Author + information
- Received November 24, 1997
- Revision received April 17, 1998
- Accepted June 22, 1998
- Published online November 1, 1998.
- Bernd Waldecker, MDa,* (, )
- Wolfgang Waas, MDa,
- Werner Haberbosch, MDa,
- Reinhard Voss, MDa,
- Heinrich Heizmanna and
- Harald Tillmanns, MDa
- ↵*Address for correspondence: Dr. Bernd Waldecker, Medizinische Klinik I Kardiologie-Angiologie, Justus-Liebig Universität, Klinikstrasse 36, 35392 Giessen, Germany
Objectives. The purpose of this study was to analyze long-term follow-up information over several years from consecutive, unselected patients treated with direct percutaneous transluminal coronary angioplasty (PTCA) for acute myocardial infarction (MI).
Background. Direct PTCA is often used in patients with acute MI. Short-term results are favorable. However, there is less information available on long-term observations over several years in these patients.
Methods. A total of 416 consecutive and unselected patients with acute MI underwent direct PTCA. Survival of the acute infarct phase was 94.2%; the remaining 392 patients—the study population—were discharged and followed for 3.3 ± 1.4 years. Mortality as well as cardiac events and reinterventions are reported. Clinical variables assessed at the time of discharge are submitted to statistical analysis to detect potential risk factors.
Results. Total cumulative mortality in the first year was 10% for the entire group and 6% for patients not presenting in cardiogenic shock. Mortality after discharge was 4.6% in the first year and dropped to <4% per year thereafter. Reinterventions after discharge were required in 16% in the first year and in <4% per year in years 2 to 4. Poor left ventricular ejection fraction (<35%), three-vessel disease and advanced age (≥75 years) were long-term risk factors for total mortality after direct PTCA.
Conclusions. The clinical benefit of direct PTCA for acute MI is maintained during follow-up with respect to mortality. However, reinterventions for restenosis or de novo stenosis are often required (10% to 20%). Although few in number (<10%), patients with severely impaired left ventricular function continue to have a poor prognosis.
Direct percutaneous transluminal coronary angioplasty (PTCA) is often used in patients with acute myocardial infarction (MI) to restore coronary flow through the infarct-related artery. It has been shown that acute, in-hospital results and short-term follow-up after direct percutaneous transluminal coronary angioplasty for acute MI are favorable (1–6). However, there is much less information focusing on long-term observations over several years in such patients (2,7,8). It is not precisely known how the high initial success rate (>90%) of direct percutaneous transluminal coronary angioplasty in recanalizing the occluded coronary artery translates into the subsequent clinical course.
Therefore, the purpose of this study was to carefully accumulate and analyze long-term follow-up information over several years from consecutive and unselected patients treated with direct percutaneous transluminal coronary angioplasty for acute MI. We emphasize not only total and cardiac mortality, but also the detailed description of the types, incidence and time course of cardiac events and reinterventions to document the course of the disease in the era of rapid and (almost) complete reperfusion. Multivariate analysis was used to investigate whether commonly accepted long-term risk factors for mortality are also pertinent to patients with acute MI but rapid and complete recanalization using direct percutaneous transluminal coronary angioplasty.
Direct percutaneous transluminal coronary angioplasty has become the exclusive approach to recanalizing coronary occlusion in all patients with acute MI at our institution 24 hr/day. Acute MI is diagnosed in the presence of typical chest pain and ST-segment elevation of >2 mm in the precordial leads or >1 mm in leads I, II, III, avR, avL and avF in patients with supraventricular rhythms and no left bundle branch block. The clinical presentation is decisive in patients with other rhythms or left bundle branch block.
A total of 416 consecutive patients were admitted with an acute MI. Mean age was 61 ± 12 years (range 30 to 91 years). There were 107 women (26%) and 309 men (74%); 172 patients (41%) had anterior MI, 77 patients (19%) had their second MI and 26 patients (6%) had had previous bypass grafting. Cardiogenic shock was present in 48 patients (11%). No patient, irrespective of age or clinical or other conditions, was excluded from the invasive approach to acute MI. Patients with contraindications for intravenous thrombolysis were also included.
Direct PTCA and subsequent hospital care
The diagnosis of acute MI prompted treatment with intravenous heparin, acetylsalicyl acid (500 mg), and beta-adrenergic blocking agents if not contraindicated. Thereafter, all patients presenting within 6 h after onset of pain underwent standard diagnostic coronary angiography. After identification of the infarct-related vessel, direct percutaneous transluminal coronary angioplasty of that vessel was attempted if, according to the Thrombolysis In Myocardial Infarction trial (TIMI) classification (9), blood flow was grade 0, 1 or 2. Direct percutaneous transluminal coronary angioplasty was considered successful if TIMI-3 flow was restored, the residual stenosis could be reduced to <50%, or both. Percutaneous transluminal coronary angioplasty was not attempted in 25 patients (6%) because of spontaneous TIMI-3 flow (n = 14), emergency bypass grafting (n = 3), peripheral occlusion of a small vessel (n = 5) or intracoronary fibrinolysis (n = 3). Direct percutaneous transluminal coronary angioplasty was successful in 373 of 391 patients (95%). Intracoronary stents were used in 13 patients (3%). Another 17 patients had spontaneous or fibrinolysis-induced reperfusion. Therefore, 390 of 416 patients (94%) left the catheterization laboratory with an open infarct-related artery. Patients were transferred to an intensive care unit. Hospital care followed conventional guidelines (10). In-hospital mortality was 5.8% (24 of 416 patients) in the entire group and 1.4% (5 of 358) in patients without cardiogenic shock. Death during initial hospitalization was caused by heart failure in 17 patients, left ventricular (LV) rupture in 2 patients, reinfarction in 1 patient and noncardiac reasons in 4 patients. Reinfarction occurred in 15 patients (3.6%). Coronary bypass grafting was scheduled in 42 patients (10%), including 9 patients (2%) who had surgery within 3 days of the index infarct.
The study group consisted of 392 patients who were discharged alive. Mean LV ejection fraction before discharge was 54 ± 15% (15% to 75%). Follow-up was 3.3 ± 1.4 years (1,192 ± 542 days). Most of the patients were seen regularly in the outpatient department of this institution. Otherwise, the patients themselves or their general physician were contacted by phone. Stress testing and recatheterization during follow-up were guided by symptoms. Five patients (foreign residents) were lost to follow-up after discharge within the first year. Causes of death during follow-up were classified as cardiac, noncardiac and unclear according to the judgment of the physician who first witnessed the patient’s death. A diagnosis of reinfarction required typical chest pain and creatine kinase rise greater than three times normal.
The statistical analysis was performed using a commercially distributed software package (PCS version 2 from TopSoft, Hannover, Germany) that provided univariate analysis of data (descriptive statistics, testing for normal distribution, unpaired ttest, chi-square test) as well as time–event analysis and the proportional-hazards model according to Cox. For chi-square tests figures are given in percentages in the text and Table 2, but the chi-square tests used absolute numbers. Time–event analysis in the following section and Figures 1 and 2applied the Kaplan–Meier method. The results of the Kaplan–Meier analysis were extended with respect to the potential risk factors that are listed in Table 3and described in the text. These (and only these) factors were entered into the proportional-hazards model to calculate the independent prognostic significance of each parameter.
A total of 41 of 392 patients (10%) died during the follow-up period of 3.3 ± 1.4 years (Table 1); 21 of these 41 patients (51%) had cardiac death. Noncardiac deaths were due to carcinoma (n = 5), pneumonia (n = 3), cerebrovascular attack (n = 3), perioperative bleeding after noncardiac surgery, septic shock, end-stage renal failure, pancreatitis and suicide.
Figure 1(Kaplan–Meier analysis, exact numbers rounded) compares all-cause cumulative mortality (including in-hospital mortality) with mortality during follow-up after discharge. All-cause mortality after discharge was 4.6%, 1%, 3% and 4%, respectively, in years 1 through 4 after discharge. Cumulative mortality including hospital mortality was 10.1% in the first year. If one excludes the 48 patients who were in cardiogenic shock, cumulative mortality was 6%, 7%, 10% and 14% after 1, 2, 3 and 4 years, respectively. Mortality after discharge in patients without initial cardiogenic shock (n = 353) was almost identical to the entire group of discharged patients (n = 392) and to patients (n = 48) with cardiogenic shock in the acute phase.
Univariate analysis of long-term risk factors is given in Table 2. Significant differences between subgroups are present in women versus men, in elderly patients, in patients with three-vessel versus single-vessel disease, in diabetic patients and in patients with a very poor LV ejection fraction. However, the proportional-hazards model identifies three-vessel disease, age ≥75 years and LV ejection fraction <35% (assessed approximately 1 to 2 weeks after MI) to be predictive of total (Table 3)and cardiac mortality. Comorbidity with diabetes mellitus falls just short of reaching statistical significance. The arrhythmia status was recorded in 287 patients before discharge. The presence of nonsustained ventricular tachycardia during Holter monitoring was not significantly related to increased total mortality in discharged patients.
A total of 22 patients had another MI after discharge; 8 of 22 MIs occurred within the first year. There were 16 true reinfarctions in the same electrocardiographic location as the index MI and 6 MIs in different locations. All follow-up MIs during year 1 were true reinfarctions. The annual reinfarction rate after discharge was 1% (Kaplan–Meier analysis). Unstable angina was diagnosed in a total of 31 patients during follow-up. Another 6 patients were readmitted in heart failure class IV (New York Heart Association classification), and 11 patients had malignant but nonfatal arrhythmias.
Unscheduled reinterventions for angina during follow-up (i.e., target vessel revascularization, percutaneous transluminal coronary angioplasty of a different vessel or surgery) were required in a total of 84 of 392 patients (21%) after discharge. Reintervention was coronary artery bypass grafting in 19 of 392 patients (5%) and heart transplantation in 1 patient. A total of 83 percutaneous transluminal coronary angioplasty procedures were performed in 70 of 392 patients (18%) after discharge, including target vessel re-PTCA in 58 of 83 cases (70%), percutaneous transluminal coronary angioplasty of a different, originally unstenosed vessel in 29 cases, and repeat percutaneous transluminal coronary angioplasty of the infarct-related artery as well as a different vessel in 4 cases. Figure 2shows the incidence of each reintervention over time during follow-up. The incidence of target vessel re-PTCA was 10% in the first year and about 2% per year thereafter. Univariate analysis of the need of reintervention with respect to gender, age ≥75 years, ejection fraction <35%, infarct location, multivessel disease, diabetes mellitus, and nonsustained tachycardia during Holter monitoring revealed a significantly higher incidence of such procedures in patients ≥75 years (p = 0.003), patients with ejection fractions <35% (p = 0.025) and patients with anterior MI (p = 0.02).
We report long-term observations (mean 3.3 years) in 392 consecutive and unselected patients discharged alive after acute MI and direct percutaneous transluminal coronary angioplasty. Hospital mortality was 5.8%, but <2% when disregarding patients presenting with cardiogenic shock in the acute phase of MI.
Follow-up mortality after discharge
Total mortality after discharge was highest within the first year after MI (4.6%) and decreased thereafter to 1% to 4% per year (Fig. 1)irrespective of whether patients were or were not in cardiogenic shock in the acute phase. Only half of these deaths were attributed to cardiac causes. Similar total and cardiac mortality rates late after direct percutaneous transluminal coronary angioplasty as well as the same trend to a lower mortality after 1 year have been reported by others (2,7). In a third series of patients (8)total mortality reached 10% after 1 year but also subsequently decreased to <5% per year. Thus, total, and particularly cardiac mortality late after direct percutaneous transluminal coronary angioplasty is low.
The only prospective and randomized long-term (≥1 year) comparison between direct percutaneous transluminal coronary angioplasty and intravenous fibrinolysis addressing long-term prognosis reported a significant reduction of cumulative cardiac mortality in the percutaneous transluminal coronary angioplasty group (7). However, detailed information on total and cardiac mortality after discharge was not provided (7). Another less systematic, retrospective, nonrandomized, multicenter analysis (8)failed to detect any differences in long-term total mortality between the thrombolysis and the percutaneous transluminal coronary angioplasty groups. After thrombolysis first-year total mortality in patients treated with intravenous thrombolysis ranged from 3% (11,12)to >4% (13), 5% (14–16)and 7% (17). These follow-up data do not suggest major differences between our observations and those of others using direct percutaneous transluminal coronary angioplasty (2,7). However, in considering these observations one has to keep in mind that the patient populations in thrombolytic trials represent a selected group with a normal to low risk profile and without contraindications to intravenous thrombolysis. Regular postinfarction care in some of the early studies did not systematically include more recent therapeutic concepts, most notably afterload reduction. However, given these significant restrictions to a long-term comparison of direct percutaneous transluminal coronary angioplasty with thrombolysis, we suspect that the favorable results of direct percutaneous transluminal coronary angioplasty in the acute infarct phase are maintained after discharge.
An LV ejection fraction <35%, three-vessel disease, and age ≥75 years are independent risk factors of a two- to fourfold increase in mortality. Although univariate comparison detects several parameters that are related to increased mortality (Table 2), only the aforementioned factors are independent indicators of less favorable survival (Table 3). The most significant predictor of mortality in our cohort of patients is an LV ejection fraction of <35%, which indicates an almost fourfold increase in mortality compared with subjects with ejection fractions >35%. The central importance of LV function in predicting long-term survival has been repeatedly recognized irrespective of acute therapy (13,18–24). Not surprisingly, this predictor prevails as an independent risk factor in patients with rapid and almost complete restoration of coronary flow during acute MI by percutaneous transluminal coronary angioplasty. However, the subset of patients with a very low LV ejection fraction (<35%) at the time of discharge was small: 8% (33 of 392) in our population. This was perhaps a result of the rapid and complete revascularization through direct percutaneous transluminal coronary angioplasty. On the other hand, >90% of our patients had a predischarge ejection fraction >35%, resulting in a 97% survival rate for the first year after discharge.
Three-vessel disease (18,20,22)and advanced age (≥75 years) (24–27)are also well-known risk factors in patients undergoing conservative or thrombolytic therapy during acute MI. The latter group of elderly patients is of particular interest because they have a high acute mortality: 20% in the Global Utilization of Streptokinase and TPA for Occluded arteries-I (GUSTO-I) and Gruppo Italiano per lo Studio della Streptochinasi nell’ Infarto Miocardico studies and 14% in our patients after direct percutaneous transluminal coronary angioplasty. In addition, the number of elderly patients is significant: 15% of our total group and 12.5% in the GUSTO-I study (12). The cumulative 1-year mortality of our patients ≥75 years was 26%. With all the caution required in interpreting our findings, we propose that direct percutaneous transluminal coronary angioplasty yields a favorable, cumulative 1-year mortality after acute MI of elderly individuals. The favorable acute survival is not compromised by excessive mortality thereafter.
The trend to a higher long-term mortality for women or for patients with anterior MI, diabetes, or both, does not reach statistical significance (Table 3). Patients with nonsustained, asymptomatic ventricular tachycardia on Holter recordings between days 7 and 90 after the MI had a 4% 1-year total mortality after discharge, which is slightly lower than the 5% mortality of patients without that arrhythmia. This is in contrast to previous studies (28,29)in patients who underwent revascularization procedures other than direct percutaneous transluminal coronary angioplasty.
Reinfarction and reinterventions during long-term follow-up after direct PTCA
Because direct percutaneous transluminal coronary angioplasty reopens occluded vessels and reduces the underlying coronary stenosis at the same time, one might speculate that clinical events, most importantly reinfarction in the same region, may occur less frequently and that subsequent target vessel revascularization procedures may be rarely required. On the other hand, because many vessels are acutely reopened, restenosis and reocclusions can occur. This dilemma has been addressed and recently validated for placebo-controlled and comparative trials (30).
Annual recurrence of MI after discharge was 1% to 2% in our series, with >50% occurring in the same location (i.e., true reinfarctions). Similar reinfarction rates of 3% and 2% at 1 year after direct percutaneous transluminal coronary angioplasty have been reported by O’Keefe et al. (2)and Zijlstra et al. (7), respectively.
Recurrent angina or MI prompted the reinterventions listed in Figure 2. Target vessel repeat percutaneous transluminal coronary angioplasty was required in 10% of patients in the first year after discharge and in <5% per year in subsequent years (Fig. 2). In interpreting these repeat percutaneous transluminal coronary angioplasty rates it must be noted that 12% of our patients had target vessel repeat percutaneous transluminal coronary angioplasty during the initial hospital phase before discharge. If one also adds all surgical procedures during follow-up that bypassed the target vessel, cumulative incidences of target vessel revascularization after 6 months and 1, 2, 3 and 4 yr were 26%, 28%, 29%, 31% and 33%, respectively (Kaplan–Meier analysis). Ongoing trials that test the general use of stents (only 3% in our series) or glycoprotein IIb/IIIa platelet receptor antagonists during direct percutaneous transluminal coronary angioplasty will have to demonstrate whether the need for repeat target vessel procedures during long-term follow-up can be reduced.
The favorable results of direct percutaneous transluminal coronary angioplasty that can be achieved in the acute phase of MI in experienced hands are not undone after discharge. Because mortality (4.6%) and recurrent ischemia peak in the first year after discharge for acute MI, frequent follow-up visits and careful reevaluation of these patients are required during that time frame. Reinterventions are often required (10% to 20%) within 1 year after discharge. Although few in number after direct percutaneous transluminal coronary angioplasty, patients with severely impaired LV ejection fraction (<35%) are at particular risk.
The authors thank M.K. Steen-Müller, MD, Medizinische Klinik I Kardiologie-Angiologie, Justus-Liebig University of Giessen, for careful review of the manuscript.
- left ventricular
- myocardial infarction
- percutaneous transluminal coronary angioplasty
- Thrombolysis In Myocardial Infarction trial
- Received November 24, 1997.
- Revision received April 17, 1998.
- Accepted June 22, 1998.
- American College of Cardiology
- Gibbons RJ, Holmes DR, Reeder GS, Bailey KR, Hopfenspirger MR, Gersh BJ, for the Mayo Coronary Care Unit and Catheterization Laboratory Groups
- Zijlstra F,
- de Boer M,
- Beukema W,
- et al.
- Chesebro J,
- Knatterud G,
- Roberts R,
- et al.
- American College of Cardiology/American Heart Association
- TIMI Investigators,
- Williams D.O,
- Braunwald E,
- Knatterud G,
- et al.
- Califf R, White H, vandeWerf F, et al., for the GUSTO-I Investigators
- Hohnloser S,
- Franck P,
- Klingenheben T,
- Zabel M,
- Just H
- McClements B,
- Adgey A
- Arnold A,
- Simoons M,
- Detry J,
- et al.
- Gottlieb S, Goldbourt U, Boyko V, et al. for the SPRINT and Thrombolytic Survey Group
- Smith S,
- Gilpin E,
- Ahnve S,
- et al.
- Weaver D, Litwin P, Martin J, et al. for the MITI Project Group
- The Multicenter Post-Infarction Group,
- Bigger J,
- Fleiss J,
- Kleiger R,
- Miller P,
- Rolnitzky L
- Farrell T,
- Bashir Y,
- Cripps T
- van de Werf F