Author + information
- Received October 28, 1999
- Revision received January 28, 2000
- Accepted March 30, 2000
- Published online August 1, 2000.
- Antoine Lafont, MD, PhD∗,* (, )
- Jean L Dubois-Randé, MD, PhD†,
- Philippe G Steg, MD‡,
- Patrick Dupouy, MD†,
- Didier Carrié, MD§,
- Pierre Coste, MD∥,
- Alain Furber, MD¶,
- Farzin Beygui, MD∗∗,
- Laurent J Feldman, MD, PhD‡,
- Saliha Rahal, MD∗,
- Christophe Tron, MD††,
- Martial Hamon, MD‡‡,
- Gilles Grollier, MD§§,
- Philippe Commeau, MD∥,
- Pascal Richard, MD¶¶,
- Patrice Colin, MD∗∗∗,
- Christophe Bauters, MD†††,
- Gaetan Karrillon, MD‡‡,
- François Ledru, MD§§,
- Bernard Citron, MD∥,
- François Noel Marié, MD∗,
- Morton Kern, MD¶¶,
- F.R.O.S.T. Study Group
- ↵*Reprint requests and correspondence: Prof. Antoine Lafont, Cardiology Department, Hopital Boucicaut, University Paris V, 78, rue de la Convention, Paris 75015, France.
We sought to make a prospective comparison of systematic stenting with provisional stenting guided by Doppler measurements of coronary velocity reserve and quantitative coronary angiography.
Despite the increasing use of stents during percutaneous transluminal coronary angioplasty, it is unclear whether systematic stenting is superior to a strategy of provisional stenting in which stents are placed only in patients with unsatisfactory results or as a bail-out procedure.
Two hundred fifty-one patients undergoing elective coronary angioplasty were randomly assigned either to provisional stenting (group 1, in which stenting was performed if postangioplasty coronary velocity reserve was <2.2 and/or residual stenosis ≥35% or as bail-out) or to systematic stenting (group 2). The primary end point was the six-month angiographic minimal lumen diameter (MLD). Major adverse cardiac events were secondary end points (death, acute myocardial infarction and target lesion revascularization).
Stenting was performed in 48.4% of patients in group 1 and 100% of patients in group 2 (p < 0.01). Six months after angioplasty, the MLD did not differ between groups (1.90 ± 0.79 mm vs. 1.99 ± 0.70 mm, p = 0.39), as was the rate of binary restenosis (27.1% vs. 21.4%, p = 0.37). Among patients with restenosis, 13/32 (40.6%) in group 1 but 100% (25/25) in group 2 had in-stent restenosis (p < 0.01). Target lesion revascularization (15.1% vs. 14.4% in groups 1 and 2 respectively, p = 0.89) and major adverse cardiac events (15.1% vs. 16.0%, p = 0.85) were not significantly different.
Systematic stenting does not provide superior angiographic results at six months as compared with provisional stenting.
Coronary stenting is one of the major advances in percutaneous transluminal coronary angioplasty (PTCA), allowing reduction of the short-term complications as well as the incidence of restenosis (1,2). This has led to a rapid and steep increase in the use of stents for PTCA. It remains unclear whether stents should be systematically implanted during PTCA or whether a strategy of “provisional stenting” is acceptable, in which stents are placed in case of unsatisfactory immediate results of balloon angioplasty. Specifically, the benefits of stenting may be partially offset by the high recurrence rate of in-stent restenosis when treated by repeat percutaneous coronary intervention (3). Conversely, the combination of quantitative coronary angiography (QCA) and coronary velocity reserve (CVR) measurements after PTCA allows identification of a subset of patients with an excellent clinical and angiographic outcome without the need for systematic stenting (4). Thus, the current trial sought to make a prospective comparison of systematic stenting with a policy of provisional stenting guided by QCA and CVR measurements with respect to angiographic outcome and major cardiac adverse events.
Selection of patients
The study population consisted of patients with proven myocardial ischemia and de novo lesions of the native coronary circulation scheduled for PTCA of a single discrete lesion of ≤15 mm length and a vessel reference diameter ≥2.7 mm by local QCA. Exclusion criteria were age >75 years, acute myocardial infarction (AMI) within the previous three weeks, left ventricular ejection fraction <50%, total occlusion of the target vessel, left main coronary artery stenosis ≥50%, abnormal wall motion in the area of the target vessel, hypertrophic cardiomyopathy and more than one lesion on the target vessel for optimal interpretation of the Doppler findings (4).
After providing written informed consent, the patients were randomly assigned to two groups using a sealed envelope system: in group 1 (provisional stenting) PTCA was performed using the local routine, and the final result was assessed by QCA and CVR measurements using a Doppler wire (Flowire, Endosonics Corporation, Rancho Cordova, California). Patients in whom the residual diameter stenosis (DS) was <35% and CVR was ≥2.2 did not undergo stent placement. The choice of CVR ≥2.2 was based upon a previous pilot study (5) and is within the range of values chosen by the Doppler End Points Balloon Angioplasty Trial Europe (DEBATE) II and Doppler End Point Stenting International Investigation Coronary Flow Reserve (DESTINI) trials (2.5 and 2.0, respectively [6,7]). Patients who failed to fulfill both of these criteria underwent stenting. The steps taken toward improving the suboptimal outcome were left to the discretion of the operator. Additionally, patients with acute or threatened closure of the target vessel received a stent as a “bail-out” procedure. In group 2 (systematic stenting), patients underwent balloon angioplasty followed by systematic stent placement using a PS-153 stent (Johnson and Johnson Interventional Systems, Warren, New Jersey) without measurement of CVR and regardless of the angiographic result achieved by balloon angioplasty alone. Stents were placed using delivery pressures >10 atmospheres.
In group 1, a 0.014 in. Doppler-tipped wire was used as the angioplasty guidewire. Baseline and hyperemic flow velocity measurements were obtained before and after angioplasty at the same guidewire location. Coronary velocity reserve was defined as the ratio between maximal velocity during hyperemia induced by an intracoronary bolus of adenosine (12 μg for the right coronary artery and 18 μg for the left coronary artery) and rest velocity as previously described (4,8). The Doppler velocity signal was continuously recorded on videotape for off-line analysis.
All patients were on aspirin (250 mg daily) and received intravenous heparin 100 IU/kg at the onset of the procedure. Patients in whom a stent was placed additionally received ticlopidine 500 mg daily for four weeks after the procedure.
All measurements were made after intracoronary injection of 1 mg of linsidomine (except in patients in whom systolic blood pressure was <100 mm Hg).
All cinefilms were analyzed by an independent core laboratory, blinded to all clinical and Doppler data (Hôpital Pitié-Salpètrière, Paris, France). Two orthogonal views of the target lesion at each step of the procedure (pre-PTCA, after balloon angioplasty and after stenting) were selected for off-line quantitative analysis. Computer-assisted quantitative angiographic analysis was performed using a validated automatic edge-detection algorithm, as previously described (9). A 6F guiding catheter free of contrast was used for calibration. The minimal luminal diameter (MLD) and reference diameter proximal to the lesion were measured as previously described (9). The inter- and intraobserver variations for QCA DS were 9 ± 12% and 7 ± 14%, respectively.
Six months after the procedure, all patients were evaluated clinically and underwent repeat coronary angiography, unless it had been clinically required after the initial three months, with repeat QCA measurements, duplicating the views and methods of the initial assessment.
The primary end point was the final MLD at the target site measured by QCA at six-month follow-up. Secondary end points were: 1) the binary restenosis rate defined as ≥50% residual stenosis at the six-month angiographic follow-up, and 2) the incidence of major adverse cardiac events (death, myocardial infarction, revascularization of the target lesion) at the six-month follow-up. Myocardial infarction was defined as two or more of the three following criteria: new Q waves, chest pain of >30 min duration, elevation of creatinine phosphokinase to more than two-fold the normal level associated with elevated CK-MB fraction within the 24 h after angioplasty.
The protocol was reviewed and approved by the Ethics Committee of the Hôpital Boucicaut (Paris, France).
Analysis was based on intent to treat with comparison of the six-month MLD of group 1 (provisional stenting) and group 2 (systematic stenting). Sample size was calculated, using two-sided testing, an alpha risk of 0.05 and a beta risk of 0.20. Assuming a difference between groups of at least 0.22 mm and a six-month MLD of 1.9 ± 0.6 mm based on previous quantitative angiographic studies of stents in similar patient populations, 117 patients were required per group (2,10,11). Assuming a 10% dropout rate, the sample size required was 251 patients.
All quantitative values were given as mean ± standard deviation (SD). Comparisons between groups used unpaired t tests for quantitative variables and chi-square test and Fisher’s exact test for qualitative variables, as appropriate. All p values were two-tailed. Data acquisition and analyses were performed using the SAS statistical software (Proc Access of SAS) in order to compare double data acquisition (Proc Compare, SAS). All differences detected were reviewed with the investigators and the Critical Event Committee and performed using the SAS statistical software (SAS 6.12, SAS Institute Inc., SAS Cary, North Carolina).
Characteristics of patients
Between November 1996 and April 1998, 253 patients were enrolled in the trial by 17 french hospitals: 126 patients were randomly assigned to provisional stenting (group 1) and 127 to systematic stenting (group 2). After randomization, two patients (group 2) were excluded from analysis for major protocol violations (one for elective PTCA of two lesion sites on the same vessel and one for recent myocardial infarction). The baseline characteristics of the patients are displayed in Tables 1 and 2. ⇓⇓ Most of the patients (64.5% of the total cohort) underwent PTCA after unstable angina. The procedural results are displayed in Table 3.
In group 1, Doppler signals were acquired in all patients but one (99.2%) in whom no correct signal could be obtained. There was no complication related to the Doppler device. Sixty-five patients (51.6%) fulfilled both the angiographic and Doppler criteria for successful angioplasty after balloon PTCA and did not receive a stent. Conversely, 61 patients (48.4%) needed a stent either as a bail-out measure (17 patients, 13.5%) or due to failure to meet the CVR and QCA criteria for success after balloon angioplasty alone (44 patients, 34.9%). Among these 44 patients, Doppler was involved in the decision to stent in 75%: 20 (45.5%) for failure to fulfil the CVR criteria alone and 13 (29.5%) for failure to fulfill both CVR and QCA criteria; 11 (25%) failed to fulfil the QCA criteria alone. In group 1, there was no significant difference in heart rate and arterial pressure before and after PTCA (124.2 ± 22.7 mm Hg vs. 125.6 ± 23.3 mm Hg; p = 0.81 and 67.0 ± 15.2 vs. 66.1 ± 11.9, p = 0.77).
In group 2, stent placement was successfully performed in 100% although two patients required another stent than the PS153, due to failure of the device to cross the lesion. Ultimately, 48.4% of the patients in group 1 received a stent as opposed to 100% in group 2 (p < 0.0001).
The baseline MLD and DS were comparable between groups 1 and 2 (Table 3). There was a major reduction in the degree of lesion stenosis in both groups; as expected, this was more marked in group 2. The larger balloon size and balloon/artery ratio in group 1 reflected attempts to optimize results of balloon angioplasty.
During the in-hospital stay, major adverse cardiac events were identical between groups: there were two myocardial infarctions (1.6%) in each group, no death and no need for target vessel revascularization (bypass surgery, re-PTCA).
In group 1 there was an increase in CVR after the procedure in both unstented patients (from 1.7 ± 0.7 to 3.0 ± 0.7, p < 0.0001) and stented patients (from 1.6 ± 0.6 to 2.5 ± 0.9, p < 0.0001).
Clinical follow-up was obtained in all the patients. Repeat coronary angiography was performed in 93.6% of the patients in group 1 and 92.1% in group 2, on average 25.5 ± 8.8 weeks after the initial procedure. At follow-up, there were three deaths in group 2, one due to sudden death, one secondary to myocardial infarction, one occurring after elective coronary artery bypass surgery. There was one myocardial infarction in group 1 and two in group 2. The six-month angiographic results are summarized in Table 4. The final angiographic outcome was similar between groups, both in terms of MLD, DS and late loss. The binary restenosis rate was similar between groups (27.1% vs. 21.4%, in groups 1 and 2 respectively, p = 0.37). This was consistent with the lack of difference in terms of target lesion revascularization between groups (15.1% vs. 14.4% in groups 1 and 2, respectively, p = 0.89). However, the study was not powered for detecting meaningful clinical differences. Among patients with restenosis, 13/32 (40.6%) in group 1 but 100% (25/25) in group 2 had in-stent restenosis (p < 0.01).
The strategy of systematic stenting did not appear to provide significantly superior angiographic results at six months compared with a strategy of provisional stenting guided by QCA and CVR assessment. The latter strategy appeared safe because it was not associated with an excess of adverse clinical events and achieved similar six-month results while obviating the need for stenting in 51.6% of the patients (without suppressing it) although the study was powered for detecting only angiographic differences. Need for stenting was comparable with DEBATE II and DESTINI trials (55% and 45%, respectively [6,7]). Consequently, there was much less in-stent restenosis in the provisional stenting group than there was in the systematic stent group (13 vs. 25 patients, that is, 40.6 vs. 100% of the restenotic patients, p < 0.01).
Restenosis occurring after PTCA remains the major limitation of interventional cardiology (12,13). While stents have been shown to reduce its incidence in selected lesion subsets, a new problem has emerged with the high recurrence rate of in-stent restenosis when treated with repeat percutaneous coronary intervention (14). In fact, in-stent restenosis appears refractory to conventional balloon angioplasty techniques, leading to research into radiation therapy, gene therapy and local drug delivery as possible therapeutic avenues to prevent its recurrence (15–18). Therefore, it is worthwhile to be able to identify patients with satisfactory PTCA results who may derive no benefit from systematic stenting.
Strategies of provisional stenting
Real-time assessment of the immediate outcome of percutaneous intervention remains difficult because angiography has important limitations (19). Even after plain balloon angioplasty, the majority of patients do not require repeat revascularization and would derive no benefit from additional stenting. This was the message proposed by DEBATE-I, an observational study including only nonstented patients; patients with a good distal CVR and minimal residual stenosis after balloon angioplasty had a low six-month restenosis rate and a clinical follow-up similar to those of stented patients from contemporary randomized trials (1,2,4). Subsequently, it has been shown that, after successful conventional balloon angioplasty, optimization maneuvers, such as stenting or using bigger balloons, were able to improve CVR even when the initial angiographic result of PTCA appeared satisfactory (6). This suggests that Doppler assessment of CVR is able to detect angiographically invisible alterations of the vessel, which impact on long-term outcome despite CVR self limitations (that is, left ventricular hypertrophy, cardiomyopathy, presence of collaterals).
This provides the background for provisional stenting. This study is the first to evaluate this strategy using physiological assessment of the PTCA result on top of QCA. The results appear to validate the concept since provisional stenting achieved an MLD at six months similar to systematic stenting. It is reassuring that the attempt to achieve a satisfactory CVR according to predefined criteria in the provisional stenting group, reflected by the higher balloon size and balloon/artery ratio, was not associated with an increased incidence of periprocedural complications or a high rate of bailout stenting. As expected, the immediate angiographic result appeared more favorable with systematic stenting than with provisional stenting. However, increased late loss with systematic stenting ultimately resulted in similar clinical outcome in the DESTINI trial and angiographic outcome at six months in both groups (7).
The MLD was selected as the primary end point, which allows the study of restenosis using a continuous parameter rather than a dichotomic one, better reflecting the biologic phenomenon. However, the trial was not powered to detect meaningful differences in binary restenosis or clinical outcomes. By design, this pragmatic trial could not be blinded. However, possible biases in evaluation of the results were limited by angiographic core laboratory analysis. Despite randomization, there may have been imbalances in patients characteristics, including nonsignificant differences in left anterior descending artery distribution. This study was designed for an angiographic primary end point. From an angiographic point of view, restenosis is independent of the location of the dilated lesion. Thus, this particular point has a minimal impact on the interpretation of the primary end point of the study. The trend towards more frequent left anterior descending artery lesions in the systematic stenting group could have favored the provisional stenting group in terms of clinical follow-up, which was a secondary end point of our study. Mean pressure for stent deployment was 12 ± 1.9 atmospheres. In a recent study evaluating the respective interest of intracoronary ultrasound and intracoronary pressure measurements to optimize stent placement, the best results were obtained with a mean inflation pressure of 11.8 atmospheres (20). We can, thus, consider that the quality of stent placement would not gain any benefit from either IVUS or pressure guidance in this study. This study did not aim to evaluate the cost/effectiveness of the respective strategy. However, it has to be pointed out that the cost of the catheters and equipment for CVR added to 48.4% stenting was obviously superior than the cost of 100% stenting since a Doppler catheter was as expensive as a stent. The procedure time of the guided strategy has to be mentioned, namely with regard to possibly longer fluoroscopic exposure. The systematic angiographic follow-up may have increased the target lesion revascularization rates although these are similar to those reported by the DESTINI trial, which did not have angiographic follow-up. In addition, any potential bias derived from the practice of systematic reangiography at six months must have affected both groups in a similar way. The relatively short follow-up may have missed some differences in outcome related to the high propensity for recurrence of in-stent restenosis (14,21,22). However, since there was more in-stent restenosis in group 2 than in group 1, presumably more late events would occur in group 2. Therefore, this short follow-up period would tend to underestimate the benefit of provisional stenting. The criteria used for patient and lesion selection were somewhat restrictive and, therefore, any extrapolation of the results to a population at higher risk and more severe lesions may be questioned. Nevertheless, it is noteworthy that the vast majority of patients had an unstable clinical presentation, and angiographic selection criteria were similar to those of contemporary stent trials (17,23).
This study evaluated the strategy of provisional stenting guided by anatomic and functional criteria. It appears that systematic stenting does not provide superior angiographic results at six months as compared with provisional stenting. There is a dramatically lower rate of in-stent restenosis occurring after provisional stenting compared with systematic stenting (40.6% vs. 100%), which may impact the long-term clinical outcome. (Appendix)
Frost study group
Steering Committee: A. Lafont (chairman), J.L. Dubois-Randé, P. Dupouy, P.G. Steg.
Critical Events Committee: S. Dufourmantelle, P.G. Steg, S. Weber.
Angiographic Assessment Committee: F. Addad, N. Danchin, G. Montalescot.
Data-Coordinating and Analysis Center and Quantitative Angiographic Core Laboratory: S. Bastugi, G. Drobinski, S. Dufourmantelle, C. Gousset, F. Veyssière.
Participating Hospitals and Investigators: Hôpital Henri Mondor, Créteil, France (J.L. Dubois-Randé, P. Dupouy), Hôpital Purpan, Toulouse, France (D. Carrié, J. Puel), Hôpital Bordeaux-Pessac, Bordeaux, France (P. Besse, P. Coste), Hôpital d’Angers, Angers, France (A. Furber), Hôpital Bichat, Paris, France (L.J. Feldman, P.G. Steg), Hôpital Boucicaut, Paris, France (A. Lafont, S. Rahal), Hôpital Necker, Paris, France (F. Beygui, J.P. Metzger), Hôpital Charles Nicolle, Rouen, France (A. Cribier, C. Tron), Hôpital Emile Muller, Mulhouse, France (M. Hamon, J.P. Monassier), Hôpital de la Côte de Nacre, Caen, France (G. Grollier), Clinique Saint Martin, Caen, France (P. Commeau, J.F. Morelle), Hôpital de Brabois, Nancy, France (N. Danchin, P. Richard), Hôpital Antoine Béclère, Paris, France (P. Colin, M. Slama), Hôpital Cardiologique de Lille, Lille, France (C. Bauters, E. van Belle), Hôpital Lariboisière, Paris, France (G.J. Karrillon, V. Stratiev), Hôpital Broussais, Paris, France (F. Ledru), Hôpital Gabriel Montpied, Clermont-Ferrand, France (B. Citron, L. Sarfati).
☆ This study was partially supported by a grant from Cordis/Johnson, Johnson Interventional Systems, France and Endosonics Corporation, Europe and University Paris V.
- acute myocardial infarction
- coronary velocity reserve
- Doppler End Points Balloon Angioplasty Trial Europe
- Doppler End Point Stenting International Investigation Coronary Flow Reserve
- diameter stenosis
- minimal lumen diameter
- percutaneous transluminal coronary angioplasty
- quantitative coronary angiography
- standard deviation
- Received October 28, 1999.
- Revision received January 28, 2000.
- Accepted March 30, 2000.
- American College of Cardiology
- Mehran R,
- Mintz G.S,
- Satler L.F,
- et al.
- Serruys P.W,
- di Mario C,
- Piek J,
- et al.
- ↵Dupouy P, Pelle G, Garot P, et al. Physiologically-guided angioplasty in support of provisional stenting: immediate and six-month outcome. Cathet Cardiovasc Intervent 2000;49:369–75.
- de Bruyne B,
- Groothuis W,
- Sousa E,
- et al.
- Di Mario C,
- Moses J,
- Anderson T,
- et al.
- Kern M.J,
- Bach R.G,
- Mechem C.J,
- et al.
- Montalescot G,
- Ankri A,
- Vicaut E,
- Drobinski G,
- Grosgogeat Y,
- Thomas D
- Serruys P.W,
- Emanuelsson H,
- van der Giessen W,
- et al.
- Currier J.W,
- Faxon D.P
- De Young M.B,
- Dichek D.A
- de Feyter P.J,
- Serruys P.W,
- Davies M.J,
- Richardson P,
- Lubsen J,
- Oliver M.F
- Hanekamp C.E,
- Koolen J.J,
- Pijls N.H,
- Michels H.R,
- Bonnier H.J
- Bauters C,
- Banos J.L,
- Van Belle E,
- Mc Fadden E.P,
- Lablanche J.M,
- Bertrand M.E
- Eltchaninoff H,
- Koning R,
- Tron C,
- Gupta V,
- Cribier A
- Ryan T.J,
- Faxon D.P,
- Gunnar R.M,
- et al.