Percutaneous Coronary Intervention in the Current Era Compared With 1985–1986
The National Heart, Lung, and Blood Institute Registries
Background—Although refinements have occurred in coronary angioplasty over the past decade, little is known about whether these changes have affected outcomes.
Methods and Results—Baseline features and in-hospital and 1-year outcomes of 1559 consecutive patients in the 1997–1998 Dynamic Registry who were having first coronary intervention were compared with 2431 patients in the 1985–1986 National Heart, Lung, and Blood Institute Registry. Compared with patients in the 1985–1986 Registry, Dynamic Registry patients were older (mean age, 62 versus 58 years; P<0.001) and more often female (32.1% versus 25.5%; P<0.001). In the Dynamic Registry, procedures were more often performed for acute myocardial infarction (22.9% versus 9.9%; P<0.001) and treated lesions were more severe (84.5% versus 82.5% diameter reduction; P<0.001), thrombotic (22.1% versus 11.3%; P<0.001) or calcified (29.5% versus 10.8%; P<0.001). Stents were used in 70.5% of Dynamic Registry patients, whereas 1985–1986 patients received balloon angioplasty alone. Procedural success was higher in the Dynamic Registry (92.0% versus 81.8%; P<0.001) and the rate of in-hospital death, myocardial infarction, and emergency coronary bypass surgery combined was lower (4.9% versus 7.9%; P=0.001) than in the 1985–1986 Registry. The 1-year rate for CABG was lower in the Dynamic Registry (6.9% versus 12.6%; P<0.001).
Conclusions—Although Dynamic Registry patients had more unstable and complex coronary disease than those in the 1985–1986 Registry, their rate of procedural success was higher whereas rates of complications and subsequent CABG were lower. Results of percutaneous coronary intervention have improved substantially over the past decade.
In 1979, the National Heart, Lung, and Blood Institute (NHLBI) established a voluntary registry to characterize coronary angioplasty, at the time an emerging technique of percutaneous coronary revascularization.1 Patients were enrolled from 1977–1980 and subsequently from 1985–1986. These 2 registries provided the first comprehensive description of technical and clinical results, and in the mid-1980s identified that angioplasty had matured to the point that clinical trials were necessary to resolve the genuine dilemma as to potential comparability to CABG.
These trials, including Bypass Angioplasty Revascularization Investigation (BARI)2 and EAST,3 found that in most patients with multivessel coronary artery disease (CAD), balloon angioplasty did not compromise survival and proved to save costs slightly relative to CABG but many patients who received PTCA required repeat revascularization. Since that time, new devices have been developed as potential adjuncts or replacements for the balloon catheter.4 Although a few randomized clinical trials have helped to clarify the value of these new devices in limited patient subgroups,5 6 7 8 little is known of the extent of their use in overall clinical practice or of their effect on patient selection and outcomes.
The primary goal of the recently established Dynamic Registry is to characterize percutaneous coronary intervention in the new device era. The Registry provides the optimal design to assess and to compare with the 1985–1986 Registry the following: (1) contemporary patients and disease characteristics selected for percutaneous coronary intervention, (2) types of devices used in various settings, and (3) changes in interventional strategy as well as in-hospital and 1-year outcomes. The present report describes the findings from these comparisons.
Design and Study Population
The Dynamic Registry includes 15 clinical centers (10 participants in the 1985–1986 Registry and 5 added to augment enrollment of minorities) and a Coordinating Center. Three sequential waves of 2000 patients, spaced 18 months apart, are to be enrolled. Each patient who is having percutaneous coronary catheter-based intervention performed by a Dynamic Registry investigator is registered. Informed consent is obtained to collect information after hospital discharge. The first wave of enrollment, which began July 1997 and was completed February 1998, recruited 2526 patients, of whom 2206 were consecutively enrolled. Because patients with prior angioplasty were excluded from the 1985–1986 Registry, for the sake of comparability, 647 Dynamic Registry patients with history of prior percutaneous coronary intervention are excluded from this report.
Registry research coordinators responsible for data collection participated in a training session before patient enrollment. Data collection included demographic information, medical history, and risk factor profile. Coronary angiographic information before and after intervention was obtained according to definitions developed in previous registries and the Bypass Angioplasty Revascularization Investigation.1 9 Procedural strategy data included device use, procedural staging, and success of each coronary lesion attempted. Successful lesion dilatation was defined as an absolute 20% reduction in lesion severity with final stenosis <50%. Angiographic success was classified as either partial (some but not all attempted lesions successfully treated) or total (all attempted lesions successfully treated). Untoward events included death from any cause, myocardial infarction (MI), or CABG. MI was defined as evidence of ≥2 of the following: (1) typical chest pain >20 minutes not relieved by nitroglycerin, (2) serial ECG recordings showing changes from baseline or serially in ST-T and/or Q-waves in ≥2 contiguous leads, or (3) serum enzyme elevation of CK-MB>5% of total CK (total CK>2× normal; LDH subtype 1>LDH subtype 2). Congestive heart failure was defined as presence of paroxysmal nocturnal dyspnea, dyspnea on exertion, or radiographic pulmonary congestion. Risk for CABG was classified as low, moderate, high, or inoperable according to judgment of the interventionist. Procedures were classified according to clinical circumstances: emergent when required immediately because of clinical instability, urgent when required within 24 hours to minimize cardiac risk, and elective when deferrable >24 hours without cardiac risk. Procedural success was defined as achievement of either partial or total angiographic success without death, Q-wave MI, or emergency CABG.
Differences between proportions were assessed by χ2 test or Fisher’s Exact Test when the number of patients in a group was small. Continuous variables were compared by Student’s t test. One-year outcomes were reported with Kaplan-Meier estimates and log rank statistics. Standard stepwise procedures were used with Cox proportional hazards models to obtain adjusted relative risks comparing the 2 registries. Consent to collect follow-up information after initial procedure hospitalization was not obtained for 253 of the 1559 Dynamic patients, who were censored at the time of hospital discharge in analyses of 1-year event rates.
Although most of the consecutively enrolled patients in both registries were white men of middle age, prevalence of certain baseline characteristics differed between the 2 (Table 1⇓). Dynamic Registry patients were older; more often female and nonwhite; more often had history of diabetes mellitus, congestive heart failure, or hypertension; and more often were considered to be at high risk for CABG. Dynamic Registry patients were less likely to have history of prior MI or cigarette smoking. No differences were noted in the extent of coronary disease.
Unstable angina was the most common procedural indication for intervention in the Dynamic Registry (Table 2⇓). Compared with the 1985–1986 Registry, procedures were more likely to be performed for acute MI (22.9% versus 9.9%; P<0.001) and less frequently for stable angina. Also, interventions were more often emergent (13.0% versus 5.8%; P<0.001) for Dynamic than for 1985–1986 Registry patients. Among the 357 Dynamic patients with AMI as indication for intervention, 48.7% had emergent intervention. Glycoprotein IIb/IIIa receptor inhibitors, not available in the 1985–1986 Registry, were administered to 25.3% of Dynamic Registry patients.
Comparison of interventional strategy revealed some interesting trends. Most patients had procedures attempted on only 1 lesion, and fewer Dynamic Registry patients had multilesion procedures than in the 1985–1986 Registry. Multivessel attempt was 9% in the Dynamic compared with 20.7% in the 1985–1986 Registry (P<0.001). Balloon angioplasty, the only percutaneous intervention at the time of the 1985–1986 Registry, was used as the sole device in 24.7% of Dynamic Registry patients, concomitantly with stent placement in 63.7%, rotational atherectomy in 3.1%, and both stent and rotational atherectomy in 5.9% of patients. (Directional atherectomy, extraction atherectomy, and laser each were used in <1% of patients.)
Although the left anterior descending coronary artery was the most common location for an attempted lesion in both registries, lesions in other locations were more often attempted in the Dynamic compared with the 1985–1986 Registry (Table 3⇓). Attempted lesion stenosis in the Dynamic Registry was more severe, and lesions were more often total occlusions. These differences were independent of sex of the patient or whether AMI was the indication for intervention.
Thrombus and calcification were reported far more frequently in the Dynamic Registry, yet angiographic success was achieved more often (93.7% versus 80.9%; P<0.001) and final lesion narrowing was less severe. Although abrupt artery closure was less common in the Dynamic Registry, side-branch occlusion and local coronary dissection were more common.
In-hospital mortality was not significantly different between the registries (Table 4⇓), whereas MI and need for emergent CABG were significantly lower in the Dynamic Registry. Both total angiographic success and procedural success were achieved significantly more often in the Dynamic Registry. In addition, mean length of hospital stay decreased significantly, from 4.1 to 2.7 days. Overall incidence of repeat revascularization during initial hospitalization was 5.6% (4.2% repeat percutaneous intervention and 1.5% CABG).
Crude mortality during 1 year of follow-up was higher in the Dynamic Registry compared with the 1985–1986 Registry (1-year rate, 5.4% versus 3.6%), whereas 1-year death or MI rate was similar (Figure 1⇓). After controlling for important baseline differences between registries (Table 5⇓), death rates became comparable. Lower adjusted mortality was not significant, although the 30% lower 1-year risk for combined-endpoint death or MI achieved significance. Similarly, 1-year CABG and repeat percutaneous procedure rates were significantly lower in the Dynamic Registry both before and after adjustment for baseline inequalities (Figure 2⇓). These findings were uniform regardless of number of lesions or vessels attempted during the initial procedure.
Within the Dynamic Registry, differences were seen in unadjusted 1-year event rates within several clinically important subgroups. Women had higher 1-year mortality (7.6% versus 4.4%; P<0.05) and need for CABG (9.5% versus 5.6%; P<0.05) than men. Patients whose indication for intervention was AMI had higher 1-year mortality (9.2% versus 4.3%; P<0.001) than those with other indications, primarily due to higher in-hospital mortality (5.9% versus 0.7%; P<0.001). Use of GP IIb/IIIa receptor inhibitor was associated with increased 1-year MI rate (9.5% versus 4.0%; P<0.001), due in part to more in-hospital events (5.1% versus 2.0%; P<0.001). No differences were seen in 1-year death, MI, or revascularization by stent use during initial procedure.
In the contemporary Dynamic Registry, patients were older and more often had extensive cardiovascular morbidity compared with patients undergoing percutaneous coronary intervention a decade ago. Some of these observed differences may relate to the larger proportion of women in the Dynamic Registry, given that women presenting with CAD and symptoms requiring revascularization tend to be older and have more coronary risk factors than men do.10 11 12
Coronary intervention was more often performed in the setting of unstable coronary disease in the Dynamic Registry. In fact, AMI was reported as the primary indication for intervention more than twice as often as in 1985–1986. Because coronary angioplasty can be performed quickly and as an immediate adjunct to coronary angiography, it is particularly well suited for coronary syndromes in which rapid revascularization is essential.13 14 These findings may indicate that percutaneous intervention has become an established treatment for patients with acute coronary syndromes and may reflect a preference for a catheterization-based approach over medical or surgical alternatives.15 16 17 18
Substantial differences were found in revascularization strategy at both the lesion and patient levels. Calcification of attempted lesions was nearly 3-fold greater than in 1985–1986. The increase in lesions determined to be calcified may reflect more attention paid to this factor because of availability of rotational atherectomy and improved visibility of modern X-ray equipment and not necessarily a true increase in calcification.
Nearly twice as many attempted lesions showed evidence of thrombus, and attempted lesions had on average more severe stenosis. Lesion access was more difficult in the Dynamic Registry because more lesions were located in arteries other than the left anterior descending. All of these observations indicate that although operators now attempt fewer lesions, they attempt substantially more challenging lesions than in the past. The feasibility of treating more advanced coronary disease by percutaneous techniques has expanded the scope of patients eligible for this approach and is probably a reflection of improvements in angioplasty techniques and equipment.
Remarkably, although most Dynamic Registry patients had multivessel coronary disease and averaged nearly 3 significant lesions each, two-thirds had only 1 lesion attempted. Similarly, more than two thirds of patients with multivessel disease had a single vessel or graft targeted for revascularization. This shift to a more selective revascularization strategy was seen among patients treated outside of the AMI setting as well as among AMI patients, for whom initial intervention is usually performed to reestablish perfusion of the infarct-related artery rather than to achieve complete revascularization.15 A potential explanation for this observation is that some lesions may cause more ischemia than others19 and that operators place strategic emphasis in treating these lesions rather than others that may be of less functional significance. Operators may now also be more astute at identifying these “culprit” lesions. Certainly, the selective use of angioplasty in the Dynamic Registry was not based on a strategy of avoiding complex lesions, which were attempted more often than in the 1985–1986 Registry.
Despite treatment of older, less stable patients with more severe and complex lesions, procedural success rates were substantially higher in the Dynamic Registry than the 1985–1986 Registry. Coronary stenosis was relieved in a larger proportion of patients in the Dynamic registry, and magnitude of improvement was also much greater. Importantly, the increased lesion success was not achieved at the expense of additional risk.
Although angiographic dissection was observed more often, abrupt artery closure was substantially reduced in the Dynamic Registry. Although the reason for more dissection in contemporary practice is not readily apparent, potential explanations include more aggressive balloon angioplasty due to availability of stents or perhaps a change in threshold for declaring when dissection is present.
In-hospital mortality was not significantly different from the 1985–1986 Registry, and incidence of periprocedural MI and emergency CABG were significantly lower than a decade ago. With the increased rate of side-branch closure, probably as a result of stent use, a higher rate of periprocedural MI might have been expected. Because this effect was not observed, side-branch occlusion was probably limited to small arterial branches or transient.
Important differences in late outcome were also observed. Although 1-year crude mortality was higher in the new Registry, this difference was explained entirely by baseline differences between the 2 cohorts. Need for additional revascularization, either percutaneous or surgical, was 41% lower during 1-year follow-up in the Dynamic Registry. CABG reduction is explained mostly by lower procedural complications in the Dynamic Registry, whereas repeat procedure rates are gradually reduced over the entire follow-up period, which indicates important reductions in restenosis. Thus, current techniques have not only enhanced safety of percutaneous coronary revascularization, but also have augmented durability of the intervention.
Others have suggested that use of intracoronary stents may be responsible for improved results with contemporary percutaneous intervention.20 21 Our analysis found no difference in outcome between stented and nonstented patients. Accordingly, a significant percentage of patients may achieve a favorable clinical outcome without routine stent use. On the other hand, we cannot exclude a beneficial effect of stents. Because stent use is based on clinical judgment, stents were probably deployed in the Dynamic Registry patients most likely to benefit from stenting. Patients who did not receive stents may have been at low risk for complications or restenosis. Also, improved clinical outcome has been observed in trials of IIb/IIIa receptor antagonists.22 23 24 The present study does not support this effect and in fact demonstrated a higher MI rate among patients who received this therapy. However, proper interpretation of our finding requires acknowledgment that patients who received IIb/IIIa receptor blockade were more likely to have features associated with increased risk of complication. Outcomes for these patients might have been worse had this therapy not been administered. Finally, other factors, including more aggressive medical therapy and philosophies regarding indications for surgical revascularization, may have affected both in-hospital and 1-year outcomes.
Certain limitations must be acknowledged in interpretation of our findings. First, although the Dynamic Registry included a broad representation of clinical centers, most Registry investigators operated at large referral centers and were experienced interventional cardiologists. Whether the results achieved in the Registry apply to other types of institutions or given less experienced operators is unknown. Second, to permit meaningful comparison with the 1985–1986 Registry, Dynamic Registry patients who had had prior coronary intervention were not included in this analysis. These patients will be included in subsequent reports.
In summary, the Dynamic Registry has documented the substantial changes that have taken place in the field of percutaneous coronary intervention in the past decade. Patients now more often present with unstable coronary syndromes. Although revascularization strategy is somewhat more conservative than in 1985–1986, substantially more complex lesions are now being attempted. Rates of success are higher and acute complications lower compared with prior experience. Of particular importance is the documentation of the striking reduction in need for subsequent repeat surgical or percutaneous procedures. These results support use of percutaneous coronary interventions as effective means for achieving coronary revascularization.
Dynamic Registry Participants include the following: Rhode Island Hospital, Providence, RI: D.O. Williams, MD (Principal Investigator), and J. Raymond (Coordinator);St. Luke’s Medical Center, Milwaukee, Wis: G. Dorros, MD (Principal Investigator), and L. Mathiak (Coordinator); Lankenau Hospital, Wynnewood, Pa: P. Coady, MD (Principal Investigator) and E. Shappell (Coordinator); Seton Medical Center, Daly City, Calif: R. Myler, MD (Principal Investigator), R. Shaw, MD (Coinvestigator), and M. Murphy (Coordinator); University of Southern California, Los Angeles, Calif: D. Faxon, MD (Principal Investigator), and R.P. Singh, MD, and W. Hill (Coordinators); University of Pittsburgh Medical Center, Presbyterian University Hospital, Pittsburgh, Pa: H. Cohen, MD (Principal Investigator), and D. Rosenfelder and S. Iovino (Coordinators); Medical College of Virginia, Richmond, Va: M. Cowley, MD (Principal Investigator), and K. Hall (Coordinator); Providence/St. Vincent Medical Center, Portland, Ore: P. Block, MD (Principal Investigator), and E. Block (Coordinator); University of Maryland Hospital, Baltimore, Md: W. Laskey, MD (Principal Investigator), and D. Beach (Coordinator); Institute for Clinical and Experimental Medicine, Czech Republic: V. Stanek, MD (Principal Investigator), and M. Hrncarek, MD (Coordinator); St. Luke’s Roosevelt Hospital Center, New York, NY: J. Slater, MD (Principal Investigator), and D. Tormey (Coordinator); University of Pennsylvania Health System, Philadelphia, Pa: R. Wilenski, MD (Principal Investigator), and S. Hanlon (Coordinator); St. Mary’s Hospital, Rochester, Minn: D.R. Holmes, MD (Principal Investigator), and S. Brevig (Coordinator); Boston Medical Center, Boston, Mass: A. Jacobs, MD (Principal Investigator), and M. Mazur and S. Murphy (Coordinator); Emory University Hospital, Atlanta, Ga: S. King, MD (Principal Investigator), and C. Friedrich (Coordinator); Montreal Heart Institute, Montreal, Quebec, Canada: M.G. Bourassa, MD (Principal Investigator), Gilles Côté, MD (Coinvestigator), and S. Taillefer (Coordinator); and Cardiovascular Medicine Associates, Houston, Tex: M. Al-Bassam, MD (Principal Investigator), and D. Lance (Coordinator).
Personnel from the Coordinating Center, University of Pittsburgh, Pittsburgh, Pa, included the following: K. Detre, MD, DrPH (Director); S. Kelsey, PhD, R. Holubkov, PhD (Codirectors); M. Brooks, PhD, W. Yeh, H. Vlachos (Statisticians); V. Niedermeyer (Project Manager); S. Lawlor (Director, Data Management); E. Passano (Data Manager); J. Bondi and T. Ledneva (Data Entry); and N. Reck (Project Secretary).
NHLBI Project Office personnel included G. Sopko, MD, and P. Desvigne-Nickens, MD.
This work was supported in part by NHLBI grant HL33292-14.
Additional coinvestigators are listed in the Appendix.
- Received May 22, 2000.
- Revision received July 20, 2000.
- Accepted July 25, 2000.
- Copyright © 2000 by American Heart Association
Baim DS, Cutlip DE, Sharma SK, et al, for the BOAT Investigators. Final results of the balloon versus optimal atherectomy trial (BOAT). Circulation. 1998;97:322–331.
Alderman E, Stadius M. The angiographic definitions of the Bypass Angioplasty Revascularization Investigation. Coron Artery Dis. 1992;3:1189–1207.
Kelsey SF, James M, Holubkov AL, et al. Results of percutaneous transluminal coronary angioplasty in women: 1985–1986 National Heart, Lung, and Blood Institute’s Coronary Angioplasty Registry. Circulation. 1993;87:720–727.
Williams DO, Ruocco NA, Forman S, and TIMI Investigators. Coronary angioplasty following recombinant tissue-type plasminogen activator in acute myocardial infarction: a report from the Thrombolysis in Myocardial Infarction Trial. J Am Coll Cardiol. 1987;10:45B–50B.
Williams DO, Braunwald E, Thompson B, et al, for the TIMI IIIB Investigators. Results of percutaneous transluminal coronary angioplasty in unstable angina and non-Q wave myocardial infarction: observations from the TIMI IIIB Trial. Circulation. 1996;94:2749–2755.
Boden WE, O’Rourke RA, Crawford MH, et al. Outcomes in patients with acute non-Q-wave myocardial infarction randomly assigned to an invasive as compared with a conservative management strategy: Veterans Affairs Non-Q-Wave Strategies in Hospital (VANQWISH) Trial Investigators. N Engl J Med. 1998;338:1838–1839.
The Global Use of Strategies to Open Occluded Coronary Arteries in Acute Coronary Syndromes (GUSTO IIb) Angioplasty Substudy Investigators. A clinical trial comparing primary coronary angioplasty with tissue plasminogen activator for acute myocardial infarction N Engl J Med. 1997;336:1621–1628.