Long-term recovery of left ventricular function after primary angioplasty for acute myocardial infarction

Transcription

1 European Heart Journal (2001) 22, doi: /euhj , available online at on Long-term recovery of left ventricular function after primary angioplasty for acute myocardial infarction J. P. Ottervanger 1, A. W. J. van t Hof 1, S. Reiffers 2, J. C. A. Hoorntje 1, H. Suryapranata 1, M. J. de Boer 1 and F. Zijlstra 1 Departments of 1 Cardiology and 2 Nuclear Medicine, Isala Klinieken, Hospital de Weezenlanden, Zwolle, The Netherlands Aims To investigate changes in left ventricular function in the first 6 months after acute myocardial infarction treated with primary angioplasty. To assess clinical variables, associated with recovery of left ventricular function after acute myocardial infarction. Methods Changes in left ventricular function were studied in 600 consecutive patients with acute myocardial infarction, all treated with primary angioplasty. Left ventricular ejection fraction was measured by radionuclide ventriculography in survivors at day 4 and after 6 months. Patients with a recurrent myocardial infarction within the 6 months were excluded. Results Successful reperfusion (TIMI 3 flow) by primary angioplasty was achieved in 89% of patients. The mean ejection fraction at discharge was 43 7% 11 4, whereas the mean ejection fraction after 6 months was 46 3% 11 5 (P<0 01). During the 6 months, the mean relative improvement in left ventricular ejection fraction was 6%. An improvement in left ventricular function was observed in 48% of the patients; 25% of the patients had a decrease, whereas in the remaining patients there was no change. After univariate and multivariate analysis, an anterior infarction location, an ejection fraction at discharge 40% and single-vessel disease were significant predictors of left ventricular improvement during the 6 months. Conclusions After acute myocardial infarction treated with primary angioplasty there was a significant recovery of left ventricular function during the first 6 months after the infarction. An anterior myocardial infarction, single-vessel coronary artery disease, and an initially depressed left ventricular function were independently associated with recovery of left ventricular function. Multivessel disease was associated with absence of functional recovery. Additional studies, investigating complete revascularization are needed, as this approach may potentially improve long-term left ventricular function. (Eur Heart J 2001; 22: , doi: /euhj ) 2001 The European Society of Cardiology Key Words: Acute myocardial infarction, primary angioplasty, left ventricular ejection fraction, stunning. See page 785 for the Editorial comment on this article Introduction In survivors of acute myocardial infarction, left ventricular function is the most important predictor of long-term prognosis [1,2]. However, left ventricular function may change during the months after myocardial infarction, by mechanisms such as remodelling and gradual relief of stunning [3 7]. Determinants of a change Revision submitted 10 June 2000, and accepted 21 June Correspondence: Dr Felix Zijlstra, Department of Cardiology, Isala Klinieken Hospital de Weezenlanden, Groot Wezenland 20, 8011 JW Zwolle, The Netherlands X/01/ $35.00/0 in left ventricular function may have prognostic significance, give insight into the mechanisms of the changes in ventricular function, and may have important implications for the therapeutic approach. Data on changes of left ventricular function in the first 6 months after myocardial infarction, in patients treated with primary angioplasty are limited. To identify predictors of left ventricular functional recovery, a large group of survivors of acute myocardial infarction, who were all treated with primary angioplasty, were studied. All patients had serial measurements of left ventricular ejection fraction by radionuclide ventriculography, both at day 4 and 6 months after the index infarction The European Society of Cardiology

2 786 J. P. Ottervanger et al. Methods Patients All patients were included prospectively from January 1994 to January 1998, with the inclusion and exclusion criteria as described previously for our primary angioplasty vs thrombolysis trial [8,9]. In short, they had symptoms of acute myocardial infarction and ST segment elevations, with presentation within 6 h after onset of symptoms (or between 6 and 24 h if there was evidence of continuing ischaemia). Both patients who presented at the emergency department of our own hospital and patients referred to our angioplasty centre from community hospitals were included. All patients were treated with primary angioplasty immediately after admission. From each patient, baseline characteristics, coronary anatomy and time to reperfusion were recorded. Only patients who died or had a recurrent infarction within 6 months were excluded in the present analysis. All patients received aspirin, nitroglycerin and heparin intravenously. Beta-blockers were given unless contraindications were present and ACE inhibitors were started in patients with clinical signs or symptoms of a large infarction and/or heart failure, in accordance with current guidelines of therapy of acute myocardial infarction. Radionuclide ventriculography Radionuclide ventriculography was performed, according to the study protocol, on day 4 or day 5 after myocardial infarction and at the 6 month follow-up. Measurements were performed by the multiple-gated equilibrium method after in vivo labelling of red blood cells with 99m Tc pertechnetate. A γ-camera (General Electric, Milwaukee, WI, U.S.A.) was used. The global left ventricular ejection fraction was calculated with the PAGE program (version 2.3). The standard deviation of the difference between repeat measurements obtained by this technique is 1 2%. A significant change in left ventricular ejection fraction was therefore defined as an increase or decrease of more than 2 4%. A left ventricular ejection fraction 40% was regarded as clinically significant depressed left ventricular function. Left ventricular ejection fraction difference was defined as LVEF 6 month LVEF discharge =ΔLVEF. A relative left ventricular ejection fraction difference was calculated: ΔLVEF/LVEF discharge. This was used as the dependent variable in a multivariate model. By this method, the discharge left ventricular ejection fraction was taken into account in the analysis. Statistical analysis In the presentation of the data, continuous variables are given as mean SD, whereas discrete variables Table 1 Clinical characteristics of 600 patients with myocardial infarction treated with primary angioplasty are given as absolute values and percents. A value of P<0 05 was considered significant. Multivariate analysis was performed by fitting a linear regression model. In the multivariate analysis, adjustments were made for differences in age (continuous variable), gender, infarct location (anterior vs non-anterior), time from symptom-onset to first balloon inflation, previous infarction, multivessel disease and enzymatic infarct size. Results Study population n or mean n 600 Male 479 (80%) Age (years) 59 ( 11) Anterior infarct location 305 (51%) Previous infarction 66 (11%) Single vessel disease 286 (48%) TIMI flow grade 3 after PTCA 515 (89%) Enzymatic infarct size (LDH, U. l 1 ) 1237 ( 1005) ACE inhibitor at discharge 266 (44%) TIMI=Thrombolysis In Myocardial Infarction; ACE=angiotensin converting enzyme; LDH=lactate dehydrogenase. During the 4 year study period, 785 patients with acute myocardial infarction were treated with primary angioplasty in our centre. Within the 6-month follow-up period the mortality rate of this cohort was 5% (n=41 patients). Of the survivors, in 124 patients (15%) no paired measurements of left ventricular ejection fraction were performed. Most of these patients (n=111, 90%) were referred from other hospitals, and returned to the referring hospital immediately after primary PTCA, before left ventricular ejection fraction measurement was performed. Twenty patients (3%) had a recurrent infarction during the 6 months follow-up and were excluded. The remaining 600 patients were included in the final analysis. Their clinical characteristics are shown in Table 1. The median total ischaemic time was 200 min. Left ventricular function at discharge The mean left ventricular ejection fraction at discharge was 43 7% (SD 11%). Two hundred and fifteen patients (36%) had a LVEF discharge 40%. Patients with a low LVEF discharge ( 40%) were compared to those with a preserved LVEF discharge (>40%). The clinical characteristics of these two groups are shown in Table 2. The mean left ventricular ejection fraction in the group with

3 Recovery of LV function after AMI angioplasty 787 Table 2 Differences between patients with low (LVEF 40%) and those with preserved (LVEF >40%) left ventricular function at discharge LVEF discharge 40% n=215 LVEF discharge >40% n=385 Male 174 (81%) 305 (79%) Age (years) 59 ( 11) 58 ( 12) Anterior infarct location 170 (80%) 135 (35%)* Previous infarction 23 (11%) 43 (11%) Single vessel disease 103 (48%) 203 (47%) TIMI flow grade 3 after PTCA 173 (85%) 342 (91%) Median ischaemic time (min) Enzymatic infarct size (LDHQ72, U. l 1 ) 1898 ( 1141) 859 ( 672)* ACE inhibitor at discharge 174 (81%) 92 (24%)* LVEF=left ventricular ejection fraction; TIMI=Thrombolysis In Myocardial Infarction; ACE=angiotensin converting enzyme; LDHQ72=cumulative release of lactate dehydrogenase. *P<0 01. Table 3 Differences between patients with low (LVEF 40%) and those with preserved (LVEF >40%) left ventricular function 6 months after myocardial infarction LVEF 6 months 40% n=164 LVEF 6 months >40% n=436 Male 135 (82%) 344 (79%) Age (years) 58 ( 11) 58 ( 12) Anterior infarct location 120 (73%) 185 (42%)* Previous infarction 24 (15%) 42 (10%) Single vessel disease 66 (40%) 220 (51%) TIMI flow grade 3 after PTCA 128 (78%) 387 (89%)* Median ischaemic time (min) Enzymatic infarct size (LDHQ72, U. l 1 ) 2040 ( 1201) 924 ( 692)* ACE inhibitor at discharge 126 (77%) 140 (32%)* LVEF=left ventricular ejection fraction; TIMI=Thrombolysis In Myocardial Infarction; ACE=angiotensin converting enzyme. *P<0 01, P<0 05. left ventricular ejection fraction 40% was 31 5%. Patients with an anterior infarct location had an increased risk of a low ejection fraction, relative risk 3 7 (95% confidence interval ). Mean enzymatic infarct size was significantly higher in the group with a low ejection fraction. As expected, patients with a low left ventricular ejection fraction more often used an ACE inhibitor at discharge. Other clinical characteristics were comparable between the two groups. Left ventricular function after 6 months After 6 months, the mean left ventricular ejection fraction was 46 3% (SD 11 5), whereas 164 patients (27%) had an ejection fraction 40%. Again, patients with a low left ventricular ejection fraction ( 40%) were compared to those with a left ventricular ejection fraction higher than 40% (Table 3). Apart from the characteristics already found at discharge, single-vessel disease and TIMI flow 3 were significantly associated with a preserved left ventricular function after 6 months and after primary angioplasty, respectively. Changes in left ventricular function during the 6 months The mean left ventricular ejection fraction increased from 43 7% (SD 11 3%) at discharge, to 46 3% (SD 11 5%) at 6 months (P<0 01). The mean relative improvement in the left ventricular ejection fraction was 6%. No significant change in left ventricular ejection fraction was observed in 164 patients (27%), an increase in left ventricular ejection fraction was found in almost half of the patients (289 patients, 48%), whereas in 147 patients (25%) there was a decrease in left ventricular ejection fraction. At discharge, 215 patients (36%) had a low left ventricular ejection fraction ( 40%) whereas after 6 months only 164 patients (27%) had a low left ventricular ejection fraction (Fig. 1, P<0 05).

4 788 J. P. Ottervanger et al. Prevalence (%) Discharge P < % 27% 6 months follow-up Figure 1 Prevalence of a low left ventricular ejection fraction ( 40%) at discharge and after 6 months follow-up. Table 4 Relative increase of left ventricular ejection fraction, (LVEF 6 months LVEF discharge )/LVEF discharge 6 months after myocardial infarction according to several clinical characteristics Relative increase of LVEF (%) Males 5 3 Females 8 0 Age >70 years 6 8 Anterior location 10 8* Previous infarction 1 5 No previous infarction 6 4 Single-vessel disease 8 9* Multivessel disease 3 1 LVEF discharge 40% 15 1* ACE inhibitor at discharge 10 3 LVEF=Left ventricular ejection fraction. *P<0 01. Predictors of left ventricular recovery In Table 4, the relative increase in left ventricular ejection fraction according to several clinical characteristics is summarized. In the total study group, after both univariate and multivariate analysis, an anterior location, single-vessel disease and a low LVEF discharge were significantly associated with left ventricular ejection fraction recovery at follow-up (Table 4). The 215 patients with left ventricular ejection fraction at discharge 40% had a relative increase of left ventricular ejection fraction of 15 1%, whereas the 385 patients with left ventricular ejection fraction at discharge >40% had only a 2 7% increase. Two hundred and sixty-six patients used an ACE inhibitor at discharge. These patients had a relative increase in left ventricular ejection fraction of 10 3%, whereas patients not using an ACE inhibitor had a relative increase of 2 7%. However, after multivariate analysis, this was not statistically significantly different. Separate analysis for patients with LVEF discharge 40% showed that in these patients only single-vessel disease remained significantly associated with an improvement in left ventricular function after multivariate analysis. In the 103 patients with LVEF discharge 40% and single-vessel disease, the relative increase in left ventricular ejection fraction was 20%. In patients with LVEF discharge >40%, single-vessel disease (relative left ventricular ejection fraction increase 4 9%) and an anterior location (relative left ventricular ejection fraction increase 6 2%) were significantly associated with an increase of left ventricular ejection fraction after multivariate analysis. There was no significant difference in the relative increase of left ventricular ejection fraction between patients with ischaemic time <6 h (6 1) and patients with ischaemic time >6 h (5 2). Discussion In our study of patients who were treated with primary angioplasty for acute myocardial infarction, almost 50% had a significant improvement in left ventricular function after 6 months. The mean left ventricular ejection fraction increased from 43 7% at discharge, to 46 3% at 6 months (P<0 01). This improvement is in contrast to a study assessing systolic left ventricular function after thrombolytic therapy [10]. An anterior infarction, single-vessel disease and a low ( 40%) LVEF discharge were independent predictors of improvement of left ventricular function in our study population. Predictors and potential mechanisms of left ventricular improvement After acute myocardial infarction, the earliest change is expansion of the infarcted zone. In the period thereafter, compensatory hypertrophy of the non-infarcted segments leads to stable enlargement of the ventricle [11 14]. The main factor affecting this remodelling process is infarct size [15,16]. In patients treated with reperfusion therapy (either thrombolytic therapy or angioplasty), infarct size is reduced when patency is achieved within the period of myocardial salvage, in particular if patency is also complete and sustained [17 19]. Reperfusion not only limits infarct size [17], but also preserves viable myocardium in the infarct zone [18,20]. However, in this potentially viable myocardium, despite restoration of perfusion, myocardial function can be depressed for a long period (stunning) [17,21]. This reversible impairment of left ventricular function has been demonstrated, in particular, in patients with anterior infarction [22]. This is in agreement with our findings, that an anterior infarction lent itself to an improvement in left ventricular function during the months after the infarction.

5 Recovery of LV function after AMI angioplasty 789 Hibernation is another cause of reversible left ventricular dysfunction [23]. Revascularization or optimal antiischaemic pharmacotherapy may cause a reversal of hibernation, and consequently an improvement in left ventricular ejection fraction [21]. In our study, all patients were treated with primary angioplasty, which achieves early and complete patency in a large number of patients [24]. Furthermore, the majority of patients were treated within the time window of myocardial salvage. The relatively high percentage of patients with an increase in left ventricular function between day 4 and 6 months could, in part, be a reflection of this rapid and complete reperfusion [21]. A low left ventricular ejection fraction at discharge was a significant predictor of left ventricular ejection fraction improvement in our study. Possibly, stunning is prolonged in larger infarctions, while in smaller infarctions this has already (partially) resolved when left ventricular ejection fraction measurement at day 4 is performed. The presence of single-vessel disease was also a strong predictor of left ventricular improvement in our analysis, in particular in patients with a low left ventricular ejection fraction at discharge. Patients with multivessel disease have more extensive coronary artery disease and probably a limited collateral blood flow. This may be associated with prolonged or profound myocardial ischaemia, causing more severe hibernation. Studies investigating the influence of complete revascularization are therefore needed to demonstrate whether revascularization of potentially ischaemic myocardium distant from the infarct size is of clinical benefit. There was only a non-significant association between a shorter ischaemic time and better recovery of left ventricular function. assessment of left ventricular function was undertaken before or immediately after angioplasty. Since an improvement in left ventricular function may have occurred during the first 4 days, the observed improvement in left ventricular ejection fraction may have under-estimated the true benefit. In our study population, we measured only serial left ventricular ejection fraction. We had no data on diastolic left ventricular dysfunction, or left ventricular dimensions, including end-diastolic or end-systolic volumes. This could have provided additional information on the remodelling process. Furthermore, follow-up left ventricular ejection fraction was performed at 6 months, and some reports have demonstrated ongoing remodelling beyond this period [11]. Moreover, we had no data on the extent of stunned but functionally viable myocardium in the infarct zone, as can be estimated by dobutamine stress echocardiography [27]. Also we did not study the possible influence of collaterals on recovery of left ventricular function. It was not possible to evaluate the exact role of drugs, in particular ACE inhibitors, in our analysis. ACE inhibitors were given mainly to patients with poor left ventricular ejection fraction at discharge, introducing selection bias. Randomized trials have demonstrated the beneficial role of ACE inhibitors [28,29], and it is indeed possible that the significant increase in left ventricular ejection fraction in patients with low left ventricular ejection fraction at discharge was, in part, a result of use of ACE inhibitors. Since we had no follow-up data after 6 months, it was not possible to evaluate whether a change in left ventricular function is also associated with a change in prognosis. Future follow-up studies should be performed to evaluate this. Deterioration of left ventricular function There was deterioration of left ventricular function between day 4 and 6 months in 25% of patients. This may be due to hyperkinesia of the non-infarcted area in the acute phase, further loss of myocardial function during the 6 months, or a combination of these mechanisms. To distinguish between these two causes, regional wall motion analyses or simultaneous measurements of flow and function should be performed. Silent restenosis of the infarct-related artery during the 6 months follow-up period could have been a cause of deterioration of left ventricular function [25], although the incidence of restenosis after primary angioplasty is not very high [26]. However, since we did not perform coronary angiograpy after 6 months routinely in our study population, we cannot rule out that restenosis or new stenosis had occurred. Study limitations In our study early left ventricular function was measured 4 days after primary angioplasty, whereas no routine Conclusions and clinical implications Recovery of left ventricular function is seen in almost half the patients with acute myocardial infarction treated with primary angioplasty, whereas in 25% a deterioration is observed. Anterior infarct location, low left ventricular ejection fraction at discharge and single-vessel disease are associated with left ventricular functional recovery. In particular, patients with a poor left ventricular function at discharge and single-vessel disease have pronounced ventricular function recovery after 6 months. Multivessel disease with the potential of additional myocardial ischaemia, is associated with a decrease in left ventricular ejection fraction. As patients with coronary artery disease and left ventricular dysfunction with non-revascularized viable myocardium may benefit from revascularization, a trial with a more aggressive approach to myocardial ischaemia, by complete revascularization in the first months after myocardial infarction, is needed.

6 790 J. P. Ottervanger et al. References [1] White HD, Morris RM, Brown MA, Brandt PWT, Whitlock RML, Wild CJ. Left ventricular end-systolic volume as the major determinant of survival after recovery from myocardial infarction. Circulation 1987; 76: [2] St. John Sutton M, Pfeffer MA, Plappert T et al. Quantitative two-dimensional echocardiographic measurements are major predictors of adverse cardiovascular events after acute myocardial infarction. Circulation 1994; 89: [3] Braunwald E, Kloner RA. The stunned myocardium: prolonged, postischemic ventricular dysfunction. Circulation 1982; 66: [4] Bolli R, Zhu WX, Thomby JE, O Neill PG, Roberts R. Time-course and determinants of recovery of function after reversible ischemia in conscious dogs. Am J Physiol 1989; 254: H [5] Patel B, Kloner RA, Przykelnk K, Braunwald E. Postischemic myocardial stunning : a clinically relevant phenomenon. Ann Intern Med 1988; 108: [6] Bolognese L, Cerisano G, Buonamici P et al. Influence of infarct-zone viability on left ventricular remodeling after acute myocardial infarction. Circulation 1997; 96: [7] Wijns W, Vatner SF, Camici PG. Hibernating myocardium. Mechanisms of disase. N Engl J Med 1998; 3: [8] Zijlstra F, de Boer MJ, Hoorntje JCA et al. A comparison of immediate coronary angioplasty with intravenous streptokinase in acute myocardial infarction. N Engl J Med 1993; 328: [9] Zijlstra F, Hoorntje JCA, de Boer MJ et al. Long-term benefit of primary angioplasty as compared with thrombolytic therapy for acute myocardial infarction. N Engl J Med 1999; 341: [10] Harrison JK, Califf RM, Woodlief LH et al. Systolic left ventricular function after reperfusion therapy for acute myocardial infarction. Analysis of determinants of improvement. The TAMI Study Group. Circulation 1993; 87: [11] Gaudron P, Eilles C, Ertl G, Kochsiek K. Compensatory and noncompensatory left ventricular dilatation after myocardial infarction: Time course and hemodynamic consequences at rest and during exercise. Am Heart J 1992; 123: [12] Pfeffer MA. Left ventricular remodeling after acute myocardial infarction. Ann Rev Med 1995; 46: [13] Golia G, Rossi A, Anselmi M et al. Opposite effects of the remodeling of infarcted and non-infarcted myocardium on left ventricular function early after infarction in humans. An echocardiographic study in patients examined before and after myocardial infarction. Int J Cardiol 1997; 60: [14] Gaudron P, Eilles C, Kugler I, Ertl G. Progressive left ventricular dysfunction and remodeling after myocardial infarction. Potential mechanisms and early predictors. Circulation 1993; 87: [15] Chareonthaitawee P, Christian RF, Hirose K, Gibbons RJ, Ruberger JA. Relation of initial infarct size to extent of left ventricular remodeling in the year after acute myocardial infarction. J Am Coll Cardiol 1995; 25: [16] Erlebacher JA, Weiss JL, Eaton LW, Kallman C, Weisfeldt ML, Bulkley BH. Late effects of acute infarct dilatation on heart size: a two-dimensional echocardiographic study. Am J Cardiol 1982; 49: [17] Lamas GA, Flaker GC, Mitchell G et al., for the Survival and Ventricular Enlargement Investigators. Effect of infarct artery patency on prognosis after acute myocardial infarction. Circulation 1995; 92: [18] Brodie BR, Stuckey TG, Kissling G, Hansen CJ, Weintraub RA, Kelly TA. Importance of infarct-related artery patency for recovery of left ventricular function and late survival after primary angioplasty for acute myocardial infarction. J Am Coll Cardiol 1996; 28: [19] Bolognese L, Antoniucci D, Rovai D et al. Myocardial contrast echocardiography versus dobutamine echocardiography for predicting functional recovery after acute myocardial infarction treated with primary coronary angioplasty. J Am Coll Cardiol 1996; 28: [20] Christian TF, Behrenbeck T, Pellikka PA, Huber KC, Chesebro JH, Gibbons RJ. Mismatch of left ventricular function and infarct size demonstrated by technetium-99m isonitrile imaging after reperfusion therapy for acute myocardial infarction: identification of myocardial stunning and hyperkinesia. J Am Coll Cardiol 1990; 16: [21] Bestetti A, Lomuscio A, Giovanella LC, Castellani M, Tarolo GL. Delayed effect of streptokinase on left ventricular function after acute myocardial infarction assessed by equilibrium gated radionuclide angiocardiography. J Nucl Biol Med 1993; 37: [22] Christian TF, Gitter MJ, Miller TD, Gibbons RJ. Prospective indentification of myocardial stunning using Technetium-99m Sestamibi-based measurements of infarct size. J Am Coll Cardiol 1997; 30: [23] Galli M, Marcassa C, Bolli R et al. Spontaneous delayed recovery of perfusion and contraction after the first 5 weeks after anterior infarction. Evidence for the presence of hibernating myocardium in the infarcted area. Circulation 1994; 90: [24] Weaver WD, Simes RJ, Betriu A et al. Comparison of primary coronary angioplasty and intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review. JAMA 1997; 278: [25] Monin JL, Garot J, Scherrer-Crosbie M et al. Prediction of functional recovery of viable myocardium after delayed revascularization in postinfarction patients: accuracy of dobutamine stress echocardiography and influence of longterm vessel patency. J Am Coll Cardiol 1999; 34: [26] Van t Hof AWJ, de Boer MJ, Suryapranata H, Hoorntje JCA, Zijlstra F. Incidence and predictors of restenosis after successful primary coronary angioplasty for acute myocardial infarction: The importance of age and procedural result. Am Heart J 1998; 1336: [27] Carlos ME, Smart SC, Wynsen JC, Sagar KB. Dobutamine stress echocardiography for risk stratification after myocardial infarction. Circulation 1997; 95: [28] Pfeffer MA, Braunwald E, Moye L et al., on behalf of the SAVE Investigators. Effect of captopril on mortality and morbidity in patients with left ventricular dysfunction after myocardial infarction. N Engl J Med 1992; 327: [29] Gruppo Italiano per lo Studio della Sopravvivenza nell infarto Miocardico. GISSI-3: effects of lisinopril and transdermal glyceryl trinitrate singly and together on 6-week mortality and ventricular function after acute myocardial infarction. Lancet 1994; 343: