Transesophageal echocardiography in patients with cryptogenic cerebral ischemia

Cerebral ischemia is among the most common causes of hospitalization, morbidity and mortality in western civilizations. Stroke databases suggest that despite intensive evaluation, approximately 40% of all patients suffering ischemic strokes have no clearly identifiable cause and 15–20% of the ischemic events occur in younger patients (<55 years). In the literature, however, there is no consensus on the percentage of ischemic strokes caused by cardioembolic events [1].

Interatrial communications (ASD, PFO with and without atrial septal aneurysms [ASA]) can lead to paradoxical stroke and increase the risk for recurrent thrombembolic cerebral events. [2‐4].

In an autopsy study, the overall prevalence of a PFO was 27.3%. [5] A meta-analysis has shown that a PFO is more common in patients younger than < 55 years with suspected cryptogenic stroke than in healthy controls [6]. Hence, a PFO is considered as a risk factor for ischemic cerebral events in younger patients. In older patients (i.e. > 55 years), the association between strokes and prevalence of a PFO is much weaker [7].

In clinical practice, it is not easy to verify paradoxical embolism. In current guidelines, there is no consensus on the clinical relevance of a PFO in cryptogenic stroke and on the indication for the closure of a PFO after a stroke [8].

The aim of this study was to determine the prevalence of echocardiographic findings by TEE in a large number of patients with cryptogenic cerebral ischemia with a focus on pre-defined subgroups (age, sex, stroke vs. TIA).

In this single center retrospective study, 702 consecutive patients with acute cerebral ischemia were examined by TEE (1996–2001). All patients underwent neurological clinical examination, electroencephalogram, cCT or cMRT, and laboratory tests. All patients had a 12-lead-ECG, Holter-ECG, extracranial color-coded sonography, transthoracic echocardiography with no explanations for cerebral ischemia.

Exclusion criteria were: hemodynamically relevant stenoses of the intra- and extracranial arteries, cerebral hemorrhage, atrial fibrillation, migraine, epilepsy, intracranial tumors, acute myocardial infarction in the previous four weeks, previously diagnosed chronic heart failure and left ventricular aneurysms and less than 18 years of age.

Ischemic stroke was defined as a cerebrovascular event with symptoms lasting longer than 24 hours. Diagnosis of ischemic stroke or TIA was confirmed by a study neurologist based on clinical syndrome and results of diagnostic tests, including CT/MRI scans.

Transesophageal echocardiography was performed with a multiplane probe (5–7 MHz) on a Vivid 5 (TEE probe MPTE 5MHz 6A, GE Vingmed, Horton, Norway) and a HP Sonos 5500 (TEE probe HP 21369A; Hewlett-Packard, Paolo Alto, California, USA).

Before transesophageal echocardiography, intraoral xylocaine spray and intravenous midazolam were administered according to the requirements of the patient. The heart rhythm was monitored by ECG during the examination.

Each TEE examination included the standard views and measurement of blood velocities in the left atrial appendage, intravenous administration of contrast agent and Valsalva manoeuvre to exclude interatrial communications (according to [9]).

The images were stored digitally and analyzed off-line by EchoPac PC Dimension (GE Vingmed, Horton Norway) and additionally on video. Echovist 300 was used as contrast medium (Galactose; Schering, Berlin, Germany). Echovist is approved for the diagnosis of intra-atrial communications.

Written consent was obtained from each patient for the TEE, and the ethics committee of the Charité University Hospital approved the protocol.

702 consecutive patients were included. 380 (54.1%) were male, 322 (45.9%) female. 383 (54.6%) of the patients had an IS and 319 (45.4%) had a TIA. 411 patients were younger than 55 years (58.5%) and 291 patients were older than 55 years (41.5%).

The stroke subgroup was significantly older and had more frequently pathological TEE findings than the patients suffering from a TIA. (Table 1).

The prevalence of abnormal echocardiographic findings is listed in (Table 2). Neither the presence of a PFO (p = 0.053), nor of an ASD (p = 0.65) or an atrial septal aneurysm (p = 0.21) was significantly different in the IS and TIA groups. However, findings attributable to atherosclerosis (valvular abnormalities including mitral valve calcification, aortic calcification, aortic valve sclerosis, aortic plaque) were significantly more frequent in the ischemic stroke group (Table 3).

In the older patients (> 55 years, n = 291, according to [2, 10]) pathological echocardiographic findings were more frequent (p = 0.0023) than in the younger patients. The prevalence of a PFO was higher in the younger patients (< 55 years: n = 78; 26.8%, > 55 years: n = 75; 18.0%, Pearson's Chi square p = 0.005). The frequency of a ASD was lower in the older patients (< 55 years: n = 28; 9.6%, > 55 years: n = 20; 4.9%, Pearson's Chi square p = 0.014); see Figure 1, for examples see Figure 2, Figure 3 and Figure 4 and Additional file 1, Additional file 2 and Additional file 3

The most frequent combination of two findings was PFO+ASA (2.63% male with IS, 2.71% male with TIA, 2.63% female with TIA, 2.71% female with IS).

This study is to our knowledge the largest TEE study in patients with cryptogenic cerebral ischemia allowing the analysis of the full spectrum of findings and especially the prevalence of a PFO.

In half of all patients, a possible cardiogenic cause was identified by TEE. The most frequent finding was a PFO. The frequency of positive TEE findings that might explain cerebral ischemia is comparable to previous smaller studies [11, 12].

Cardiogenic causes of ischemic stroke can be divided into major (annual incidence of embolic events > 1%) and minor risk factors (annual incidence < 1%) [13]. The major risk factors are atrial fibrillation (1–12%/year), intracardiac thrombi (0–35%), atrial myxoma (30–40%), mitral valve stenosis (8–14% in sinus rhythm, 31–65% with concomitant atrial fibrillation), recent myocardial infarction (1–2%), anticoagulated mechanical heart valves (1.5–3%), infective endocarditis (12–40%), dilated cardiomyopathy (4%) and aortic arch atheromatous plaques (4–16% especially if ≥ 4 mm in diameter, [14]). The minor risk factors include mitral valve prolapse (<0.02%) and left ventricular aneurysm (<1%). Rare causes of cardioembolism include Chiari network, Lambl's excrescences [15], and valvular abnormalities [16, 17]. Spontaneous echo contrast (SEC) is seen in areas of blood stasis with a slowly moving, cloud-like swirling pattern of "smoke" or increased echogenicity recorded in TEE. However, there is little data on the precise embolic risk of mitral valve annular calcification, spontaneous echo contrast, atrial septal aneurysm, and calcific aortic stenosis [13].

A substantial subset of the cardiogenic factors diagnosed by TEE could have been found with TTE (in combination with trans-cranial Doppler) as well. However, there are no prospective studies showing equivalent diagnostic accuracy of TTE compared to TEE in this setting. Despite the semi-invasivity of TEE, it is still the gold standard for the detection of inter-atrial communications.

In conclusion, in this large study in patients with cryptogenic stroke, a TEE reveals in about half of the patients cardiogenic factors that might explain the stroke. Especially, the prevalence of a PFO is higher in the younger patients (< 55 years of age). Therefore, a TEE seems indicated in all patients with cryptogenic stroke – irrespective of age – because of specific therapeutic consequences.

This study was observational. The quality of observational studies seems to be comparable to randomized controlled trials [29, 30]. There was no follow-up of the patients concerning recurrence of cerebral ischemia or mortality.

We did not analyze the prevalence of abnormal TEE findings in single-vessel and lacunar strokes in comparison to multi-vessel strokes. The data of our study is historical (1996–2001). Currently, the indication for a TEE in cryptogenic stroke is more selective.

FK and FM equally contributed to the study. FK, FM, WVH, ACB have designed the study, and have performed the examinations. FK, ACB, FM analyzed the data and have written the manuscript. MY, VR, TW, HD performed the TEE examination, collected and interpreted the acquired data. GB has supervised the study and contributed by revising the manuscript critically.