METHODS

Patients

Nine patients (58 (3) years, eight women) with documented microvascular angina (typical anginal pain, positive electrocardiography on exercise with ST-segment depression >1 mm, abnormal perfusion scans, normal coronary arteriography) were assessed at rest, and then during an infusion of adenosine through a peripheral vein (50 μg/kg/min, then 100 μg/kg/min, and then 140 μg/kg/min, each for six minutes). The infusion was stopped in all patients due to the onset of chest pain. All patients were in sinus rhythm. All drugs were stopped 24 hours before the study. The protocol was approved by the locally appointed ethics committee and the informed consent was obtained from all subjects.

Radionuclide angiography

At rest and peak stress, global left ventricular function was assessed by ^99mTc blood pool imaging, using an Elscint Gamma Camera. Best septal separation views (40° left anterior oblique) were obtained for 10 million counts/frame, in list mode. Images were formatted with 32 frames/cardiac cycle. The left ventricular region of interest was drawn semiautomatically but checked individually, with atrial filling indicated manually.

Echocardiography

Images were obtained during two minutes, at the end of each stage, using an apical four chamber view (Vingmed System 5, 2.5 MHz transducer). All echocardiographic data were obtained and processed by one investigator, blinded to the radionuclide data.

Global diastolic function was assessed by pulsed wave Doppler of the transmitral flow. The sample volume was placed at the tips of the mitral leaflets and peak early (E) and atrial (A) velocities were measured. E/A velocity ratio was calculated.

Left ventricular long axis function was assessed by tissue Doppler of the lateral mitral annulus. A 6 mm pulsed sample volume was placed within the lateral mitral annulus in systole. Every effort was made to align the pulsed wave cursor so that the Doppler angle of incidence was as close to 0° as possible to the direction of motion of the lateral mitral annulus. From the spectral traces we measured peak systolic velocity, and peak diastolic velocities during early filling (E_TDE) and atrial contraction (A_TDE). E_TDE/A_TDE ratio was calculated.

Statistical analysis

Statistical analysis was performed with SPSS software (version 9.0). Results are presented as mean (SD). The change of parameters from baseline to peak adenosine was compared by paired samples t test or by Fisher's exact test for proportions. A p<0.05 for a two tailed test was considered significant.

RESULTS

General resting echocardiographic characteristics of the study group are given in table 1.

Seven patients completed two stages and two patients all three stages of the adenosine stress protocol. Heart rate increased from a 81 (7) beats/min at baseline to 113 (11) beats/min at peak adenosine (p<0.01). Blood pressure did not change significantly during adenosine infusion: 156 (15)/87 (8) mm Hg at rest, and 150 (16)/82 (7) mm Hg at peak stress.

There were no significant changes in global or regional systolic function (table 2). Adenosine induced global diastolic dysfunction, demonstrated by ^99mTc blood pool imaging (first third filling) and by conventional Doppler of the transmitral flow (E/A ratio).

All patients had long axis diastolic dysfunction at peak adenosine, revealed by the development of a ratio of early-to-late diastolic velocity of the lateral mitral annulus of <1 (fig 1). Reversal of the E_TDE/A_TDE ratio (which decreased by 35%) was due mainly to an increase of the A wave velocity (by 47%), rather than a decrease of the E wave velocity (which fell by 11%).

DISCUSSION

In nine patients with microvascular angina adenosine provoked regional and global diastolic dysfunction. This may have been a consequence of subendocardial ischaemia, which is suggested by abnormal left ventricular long axis function assessed by tissue Doppler imaging.

Definition of microvascular angina

There is no accepted definition of this condition. We have employed two main criteria: (1) arteriographically entirely normal coronary arteries; and (2) evidence of reduced perfusion. Some authors accept a diagnosis of microvascular angina in patients with non-flow limiting stenosis (that is, less than 50%).⁹ However, since data from intravascular ultrasound studies suggest that the degree of stenosis might be underestimated angiographically,¹⁰ we have selected only those patients with entirely normal arteriograms. We used clinical and electrocardiographic markers of ischaemia, combined with a positive myocardial perfusion scan.

Adenosine in microvascular angina

Microvascular angina is associated with patchy perfusion abnormalities.¹¹ No single underlying basis has been defined. Small vessel coronary artery disease, abnormal coronary vascular resistance, and subendocardial ischaemia, have each been invoked as a possible mechanism, although none is universally accepted. Several reports suggest that adenosine may cause myocardial ischaemia in microvascular angina, probably as a result of an increase in the regional inhomogeneity of flow.⁸ Ischaemia provoked by adenosine might be localised to the subendocardial layers, causing only diastolic dysfunction with preserved systolic function. All investigated patients demonstrated subendocardial diastolic dysfunction. We did not have a control group, but it has been reported that in normal subjects adenosine infusion induces increases in early filling and E/A ratio, as assessed by Doppler of the transmitral flow.¹²

We used the protocol described because, in our experience, patients with microvascular angina are particularly sensitive to this form of stress, often experiencing their normal chest pain at a lower infusion rate. Moreover, there are now extensive data suggesting that maximal coronary vasodilatation is achieved at 140 μg/kg/min, heart rate−systolic blood pressure product being less important than the selective effect on the coronary vasculature.¹²

Diagnosis of subendocardial dysfunction by tissue Doppler

The subendocardial fibres are aligned longitudinally and connected to the mitral annulus. Long axis contraction and relaxation result in its displacement, which can be measured by the new method of tissue Doppler echocardiography, in terms of velocity. There are three distinct velocity waves during each cardiac cycle: a systolic wave, starting just after the QRS complex, an early diastolic wave (E), and a late diastolic wave (A), starting after the P wave. A ratio of early-to-late velocities of mitral annular motion of <1 has good sensitivity and specificity (>70%) for detecting diastolic dysfunction, even in patients with a pseudonormal E/A ratio.¹³

The method of recording velocities of the mitral annular motion has been reported before.¹⁴ We used the apical window because it provides the best alignment of the Doppler beam for both transmitral flow and mitral annular movement. The intraobserver and interobserver variability of both acquiring and measuring the tissue Doppler velocities of the lateral mitral annular motion were ≤16% and ≤20%, respectively.¹⁴

CONCLUSION

Long axis diastolic dysfunction revealed by tissue Doppler of lateral mitral annular motion was present at peak adenosine in nine investigated patients with microvascular angina.

We have described a small, but very homogeneous group of patients. These patients are difficult to identify, resulting in limited numbers in most studies. This pilot study shows that adenosine tissue Doppler stress echocardiography might be useful as a diagnostic test for microvascular angina, if confirmed in large series of patients and in comparison with control subjects; it also demonstrates a possible mechanism for the chest pain in patients with normal coronary arteries.