Number 20, 2003
Hibernation preconditioning

Warmup angina: Ischemic preconditioning in a patient with no detectable coronary collaterals

Back to the Summary

Pier Lambiase
Department of Cardiology, Rayne Institute, St Thomas’ Hospital, Kings College London,
London, UK
Correspondence: Dr Pier Lambiase, Department of Cardiology, Rayne Institute, St Thomas’ Hospital, KCL, London SE1 7EH, UK.
Tel: +44 207 9228191, fax: +44 207 9605659, e-mail: pierlambiase@yahoo.co.uk

Introduction
In 1772 William Heberden read a letter to the Royal College of Physicians in London that he had received from a patient. The patient described the ability to exercise to angina and then to continue exertion with few or no symptoms, giving the first detailed clinical description not only of angina pectoris but also clinical surrogates of preconditioning: walk-through and warmup angina. Interest in these phenomena has waned with the advent of long-lasting prophylactic antianginal therapy and investigative exercise protocols that do not allow patients to proceed at their own pace. In the past it was thought that the warmup phenomenon was due to vasodilatation of coronary collaterals. However, documentary proof was lacking, whilst more recent attempts to demonstrate collateral recruitment failed to show the expected associated increase in coronary vein flow or radiographic opacification of collateral vessels [1, 2]. However, these techniques are insensitive and may have failed to detect significant collateral flow [1, 3]. This case report illustrates the warmup phenomenon in a patient with single-vessel coronary disease in whom collateral perfusion was accurately quantified by the pressure wire technique [4].

Case report
A 53-year-old man began to develop exertional chest tightness during a golfing holiday abroad. He noticed that he had angina on walking from the first tee along the fairway. His symptoms subsequently waned as he continued his game. On returning to the UK, he presented to his local accident and emergency department following an episode of pain whilst walking to work. Risk factors for coronary disease included diet-controlled diabetes, a positive family history, and smoking 20 cigarettes/day. Clinical examination was unremarkable. The resting ECG was normal. Cardiac enzymes including CKMB and LDH were normal. The patient was commenced on aspirin and discharged to undergo an outpatient exercise ECG.
He completed 10 minutes of a Bruce protocol (achieving 95% of age-specific target heart rate) with 0.22 mV of ST-segment depression at peak exercise. In order to assess his angina symptoms in more detail, he underwent two further exercise tests 15 minutes and 90 minutes after the first.

Figure 1. Serial ECG recordings taken at identical time points during each of the three exercise tests. ST-segment depression occurs in tests 1 and 3 but is less marked at the identical time point in test 2 despite a similar rate-pressure product.

As Figure 1 shows, the degree of ischemia at 9 minutes was significantly less at the same point during the second exercise test. This effect was lost at 90 minutes, indicating that the reduction in ischemia could not be explained by a training effect. Coronary angiography (Figure 2) revealed a 90% stenosis in the LAD beyond the first diagonal vessel with no evidence of collaterals on right coronary artery injection.

Figure 2. Measurement of collateral flow index (CFI). (A) A critical mid-LAD lesion is visible beyond the first diagonal demonstrated in the anteroposterior (AP)-cranial projection. (B) A pressure wire has been passed beyond this stenosis and the angioplasty balloon inflated. CFI is calculated from the coronary occlusion pressure (Poccl), mean aortic pressure (Ao), and right atrial pressure (RA) measured at the positions shown in B.

During subsequent percutaneous coronary intervention and stenting of the LAD, a pressure wire was passed into the distal LAD vessel. The coronary occlusion (Poccl), mean aortic (Ao), and right atrial pressures (RA) were recorded simultaneously (Figure 3).
In recent years, clinical assessment of the coronary circulation has focused on the physiological assessment of the severity of coronary stenoses using Doppler wires to measure epicardial blood flow. Development of the pressure wire to measure a trans-stenotic pressure gradient in the presence of maximal hyperemia has allowed the characterization of intermediate (50% to 70%) stenoses to guide the use of percutaneous coronary intervention in patients where the evidence of myocardial ischemia is equivocal. This technology has been extended to assess the degree of collateral perfusion in a territory distal to balloon occlusion.
The collateral flow index (CFI) evaluates collateral blood flow as a proportion of normal myocardial blood flow by expressing the transmyocardial pressure gradient gener-ated by collaterals as a fraction of the gradient in the absence of an epicardial obstruction [3].

CFI = (Poccl - RA)
------------------------
   (Ao - RA)

 CFI = (12 - 12)/(102 - 12) = 0

Figure 3. (A) Records of aortic and intracoronary pressures during serial coronary artery occlusion. Flow across the myocardial bed is largely determined by the difference between the venous and arterial pressures. Right atrial pressure was equal to the coronary occlusion pressure (12 mm Hg) after 3 minutes of balloon occlusion of the LAD, resulting in a calculated collateral flow index (CFI) of 0. Identical CFI measurements were obtained on a second 3-minute LAD occlusion performed 5 minutes later. (B) The degree of ST-segment elevation on inflation 1 at 3 minutes is significantly greater than that on inflation 2 at the same time point of 180 seconds despite no measurable recruitable collateral flow on the second occlusion.

In this case, CFI was calculated as 0 (Figure 3A), indicating that there was no measurable collateral perfusion to the anterior wall despite maximal hyperemia induced by two serial balloon occlusions of the LAD. There was also a significant reduction in ST-segment elevation seen on the second occlusion of the LAD (Figure 3B) despite no measurable collateral flow. Therefore, the warmup effect seen on exercise and the reduction in ST-segment elevation during serial balloon occlusion cannot be explained by collateral recruitment in this patient.

Discussion
Ischemic preconditioning classically refers to the reduction of myocardial infarct size following preceding ischemic stress [5]. The warmup phenomenon refers to the improved performance exhibited by more than half of patients with ischemic heart disease following a first exercise test [6–9]. However, the mechanisms underlying this phenomenon remain only partially understood and somewhat controversial.
Since Heberden’s original description of the warmup angina phenomenon in 1772, the vasodilatation of coronary collaterals has been proposed as a possible mechanism [10, 11]. However, recent attempts to demonstrate collateral recruitment failed to show the expected associated increase in coronary vein flow or radiographic opacification of collateral vessels [1]. However, these techniques are insensitive [3] and may have failed to detect significant collateral flow.
There is supporting evidence that the warmup phenomenon is a manifestation of ischemic preconditioning, including the time course of protection (60 to 90 minutes’ duration), which is commensurate with classic preconditioning. Okazaki et al [9] demonstrated that in patients with a single LAD coronary artery lesion, great cardiac vein flow is similar during the first and second exercise stress tests, suggesting that the warmup phenomenon is not accompanied by an increase in total myocardial blood flow. However, this methodology is subject to variation and cannot assess changes in regional myocardial blood flow, ie, the local opening of collateral vessels. Myocardial oxygen consumption was reduced during the second exercise test, suggesting increased metabolic efficiency — a typical feature of preconditioning. Further evidence to support the contention that the warmup phenomenon is a manifestation of ischemic preconditioning includes lower lactate production and rate-pressure product — a measure of myocardial oxygen consumption on the second versus the first exercise test [12]. However, there are conflicting mechanistic data. Inhibition of adenosine receptors prior to exercise fails to abolish the warmup phenomenon [13, 14]. Investigation of the role of KATP channels in mediating this form of protection has produced conflicting results [15, 16]. Therefore, it remains unclear whether the adaptation observed during repeated exercise is a representation of preconditioning or whether other mechanisms are involved; particularly collateral recruitment, which has not been fully evaluated. The advent of the pressure wire to precisely measure collateral flow during balloon occlusion has enabled one to determine whether the warmup phenomenon can occur independently of collateral flow.
Evidence from our own laboratory shows that the warmup phenomenon exhibited during early preconditioning occurs independently of collateral recruitment [17]. In this study, nine patients (25%) with no demonstrable collaterals despite 6 minutes of coronary artery occlusion had reduced time to 1-mm ST-segment depression on a second exercise test 15 minutes after the first, as well as an increased rate-pressure product with this degree of ischemia. Collateral recruitment is determined by the blood pressure gradient between adjacent myocardial vascular beds [18]. The absence of collateral recruitment after 6 minutes of coronary artery occlusion almost certainly predicts the absence of collateral recruitment during exercise when the coronary artery remains patent. These patients also exhibited at least a 50% reduction in ST-segment elevation on a second balloon occlusion compared with the first despite no documented recruitable collateral flow. The results from this study, as illustrated by the present case report, indicate that in warmup angina initiated either by exercise, when collaterals were not assessed, or by coronary artery occlusion, when collaterals were assessed, protection is independent of collateral recruitment. However, in patients with well-developed collaterals, collateral recruitment has been documented to contribute to the attenuation of ischemia on subsequent balloon inflations [19, 20].
The protective capacity of collaterals against myocardial infarction is well established. The advent of percutaneous transluminal coronary angioplasty has enabled visualization of recruitable collaterals during infarction and produced evidence that collaterals reduce several ischemic parameters during coronary occlusion, ie, the degree of ST-segment elevation [21, 22], lactate release [21], angina pain [22], and wall motion abnormalities [23–25]. This translates into a mortality reduction in patients receiving thrombolysis [26]. This benefit occurs principally through a reduction in cardiogenic shock and malignant arrhythmia.
However, the capacity of collaterals to reduce the degree of exercise-induced ischemia is less clear. There is good evidence that collaterals can protect against resting ischemia and maintain normal resting function even in the context of complete coronary occlusion. Vanoverschelde et al [27] performed positron emission tomography in 26 angina patients with chronic occlusion of a major coronary artery but without previous infarction. In nine patients the collateral supply was able to support normal wall motion at rest as well as similar blood flow and oxidative metabolism in comparison to normal segments. However, although the collateral flow reserve was up to three times the normal basal flow in three patients, this only represents 50% of the coronary flow reserve in normal myocardium and these patients remained symptomatic at peak stress.
Previous studies describing the protective capacity of collaterals in stable angina using exercise ECG and perfusion scanning are limited. Information on both the occurrence and function of collaterals was often biased by: (1) great heterogeneity of the populations studied; (2) multivessel disease that might have jeopardized the collateral circulation due to atheromatous lesions in supplying donor arteries; and (3) only taking spontaneously visible collaterals into account, ignoring the 60- to 200-mm collaterals which can make a significant contribution to overall collateral perfusion. The majority of evidence from these earlier studies suggested that collaterals did not significantly improve the manifestations of ischemia evaluated by the left ventricular ischemic dysfunction [28], severity [29], and duration [30] of angina, nor degree of ST-segment shift during exercise [31]. However, some studies found that patients with more severe coronary disease and a well-developed collateral supply could exercise to the same level as patients without a visible collateral supply and less severe disease [32]. The diversity of these results is explained by the deficiencies delineated above and confounded by both selection bias and the retrospective nature of much of the early research.
This case illustrates that the warmup angina phenomenon in a patient without demonstrable collateral flow indicates that this phenomenon is probably a manifestation of ischemic preconditioning.

Back to the Summary

REFERENCES

Back to the Summary