1. Introduction

Over the last 10 years it has become evident that for patients with coronary heart disease, episodes of ischemia, either silent or painful, carry important prognostic information. Particularly for patients suffering acute coronary syndromes (unstable angina and acute myocardial infarction), continuous ST-segment monitoring provides noninvasive, quantitative information to support both early diagnosis and ongoing assessment of response to therapy.

Multilead, realtime, continuous ST-monitoring has become technically feasible for clinical use [Dellborg et al., 1990; Krucoff et al., 1990]. It is now well appreciated that acute coronary syndromes are highly dynamic clinical and pathophysiological entities. Occasional static ECGs may undersample important information, either to correlate with patient symptoms or to raise alerts when ischemia is not accompanied by patient symptoms. Thus any device with real time capability and reasonably modern signal fidelity that provides continuous ECG monitoring will yield important information above that encountered in routine practice.

While ST-segment changes may be observed with postural change, hyperventilation, left ventricular hypertrophy, changes in potassium concentration and use of digoxin in the absence of coronary artery disease or myocardial ischemia. However, well characterized ST-segment shifts in patients highly likely to have coronary artery disease who have normal atrio-ventricular conduction patterns at rest have been shown to be both sensitive and specific for myocardial ischemia [Chierchia et al., 1983; Deanfield et al., 1984]. Continuous multilead ST-segment monitoring is in the third millenium a tool to measure the electrocardiologic activity of patients with highly suspicious clinical presentations or known coronary disease and its response to directed therapy.

2. Technical Aspects

Today we have several kinds of devices available for multilead continuous ST-segment monitoring. While much of our knowledge comes from older studies using Holter technology that is well established and widely available the retrospective nature of Holter technologies implicitly limits use in acute settings where real time information is critical to optimal patient management. There are currently two different techniques used for real-time, multilead ST-segment monitoring.

2.1. Continuous 12-Lead ECG

A number of manufacturers have developed continuous 12-lead ECG monitoring devices. Reports using continuous 12-lead monitoring have been published based on studies of patients undergoing angioplasty, in patients treated for ST-segment elevation infarction, and, in patients with chest pain syndromes, unstable angina and non-Q wave infarction. To minimize artifact limb leads are placed on the torso lead positions, with a notable effect on QRS axis but very little effect on ST deviation measures [Krucoff et al., 1994]. Depending on software and memory structure, systems may differ on how many ECGs can be stored and how frequently they are assessed. Trends of ST level vs. time for individual leads, lead groups and summated ST deviation are all accessible at the bedside and/or at a central review station.

2.2. Continuous Vectorcardiography

Using 8-leads placed according to Frank [Frank 1956] a continuous 3-lead ECG, X, Y and Z is recorded. After filtering and rejection of ectopic beats, incoming beats are averaged over 30 -120 second periods and each such average is stored. Comparison is made with the first 2 minutes of recording i.e. the reference. Changes are displayed as trends over time where summarized ST parameters such as ST vector magnitude (ST-VM) and differential QRS parameters such as QRS vector difference (QRS-VD) are displayed [Dellborg et al., 1990]. The ST-VM is the summarized deviation of the ST-segment for lead X, Y and Z, added quadratically i.e. thereby equalizing ST-elevation and ST-depression. Many reports are based on ST-VM data measured at J+20 milliseconds while others have used J+60 milliseconds. While QRS changes clearly occur during myocardial ischemia and infarction, the clinical usefulness is limited by the low specificity of QRS-changes, particularly with changes in body position.

3. Overview of Published Data

3.1. What is Gained by Multi-Lead Monitoring as Compared to Standard Arrhythmia Monitoring?

Ischemia, unlike arrhythmias, can be very focal in where it causes ST deviation. It has been clearly shown that the more leads you monitor, the more ischemic episodes you are likely to pick up [Klootwijk et al., 1997; Krucoff 1988]. The widespread use and acceptance by coronary care unit nurses and patients of the 8-10 leads used in standard 12-lead and continuous vectorcardiographic monitoring, respectively, is feasible in critically ill patients. With these tools high quality data can be obtained in multicenter studies [Andersen and Dellborg 1998; Langer et al., 1995].

3.2. Which Patients Should be Monitored?

ST Segment Elevation Infarction
Continuous ST-monitoring with 12-lead ECG and vectorcardiography have both been extensively studied as noninvasive markers of the quality and stability of reperfusion in ST elevation infarction [Dellborg et al., 1995; Klootwijk et al., 1996; Krucoff et al., 1993]. Both techniques are suitable for real time identification of patients with failed reperfusion after thrombolytic therapy, who might be candidates for rescue angioplasty or other additional therapy, as well as for triage in high or low risk categories. However, it is important to realize that seeing evidence of high risk, such as persistent ST elevation or ST re-elevation after transient ST recovery, does not mean that any or every available therapeutic option will lower that risk profile. So while continuous ST-segment monitoring is uniquely useful for ongoing risk stratification and assessment of response to therapy, its integration into the clinical care of patients with ST elevation infarction must be individualized by the decision making of the clinician.

Unstable Angina/nonQ-Wave Infarction
In non- ST-elevation and unstable angina ST-segment monitoring clearly can be used to identify patients at increased risk for recurrent infarction, urgent revascularization or death [Johnson et al., 1982; Langer et al., 1989; Patel et al., 1996]. This prognostic information is additive to that obtained from a careful clinical history, standard 12-lead ECG, as well as biochemical monitoring such as serial Troponin-T measurements [Andersen et al., 1997; Holmvang et al., 1999; Norgaard et al., 1999]. However, no clinical trials have used ST-segment monitoring as a real time basis for intervention. Therefore, ST-segment monitoring should be considered a powerful tool to identify high risk and be used to asign intensive care vs. regular ward beds facilities and/or to otherwise more appropriately distribute interventional resources when supply is limited.

Chest-Pain Units
Multi-lead ST-segment monitoring has also been used in chest pain units for early diagnosis of ischemia and infarction [Fesmire and Smith 1993; Gustafsson et al., 1996; Lundin et al., 1992]. It is not infrequent for patients who experience chest pain at home to be pain free by the time they arrive in the emergency department. Continuous ECG monitoring may document unequivocal episodes of ischemia in the absence of symptoms or may capture ECG changes at the time of recurrent symptoms, assisting in more timely and definitive diagnosis. Continuous monitoring of patients who are very unlikely to have ischemic heart disease will yield significantly more nonspecific ST deviation, while monitoring patients with multiple risk factors for coronary disease, or patients who have known coronary disease who present with chest pain will detect transient episodes of ST deviation more specific for ischemia.

Bundle-Branch Block/Pacemaker
In patients with suspected acute coronary artery disease and bundle branch block or pacemaker ST-segment monitoring has been assessed, using continuous 12-lead technique and using continuous vectorcardiography [Eriksson et al., 1997; Stark et al., 1991]. Small changes in heart rate and body position create such marked changes on the surface ECG that ST segment monitoring by and large is not yet clinically useful in these patients.

3.3. For How Long Should Monitoring be Performed?

In principle, the longer a patient is being monitored, the more information is obtained. In the ideal situation multilead ST-monitoring is begun in the prehospital setting, continued during transport and for the first 24 hours in the coronary care unit. If the patient is transferred to the cathlab monitoring is continued also during and after intervention. There are today vectorcardiographic systems that makes this kind of monitoring feasible, enabling the clinician to have an unbroken chain of monitoring. For patients with unstable angina/nonQ wave infarction 24 hours of monitoring is desirable; even after 24 hours without recurrent ischemia, a significant proportion of patients will experience recurrent ST-changes [Klootwijk et al., 1997].

However, early ambulation and shortage of coronary care beds mostly limits the feasible duration of monitoring to about 24 hours. For patients with a limited suspicion of acute coronary disease, a 12 hours period of monitoring in combination with serial determinations of cardio-specific cardiac markers such as troponin-T or I is sufficient to identify a low-risk group suitable for early ambulation, stress-testing or discharge [Andersen et al., 1996; Gustafsson et al., 1996; Holmvang et al., 1999; Jernberg et al., 1999].

3.4. Quantitative Information From Multilead ST-Monitoring

ST-Elevation Infarction
There are several reports where it has been shown that rapid restoration of adequate flow corresponds to rapid regression of ST-elevation. In most studies, the overall agreement between ST-monitoring and early (typically 90 minut after start of lytic) angiography is about 80% [Dellborg et al., 1995; Krucoff et al., 1993; Langer et al., 1995]. Considering the anatomic nature of angiography and the physiologic measurements obtained from ST-montoring, the discrepancy is not unexpected. The reasons are probably two-fold i.e. the presence of protective collaterals (ST-regression with an occluded vessel) and/or no-reflow phenomenon (open vessel but no ST-regression). Even after primary angioplasty regression of ST-elevation is not always observed; these patients do have a worse outcome as compared to patients with similar angiographic flow but rapid ST-elevation decline [Claeys et al., 1999]. Brief episodes of recurrent ST-change after successful thrombolytic treatment is not associated with an impaired outcome unless repeated and/or longer lasting episodes are observed [Langer et al., 1998].

Unstable Angina/nonQ-Wave Infarction
Recurrent ischemia at rest during the initial 24 hours of coronary care, as determined by continuous multilead ST-segment monitoring, is an important predictor of increased risk in this patient group [Andersen et al., 1996]. While a large number of episodes signal a definately increased risk of early (re-)infarction or death, the absence of recurrent ST-change clearly identifies a lowrisk group of patients. The combination of ST-monitoring and biochemical markers and ST-depression of admission resting ECG further helps in identifying low- and high-risk groups [Andersen et al., 1997; Holmvang et al., 1999; Norgaard et al., 1999; Patel et al., 1996]. In contrast, the more episodes and the larger the episodes of ST-change, the more at risk the patient appears to be [Abrahamsson et al., 1999; Andersen and Dellborg 1998].

4. Discussion

While ST-segment changes may be observed in a variety of conditions, it is reasonable to assume that the occurrence of repeated and brief episodes of reversible ST-change in patients with a high likelyhood of acute coronary artery disease represents bousts of myocardial ischemia. Since most such episodes are asymptomatic, continuous multilead ST-segment monitoring is of substantial value in caring for these patients. Continuous ST-segment monitoring by either continuous 12-lead ECG or continuous vectorcardiography is today available in reliable, user-friendly monitoring systems. ST-segment monitoring enables the clinician to continuously follow the dynamic changes that characterize unstable angina and acute myocardial infarction syndromes. After thrombolytic therapy it may help in identifying patients with failed reperfusion or recurrent occlusion in addition to providing important prognostic information. In unstable angina/non Q wave infarction ST-monitoring clearly identifies high risk patients but can also be used to select lowrisk patients for early ambulation and discharge, particularily in combination with repeated determinations of cardiospecific markers such as troponin-T. While formal trials are lacking, it is reasonable to assume that most of the benefit derived from early revascularization or novel potent drugs such as GP IIb/IIIa inhibitors, comes from patients with an increased risk. Although ST-monitoring possesses some unique characteristics it should be regarded as a complimentary tool to biochemical monitoring and careful history including resting 12-lead ECG. It provides important information for risk stratification in unstable coronary syndromes and helps in differentiating between extra-cardiac chest pain and acute coronary disease. Futher evolution in semiautomatic evaluation software, body position detectors and ongoing clinical trials will further develop this very important clinical tool.

References