Keywords
INTRODUCTION
Elderly acute myocardial infarction (AMI) patients present specific clinical characteristics which differ from younger patients. These include a higher proportion of women, high comorbidity, longer delay before using the emergency service, and very high hospital and short-term mortality.1-6 Despite the poor prognosis, very elderly patients, ≥75 years old, normally receive less aggressive or complete medical treatment.1 Furthermore, the efficacy of reperfusion therapy in these patients remains a matter of dispute. This is particularly true regarding fibrinolysis.7-10 On the other hand, there is little data regarding the advantages of primary angioplasty (PA) versus fibrinolysis.11-13 For these reasons, the need for a randomized study comparing both therapeutic choices is patent.13 The Ischemic Cardiopathy and Hemodynamics Sections of the Spanish Society of Cardiology, being aware of this need, considered it necessary to investigate the clinical characteristics, treatment and prognosis of elderly AMI patients admitted to Spanish hospitals with an active PA program. The final aim was to study the feasibility of carrying out a national clinical trial comparing fibrinolysis and PA in these patients. Thus, the TRatamiento del Infarto Agudo de miocardio eN Ancianos (TRIANA) Registry was set up and divided into 2 subregistries: TRIANA 1, which included all the AMI patients of any age who underwent PA or rescue angioplasty, and TRIANA 2, which included all the patients >75 years old who did not undergo early coronary angioplasty (primary or rescue). This paper presents the outcomes of the patients ≥75 years old included in both registries.
PATIENTS AND METHODS
Patients
The TRIANA Registry is a prospective multicenter study in which all the Spanish hospitals that performed a minimum of 50 PA per year were invited to participate. A total of 26 centers accepted the invitation (1 hospitals did not include patients in TRIANA 1 and 1 did not include them in TRIANA 2). Between 18 March and 31 July 2002, the TRIANA 2 subregistry included all the patients ≥75 years old admitted with ST-segment elevation AMI or complete left branch bundle block who did not receive PA or rescue angioplasty. The TRIANA 1 subregistry included 459 patients on similar dates but with 1 month less recruitment time. In the present article the data from the 306 patients included in the TRIANA 2 subregistry are analyzed plus those from the 104 patients ≥75 years old included in the TRIANA 1 subregistry (92 received PA and 12 rescue angioplasty).
Variables
The variables analyzed were: a) baseline and demographic characteristics (Table 1); b) infarction characteristics (Table 2); c) medical treatment (Table 3); d) diagnostic tests: echocardiography, non-invasive ischemia detection tests (ergometry, echocardiography stress test or tests with radioactive isotopes), emergency or elective catheterization, and angioplasty; e) hospital evolution (Table 4); and f) 30-day follow-up.
Statistical Analysis
The results of the continuous variables are expressed as mean±SD, or as median and 25/75 percentiles when they did not follow a normal distribution. The means between the three groups were compared with ANOVA. Between-group categorical variables were compared with the χ² test. The predictive value of the variables contributing to mortality was analyzed using a multiple logistic regression model constructed by introducing all the variables with a significant relationship to mortality (P<.1). Subsequently, those which did not contribute additional information to the model were excluded step-by-step. In addition, the variable reperfusion was introduced (although not in contrast to fibrinolysis/PA). All the analyses were done with SPSS software version 11.0 (SPSS, Inc., 2002) and expressed as 2-tailed P-values. Mortality risk was adjusted for age, sex, systolic blood pressure at the time of admission, fibrinolysis, PA, Killip class and AMI location.
RESULTS
Characteristics of the Population
Four hundred and ten patients were recruited over 4.5 months. Mean age was 80±4.3 years (range, 75-98 years) and 46.3% were women. Out of the total, 146 (35.6%) received fibrinolytic treatment--12 of whom (9%) also underwent rescue angioplasty--, 92 (22.4%) PA, and 172 (42%) did not undergo any type of reperfusion therapy. Treatment was decided according to the criterion of the acting specialist and the current protocol in each hospital. The majority of the patients treated with fibrinolytics (86.9%) received the drug in the first 6 h from onset of symptoms; tenecteplase was the most frequently administered agent. No reason was given for non-administration of fibrinolytic therapy in 21.6% of cases, whereas in the remainder the researchers considered that there was some type of contraindication (51.1% delayed admission, 12.4% advanced age, 10.9% comorbidity, 5.8% hemorrhagic risk, 1.5% previous stroke, and 18.2% other causes).
The patients who received reperfusion therapy, versus those who did not, were younger, presented pain as the main symptom and anterior AMI more often, were admitted with less delay and presented fewer signs of pulmonary congestion. Furthermore, there were fewer women and less background of kidney failure or dependency during daily activities. Tables 1 and 2 show differences in baseline and AMI characteristics between the 3 groups of patients.
Patients receiving some type of reperfusion therapy were more often treated with heparin, clopidogrel, glycoprotein IIb/IIIa inhibitors, beta-blockers, and ACE inhibitors, whereas those who did not receive reperfusion therapy required diuretics more often. Table 3 shows the between-group differences.
The patients who did not receive reperfusion therapy presented a higher incidence of heart failure and postinfarction angina, a trend toward greater incidence of shock and mechanical complications, but less stroke (1.2 vs 4.2%; P=.07) and hemorrhage (2.9 vs 8.8%, P=.02). The incidence of stroke was slightly higher in the group treated with fibrinolytics (six out of the eight cases involved hemorrhagic stroke). Table 4 shows evolution according to type of treatment. Echocardiogram was done in 75.1% of the patients, with a left ventricular ejection fraction lower than 0.40 in 32%. A screening test for ischemia (ergometry, echocardiography stress test or tests with radioactive isotopes) was only done in 13% of the patients. Elective catheterization was done in 18%.
Hospital and 30-Day Mortality
Global hospital mortality was 24.1%, mainly due to cardiogenic shock (58.6%). There were no differences in hospital mortality regarding reperfusion therapy (22 vs 26.7%; P=.30), although the cause of death differed (69.6% of deaths were due to cardiogenic shock in the untreated group vs 49.1% in treated patients; P=.04). There were 32 rehospitalizations (7.8%) over 1 month and total mortality increased to 24.9% with five deaths; 23% among the patients who received reperfusion therapy and 28.2% among those who did not (P=.23). There was no significant difference in hospital and 1-month mortality between the three treatment modalities (Table 4 and Figure 1).
Figure 1. Survival curves of the 410 patients >=75 years old included in the TRIANA 1 and 2 subregistries according to treatment received: fibrinolysis, primary angioplasty, or no reperfusion therapy.
Table 5 and Figure 2 show predictors of mortality at 1 month (univariable and multivariable analysis). Greater age, lower blood pressure at admission and, in particular, advanced Killip class were the predictors of 30-day mortality, but not the use of fibrinolytics or primary angioplasty.
Figure 2. Independent predictors of 30-day mortality in the total population obtained via logistic regression.
OR indicates odds ratio; CI, confidence interval; SBP, systolic blood pressure; PA, primary angioplasty; AMI, acute myocardial infarction.
DISCUSSION
The elderly AMI patients treated in Spanish hospitals with an active PA program are high-risk patients presenting a high rate of in-hospital complications associated with increased mortality. Despite this, almost half of these patients do not receive reperfusion therapy and most of those receiving it are treated with fibrinolysis.
This registry contains some data worthy of comment. On the one hand, the patients studied present similar characteristics to those described for this age group (high prevalence of diabetes, arterial hypertension, stroke, previous chronic kidney failure, and long delays before arrival at hospital, i.e., high-risk criteria) and, in fact, evolve unfavorably with a high incidence of complications and mortality during admission. However, the general management of these patients in Spanish hospitals, which we can consider as being elite centers from the viewpoint of therapeutic resources for AMI, does not seem to differ greatly from that in the general population.14 Thus, 42% of the patients did not receive any reperfusion therapy, which in approximately half of these cases was due to the delay in hospital admission, but in 2 out of every 5 cases the reason for the lack of reperfusion was not given. Furthermore, a certain number of patients with other reported contraindications had been candidates for PA although not for fibrinolysis. Still more noteworthy is the fact that, in the hospitals with PA available, the reperfusion therapy of choice for the most elderly AMI patients is fibrinolysis, a fact which is quite difficult to explain. Although no definitive evidence exists regarding the ideal reperfusion therapy for elderly AMI patients, some evidence suggest that PA is probably better than fibrinolysis. It is known that PA is better than fibrinolysis in the general population and, in particular, in higher-risk patients.15 In the only randomized study conducted specifically in patients >75 years old comparing PA with fibrinolysis, although this only included 77 patients over 4 years in a single hospital, there was a considerable reduction in early and long-term mortality in the patients treated with PA compared to those who received streptokinase.11 These data are in line with other clinical trials and with observational studies which suggest that primary angioplasty specifically benefits elderly patients.12,13-16 In addition, the efficacy of fibrinolysis in elderly patients is a matter of controversy, since some published results suggest that it could be more effective in patients ≥75 years than in younger patients,7 less effective8 or could even be associated with greater 30-day mortality.9-17 Most patients treated with fibrinolysis received tenecteplase, but the ideal fibrinolytic for elderly patients remains unknown. In the GUSTO I study, alteplase was associated with greater benefit than streptokinase, despite a slightly higher risk of stroke.18 In the ASSENT 2 study, tenecteplase was slightly better than alteplase, especially in women >75 years old and in the patients who presented with a delay >4 h,19 a frequent situation among the elderly. However, although the most widely used drug in our patients (tenecteplase) might have some advantages in this special population group, the ASSENT 3 Plus study demonstrated a prohibitive rate (6.7%) of intracranial hemorrhages in the patients >75 years old treated with tenecteplase and standard doses of enoxaparin.20 Two-thirds of the patients in the TRIANA registry treated with tenecteplase received treatment with enoxaparin, which could explain the high rate of intracranial hemorrhages observed. The fundamental reason why fibrinolysis is the treatment of choice in elderly AMI patients in hospitals that employ primary angioplasty remains unknown since, in addition to their probable greater efficiency, these hospitals do not suffer the two main limitations to PA: logistic difficulties regarding its implementation and the team's lack of experience.
Finally, there was excessive delay before beginning reperfusion, especially regarding fibrinolysis, where the recommended 30 min was exceeded and even the time recorded in other series of Spanish hospitals was surpassed. The Acute Myocardial Infarction Hospital Registration Project (PRIAMHO II) showed that in patients with a mean age of 65.4 years in the year 2000 there was a median of 45 min14 from the time of arrival at the emergency unit until fibrinolysis was done, which is slightly less time than that found in the TRIANA registry (48 min). In addition, the median door-guidewire time in the patients treated with PA was 90 min in our series, which compares negatively with the 80 min in the PRIAMHO II study. It is known that, in elderly patients, delays in arrival at the Emergency Unit is associated with a delay in establishing any reperfusion therapy when compared with younger patients,21 even though they have greater absolute risk. The TRIANA study demonstrates that hospital delay in elderly AMI patients is also very high.
Clinical Implications
Elderly AMI patients treated in Spanish centers with active PA programs form a selected subgroup with a high rate of complications and mortality. The TRIANA registry data suggest deficiencies and inconsistencies in the treatment of AMI in patients ≥75 years old and highlights the pending issue concerning which reperfusion therapy is optimal in elderly patients. It provides valuable information regarding the feasibility of a study to compare both reperfusion therapy options.
APPENDIX. Researchers and Hospitals Participating in the TRIANA 1 and 2 Registries
TRIANA 1. See p. 349.
TRIANA 2. N. Alonso, Complejo Hospitalario de León Hospital Princesa Sofía, León. J. Ángel, Ciudad Sanitaria Vall d'Hebron, Barcelona. J.J. Araiz, Hospital Clínico Universitario Lozano Blesa, Zaragoza. A. Bethancourt, Hospital Universitario Son Dureta, Palma de Mallorca. R. Blanco, Hospital de Cruces, Baracaldo. H. Bueno, Hospital General Universitario Gregorio Marañón, Madrid. F. Calvo, Hospital Do Meixoeiro, Vigo. R. Camarasa, Hospital General Universitari d'Alacant, Alicante. E. Esplugas, Ciudad Sanitaria y Universitaria de Bellvitge, L'Hospitalet de Llobregat, Barcelona. F. Fernández Avilés, Hospital Clínico de Valladolid, Valladolid. A. Fernández-Ortiz, Hospital Clínico San Carlos, Madrid. F. Hernández, Hospital 12 de Octubre, Madrid. J.M. Hernández, Hospital Virgen de la Victoria, Málaga. R. Hidalgo, Hospital Virgen de la Macarena, Sevilla. M. Jaquet, Hospital Clínico Universitario de Santiago, Santiago de Compostela. J. Martínez. Hospital de Navarra, Pamplona. L. Rodríguez Padial, Hospital Virgen de la Salud, Toledo. S. Romaní, Hospital Clínic, Barcelona. C. Romero, Hospital Universitario de la Princesa, Madrid. J. Romero, Fundación Jiménez Diaz, Madrid. F. Ruiz, Hospital Universitario Virgen de las Nieves, Granada. I. Sánchez, Hospital Central de Asturias, Oviedo. C. San José, Hospital Universitario Marqués de Valdecilla, Santander. C. Vázquez, Hospital Universitario Miguel Servet, Zaragoza. N. Vázquez, Hospital Juan Canalejo, A Coruña.
This study was organized by the Ischemic Cardiopathy and Hemodynamics and Interventional Cardiology Sections and Coronary Units of the Spanish Society of Cardiology and partly financed with unconditional grants from Sanofi, Medtronic, Boston Scientific, Guidant, and Johnson & Johnson.
*The researchers and hospitals participating in the TRIANA 1 and 2 subregistries are listed in the Appendix.
Correspondence: Dr. A. Bardají.
Servicio de Cardiología. Hospital Universitario Joan XXIII.
Dr. Mallafré Guasch, 4. 43007 Tarragona. España.
E-mail: abardaji@comt.es
Received February 25, 2004.
Accepted for publication January 20, 2005.