Author + information
- Received January 6, 1997
- Revision received June 17, 1997
- Accepted June 26, 1997
- Published online October 1, 1997.
- Michel E. Bertrand, MD, FACCA,* (, )
- Eugène P. McFadden, MRCP, FACCA,
- Jean-Charles Fruchart, PhDA,
- Eric Van Belle, MDA,
- Philippe Commeau, MDB,
- Gilles Grollier, MDC,
- Jean-Pierre Bassand, MDD,
- Jacques Machecourt, MDE,
- Jean Cassagnes, MDF,
- Jean-Marie Mossard, MDG,
- André Vacheron, MDH,
- Alain Castaigne, MDI,
- Nicolas Danchin, MD, FACCJ,
- Jean-Marc Lablanche, MD, FACCA,
- for the PREDICT Trial Investigators1
- ↵*Dr. Michel E. Bertrand, Division of Cardiology B, Hôpital Cardiologique, Boulevard du Professeur Leclercq, 59037 Lille, France.
Objectives. This study sought to determine whether pravastatin affects clinical or angiographic restenosis after coronary balloon angioplasty.
Background. Experimental data and preliminary clinical studies suggest that lipid-lowering drugs might have a beneficial effect on restenosis after coronary angioplasty.
Methods. In a multicenter, randomized, double-blind trial, 695 patients were randomized to receive pravastatin (40 mg/day) or placebo for 6 months after successful balloon angioplasty. All patients received aspirin (100 mg/day). The primary angiographic end point was minimal lumen diameter (MLD) at follow-up, assessed by quantitative coronary angiography. A sample size of 313 patients per group was required to demonstrate a difference of 0.13 mm in MLD between groups (allowing for a two-tailed alpha error of 0.05 and a beta error of 0.20). To allow for incomplete angiographic follow-up (estimated lost to follow-up rate of 10%), 690 randomized patients were required. Secondary end points were angiographic restenosis rate (restenosis assessed as a categoric variable, >50% stenosis) and clinical events (death, myocardial infarction, target vessel revascularization).
Results. At baseline, clinical, demographic, angiographic and lipid variables did not differ significantly between groups. In patients treated with pravastatin, there was a significant reduction in total and low density lipoprotein cholesterol and triglyceride levels and a significant increase in high density lipoprotein cholesterol levels. At follow-up the MLD (mean ± SD) was 1.47 ± 0.62 mm in the placebo group and 1.54 ± 0.66 mm in the pravastatin group (p = 0.21). Similarly, late loss and net gain did not differ significantly between groups. The restenosis rate (recurrence >50% stenosis) was 43.8% in the placebo group and 39.2% in the pravastatin group (p = 0.26). Clinical restenosis did not differ significantly between groups.
Conclusions. Although pravastatin has documented efficacy in reducing clinical events and angiographic disease progression in patients with coronary atherosclerosis, this study shows that it has no effect on angiographic outcome at the target site 6 months after coronary angioplasty.
Over the past 15 years, it has becoming increasingly evident that percutaneous transluminal coronary angioplasty (PTCA) is an effective method of myocardial revascularization. Compared with medical treatment in patients with left anterior descending coronary artery stenosis, PTCA has demonstrated its superiority in terms of decreasing symptoms and improving exercise performance . Six multicenter randomized trials comparing PTCA with coronary artery bypass graft surgery (CABG) have shown that survival without myocardial infarction was similar for the two treatment modalities [2–9]. However, patients treated by PTCA had more revascularization procedures than patients treated by CABG, mainly related to the occurrence of restenosis. This phenomenon remains the Achilles’ heel of PTCA. Restenosis is relatively frequent, occurring in 35% to 40% of dilated lesions, is often associated with recurrence of symptoms requiring further revascularization and has a major economic impact. The mechanisms of restenosis have been extensively studied, and four major mechanisms have been identified: immediate recoil, incorporation of thrombus and a healing process in response to arterial injury that involves myointimal proliferation and vascular remodeling.
When the present study was designed, myointimal proliferation was considered to play a dominant role in the pathogenesis of restenosis. Furthermore, several lines of evidence suggested that the phenomenon of restenosis and that of atherosclerosis progression had some pathophysiologic characteristics in common. Several angiographic studies had described less progression and even some regression of atherosclerosis when lipid levels were lowered [10, 11]. In addition, experimental and clinical data showed a trend toward a lower restenosis rate when patients undergoing PTCA received fish oil supplementation [12–17]. In this context, the Prévention des Restenoses par Elisor après Dilatation Coronaire Transluminale (Prevention of Restenosis by Elisor After Transluminal Coronary Angioplasty [PREDICT]) study∗ was designed to determine whether treatment with pravastatin, an HMG coenzyme A reductase inhibitor, was able to reduce restenosis after PTCA.
1.1 Study design
From 1992 to 1994, 695 patients were enrolled in this prospective multicenter, randomized, double-blind, placebo-controlled trial. The protocol was approved by the ethics committee of the University of Lille. The last angiographic follow-up procedure was performed in December 1994. Twenty-nine French centers (see Appendix A) participated in the study.
Patients 25 to 75 years old, with a left ventricular ejection fraction, assessed by angiography, that exceeded 40% were eligible for inclusion. In addition, all patients were required to have total cholesterol levels between 200 and 310 mg/dl and triglyceride levels <500 mg/dl. Patients with a recent myocardial infarction (within 15 days) and patients who had previously undergone PTCA or CABG of the target vessel were excluded. Patients were also excluded if they received drugs not allowed by protocol (fish oil or other lipid-lowering agents within 1 month of the procedure), corticosteroids or immunosuppressive drugs.
Patients were recruited from those who had undergone successful, uncomplicated PTCA of one or more coronary stenoses in the participating institutions. Randomization was performed after PTCA and within 24 h of the procedure. After giving written informed consent, patients were randomly assigned to receive either pravastatin (40 mg/day) or placebo. All patients received aspirin (100 mg/day).
Two months after the procedure, patients returned for outpatient assessment and blood sampling for lipid measurements. At 6-month follow-up, lipid measurements and coronary angiography were performed. Follow-up angiography was performed earlier if there was a clinical indication. If follow-up angiography performed <4 months did not demonstrate restenosis, the patient was encouraged to return for further angiography at 6 months.
1.2 Lipid measurements
The following measurements were performed at baseline and at 2 and 6 months: total, low density lipoprotein (LDL), high density lipoprotein (HDL) and HDL subfraction E (HDL-E) cholesterol; triglycerides; lipoprotein(a) [Lp(a)]; lipoprotein E-B; and apolipoprotein A1 and B.
All lipid measurements were performed at a core laboratory (Service d’Etude et de Recherche sur les Lipoproteines et l’Atherosclerose [SERLIA], J. C. Fruchart, PhD, Institut Pasteur, Lille, Director). The technicians performing the measurements were unaware of the treatment allocation.
1.3 Angiographic measurements
Catheterization and PTCA were performed according to standard techniques. Isosorbide dinitrate (2 mg) was injected into the coronary artery before each angiogram in an attempt to standardize vasomotor tone. The angiograms were recorded on standard 35-mm film. Three views of the stenosis were obtained at the time of PTCA and were recorded on a worksheet to allow them to be duplicated exactly at the time of follow-up angiography. An attempt was made to obtain two orthogonal views for each lesion.
At the end of the study, films were sent to the core laboratory at the University of Lille (J. M. Lablanche, MD and E. McFadden, Directors) for qualitative and quantitative analysis. Angiographic analysis was performed without knowledge of treatment allocation or of clinical data. Quantitative analysis was performed on sequential angiograms filmed in the same projection. The frames were selected by the cardiologist who performed the quantitative analysis from the projection in which the stenosis appeared most severe just before angioplasty. Quantitative analysis was performed with the Computer-Assisted Evaluation of Stenosis and Restenosis (CAESAR) system, a computerized automatic analysis system that has been described in detail elsewhere . Briefly, the 35-mm cine film was projected with a 35AX projector (Tagarno, Denmark), and the cine frame selected for analysis was scanned with a high resolution video camera. The signal produced by the video camera was digitized and displayed on a video monitor. Regions of interest were chosen in the vessel, and a centerline was manually traced with use of a light pencil. The contours of the vessel were then automatically detected on the basis of the weighted sum of first- and second-derivative functions applied to the digitized brightness information. The diameter of the empty coronary catheter was used to convert the imaging data from pixels to millimeters. The mean diameters of proximal and distal reference segments and the minimal diameter of the stenotic segment were measured. The accuracy (defined as the signed difference between the measured and the true value) and the precision (defined as the standard deviation of these differences) of the CAESAR system were previously determined in a study analyzing cine films of Plexiglas blocks containing precision-drilled models of coronary arteries filled with contrast medium. The accuracy was 0.07 mm and the precision 0.14 mm. To assess the intraobserver and interobserver variability of the system, 90 arterial segments from patients undergoing PTCA were analyzed by two independent observers (J.M.L., E.P.M.) and reanalyzed at a remote time. The mean intraobserver variation, expressed as the standard deviation of the differences, was 0.10 mm, and the interobserver variation was 0.11 mm.
1.4 End points
Predetermined clinical and angiographic end points were assessed. The primary angiographic end points were minimal lumen diameter (MLD) at follow-up angiography, net gain and late loss in MLD at the dilated site. When more than one lesion was dilated in the same patient, the first lesion dilated was used for assessment of the primary angiographic end point. Secondary end points were the percent of patients with recurrence >50% stenosis at follow-up (at any dilated lesion) and the following clinical end points: occurrence of death, nonfatal target lesion myocardial infarction, CABG or repeat PTCA of the target lesion. Target lesion myocardial infarction was defined clinically at the participating site. These clinical end points constitute clinical restenosis.
Acute gainwas defined as the MLD at the dilated site immediately after PTCA minus the MLD just before the procedure. Late losswas defined as the MLD at the dilated site immediately after the procedure minus the MLD at the dilated site 6 months after PTCA. Net gainwas defined as the MLD at the dilated site 6 months after angioplasty minus the MLD at the dilated site just before angioplasty. Angiographic results given at follow-up are those obtained at 6 months, or earlier in case of early restenosis.
1.6 Data management and statistical analysis
The study adhered to the European Guidelines for Good Clinical Research Practice. Data were prospectively recorded by the investigators at each site on case record forms and were monitored by clinical research associates before data entry. The forms were verified by range and consistency checks, with queries returned to the investigators for any missing or inconsistent data. Clinical safety of the study drugs was evaluated, and adverse events occurring during the study were recorded in the case record form. At each center, research assistants monitored patient compliance with assigned therapy.
Statistical analysis was performed with use of SAS PC software (Version 6.04, SAS Institute). All tests were two-tailed, and p values <0.05 were considered significant. The predetermined primary end point of the study was the MLD at the dilated site 6 months after angioplasty. For this end point, it was calculated that a sample size of 313 patients/group was required to demonstrate a difference of 0.13 mm in MLD between the groups (allowing for a two-tailed alpha error of 0.05 and a beta error of 0.20). To allow for incomplete angiographic follow-up (estimated lost-to-follow-up rate of 10%), it was decided to randomize 690 patients. Two groups were defined: 1) all randomized patients (intention to treat group); 2) patients who had a follow-up angiogram that could be analyzed by the continuous MLD approach (per-protocol group) and who met angiographic criteria.
Baseline characteristics were compared in the two groups using the ttest, chi-square test or Fischer exact test, as appropriate. Clinical events occurring during follow-up were compared with the Mantel-Haenszel test on ordered categories. When more than one clinical event occurred per patient, the most severe event was used for the analysis with the following decreasing order of severity: death, nonfatal myocardial infarction, CABG and target vessel repeat PTCA. MLD and changes in MLD and percent diameter stenosis (immediate gain, late loss and net gain) were compared between groups by ttests. Two-way analysis of variance was performed to test changes in lipid variables. Continuous variables are expressed as mean value ± SD.
In total, 695 patients were randomized and make up the intention to treat cohort; 347 were assigned to pravastatin therapy and 348 to placebo. Because of adverse events (1 patient) and withdrawal of consent, 661 patients were seen at the 2-month follow-up visit. During subsequent follow-up, five patients had adverse events and stopped the treatment. Seventeen patients were lost to follow-up, and 14 refused the angiographic follow-up. In total, 625 patients (90% of the study group) had angiographic follow-up.
Taking into account patients excluded by the core laboratory (significant [>50%] residual stenosis after PTCA) or those unsuitable for quantitative analysis (n = 69), 556 patients (283 in the placebo group, 273 in the pravastatin group) were available for the per-protocol angiographic analysis. On average, the patients underwent follow-up angiography 184 ± 52 days after the procedure. There was no significant difference between the two groups (184 ± 55 days in the placebo group, 183 ± 47 days in the pravastatin group, p = 0.83).
2.2 Baseline clinical and angiographic characteristics
The two groups were well matched in terms of baseline clinical characteristics (Table 1). Mean age and the distribution of risk factors for coronary disease did not differ significantly between groups. The proportion of patients with a history of myocardial infarction was similar in both groups. A similar proportion of patients in both groups had stable angina or unstable angina (defined as severe chest pain at rest with electrocardiographic changes) or were asymptomatic.
The anatomic distribution and qualitative angiographic characteristics of the target lesions (assessed by the American College of Cardiology/American Heart Association classification) did not differ significantly between groups (Table 2).
2.3 Lipid measurements
At baseline, the mean cholesterol level was 231 ± 36 mg/dl in the placebo group and 228 ± 38 mg/dl in the pravastatin group (p = 0.42). The mean level of LDL cholesterol was 157 ± 29 mg/dl in the placebo group and 155 ± 32 mg/dl in the pravastatin group (p = 0.3); mean HDL cholesterol was 47 ± 12 mg/dl in the placebo group and 47 ± 13 mg/dl in the pravastatin group (p = 0.8) (Table 3).
During the study there were no significant changes in lipid variables in the placebo group. In contrast, there was a significant decrease in mean total cholesterol in the pravastatin group after 2 and 6 months of treatment compared with the mean baseline value. Mean LDL cholesterol was also significantly lower at 2 and 6 months than at baseline. There was a significant increase in HDL cholesterol at 2 and 6 months compared with baseline values (Table 3). Apolipoprotein A1 significantly increased in the pravastatin group, whereas apolipoprotein B significantly decreased. Lp(a) levels were slightly but not significantly greater at 6 month follow-up (Table 4).
2.4 Angiographic results
The major results of the quantitative angiographic analysis are presented in Table 5. There were no significant differences at baseline in reference diameter or in MLD between groups. The MLD was 0.84 ± 0.28 mm in the placebo group and 0.82 ± 0.29 mm in the pravastatin group (p = 0.37).
At follow-up, the MLD, which was the primary angiographic end point, was 1.47 ± 0.62 mm in the placebo group and 1.54 ± 0.66 mm in the pravastatin group (p = 0.21). The cumulative distribution curves for MLD are shown in Fig. 1. The late loss in MLD was similar in both groups (0.48 ± 0.56 mm in the placebo group, 0.46 ± 0.58 mm in the pravastatin group, p = 0.54), as was the net gain (0.62 ± 0.59 mm in the placebo group, 0.71 ± 0.62 mm in the pravastatin group, p = 0.07).
A secondary angiographic end point was percent stenosis. At baseline, mean stenosis severity at the PTCA site was similar (72 ± 8.8%) in the pravastatin and placebo groups (71.1 ± 8.1%, p = 0.11). After PTCA, mean residual stenosis was 33 ± 8.5% in the placebo group and 33 ± 8.5% in the pravastatin group (p = 0.49). At angiographic restudy, there was no significant difference between the two groups, with mean stenosis severity 49.7 ± 20% in the placebo group and 48.4 ± 20% in the pravastatin group (p = 0.44). Furthermore, the proportion of patients with recurrence of significant (>50%) stenosis did not differ significantly (p = 0.26) between groups (43.8% in the placebo group, 39.2% in the pravastatin group).
2.5 Clinical restenosis
The analysis, based on the intention to treat cohort, included 695 patients (348 in the placebo group, 347 in the pravastatin group). There were four deaths in the pravastatin group: two sudden deaths plus one related to myocardial infarction and one related to a cerebrovascular accident. One sudden death occurred in the placebo group. The mortality rate did not differ significantly between groups (p = 0.21).
Four nonfatal myocardial infarctions occurred in each group. Target vessel revascularization during follow-up was performed in 75 patients (21.6%) receiving placebo and 66 (19%) treated with pravastatin. Thus, there was no overall difference in clinical restenosis between the groups (Table 6).
2.6 Relation between lipid measurements and angiographic restenosis
There was no relation between late loss in MLD and changes in lipid variables associated with pravastatin treatment.
The present study was designed to determine whether treatment with pravastatin could reduce angiographic restenosis after coronary balloon angioplasty. Angiographic restenosis was assessed using both a continuous approach (analysis of MLD) and a categoric approach (recurrence of significant [>50%] stenosis after initially successful PTCA). Both analyses demonstrated conclusively that pravastatin had no significant effect on angiographic restenosis.
3.1 Present study: clinical and angiographic outcome
Pravastatin treatment was associated with a significant reduction in total and LDL cholesterol and with a slight but significant increase in HDL cholesterol. However, pravastatin treatment had no significant effect on either angiographic restenosis or rate of clinical events after PTCA. The incidence of death, nonfatal myocardial infarction or target lesion revascularization was similar in both groups. The calculations of sample size were based on the number of patients required to demonstrate, with adequate power, an effect of treatment on angiographic restenosis after PTCA; a much larger sample size, as well as a longer period of follow-up, would be required to assess a potential effect on clinical outcome. Pravastatin did not reduce clinical restenosis in the present trial, but this finding is not at variance with the results of previous studies.
3.2 Statins and progression of atherosclerosis: previous studies
Previous studies that examined the effects of the statins on the occurrence of clinical events in patients with coronary atherosclerosis demonstrated a reduction in clinical events in patients receiving statin therapy. This benefit was observed in different populations and with different molecules, notably in the Scandinavian Simvastatin Survival Study (4-S) , the West of Scotland Coronary Prevention Study (WOSCOPS) and the CARE study . In the 4S study, the actuarial survival curves diverged only after 12 months and in the CARE study after 2 years.
Several angiographic trials have found that lipid-lowering drugs were able to limit progression and even to induce regression of atherosclerotic lesions. This was demonstrated in the Regression Growth Evaluation Statin Study (REGRESS) study that compared pravastatin with placebo over a longer time period. This question was not addressed in the PREDICT trial, which was designed to assess the effect of pravastatin on restenosis and not its effects on progression or regression in nondilated segments.
3.3 Lipid-lowering therapy: effects on restenosis
Previous studies that examined the effects of lipid-lowering drugs on restenosis have produced conflicting results. A potential beneficial effect of the statins on the occurrence of restenosis was suggested by a relatively small study that included 157 patients who received lovastatin or conventional care. That study , which was neither blinded nor randomized, reported that the rate of restenosis was 12% with lovastatin compared with 44% with conventional care; however, angiographic follow-up was incomplete. Subsequently, a small randomized trial with pravastatin in a Japanese population demonstrated no effect on angiographic restenosis.
A large randomized trial, the Lovastatin Restenosis Trial (LRT) , examined the effect of lovastatin on restenosis. This well conducted and well designed study showed that treatment with a relatively high dose of lovastatin had no effect on the occurrence of restenosis at 6 months after PTCA. The results of the present study are similar to the results of the LRT. There were two methodologic differences between the studies, namely, the fact that in the LRT treatment was begun between 7 and 10 days before angioplasty, and the dose of statin used was higher (40 mg of lovastatin orally twice daily) compared with 40 mg/day of pravastatin in the present study. The population studied was somewhat larger in the PREDICT study (695 patients) than in in the LRT (404 patients). The reference diameters in the dilated vessels were slightly smaller in the LRT than in the PREDICT study, but the conclusions concerning the continuous and categoric variables were essentially the same. No trials to date have dealt with the effects of simvastatin or fluvastatin, statins that have been shown in vitro to have smooth muscle cell migration inhibitory properties on restenosis.
Other recent studies [25–27]have demonstrated a lack of association between serum lipid levels and restenosis and the lack of effect of lipid-lowering agents or fish oil supplementation on restenosis.
The results of the present study, together with those of the other studies discussed, demonstrate that despite its positive effects on the angiographic and ultrasound progression of atherosclerosis, pravastatin has no effect on angiographic or clinical restenosis 6 months after coronary angioplasty despite a significant reduction in total and LDL cholesterol. These results should not be interpreted to mean that statin therapy is of no benefit in patients undergoing angioplasty. Although treatment with statins appears to have no effect on restenosis, their effect on the progression of atherosclerosis and on the clinical events associated with such progression now appears to have been established beyond reasonable doubt. Owing to the contrasting effects of lipid lowering on atherosclerosis and restenosis, it seems unlikely that restenosis and the progression of atherosclerosis have a similar pathophysiologic basis. Current evidence suggests that restenosis is related to an exaggerated healing process combined with vascular remodeling.
A.1 Participating Centers and Principal Investigators for the PREDICT Trial3
Clinique St. Martin, Caen(83): P. Commeau; Centre Hospitalier Universitaire (CHU), Caen(72): G. Grollier; CHU, Besançon(61): J. P. Bassand; CHU, Grenoble(49): J. Machecourt; CHU, Clermont-Ferrand(48): J. Cassagnes; CHU, Strasbourg(38): J. M. Mossard; CHU Necker, Paris(35): A. Vacheron, J. P. Metzger; CHU, Creteil(31): A. Castaigne; CHU, Nancy(30): N. Danchin; CHU, Broussais, Paris(27): J. L. Guermonprez, S. Makowski; CHU, Lille(26): M. Bertrand, C. Bauters; CHU, Brest(24): J. Boschat; CHU Ambroise Paré, Paris(21): J. P. Bourdarias; Centre Hospitalier Régional, Mulhouse(20): J. P. Monassier; CHU, Amiens(19): J. C. Quiret; CHU, Tours(16): P. Raynaud; Institut Arnaut Tzanck, Nice(15): R. Schmitt; CHU, Nantes(15): D. Crochet; CHU, Rennes(13): J. C. Pony, H. Le Breton; CHU Timone, Marseille(11): M. Bory; CHU, Pitié Salpétrière, Paris(11): G. Drobinski; CHU, Rouen(9): B. Letac; CHU Bichat, Paris(8): M. C. Aumont, D. Himbert; CHU, Angers(6): P. Geslin; CHU, Dijon(3): J. E. Wolf; CHU Purpan, Toulouse(3): P. Bernadet; CHU, Bordeaux(2): P. Besse; Clinique Volney, Rennes(2): P. Descaves; CHU Tenon, Paris(1): A. Vahanian.
↵2 Elisor is the registered trade name of Pravastatin.
↵fn1 This study was supported by a grant from Bristol Myers Squibb Laboratories, Paris.
↵3 Numbers in parentheses are number of study patients; all participating centers are in France.
- coronary artery bypass graft surgery
- Cholesterol and Recurrent Events (trial)
- Computer-Assisted Evaluation of Stenosis and Restenosis system
- high density lipoprotein
- low density lipoprotein
- Lovastatin Restenosis Trial
- minimal lumen diameter
- Prevention of Restenosis by Elisor After Transluminal Coronary Angioplasty (trial)
- percutaneous transluminal coronary angioplasty
- Received January 6, 1997.
- Revision received June 17, 1997.
- Accepted June 26, 1997.
- The American College of Cardiology
- RITA Trial Participants
- Rodriguez A,
- Boullon F,
- Perez-Balino N,
- et al.
- Jukema JW,
- Bruschke AV,
- van Boven AJ,
- et al.
- Leaf A,
- Jorgensen MB,
- Jacobs AK,
- et al.
- Grigg LE,
- Kay TW,
- Valentine PA,
- et al.
- Bairati I,
- Roy L,
- Meyer F,
- et al.
- Violaris AG,
- Melkert R,
- Serruys PW,
- et al.