|
|
| |||||||||||||||||||||||||||||||
Background Restenosis is a major limitation of coronary angioplasty. Directional coronary atherectomy was developed with the expectation that it would provide better results than angioplasty, including a lower rate of restenosis. We undertook a randomized, multicenter trial to compare the rates of restenosis for atherectomy and angioplasty when used to treat lesions of the proximal left anterior descending coronary artery.
Methods Of 274 patients referred for first-time, nonsurgical revascularization of lesions of the proximal left anterior descending coronary artery, 138 were randomly assigned to undergo atherectomy and 136 to undergo angioplasty; 257 of 265 eligible patients (97 percent) underwent follow-up angiography at a median of 5.9 months. Computer-assisted quantitative measurements of luminal dimensions were determined from the angiograms obtained before and immediately after the procedure and at follow-up. The primary end point of restenosis was defined as stenosis of more than 50 percent of the vessel's diameter at follow-up.
Results Quantitative analysis showed that the procedural success rate was higher in patients who underwent atherectomy than in those who had angioplasty (94 percent vs. 88 percent, P = 0.061); there was no significant difference in the frequency of major in-hospital complications (5 percent vs. 6 percent). At follow-up, the rate of restenosis was 46 percent after atherectomy and 43 percent after angioplasty (P = 0.71). Despite a larger initial gain in the minimal luminal diameter with atherectomy (mean [±SD], 1.45 ±0.47 vs. 1.16 ±0.44 mm; P<0.001), there was a larger late loss (0.79 ±0.61 vs. 0.47 ±0.64 mm, P<0.001), resulting in a similar minimal luminal diameter in the two groups at follow-up (1.55 ±0.60 vs. 1.61 ±0.68, P = 0.44). The clinical outcomes at six months were not significantly different between the two groups.
Conclusions the role of atherectomy in percutaneous coronary revascularization remains to be fully defined. However, as compared with angioplasty, atherectomy did not result in better late angiographic or clinical outcomes in patients with lesions of the proximal left anterior descending coronary artery. (n Engl J Med 1993;329:228-33.).
Methods
Participating Centers and Investigators
The coordinating center and angiographic core laboratory were at Mount Sinai Hospital, Toronto. All interventional cardiology centers in Canada performing more than 350 coronary angioplasties annually were invited to participate. All investigators had at least two years of experience with angioplasty and had used the Simpson atherectomy device for at least four months for 20 or more procedures, with a success rate exceeding 80 percent and a complication rate below 10 percent. All study procedures were carried out by the investigators at the nine participating centers listed in the Appendix. The study was approved by the institutional review board at each site.
Patient Selection, Recruitment, and Randomization
Prospective subjects for the trial were identified at the time of referral for percutaneous revascularization; decisions regarding eligibility were made by the local investigators on the basis of visual analysis of the diagnostic angiogram. Eligible patients included those with angina or objective evidence of myocardial ischemia and a stenosis of 
60 percent of the vessel's diameter in the proximal third of the left anterior descending coronary artery that was suitable for either atherectomy or angioplasty. Patients with restenosed lesions were not considered in this study. Specific angiographic exclusion criteria related to the characteristics of the lesion (length of >10 mm; involvement of the ostium or of a branch vessel measuring 2.5 mm in diameter) and the vessel (total occlusion; size of <3 mm; heavy calcification or severe tortuosity; or stenosis of the left main coronary artery exceeding 25 percent). Patients with acute myocardial infarction (within one week of the procedure), severe left ventricular dysfunction, or cardiogenic shock were excluded. In addition, patients with medical conditions likely to preclude follow-up angiography (such as renal insufficiency), those participating in a concurrent study, and those unable to give informed consent were not enrolled.
A log was kept of all eligible, unenrolled patients at each site. In addition, a registry was maintained documenting the angiographic and clinical reasons for excluding patients undergoing dilation of stenoses of the left anterior descending coronary artery. All atherectomy procedures at participating sites were recorded during the study. Randomization, stratified according to center with a block design, was carried out by means of sealed envelopes in the catheterization laboratory after the initial set of angiograms was obtained. The actual treatment assignments were cross-checked against the computer-generated randomization sequence.
Angiographic Protocol
Because of recognizable differences in the guiding catheters used for atherectomy and angioplasty, angiograms performed before and immediately after the procedures were obtained with 7-French or 8-French diagnostic catheters to ensure that the selection of frames for quantitative coronary analysis would be blinded. At least two orthogonal views of the target lesion were obtained, including the view demonstrating the most severe stenosis. The obliquity and angulation of each view were recorded for later duplication in the post-procedure and follow-up angiograms. The distal 15 cm of the diagnostic catheter was sterilized and retained for measurement. If multiple stenoses were to be dilated, the lesion that was the focus of this study was dilated first. The first three study angiograms from each site were reviewed by the angiographic core laboratory to monitor adherence to the protocol.
Atherectomy and Angioplasty
All the patients received aspirin, a calcium-channel blocker, and nitrates beginning at least 12 hours before the procedure and continuing for 24 hours afterward. Intracoronary nitroglycerin (100 to 200 µg) was injected just before angiography, both before and after the procedure and at follow-up. Heparin was given after the sheath was inserted and was repeated as necessary in doses sufficient to maintain an activated clotting time of more than 300 seconds.
For patients undergoing atherectomy, the introducer sheath was changed to accommodate either a 10-French or an 11-French guiding catheter. Balloon dilation to facilitate atherectomy was discouraged, but if required it was limited to a balloon 
2.0 mm in size, inflated to no more than 6 atmospheres of pressure. The SCA-I catheter (Devices for Vascular Intervention, Redwood City, Calif.) was used throughout the trial. The use of the 5-French size, though not prohibited, was strongly discouraged. A minimum of five passes of the cutter across the lesion was recommended during the first insertion of the device; the operators were encouraged to perform additional passes of the cutter and further insertions as required, in an attempt to achieve a final lesion diameter as close to the normal size of the vessel as possible. If balloon angioplasty was selected, any approved balloon-dilation system could be used. Perfusion balloons were permitted either as a primary or a rescue device. The balloon size and the inflation protocols were chosen by the operator to achieve optimal angiographic results.
Operators were encouraged to use only the technique selected by randomization, although crossover was permitted if this did not yield an adequate result. The timing of sheath removal and the dosage and duration of heparin treatment after the procedure were at the discretion of the operator. The patients were hospitalized for at least 18 hours after the procedure.
Follow-up
After the patients were discharged from the hospital, their referring physicians were responsible for routine clinical follow-up. Antianginal medications were discontinued unless specifically indicated. The use of n-3 fatty acids was prohibited; the use of all other medications was at the discretion of the treating physicians. The local study nurses contacted each patient by telephone every month for six months to ascertain clinical status. Angiography was repeated as close to six months after the procedure as possible, although a range of four to seven months was allowed for logistic reasons. Angiography was allowed before four months had elapsed on the basis of symptoms or the results of noninvasive tests, although if restenosis was not found, a subsequent angiogram was required in the four-to-seven-month period. All follow-up angiograms were performed in the same catheterization laboratory as the original procedure, according to the angiographic protocol described above.
Quantitative Coronary Analysis
Quantitative coronary analysis was carried out in a predetermined order at the coordinating center by investigators who had no knowledge of the specific intervention. In order to avoid any influence of the technique used or the procedural outcome on the analysis, the angiograms obtained before the procedure were analyzed first, those obtained at follow-up were analyzed second, and those obtained after the procedure were analyzed last. With the use of a Tagarno cine projector (Tagarno of America, Dover, Del.) optimal preprocedure, post-procedure, and follow-up frames were selected by a single investigator from identical radiographic projections that best demonstrated the maximal severity and anatomical features of the lesions. Motion artifacts were avoided by selecting frames at end-diastole and excluding those with overlapping branch vessels. All measurements of coronary stenosis were generated with the Cardiac Measurement System (Medical Imaging Systems, Neunen, the Netherlands) by a single trained research technician. The untapered section of the tip of the diagnostic catheter was used for absolute calibration after individual measurements (in microms) of the external diameter had been obtained. The coronary segment of interest was identified by the operator. An interactive, automated edge-detection system outlined the coronary lumen and measured the absolute minimal and reference diameters, the latter derived by a mathematical interpolation function. Validation studies on this system by the trial personnel demonstrated an intraobserver variability of 0.07 mm on immediate reanalysis and of 0.20 mm on 25 paired angiograms analyzed six months apart.
Outcome Assessment
The determination of all angiographic outcomes was based on quantitative measurements made at the angiographic core laboratory. However, eligibility for enrollment was determined by visual assessment of the angiogram at the participating sites. Angiographic success was defined as stenosis of 50 percent after the procedure. Procedural success was defined as the occurrence of angiographic success without a major complication (death, myocardial infarction [new diagnostic Q waves or an increase in the serum creatine kinase level to two times the local normal value], or coronary artery bypass surgery) during the index hospitalization. Electrocardiograms and enzyme levels were reviewed by an experienced cardiologist who had no knowledge of the technique used or the angiographic outcome. Other adverse events included abrupt vessel closure and injury to the vascular access site (requiring transfusion or surgery). Major complications were reviewed on an ongoing basis by an independent safety monitor. As a dichotomous variable, restenosis was defined as stenosis of more than 50 percent at follow-up. Luminal dimensions were also examined as continuous data.
Statistical Analysis
The prespecified primary analysis of angiographic and procedural outcomes followed the intention-to-treat principle. Categorical data were compared with continuity-adjusted or exact chi-square statistics as appropriate, and continuous data with analysis-of-variance techniques. Confidence intervals for the observed restenosis rates and their differences were calculated with the normal approximation. Logistic regression was used to identify variables contributing to restenosis from among those known before the procedure. In addition to the randomization assignment, the variables considered included patient age, sex, symptom status (stable vs. unstable angina), and all two-way interactions. Because the effects of atherectomy and angioplasty on restenosis are not independent of their effects on immediate procedural outcomes, all the patients who underwent follow-up angiography were included in the primary analysis. However, patients with stenosis exceeding 50 percent after the procedure were excluded in a secondary analysis. Continuous data are given as means ±SD; categorical data are presented as rates, with 95 percent confidence intervals. Differences between the groups were considered significant if the P value was less than 0.05 for a two-tailed test.
Results
Characteristics of the Patients
Between July 1991 and August 1992, 274 patients were selected for the study: 138 were randomly assigned to undergo atherectomy and 136 to undergo angioplasty. Two of the randomization envelopes (at separate sites) were inadvertently used out of sequence, but this did not affect the overall treatment allocation. The 274 procedures constituted 7 percent of all coronary interventions and 16 percent of procedures performed on the left anterior descending coronary artery at participating sites during this period. Fifty-eight percent of all atherectomy procedures performed at the study centers were included in the trial. Of 4024 patients screened, 1718 had lesions of the left anterior descending coronary artery; of these patients, 81 percent had angiographic and 12 percent clinical reasons for exclusion. Among the patients who were angiographically and clinically eligible, 93 percent were randomized, 3 percent were already participating in another trial, and 3 percent declined to be enrolled (the reasons for exclusion were unknown in 1 percent).
The base-line characteristics of the patients are shown in Table 1. The patients randomly assigned to undergo atherectomy were more likely to be female and were slightly older, whereas the angioplasty group had a higher proportion of patients with unstable angina. Other clinical and angiographic characteristics were reasonably well matched. Age and sex were highly correlated and probably represent a single imbalance in the base-line characteristics. After quantitative analysis, eight patients (three in the atherectomy group and five in the angioplasty group) were found to have had stenosis of less than 50 percent (mean [±SD], 46 ±5 percent) before the procedure; the visually estimated stenoses for these patients ranged from 60 to 95 percent.
|
Atherectomy was attempted and abandoned in 15 of the 138 patients randomly assigned to undergo this procedure (11 percent) because of guiding-catheter problems in 5 and an inability to cross the lesion in 10, despite balloon dilation before the procedure in 4 of the 10. In four other patients, dilation before the procedure successfully facilitated atherectomy. An additional 11 patients (8 percent) required adjunctive balloon dilation after atherectomy. In the group randomly assigned to undergo angioplasty, 5 of the 136 patients (4 percent) were crossed over to an alternative technique -- 3 underwent atherectomy and 2 received intracoronary stents. Tissue was retrieved from all the patients who underwent atherectomy except the 15 in whom the procedure was abandoned.
Procedural outcomes and the complications that occurred during hospitalization are shown in Table 2. Quantitative analysis revealed that angiographic success was achieved in 135 of 138 patients undergoing atherectomy (98 percent) and 124 of 136 patients undergoing angioplasty (91 percent) (P = 0.017); the procedural success rates were 94 percent and 88 percent, respectively (P = 0.061). There were no in-hospital deaths and only one Q-wave myocardial infarction. The incidence of other major complications, including in-hospital coronary bypass surgery and non-Q-wave myocardial infarction, was similar in the two groups (composite outcome: 5 percent for the atherectomy group and 6 percent for the angioplasty group; P = 0.98). Ninety-one percent of the patients in each group were free of any adverse event during hospitalization.
|
Angiographic Restenosis
Angiography was performed a median of 5.9 months after the procedure in 257 of the 265 patients (97 percent) who had not received a stent or who had not had coronary bypass surgery during the index hospitalization (133 of 136 in the atherectomy group and 124 of 129 in the angioplasty group). Follow-up angiograms were obtained within four months of the interventional procedure in 38 patients, 32 of whom had restenosis according to quantitative angiographic analysis. The remaining six (three in the atherectomy group and three in the angioplasty group) were considered to have restenosis on the basis of visual assessment but were subsequently found to have stenosis of less than 50 percent on quantitative review.
The rate of restenosis was 46 percent in the atherectomy group (95 percent confidence interval, 37 to 54 percent) and 43 percent in the angioplasty group (95 percent confidence interval, 34 to 52 percent) (P = 0.71). When the subjects with stenosis of more than 50 percent after the procedure were excluded from the analysis (3 in the atherectomy group and 10 in the angioplasty group), the resulting rates of restenosis were 45 percent in the atherectomy group (95 percent confidence interval, 37 to 54 percent) and 39 percent in the angioplasty group (95 percent confidence interval, 30 to 48 percent) (P = 0.31). A separate analysis of the patients who were treated successfully with use of the assigned technique alone also yielded similar results (44 percent for the atherectomy group [110 patients] and 39 percent for the angioplasty group [112 patients], P = 0.59).
Stepwise logistic regression revealed that only unstable angina remained as a predictor of restenosis. Among the patients with unstable angina, the rate of restenosis was 49 percent after atherectomy (95 percent confidence interval, 35 to 63 percent) and 49 percent after angioplasty (95 percent confidence interval, 37 to 62 percent); among the patients with stable angina, the rates were 44 percent after atherectomy (95 percent confidence interval, 33 to 55 percent) and 35 percent after angioplasty (95 percent confidence interval, 23 to 49 percent). Despite the differences in the rates of restenosis, there was no significant interaction between angina status and treatment assignment after we controlled for age and sex (P = 0.29 for the interaction term). After adjustment for angina status (stable vs. unstable), the rates of restenosis were 46 percent in the atherectomy group and 42 percent in the angioplasty group. These adjusted rates exclude, with 90 percent certainty, a difference in the restenosis rates of more than 7 percent in favor of atherectomy or of more than 15 percent in favor of angioplasty.
The luminal dimensions before and after the procedure and at follow-up are given in Table 3. The mean reference diameters before the procedure were similar in the two groups. The increase in the minimal luminal diameter was greater with atherectomy than with angioplasty (1.45 ±0.47 vs. 1.16 ±0.44 mm, P<0.001); however, the gain was offset by a greater loss during the follow-up period (0.79 ±0.61 vs. 0.47 ±0.64 mm, P<0.001). Consequently, there was no significant difference between groups in the minimal luminal diameter at follow-up (1.55 ±0.60 vs. 1.61 ±0.68 mm, P = 0.44). The cumulative-frequency distributions of the minimal luminal diameters are shown in Figure 1.
|
|
Clinical data were available for all 265 patients eligible for follow-up. There were no significant differences between the two groups with regard to clinical events. One sudden death occurred in a patient who had undergone atherectomy, and there were two myocardial infarctions, both in patients who had undergone angioplasty. Revascularization was repeated in 39 patients (by percutaneous methods in 32 and surgery in 7) in the atherectomy group and 36 patients (by percutaneous methods in 30 and surgery in 6) in the angioplasty group. Seventy-one percent of the patients in each group had no late adverse events. The proportion of patients with Canadian Cardiovascular Society class III or IV angina at any time during the follow-up period was not significantly different between groups (30 percent in the atherectomy group and 20 percent in the angioplasty group, P = 0.11).
Discussion
Although there was a higher procedural success rate with atherectomy than with angioplasty in this trial, a comparison of the late angiographic and clinical outcomes of the two techniques disclosed no advantage of atherectomy over angioplasty when used for the initial treatment of lesions in the proximal left anterior descending coronary artery. The angiographic success rate on quantitative analysis was significantly higher with atherectomy (98 percent vs. 91 percent); however, about 10 percent of these "successful" atherectomies resulted from a crossover to conventional balloon angioplasty after atherectomy was abandoned, and another 10 percent required balloon dilation before or after the atherectomy to complete the procedure. In contrast, operators resorted to an alternative technique in fewer than 5 percent of the patients randomly assigned to undergo angioplasty. The rates of adverse events in the hospital were similar in both groups. Procedural results and the incidence of complications with atherectomy in this trial were comparable to or better than those reported in earlier series in the literature8,9,10,11,12,13.
Lesions of the proximal left anterior descending coronary artery have been associated with restenosis rates of 40 to 45 percent after angioplasty14,15. Conversely, a post hoc analysis of atherectomy procedures performed in the same anatomical location has suggested a much lower rate (<25 percent)16. Such lesions are frequently eccentric and bulky -- morphologic features generally considered to be less than ideal for angioplasty but suitable for atherectomy. Furthermore, access of the stiffer, higher-profile atherectomy catheter to these lesions is facilitated by their proximity and the relative absence of vessel tortuosity. These considerations formed the rationale for restricting the target lesion in this trial to stenoses of the proximal left anterior descending coronary artery; if atherectomy were superior to angioplasty, it should be most evident for lesions in this location.
Despite this strategy, quantitative analysis of follow-up angiograms in 97 percent of the eligible patients failed to demonstrate a significant difference in the prespecified dichotomous outcome of restenosis. The actual rates of restenosis in both groups in this trial are consistent with those in contemporary reports12,13,14,15. The rates of the composite clinical outcomes (including death, myocardial infarction, coronary bypass surgery, the need for additional coronary intervention, and recurrence of angina), which are of primary concern to patients and their physicians, were also similar for both procedures. However, atherectomy procedures consumed more catheterization-laboratory resources and subjected patients to more radiation than angioplasty procedures.
As shown by other investigators,17,18,19,20,21 atherectomy was associated with larger minimal luminal diameters and with less residual stenosis after the procedure than angioplasty. These favorable angiographic features were offset, however, by a proportionally greater deterioration during the follow-up period, so that the net gain in either luminal diameter or percent diameter was virtually the same for both procedures. This is shown by the cumulative-frequency distribution curves of the minimal luminal diameter, which were nearly identical at follow-up. These findings are supported by data from both laboratory and clinical studies demonstrating that the extent of subsequent intimal hyperplasia is proportional to the depth of arterial injury22 or to the improvement in luminal diameter at the time of the procedure23,24; in other words, the greater the gain, the greater the loss. Unresolved is the question of whether the ratio of loss to gain is constant25,26 or varies according to the device used or the amount gained18,19; our data, showing proportionally greater loss after atherectomy, would tend to support the latter possibility.
In addition to compressing and reshaping atheroma, atherectomy excises tissue and debulks plaques; it was anticipated that this mechanism would result in a lower rate of restenosis than that occurring after angioplasty. These expectations were not fulfilled in this trial. Nonetheless, coronary atherectomy is a new technique that continues to evolve. A more aggressive approach to atherectomy that results in larger luminal diameters has recently been advocated by some investigators24,25,26. Whether this would have resulted in lower rates of restenosis without a concurrent increase in complications cannot be determined from our study. The role of atherectomy in dealing with restenosed lesions or specific anatomical situations considered unfavorable for angioplasty, such as ostial, bifurcated, or bypass-graft lesions, is still being assessed. Furthermore, it is conceivable that its application in vessels other than the left anterior descending coronary artery, future enhancements in its design, or greater operator experience in its use may yet affect restenosis. However, the similar rates of restenosis and the similar luminal dimensions found at follow-up after atherectomy or angioplasty in this trial may well represent an inevitable consequence of the vascular response to intervention.
Supported by a grant from the Medical Research Council of Canada and by Devices for Vascular Intervention, Redwood City, Calif., and Advanced Cardiovascular Systems, Temecula, Calif.
Source Information
From Mount Sinai Hospital, Toronto (A.G.A., A.G.L.); Sunnybrook Health Science Centre, Toronto (E.A.C.); Toronto Hospital, Toronto (B.P.K., L.S.); Institut de Cardiologie de Montreal, Montreal (R.B.); Vancouver General Hospital, Vancouver, B.
(D.R.R.); St. Paul's Hospital, Vancouver (J.G.W.); Ottawa Heart Institute, Ottawa, Ont. (L.L.); Institut de Cardiologie, Hopital Laval, Quebec, Que. (G.B.); Foothills Hospital, Calgary, Alta. (M.T.); and New Brunswick Heart Centre, St. John, N.B. (B.N.C.) -- all in Canada.
Address reprint requests to Dr. Adelman at the Cardiovascular Clinical Research Laboratory, Mount Sinai Hospital, 1609-600 University Ave., Toronto, ON M5G 1X5, Canada.
References
The following institutions, investigators, and staff members participated in the Canadian Coronary Atherectomy Trial or acted as advisors: Participating centers and investigators -- Toronto Hospital, Toronto: A.G. Adelman, B.P. Kimball, and J. Richards; Institut de Cardiologie de Montreal, Montreal: R. Bonan and C. Berube; Vancouver General Hospital, Vancouver: D.R. Ricci, C. Buller, C. Dupuis, and A. McCarthy; St. Paul's Hospital, Vancouver: J.G. Webb, R. Carere, and D. Heinrich; Ottawa Heart Institute, Ottawa: L. Laramee, J.F. Marquis, and H. Dowell; Institut de Cardiologie, Hopital Laval, Quebec: G. Barbeau and M.M. Lariviere; Foothills Hospital, Calgary: M. Traboulsi, D. Galbraith, K. Hildebrand, and D. Houston; New Brunswick Heart Centre, St. John: B.N. Corbett and J. Creighton; and Sunnybrook Health Science Centre, Toronto: E.A. Cohen, S. Naqvi, S. Dolman, and N. Cooper; Trial advisors -- Mount Sinai Hospital, Toronto: A. Logan (Clinical Epidemiology); Toronto Hospital, Toronto: L. Schwartz (Clinical Events Adjudication), W. Mahon (Clinical Trials), and P. Liu and A. Gotlieb (Tissue Pathology); Institut de Cardiologie de Montreal, Montreal: J. Lesperance (Quantitative Coronary Analysis); and Royal Columbian Hospital, New Westminster, B.C.: M. Henderson (Safety Monitoring); Coordinating center and core-laboratory staff -- Mount Sinai Hospital, Toronto: P. Slaughter (Study Coordinator), K. Sykora (Study Biostatistician), S. Bui (Quantitative Coronary Analysis Technician), J. King, C. Ellis, K. Kwok, Y. Wu, M. Landry, and G. Prasad.
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
