Women with angina but no obstructive coronary artery disease (CAD) at coronary angiography have long been considered a low-risk population.1–3 Instead, data from the Women‘s ischaemia syndrome evaluation (WISE) study4 demonstrated that persistent chest pain (PChP) – also in the absence of CAD – is not a benign condition. These patients experienced major cardiovascular (CV) events, such as myocardial infarction and stroke at approximately double the rate found in women with neither PChP/CAD. However, the misinformation among physicians about these issues may keep women from receiving appropriate cardiac care.2,5 Coronary microcirculation abnormalities, defined as microvascular endothelium dysfunction, play a pathophysiological role in patients with angina and no significant coronary stenosis.6,7 In clinical practice, the presence of myocardial perfusion abnormalities can be highlighted as inducible, using magnetic resonance imaging (MRI).8,9 The above-cited study was conducted on populations with typical ST segment depression during an exercise stress test, in the absence of significant CAD (so-called ‘cardiac syndrome X’).10 In the present study, we included a population of post-menopausal women with PChP, referred for coronary angiography to further evaluate the basis for suspected ischaemia, in order to investigate the relationship between myocardial perfusion at rest and during a stress test, using MRI. Our goals were to document if transient myocardial perfusion are induced by dipyridamole infusion and if perfusion defects are also present at rest. The presence of delayed enhancement (DE) as an indicator of myocardial necrosis was also analysed during MRI study.
The study population consisted of 45 consecutive women (mean age 57.6±8.7 years) with PChP, according to the WISE study.11 These women reported at least one year of chest pain. One year was chosen because a persistent chest pain for this period is necessary for the diagnosis of PChP. All patients underwent clinical examination, electrocardiogram and echocardiogram at the first clinic. Within 15 days of the first clinical approach, all patients underwent coronary angiography and a perfusion CV MRI.
The studies from all these modalities were reviewed and analysed using the vendor-neutral solution Synapse™ Cardiovascular from Fujifilm. Exclusion criteria included emergency referral, known cardiomyopathy, New York Heart Association (NYHA) class III/IV congestive heart failure, acute ischaemic syndrome (defined as acute myocardial infarction or unstable angina) within one month prior to study entry, coronary revascularisation within six months prior to study entry and any contraindication to provocative myocardial stress testing. Particular care was taken to exclude patients with diseases known to give non-ischaemic chest pain, such as recurrent pericarditis and pulmonary hypertension. On enrollment, evaluation included risk factors for CAD, drug use, a comprehensive symptom history, a physical examination, an activity status assessment, the collection of all demographic information and a sampling of blood. All women provided signed informed consent prior to entry into the study. The protocols of the present study were in accordance with the Declaration of Helsinki and were approved by the Ethic Committee of our Institution (Policlinico hospital, Modena, Italy).
Definition of Chest Pain
The only criterion for inclusion in the present study was that patients were suffering from PChP. All women completed a symptom questionnaire to assess the presence of ‘typical’ angina, defined as pain that is substernal, precipitated by exertion or emotional stress and relieved within 10 minutes by rest or nitroglycerin. Consistent with the WISE findings,11 PChP was defined as symptoms at one-year follow-up as assessed by the same question. Women were required to also specify the mean number of pain episodes per month.
Two investigators examined the angiograms. In our patients, coronary angiography was performed by injection of contrast medium via 6Fr catheters. At least two orthogonal views of every major coronary vessel and its side branches were acquired. All coronary segments were considered and quantification of lumen narrowing was performed using the Philips Automated Coronary Analysis-package (Philips Medical System, Best, The Netherlands) and the Synapse Cardiovascular Quantitative Coronary Analysis Software Package (Fujifilm's CV information system [CVIS]). The severity of CAD was determined by angiographic assessment of luminal diameter narrowing. ‘Obstructive’ or ‘significant’ CAD was defined as ≥50% reduction of the luminal diameter in ≥1 major epicardial coronary artery.
Magnetic Resonance Imaging Evaluation – Rest Magnetic Resonance Imaging
MRI images were acquired using a 1.5 Tesla superconductive magnet (Achieva, Philips Medical System, Best, the Netherlands). A five-element phased-array sinergy-cardiac coil was used and cardiac and respiratory triggers were employed. Two-chamber, four-chamber and multiple short axis views were acquired using the balanced fast field echo technique. Acquisition parameters were: multislice acquisition with 8mm slice thickness; 0.8mm interslice gap; repetition time=4.5 milliseconds; echo time=2.2 milliseconds; flip angle=55 degrees; acquisition matrix=192x256 with an acquired voxel size of 2.03x4.14x8.0mm and reconstructed voxel size=1.52x1.53x8.0mm. Short axis breath hold acquisition was completed with parallel imaging using a sense factor of 1.6. Two slices were acquired in each breath hold. Geometry settings used for baseline scans were stored to be repeated during perfusion and DE imaging to obtain comparable slices.
Magnetic Resonance Imaging Evaluation – Stress Perfusion Magnetic Resonance Imaging
Dipyridamole was administered intravenously under continuous electrocardiogram (ECG), blood pressure and visual monitoring with Synapse Cardiovascular. The stress imaging was performed using a double bolus of dipyridamole (0.56mg/kg over six minutes). After dipyridamole infusion, perfusion imaging was evaluated in three different short axis planes by means of a fast gradient echo sequence during the rapid injection of 0.05mmol of gadolinium DTPA/kg body bolus (Dotarem, Guerbet SA, Cedex, France), at a rate of 5ml/s and followed by 20ml of rapidly injected saline. Sixty dynamic images (one image per heart beat) were acquired for each slice of 8mm thickness during the first and second passes of the contrast agent. The dipyridamole infusion was discontinued prematurely on patient request or when progressive or severe angina, dyspnoea, decrease in systolic pressure >40mmHg, severe arrhythmias, or other adverse effects occurred. Aminophylline was administered as required. Slices were placed to acquire short axis images at 1/4, 1/2, and 3/4 of the left ventricular long axis. The pulse sequences used a non-slice selective 90-degree preparation with an echo-planar saturation (read-out delay time of 120 milliseconds). Twenty minutes after stress perfusion a second perfusion study was performed to allow for rest perfusion assessment. The second acquisition allows for differentiation between ‘fixed’ and ‘reversible’ perfusion defect and the perfusion abnormalities were reported, as previously recommended.12
DE images were acquired approximately 15 minutes after the second gadolinium-DTPA bolus injection and were used to differentiate between ischaemia and scar tissue since the latter condition is known to be positive during DE imaging.13,14 DE presence and distribution was evaluated using a 3D inversion recovery gradient turbo field echo sequence and prepulse-delay was optimised for maximum myocardial signal suppression.15 Inversion time (varying from 220 to 270 milliseconds) was optimised for each patient using an inversion recovery single slice sequence with variable inversion times. The geometry orientation of the DE sequence was the same as functional short axis acquisition in order to obtain comparable images. Imaging parameters were: thickness=5mm; echo time=1.32 milliseconds; repetition time=4.4 milliseconds; flip angle=15 degrees; acquired voxel=1.37x2.14x10.0mm; reconstructed voxel=1.37x1.37x5.0mm; turbo factor=37; and acquisition matrix=256x56.
Magnetic Resonance Imaging Analysis
Two experienced investigators evaluated all CMR studies in consensus. If consensus could not be achieved, a third opinion was included. Image analysis was performed visually with Synapse Cardiovascular for reducing the rate of false positive results due to rim artifacts, as previously reported.12 We compared stress to rest perfusion to reduce the potential rate of artifacts. If a deficit did not follow the subendocardial border, if ghosting artifacts could be seen, or if it ‘blinked’ bright and dark it was not regarded as an evident hypoperfusion, but a potential artifact. Such cases were not included in the study.
Continuous parameters are expressed as the mean value ±SD; dichotomous parameters as frequency (percentage). Comparison between groups was performed using analysis of variance or chi-square test, as indicated. Statistical significance was inferred at p<0.05.
All the 45 study patients had previous stress testing (mainly exercise-ECG [30 of 39 patients; 76.9%]), all of these with a ‘doubt’ , or at least an inconclusive result. Of this ‘particular’ population, 34 women of the 45 enrolled in the study (77.3%) did not have an obstructive CAD at coronary angiography; and the remaining 11 (22.7%) were found to have ‘significant’ coronary stenosis. Compared to the other group, the 11 women with PChP and obstructive CAD were much older and had more conventional CV risk factors, and a higher risk score according to the Framingham equation (see Table 1). MRI examination was performed in all patients without relevant complications or adverse events. Image quality was sufficient for further analysis in all patients, with the investigator’s consensus being reached in 43/45 cases and inclusion of a third opinion in only two cases. The MRI study allowed us to demonstrate that all patients in our population had normal left ventricular performance (i.e. left ventricular ejection fraction 65% or greater). In addition, no patient had MRI signs of infiltrative cardiomyopathy and/or MRI patterns consistent with acute myocarditis and/or unusual ventricular morphology (spade ventricle, etc.), to suggest a prior episode of ‘tako-tsubo’ cardiomyopathy. The results of the dipyridamol-perfusion MRI study are shown in Table 2. In the group of women with PChP and non-obstructive CAD, 16 of 34 women (47.0%) showed a well evident left ventricular perfusion defects at baselin; number and localisation of defects are shown in Table 3. These defects were ‘permanent’ or ‘fixed’, i.e. were present at rest, and were neither induced nor modified by the administration of dipyridamole (see Figure 1). Among the non-obstructive CAD patients, all of the women with either induced angina or ECG changes had at least one ‘fixed’ perfusion defect at MRI. In none of the women with CAD did we find these anomalies.
The most important feature of our work is that among post-menopausal women suffering from PChC, we can distinguish two different groups of patients. The first group is represented by patients with angiographic significant CAD, characterised by (in respect to the other group): an older age; the presence of inducible ischaemia; and the presence of myocardial areas of delayed enhancement. In addition to this group, we can define patients with ‘classic’ CAD.
We found a subgroup of women who had very unique characteristics that were abnormal but apparently different from those with classic CAD: that of women with angiographically normal epicardial coronary vessels; presence (often associated) of hypertension and diabetes; and ‘fixed’ myocardial perfusion defect at MRI. In other words, myocardial perfusion defects are present in these patients at rest and a reduced microvascular reserve was not proved by dipyridamole infusion. For these reasons, we hypothesise a structural, permanent damage of the coronary microcirculation. The latter group of patients is different from a group with classic CAD, because it presents a younger age and a pattern of CV risk characterised by the invariable presence of hypertension, diabetes or both. Some other risk factors, such as family history of CAD, hypercholesterolaemia, smoking habits etc., are far less represented. The patients that we identified are also different from those with classic syndrome X. The latter requires the presence of pain associated with changes (ST-segment depression) on an exercise ECG. In this condition, defects in the perfusion MRI and other imaging techniques are common, but in this case the microcirculatory disturbances are inducible by various stressors, such as dobutamine, adenosine and dipyridamole.5–10,16 On the contrary, the presence of perfusion defects at rest are not typical for cardiac syndrome X.
The idea that some forms of ischaemic heart disease may be caused by abnormalities of the microcirculatory vessels is not new. It was proposed 22 years ago as a cause of angina pectoris.17 Although the importance of microcirculation for the regulation of coronary blood flow has become clearer, pathophysiological explanations of the disease process are still poorly understood. Many authors indicate the cause of the microcirculatory dysfunction in a functional endothelium-mediated phenomenon.18,19
Many authors have studied the importance of perfusion defects in patients of diverse settings using different methods. Impairment of coronary microcirculation during stress MRI has been documented in hypertensive patients20 and in those with non-obstructive CAD.21 Other studies have been performed using nuclear medicine or ultrasound techniques. Scognamiglio and colleagues22 and De Lorenzo and colleagues23 have shown – in diabetic patients – that perfusion defects can be ‘fixed’ or, in alternative, ‘inducible’ by dipyridamole infusion.
The first of these authors, using echostress, places such defects in a view of ‘early’ ischaemic heart disease. The second, using gated-single photon emission computed tomography (SPECT), detects reversible perfusion defects in 15% and fixed perfusion defects in 5% of diabetic patients. The presence of perfusion defects, both fixed and reversible, indicated a poor prognosis.22,23 Our results do not contradict this, but complete this point of view, with the acquisition of a new element, namely that in post-menopausal women with PChP we can find perfusion structural defects, not inducible by dipyridamole.
Limitations of the Study
Coronary microvascular dysfunction was postulated as the most likely pathophysiological mechanism able to explain the presence of persistent perfusion defects in patients with normal angiographic coronary arteries. However, since no evidence of dipyridamole-induced myocardial ischaemia was present in women with normal coronary arteries, this would mean that the persistent myocardial perfusion defects were the result of a previous severe abnormality of the coronary microcirculation, but that no further coronary microvascular alteration was present at the time of investigation. This is the hypothesis that we consider more reasonable. However, an alternative explanation exists for the fixed perfusion defects found on cardiac MRI in these patients. Specifically, the defects might have been determined by myocardial cell death related to other mechanisms (i.e. metabolic alterations), with perfusion defects being secondary to myocardial cell loss. Even if this hypothesis is unlikely – because all our patients have a negative history for any extracardiac disease and the patients in the other group manifested metabolic abnormalities (i.e. diabetes in 18.1% of cases) but none of them showed fixed perfusion defects at MRI – it remains potentially possible.
In clinical practice it is truly important to separate normal patients from those with non-obstructive CAD, since treatment and prognosis may be different in these two groups of patients. In the population of post-menopausal women with PChP, where the stress tests are doubtful or inconclusive (after excluding cases of non-ischaemic chest pain) we can get patients, often with hypertension, diabetes, or both, with ‘fixed’ myocardial perfusion defects at MRI. These cases suggest a disease state typical for post-menopausal women in which the pathological involvement of the coronary microcirculation is presumable.