesions. There was no difference in the post-PCI minimal luminal diameter in the lesions treated inside and outside the U.S.; however, the postprocedure reference vessel diameter remained slightly smaller in the U.S.-treated patient lesions.
Table 3shows the rates of ischemic TVR, MI, death, and TVF in patients treated in the U.S. and other countries. It also shows the unadjusted and adjusted odds ratios for these outcomes. Ischemia-driven TVR occurred in a greater proportion of patients treated in the U.S. than those treated elsewhere (18% vs. 11%, p < 0.01). The rate of TVF was also significantly higher among U.S. patients (20%) as compared with those patients treated in other countries (12%), although this was predominantly driven by TVR rates. U.S. patients also had higher rates of death and MI, although these events were rare. After adjustment for the aforementioned potential confounders, U.S. patients remained at significantly increased risk of all events with the exception of death. The increased risk of TVR, TVF, and MI ranged from approximately 40% to 60%.
Table 4demonstrates the restenosis rates amongst patients treated in the U.S. and patients treated in other countries who underwent protocol-driven follow-up angiography. Before and after adjustment for gender, diabetes, current unstable angina, smoking status, previous PCI, any lesion of ACC/AHA type C, length of longest lesion, and number of lesions, the likelihood of restenosis was significantly higher in the U.S.-treated patients.
Table 5shows the number of enrolling sites, number of patients enrolled, mean days from hospital admission to PCI, and TVR rates by the participating countries. Eighteen non-U.S. countries participated in the PRESTO trial, with the number of enrolling sites by country ranging from 1 to 37, and patient enrollment from 48 to 1,022. Mean time from hospital admission to PCI ranged from 1.0 to 5.0 days in non-U.S. sites, with the U.S. having the lowest mean (0.8 days). The rates of TVR at non-U.S. sites ranged widely, from 4% in the Czech Republic and Estonia to 16% in Germany, with the highest rate of 18% occurring in the U.S.
This analysis shows that the angiographic restenosis, ischemia-driven TVR, and TVF rates after PCI were all significantly greater for patients treated in the U.S. than those patients who were treated in other countries. Previous studies have reported differences in countries’ assessments, treatments, and outcomes for patients with acute MI and cardiogenic shock (1–4). Limited data are available on the geographical differences in restenosis rates or rates of ischemia-driven TVR. Our data suggest important baseline, angiographic, and procedural differences between countries, but these differences between countries did not fully account for the geographical differences in the rates of restenosis, TVR, and TVF. This study is the largest reported to date comparing the angiographic rates of restenosis and ischemia-driven TVR on the basis of geographical differences.
There are several potential explanations for the higher incidence of both angiographic restenosis and ischemia-driven TVR rates in the U.S. First, it would be expected that restenosis would have been predictably higher for U.S. patients based on the increased frequency of baseline and some angiographic variables associated with higher risk for restenosis and TVR in previous reports (10–13). It is not surprising, however, that these previously described covariates do not account for all of the increased risk among U.S. patients noted in our models because most previously reported models have limited discriminatory power in identifying restenotic events (14).
Changes in treatment strategy that are not easily accounted for in the multivariable models also may offer part of the explanation. Unstable angina is known to be associated with higher rates of revascularization and can partly explain the higher incidence of both restenosis and ischemia-driven TVR. Approximately two-thirds of patients treated in the U.S., and fewer (39%) in other countries, had this diagnosis. Perhaps more importantly, there is a distinct difference in the invasive approach to patients with acute coronary syndrome in the U.S. as compared with other countries. Very early revascularization, as practiced in the U.S., has been shown to increase the rates of repeat revascularization (5). In a retrospective analysis from the PURSUIT trial, it was demonstrated that rates of repeat revascularization in patients with unstable angina were notably higher for PCI within 24 h (19%) than for PCI at 24 to 72 h (17%), 3 to 7 days (13%), and 8 to 30 days (8%) (5). In our study, the mean time from hospital admission to PCI at U.S. sites was 0.8 days as compared with 2.1 days at the non-U.S. sites. These rates match the practice patterns of different countries in the present study, with an ischemic TVR rate of 18% in the U.S., where aggressive intervention is practiced, and a rate of 11% outside the U.S., including the United Kingdom, Western Europe, and South Africa, where greater delays for invasive management occur after initial medical stabilization. However, contrary to the PURSUIT study, our analysis did not find any significant effect of time to treatment on TVR in patients treated with a diagnosis of acute coronary syndrome (data not shown).
Higher amounts of alcohol consumption in patients treated outside the U.S. might have conferred some protection from repeat revascularization. Alcohol consumption in small-to-moderate amounts has been shown to reduce mortality and MI (15). Some animal data support a beneficial role of alcohol in terms of reducing the smooth muscle hyperplasia in response to balloon injury of the coronary arteries (16,17), and it is conceivable that this mechanism might be partly responsible for lower events in the patients treated outside the U.S. Similarly, a higher prevalence of previous MI could be a factor associated with lower subsequent revascularization, as observed in patients treated outside the U.S. (10,18).
A major problem in the analysis of restenosis and TVR rates is the modest-to-poor predictive ability of the known demographic and angiographic variables. We previously demonstrated poor discriminatory ability of models derived from the PRESTO trial in predicting angiographic restenosis and ischemic TVR (9,14). It implies, therefore, that there are some important, unmeasured determinants of restenosis and TVR that are not considered in the present study. Knowledge and addition of these variables is paramount in determining the reasons for geographic disparity in the rates of restenosis and TVR (19,20).
The need for TVR was based on trial recommendations and hence less likely reflected the discretion of the attending physician, or differences in the practice patterns of different investigators and hospitals. However, chest pain was included as one of the reasons for TVR, which could have led to higher revascularization procedures in the U.S. In addition, the attitudes of patients, physicians, and society toward treatment for patients also may be important factors. The aggressiveness of medical care and access to health care are important confounders that were difficult to consider properly in the present study. The angiographic study was prespecified in the trial; however, we cannot totally exclude withdrawal bias due to incomplete angiographic follow-up.
In conclusion, geographical differences in angiographic restenosis and ischemia-driven TVR were observed. Patients treated in the U.S. have a higher prevalence of many baseline demographic and angiographic features associated with higher angiographic restenosis and need for repeat revascularization, but these differences between countries did not account completely for the geographical differences in restenosis and TVR.
Richard Edward Kuntz, MD, acted as guest editor for this paper.
- Abbreviations and Acronyms
- American College of Cardiology/American Heart Association
- myocardial infarction
- percutaneous coronary intervention
- target vessel failure
- target vessel revascularization
- Received November 29, 2004.
- Revision received May 27, 2005.
- Accepted July 11, 2005.
- American College of Cardiology Foundation
- Ronner E.,
- Boersma E.,
- Laarman G.J.,
- et al.
- Wallentin L.,
- Lagerqvist B.,
- Husted S.,
- Kontny F.,
- Stahle E.,
- Swahn E.
- Holmes D.R. Jr..,
- Savage M.,
- LaBlanche J.M.,
- et al.
- Singh M.,
- Gersh B.J.,
- McClelland R.L.,
- et al.
- Cutlip D.E.,
- Chauhan M.S.,
- Baim D.S.,
- et al.
- Kastrati A.,
- Schomig A.,
- Elezi S.,
- et al.
- Mercado N.,
- Boersma E.,
- Wijns W.,
- et al.
- Weintraub W.S.,
- Kosinski A.S.,
- Brown C.L. 3rd.,
- King S.B. 3rd.
- Singh M.,
- Gersh B.J.,
- McClelland R.L.,
- et al.
- Liu M.W.,
- Anderson P.G.,
- Luo J.F.,
- Roubin G.S.