Author + information
- Received October 30, 1997
- Revision received January 6, 1999
- Accepted January 21, 1999
- Published online May 1, 1999.
- Bjarne L Nørgaard, MDa,* (, )
- Karl Andersen, MD∗,
- Mikael Dellborg, MD∗,
- Putte Abrahamsson, MD∗,
- Jan Ravkilde, MDa,
- Kristian Thygesen, MD, FACCa,
- for the TRIM study group
- ↵*Reprint requests and correspondence: Dr. Bjarne Linde Nørgaard, Department of Medicine and Cardiology, Aarhus University Hospital, Tage Hansens Gade 2, DK-8000 Aarhus C, Denmark
We investigated whether the addition of 24 h of continuous vectorcardiography ST segment monitoring (cVST) for an early (within 24 h of the latest episode of angina) determination of cardiac troponin T (cTnT) could provide additional prognostic information in patients with unstable coronary artery disease (UCAD), i.e., unstable angina and non–Q wave myocardial infarction.
Determination of cTnT at admission and cVST are individually reported to be valuable techniques for the risk assessment of patients with UCAD.
Two hundred and thirty-two patients suspected of UCAD were studied. Patients were followed for 30 days, and the occurrence of cardiac death or acute myocardial infarction (AMI) were registered.
One ST segment episode or more (relative risk [RR] 7.43, p = 0.012), a cTnT level ≥0.20 μg/liter (RR 3.85, p = 0.036) or prestudy medication with calcium antagonists (RR 3.31, p = 0.041) were found to carry independent prognostic information after multivariate analysis of potential risk variables. By combining a cTnT determination and subsequent cVST for 24 h, subgroups of patients at high (25.8%) (n = 31), intermediate (3.1%) (n = 65) and low risk (1.7%) (n = 117) of death or AMI could be identified.
Twenty-four hours of cVST provides additional prognostic information to that of an early cTnT determination in patients suspected of having UCAD. The combination of biochemical and electrocardiographic methods provides powerful and accurate risk stratification in UCAD.
Patients with unstable angina pectoris (UAP) and non–Q wave myocardial infarction (nQMI) are often grouped together due to similarities in pathophysiology and clinical presentation (1,2). In the present study, UAP and nQMI are referred to as unstable coronary artery disease (UCAD). Within days or weeks after the clinical presentation, patients with UCAD are at high risk of subsequent cardiac death, myocardial infarction, reinfarction and disabling angina (3,4). However, these patients are heterogeneous regarding the severity of the disease, the clinical presentation and the prognosis (1–5). Therefore, early risk assessment is essential in cases of UCAD to allocate treatment resources appropriately.
An elevated plasma level of the cardiac structural protein troponin T (cTnT) at admission has been associated with a poor clinical outcome in UCAD (6–11). Episodes of transient myocardial ischemia detected by ambulatory electrocardiogram (ECG) recording (12–14)or continuous vectorcardiography ST segment monitoring (cVST) (15–18)have been associated with adverse clinical outcome in such patients as well. In contrast to ambulatory ECG recording, cVST has the potential for real-time analysis without the need for time-consuming retrospectively performed analysis. Because cTnT reflects previous myocardial damage, and ongoing active coronary artery disease is evidenced by cVST, these methods are chronologically complementary. Therefore, the aim of the present study was to determine whether cVST for 24 h could provide additional prognostic information for an early cTnT determination in patients with UCAD.
The present investigation was carried out as a substudy of the Thrombin Inhibition in Myocardial Ischemia (TRIM) trial, a prospective double-blind randomized multicenter trial conducted in Scandinavia from December 1994 through July 1995. The purpose of the main study was to investigate the efficacy of a new specific thrombin inhibitor (inogatran) in patients with UCAD (19). Patients between 25 and 80 years of age suspected of having UCAD were eligible. Patients were required to have new onset of ischemic chest pain or rapid deterioration of previously stable effort angina or onset of angina at rest. Time for inclusion was within 24 h of the latest episode of angina (qualifying episode of angina). At least one of the following criteria had to be present in order to comply with the diagnosis of UCAD: ECG changes compatible with myocardial ischemia, i.e., ST depression or T wave inversion (≥0.1 mV in at least two contiguous leads), or a history of previous myocardial infarction, positive coronary angiography, positive myocardial scintigraphy or a positive exercise test. The main exclusion criteria were any condition considered to increase the risk of bleeding. Patients who had been treated with oral anticoagulants, heparin or fibrinolytic drugs within one week before screening were excluded from participation, as were patients with impeding acute myocardial infarction (AMI) and indication for thrombolytic therapy. Patients with percutaneous transluminal coronary angioplasty (PTCA) or coronary artery bypass graft surgery (CABG) performed within three months were also excluded. Moreover, patients with uncontrolled heart failure, uncontrolled cardiac arrhythmia, serum creatinine >150 μM, hemoglobin <100 g/liter (6.21 mM), drug or alcohol abuse or known liver disease were excluded.
A total of 1,209 patients from 60 centers were included in the TRIM study. Of these, 232 patients from 10 centers had blood samples drawn and cVST performed for the present substudy. Written informed consent was obtained from all patients. The protocol was approved by the regional Scientific Ethical Committees of the participating centers.
Evaluation of index events
An index event was retrospectively diagnosed as either UAP or nQMI based on the clinical course, ECG changes and maximum plasma levels of cardiac markers available at each participating center (either total creatine kinase [CK], CK isoenzyme MB catalytic activity or CK-MB mass) obtained within 6 h of inclusion.
Evaluation of end-points
The end point was a composite of the events cardiac death and AMI. Patients were followed for 30 days. Acute myocardial infarction was considered present when two of the following criteria were complied with: 1) typical anginal chest pain of at least 20 min duration; 2) a diagnostic ECG (new Q-wave ≥0.04 s and ≥1 mm or Q/R >1/3, or ST elevation ≥2 mm in precordial or ≥1 mm in other leads followed by T wave inversion in at least two leads in two consecutive recordings); or 3) an increase in cardiac enzymes above the upper reference occurring separately from any elevation taking place at inclusion (reinfarction). Cardiac events were evaluated by an independent End point Committee.
Patients were randomized to one of three dose levels of inogatran or heparin (19). Treatment was given as an intravenous infusion for 72 h. All patients received low-dose aspirin daily if tolerated. Nitrates, beta-blockers and calcium channel blockers were administered at the discretion of the treating physician. Patients’ medications before inclusion are listed in Table 1. The need for coronary angiography, coronary angioplasty or CABG was determined by the treating physician.
An ECG was obtained at inclusion. The ECGs were interpreted by the local investigators and categorized according to the predominant ST/T segment status (ST elevation, ST depression, T wave inversion or any ST/T change). Blood specimens were obtained according to a predefined time schedule in the main TRIM trial (19). For the purpose of this study, a cTnT level was determined from a blood specimen obtained at inclusion. Venous whole blood was drawn into heparin-containing tubes that were centrifuged at 2,000 gfor a period of 20 min. The plasma was kept at −80°C and sent for analysis at the core laboratory of Aarhus University Hospital (Aarhus, Denmark). All measurements were performed by laboratory personnel without knowledge of the patients. Cardiac troponin T was measured using an enzyme-linked immunosorbent assay [ELISA Troponin(e) T] on the ES 300 analyzer (Boehringer Mannheim GmbH, Mannheim, Germany). The method is based on a single-step sandwich principle with streptavidin-coated tubes as the solid phase and two monoclonal antihuman cTnT antibodies. Detection limit of the assay is 0.012 μg/liter. The method is described in detail elsewhere (20).
Continuous vectorcardiography ST segment monitoring was performed for 24 h after inclusion in all patients, using the MIDA 1000 or Coronet systems (Ortivus Medical AB, Täby, Sweden). The method of ischemia monitoring using vectorcardiography has previously been described in detail (21). Briefly, the system consists of a data acquisition module and an IBM-compatible personal computer. Electrocardiographic signals are continuously collected from eight conventional body surface electrodes positioned according to Frank (22). Electrocardiographic complexes are sampled for 1-min periods and averaged to form mean vectorcardiographic complexes in three orthogonal leads: X, Y and Z. The ST segment parameter studied was ST vector magnitude (ST-VM), which is the summarized deviation of the ST segment from the isoelectric level 20 ms after the J point. At the end of each averaging period, ST-VM is calculated and presented as a trend curve updated continuously during the whole recording period. An illustrative trend curve is shown in Figure 1. Recordings from patients with bundle branch block, pacemaker in situ or with less than 16 h recording time were not analyzed because cVST had not been validated in such patients. Only the first 24 h were analyzed from each recording. Recordings were stored on floppy disks and sent for central analysis at Sahlgrenska University Hospital (Gothenburg, Sweden). Analysis was performed by personnel without knowledge of the patients.
Discriminator values for the differentiation between high- and low-risk patients
In earlier studies, a cTnT level of 0.20 μg/liter (8,23)was used as the discrimination limit. However, a lowering of the cTnT cut-off value has occurred in parallel with an advance in the immunoassay technology, and today several studies have confirmed the applicability of 0.10 μg/liter as the discriminator value in the prognostic evaluation of patients with UCAD (7,24). Therefore, in the present study, cTnT discriminator values of 0.10 and 0.20 μg/liter, respectively, were studied. There is no gold standard for the diagnosis of myocardial ischemia by ST segment monitoring. However, based on previous investigations, assessment of ischemic changes was performed as one or more transient increases of ST-VM ≥50 μV from the individual baseline for >1 min (15,17,18).
In the statistical calculations of the composite end-point, each patient was represented by one cardiac event only. Due to nonnormal distribution, cTnT and cVST data were presented as medians (interquartile range [IQR], range). Group comparisons of continuous variables were made by the Mann-Whitney Utest, and categorical variables were compared using the Chi-square or Fischer exact test as appropriate. The Spearman rank-correlation coefficient relating the number of ST-VM episodes to the level of cTnT was calculated. The variables in Table 3and Table 6were evaluated by the Kaplan-Meier method and the log-rank test to detect univariate predictors of the composite end-point. Forward “stepwise” Cox regression analysis was performed in order to detect independent predictors of the composite end point using all variables of potential significance (p < 0.10) identified in the univariate analysis. At first, baseline variables of potential significance were evaluated in this multivariate model. Once these independent baseline predictors were accounted for, the added value of the cTnT level and ST segment monitoring, respectively, were explored in a second multivariate analysis. Two separate models were built to evaluate the independent prognostic significance of each cTnT cut-off value. Accordingly, three separate multivariate models were built. In the multivariate models, a p value <0.20 was used as the criterion for selection. A p value <0.05 was considered statistically significant.
Characteristics of patients
A total of 232 patients were included in the study. The index events were classified as UAP in 152 (66%), nQMI in 74 (32%) and other diagnoses in 6 (3%) patients. All patients with other diagnoses had a Q wave myocardial infarction. In contrast to UAP, nQMI was a more frequent index event among patients with a cTnT level ≥0.20 μg/liter or one ST-VM episode or more, as compared with the total substudy population (Table 1). Moreover, among patients with a cTnT level ≥0.20 μg/liter, fewer had previous angina, or used aspirin or nitrates, at inclusion as compared with patients with one ST-VM episode or more and the total substudy group, respectively. Otherwise, the two groups of patients had similar baseline characteristics and were also representative of the total substudy population. The median (IQR) time from the qualifying episode of angina to inclusion was 12 h (7 to 18).
Cardiac troponin T and continuous ST segment monitoring
Two hundred and thirty-one, 223 and 214 patients, respectively, had determination of the cTnT level, ECG analysis and cVST analysis performed. The blood sample was lost in one (0.4%) patient. Nine (3.9%) patients were excluded from the ECG data analysis due to the presence of bundle branch block. Eighteen (7.8%) patients were excluded from the ST segment monitoring analysis due to the presence of bundle branch block (n = 9), <16 h of cVST performed (n = 4) and technical problems (loss of backup or electrical interference) (n = 5). Complete clinical, electrocardiographic, biochemical marker and ST segment monitoring data were obtained in 213 (92%) patients. Eighty-seven (38%) patients had a cTnT value ≥0.10 μg/liter, 65 (28%) had a cTnT value ≥0.20 μg/liter and 69 (32%) patients had one ST-VM episode or more. The median level (IQR; range) of cTnT was 0.02 (0.00–0.23; 0.00–3.69) μg/liter. The median number (IQR; range) of ST-VM episodes was 0 (0–2; 0–24). The median time (IQR; range) of cVST was 24.00 (23.59–24.00; 16.14–24.00) h. Patients treated with inogatran had more ST-VM episodes as compared with patients treated with heparin. Treatment effect of the study drugs on cVST has previously been described in detail (18). Among patients with one ST-VM episode or more, the median level (IQR; range) of cTnT was 0.15 (0.02–0.68; 0.00–3.69) μg/liter as compared with 0.01 (0.00–0.13; 0.00–2.91) μg/liter in patients without ST-VM episodes (p = 0.0003). The numeric relationship between the number of ST-VM episodes and the cTnT level at inclusion is presented in Table 2. There was a moderate positive relationship between the number of ST-VM episodes and the cTnT level (Spearman correlation, rho = 0.45, p < 0.0001).
Three (1.3%) patients died and 12 (5.2%) patients had a nonfatal AMI during the 30 days of follow-up. A coronary angioplasty or CABG was performed in 57 (25%) patients. A revascularization procedure was performed in nine (of the 12) patients with an AMI. One patient had an AMI 15 h after a CABG, otherwise, all AMIs occurred before coronary angioplasty (median time [range] from event to procedure: 7 [2 to 82] h) or CABG (median time [range] from event to procedure: 84 [3 to 217] h). Revascularization procedures were not performed in any of the patients who died. Among 18 patients excluded from the cVST analysis, two died (exclusion due to left bundle branch block and <16 h of monitoring time, respectively) and one had an AMI (exclusion due to <16 h of monitoring time).
Predictors of cardiac death or acute myocardial infarction
The univariate predictive values of several risk variables are presented in Table 3. Of these, prestudy medication with calcium antagonists, an index diagnosis of nQMI, inclusion ECG ST segment depression, a cTnT level ≥0.10 or ≥0.20 μg/liter at inclusion or one ST-VM episode or more at 24 h of cVST were significantly associated with the composite end-point. In one patient with a CABG-related end-point (AMI), neither cTnT testing nor cVST were of predictive value. In Figure 2, the cumulative percentage of patients with event (death or AMI)-free survival is illustrated by Kaplan-Meier curves according to the level of cTnT (cutoff value: 0.20 μg/liter) and the number of ST-VM episodes (cutoff value: 1). There was no difference in median time (range) from inclusion to the composite end-point among those patients predicted by cTnT testing as compared with those predicted by cVST: 80 (11 to 357) versus 79 (23 to 199) h (p = 0.84). By multivariate analysis of baseline variables of potential significance, inclusion ECG ST segment depression (risk ratio [RR] 3.09, 95% confidence interval [CI] 0.97–9.77; p = 0.055) and prestudy medication with diuretics (RR 2.69, CI 0.84–8.53; p = 0.093) or calcium antagonists (RR 2.35, CI 0.75–7.33; p = 0.14) were retained in the final model. Table 4shows the risk ratios, confidence intervals and p values of risk variables retained in a second multivariate model exploring the relative value of the inclusion cTnT level (cut off value 0.20 μg/liter) and subsequent 24 h of cVST once the baseline variables have been accounted for. One ST-VM episode or more, a cTnT level ≥0.20 μg/liter and prestudy medication with calcium antagonists independently predicted death or AMI at 30 days. By evaluating the cTnT cutoff of 0.10 μg/liter in a similar manner, only one ST-VM episode or more (RR 8.48, CI 1.75–40.86; p = 0.008) carried independent prognostic information; also included in this final model were prestudy medication with calcium antagonists (RR 3.12, CI 0.98–9.85; p = 0.053) and a cTnT level ≥0.10 μg/liter (RR 2.32, CI 0.65–8.21; p = 0.19).
Figure 3shows the flow of all patients according to the cTnT level at inclusion, the presence or absence of ST-VM episodes within 24 h of inclusion and the occurrence of death or AMI during 30 days of follow-up. The patients with both a cTnT level ≥0.20 μg/liter and one subsequent ST-VM episode or more had a 25.8% risk of death or AMI. Among patients with a cTnT level ≥0.20 μg/liter and 0, 1 to 3, 4 to 8, or ≥9 ST-VM episodes, the event rate was 0% (0/27), 11% (2/18), 44% (4/9) and 50% (2/4), respectively. The event rate in patients with either of the methods positive was 3.1% (2/65). In patients with both a cTnT level <0.20 μg/liter and without ST-VM episodes, the risk of an event was 1.7%. This low-risk group comprised 50.4% of the total study population. The event rate for patients ineligible of cVST analysis was 16.7% (3/18). Among this subgroup of patients, those with a cTnT level ≥0.20 μg/liter had an event rate of 28.6% (2/7) as compared with 9.1% (1/11) in patients without an elevated cTnT level.
Unstable angina pectoris
Among patients in whom an index event was categorized as UAP (n = 152), those with one ST-VM episode or more had a median cTnT level (IQR; range) of 0.02 (0.00–0.07; 0.00–2.00) μg/liter as compared with 0.00 (0.00–0.02; 0.00–1.09) μg/liter in patients without ST-VM episodes (p = 0.02). The numeric relationship between the number of ST-VM episodes and the cTnT level at inclusion among patients with UAP is shown in Table 5. The positive relationship between the number of ST-VM episodes and the cTnT level was retained in this subset of patients (Spearman correlation, rho = 0.44, p < 0.0001). Both cTnT and ST-VM retained their prognostic value among patients with UAP (Table 6).
This prospective study demonstrated that both early determination of cTnT and subsequent cVST for 24 h possess independent prognostic information in the same cohort of patients suspected of having UCAD. However, by combining early determination of cTnT and subsequent continuous ST segment monitoring, we found that patients with both a cTnT level ≥0.20 μg/liter and one ST segment (ST-VM) episode or more were at particularly high risk (25.8% event rate), while patients with either one of the methods positive were at intermediate risk (3.1% event rate) of death or AMI at 30 days. As the treatment of UCAD is cumbersome, expensive and occasionally risky, treatment should be tailored according to thorough considerations regarding the underlying pathophysiology and prognosis. However, whether specific treatment regimens will change the natural course of the underlying coronary artery disease, and thus the prognosis, in such patients identified as being at high or intermediate risk is not known at present. Recent results however, indicate that the beneficial effect of prolonged antithrombotic treatment in cases of UCAD is more pronounced in patients with elevated cTnT levels as compared with those without (24). It is equally important to identify patients at very low risk in whom early transmission to a low-cost facility or even discharge from hospital may be considered. In the present study, 3% of the patients with a cTnT level <0.20 μg/liter suffered the composite end-point. This comparatively high event rate among patients without an elevated cTnT has been emphasized as a limitation for the use of this method alone in the prognostic assessment of patients with UCAD (25–27). Among patients of a similar category as the present study population, Lindahl et al. (28)identified a group with 1% risk of death or AMI at 5 months by combining a cTnT level <0.06 μg/liter and a low-risk predischarge exercise test response. However, this very low-risk subgroup only constituted 9% of the total study population. In the same series, an event rate of 4.1% was found (in 18% of the total study population) by combining a cTnT level <0.20 μg/liter and a low-risk predischarge exercise test response. Recently, Morrow et al. (11)investigated the predictive value of combining an early rapid bedside qualitative assay for cTnT and determination of CRP among patients comparable with the present study population. Although the number of clinical events was low in that study (event rate 1.6%), the combination of a negative cTnT (<0.20 μg/liter) and CRP <1.55 mg/dl identified a subgroup of patients (63% of the total study population) with a 0.36% risk of death within 14 days. In the present study, the combination of an early cTnT level <0.20 μg/liter and no ST segment episodes identified, within 24 h of inclusion, a subgroup of patients with a 1.7% risk of death or AMI. The fact that this low-risk group comprised more than 50% of the total study population is of considerable importance, as these patients may be considered for early transmission to a less costly facility than the intensive care setting.
Relationship between ST segment changes and the level of troponin T
Previous studies have demonstrated that the risk of a poor clinical outcome in patients with UCAD increases with the detection of increasing blood levels of sensitive and specific markers of myocardial damage such as the troponins (7,10,29,30). This quantitative relationship, however, in particular among patients in whom myocardial infarction has been ruled out by classic markers such as CK-MB, is poorly understood. In the present study, it was demonstrated that patients with episodes of transient myocardial ischemia detected by cVST had higher levels of cTnT than patients without. Therefore, it might be speculated that elevated cTnT levels represent a summation of prolonged myocardial ischemia leading to a poor clinical outcome. However, in the present study, there was only a moderate relationship between the number of transient ischemic episodes and the level of cTnT at inclusion (rho = 0.45). This might be explained by the fact that the signals detected by the two methods reflect different time phases of the ischemic process. Moreover, in the present study, the number of ischemic episodes during cVST was influenced by the antithrombotic therapies (18). Hence, these subjects need further investigation in future studies.
Patients ineligible for continuous vectorcardiography ST segment monitoring analysis
An obvious limitation of cVST is the loss of data due to technical problems. In the present study, five (2.2%) patients could not be adequately monitored for this reason. In addition, 13 (5.6%) patients were prospectively excluded from cVCT analysis due to the presence of bundle branch block or monitoring time <16 h, as cVST has not yet been validated in such patients. These shortcomings are in accordance with previous observations (16,17). Our results indicate that patients ineligible of cVST analysis are at higher risk for subsequent serious cardiac events, and that early risk assessment of these patients may be performed by the use of sensitive biochemical markers. However, other clinical characteristics associated with a poorer prognosis, e.g., bundle branch block, may be present in patients ineligible of cVST analysis as compared with the rest of the study population. Thus, the role of an elevated cTnT level as an independent predictor of adverse clinical outcome in patients ineligible of cVST analysis should be investigated in future and larger trials.
The prognostic value of ECG at inclusion
After multivariate analysis of potential baseline risk variables, ST segment depression was found to be the strongest predictor of adverse outcome in the present study. The association between ECG ST segment depression and a poor clinical outcome in patients with UCAD is consistent with previous findings (7,9,13,14,17,23,29,30). However, we found that ST segment depression at inclusion did not add additional prognostic information once the prognostic value of cTnT and cVST, respectively, was considered. Conflicting results have been reported in cases of UCAD with regard to the independent prognostic information of the cTnT level as compared with ST segment depression at the admission ECG (9,23,29). In contrast, ST segment changes on ambulatory ECG monitoring seem to be more prognostically powerful than ECG ST segment depression at admission (13,14).
The present investigation was performed as a substudy in a multicenter trial with several exclusion criteria, thus, our findings may be less representative in a more general UCAD population. The prognostic value of cardiac troponin T in particular might be favored by the exclusion of patients with renal failure (9,20,31). The interpretation of ECG changes in the present study was carried out by the local investigators, therefore, a central evaluation with strict prospectively defined ECG criteria could have altered the prognostic value of the ECG findings.
Employment of cTnT and cVST are valuable for the risk assessment of patients with UCAD. This study demonstrates that the risk assessment of patients suspected of having UCAD can be substantially improved by adding 24 h of cVST for an early determination of cTnT. Cost benefit and safety in patients triaged to aggressive antithrombotic or early referral to a low-cost fascility (or even discharge), according to the information derived from the combination of biochemical markers and cVST, needs delineation in future prospective trials.
We appreciate the skillful assistance from the staff of each participating center. We also thank Lars Frison, PhD., for his assistance with the statistical analyses.
☆ This study was supported by a grant from the Danish Heart Foundation, Copenhagen, Denmark (96-2-3-36-22426), and Astra Hässle AB, Gothenburg, Sweden.
- acute myocardial infarction
- coronary artery bypass grafting
- cardiac troponin T
- continuous vectorcardiography ST segment monitoring
- non–Q wave myocardial infarction
- ST vector magnitude
- unstable angina pectoris
- unstable coronary artery disease
- Received October 30, 1997.
- Revision received January 6, 1999.
- Accepted January 21, 1999.
- American College of Cardiology
- Klein W,
- Buchwald A,
- Hillis S.E,
- et al.
- Betriu A,
- Heras M,
- Cohen M,
- Fuster V
- Stubbs P,
- Collinson P,
- Moseley D,
- Greenwood T,
- Noble M
- Antman E.M,
- Sacks D.B,
- Rifai N,
- McCabe C.H,
- Cannon C.P,
- Braunwald E
- Morrow D.A,
- Rifal N,
- Antman E.M,
- et al.
- Langer A,
- Freeman M.R,
- Armstrong P.W
- Patel D.J,
- Holdright D.R,
- Knight C.J,
- et al.
- Dellborg M,
- Malmberg K,
- Ryden L,
- Svensson A.-M,
- Swedberg K
- Andersen K,
- Eriksson P,
- Dellborg M
- Thrombin Inhibition in Myocardial Ischemia (TRIM) study group
- Müller-Bardoff M,
- Hallermayer K,
- Schröder A,
- et al.
- Frank E
- Ravkilde J,
- Nissen H,
- Hørder M,
- Thygesen K
- FRISC study group,
- Lindahl B,
- Venge P,
- Wallentin L
- Pitt B
- Rodak D.J,
- Gersh B.J
- Amsterdam E.A,
- Lewis W.R
- FRISC study group,
- Lindahl B,
- Andren B,
- Ohlsson J,
- Venge P,
- Wallentin L
- FRISC Study Group,
- Lindahl B,
- Venge P,
- Wallentin L
- Bhayana V,
- Gougoulias T,
- Cohoe S,
- Henderson A.R