Risk of Cerebral Embolism After Interventional Closure of Symptomatic Patent Foramen Ovale or Atrial Septal Defect: A Diffusion-Weighted MRI and Neuron-Specific Enolase-Based Study

Hande Oktay Tureli, MD1, Ismail Ungan, MD1, Derya Tureli, MD2, Bulent Demir, MD1, Osman Pirhan, MD1, Halil Ibrahim Bayrak, MD1, Ilker Murat Caglar, MD1, Osman Karakaya, MD1, Ercan Inci, MD3

Hande Oktay Tureli, MD1, Ismail Ungan, MD1, Derya Tureli, MD2, Bulent Demir, MD1, Osman Pirhan, MD1, Halil Ibrahim Bayrak, MD1, Ilker Murat Caglar, MD1, Osman Karakaya, MD1, Ercan Inci, MD3

Abstract: Objectives. The aim of this single-center prospective study is to investigate the silent and clinically apparent cerebral embolic events after transcatheter closure of atrial septal defect (ASD) and patent foramen ovale (PFO). Background. Although transcatheter closure of ASD and PFO is a widely accepted technique and has been proven to be safe and effective with different kinds of devices, there are few studies in the literature that report the peri-interventional cerebral embolism risk and neurological complications. In this study, we investigated the peri-interventional cerebral embolism incidence with diffusion-weighted magnetic resonance imaging (DW-MRI) and its relation to patients’ clinical neurologic examination findings and plasma neuron-specific enolase (NSE) levels. Methods. Sixteen patients with hemodynamically significant ASD and 14 symptomatic PFO patients underwent transcatheter closure procedures with new-generation PFO or ASD occluder devices. All cases were examined with DW-MRI before and after the transcatheter closure procedure. Patients were clinically examined for any signs of neurologic deficit at the time of MRI studies. Blood samples for NSE, a marker of brain tissue damage involved in an ischemic event, were taken before the procedure and at 12 and 24 hours after the procedure. Results. Successful transcatheter closure of PFO or ASD was achieved in all patients. In the DW-MRI exam following the procedure, a new microembolic lesion was found in only 1 of 30 patients (3.3%). None of the patients had positive clinical neurological exam findings. NSE levels after the procedure were found to be not correlated with presence of DW-MRI lesion and intervention times. Conclusion. With the new-generation ASD and PFO occluder devices, the incidence of clinically silent peri-interventional cerebral embolic lesions after transcatheter closure of ASD and PFO is low. Plasma NSE levels offered no additional benefit for monitoring ischemic events after ASD and PFO transcatheter closure procedures.

J INVASIVE CARDIOL 2013;25(10):519-524

Key words: cerebral embolic events, ASD closure, PFO closure


Transcatheter closure of atrial septal defect (ASD) and patent foramen ovale (PFO) is a widely accepted alternative therapy to open heart surgery. Since the very first attempt of King and Mills in 1974 at transcatheter closure of an ASD, many new devices with different materials, closure mechanisms, and implantation methods have been introduced.1 Although excellent results have been reported with different devices, there are few studies in the literature that report the peri-interventional cerebral embolism risk and neurological complications. In this single-center study, we investigated the cerebral embolism risk after transcatheter closure of ASD and PFO with new-generation Occlutech Figulla devices (Occlutech GmbH).


Patients. Transcatheter closure of PFO (n = 14) or ASD (n = 16) was performed in 30 consecutive patients between January 2011 and  March 2012. Patients (20 female and 10 male) were between 21 and 67 years old (mean, 42 ± 13 years). The study was approved by the local medical ethics committee and written informed consent was obtained from each patient. All patients underwent cerebral diffusion-weighted magnetic resonance imaging (DW-MRI) examination before and 72 hours after the transcatheter closure procedure. At these time points, focal neurologic impairment was assessed by clinical neurologic exam and plasma neuron-specific enolase (NSE) levels were determined before and at 12 and 24 hours after the procedure (h-NSE; DiaMetra). Plasma NSE levels of 0-12 ng/mL were the normal range and levels >12 ng /mL were accepted as pathologic.

Indications for closure of PFOs included patients age 18 years, who suffered from cryptogenic stroke and had evidence of a PFO in contrast transesophageal echocardiography (TEE) studies. All patients were under antiplatelet therapy with 100-300 mg aspirin. The exclusion criteria included atrial fibrillations, significant stenosis of the carotid arteries, known thrombophilic disorders, recent myocardial infarction (MI), history of coronary revascularization within the last 30 days, pregnancy, and known allergic reactions to clopidogrel, aspirin, or nickel.

ASD closure inclusion criteria were ASD with a pulmonary-to-systemic flow ratio (Qp/Qs) >1.5:1, volume overload, or paradoxical embolism. Excluded ASD patients were those with primum or sinus venosus type ASD, reversal of right-to-left shunt, a shunt volume of Qp/Qs <1.5:1, severe pulmonary hypertension (peak pulmonary pressure exceeding 70% of systemic systolic pressure), defect diameter >40 mm, pregnancy, recent MI or coronary revascularization within the last 30 days, acute infection, presence of intracardiac thrombi, permanent contraindications to platelet therapy, and known allergic reaction to aspirin, clopidogrel, or nickel.

Both ASD and PFO patients who had contraindications for magnetic resonance imaging (MRI) examination were also excluded from the study; these contraindications included patients with internal cardiac defibrillators, cochlear implants or brain aneurysm clips, as well as MRI-incompatible artificial heart valves, infusion ports and catheters, intrauterine devices, artificial limbs and metallic joint prostheses, implanted nerve stimulators, metallic pins, plaques and surgical suture materials. 

Magnetic resonance imaging. Quantitative DW-MRI was performed with a 3 Tesla whole body system (Verio; Siemens). The imaging protocol included transverse and coronal diffusion-weight imaging (DWI), transverse fluid-attenuated inversion recovery sequences (FLAIR; reperfusion time [TR]/echo time [TE], 12000/140 ms; inversion time [TI], 2800 ms). DWI was performed with a spin-echo echo-planar pulse sequence (TE, 90 ms; TR, 6800 ms; field of view, 240 mm; matrix, 128 x 256; section thickness, 5 mm; intersection gap, 1 mm) with diffusion sensitization b-values of 0, 500, and 1000 s/mm2. For all DWI examinations, apparent diffusion coefficient (ADC) maps were obtained. 

Preexisting brain abnormalities, such as infarction, microangiopathy, or atrophy, and the appearance of new hyperintense lesions in DWI on postintervention scans were evaluated. Only diffusion abnormalities consistent with embolic lesions were included in the analysis. Diffuse alterations in the DWI or patterns of watershed ischemia were excluded. Postinterventional new lesions were determined on the DWI with maximum contrast between lesion and normal tissue signal. For volume quantification of new hyperintense lesions in DWI, the images were magnified fivefold, the area of lesion was manually delineated in each image slice by region of interest. For image analysis, we used the commercially available software of the MRI unit (Siemens).

Scans were read separately by two experienced radiologists blinded to the timing of the imaging and the neurological status of the patients. In the case of discrepancy between radiologists, a consensus reading was held.

Implantation procedure for secundum ASD and PFO. Transcatheter closure procedures were performed by three equally experienced interventional cardiologists. Under local anesthesia, the right femoral vein was punctured and soft-tipped 0.035˝ guidewire was inserted and advanced through the atrial defect and positioned at the left upper pulmonary vein. Intravenous heparin was administered to maintain an activated clotting time (ACT) >200 seconds. Then, an appropriately sized Cook delivery sheath (Cook Medical) was advanced to the left atrial (LA) side over the guidewire and the Occluder was subsequently loaded in the sheath and advanced by means of the delivery system to the LA side. After opening the LA disc, the system was retracted until the LA disc was positioned opposite the left interatrial septum. The right atrial disc was then deployed. Under the guidance of fluoroscopy and transesophageal echocardiography, an initial residual right-to-left shunt was ruled out and correct positioning was confirmed. Device stability was controlled by the Minnesota maneuver; when the device was positioned properly, it was released by opening the locking mechanism. Balloon sizing was not performed in any of the patients.

Medication. Dual-antiplatelet therapy with aspirin 100-300 mg and clopidogrel 75 mg was started before the procedure and extended for 3 months for clopidogrel and 6 months for aspirin. Infective endocarditis prophylaxis was recommended during the first 6 months in concordance with the guidelines.

Statistical analysis. The statistical analysis was performed using  a software package (SPSS for Windows, release 17; SPSS, Inc). Continuous variables are presented as mean ± standard deviation. The data distribution was controlled with Kolmogorov-Simirnov test. Student’s t-test, paired-sample t-test, and Mann-Whitney U-test were used to compare between continuous variables. Categorical data were evaluated by the chi-square test or Fisher’s exact test. Correlations were analyzed via the Pearson’s correlation test.


A total of 30 patients were enrolled in the study between January 2011 and March 2012. Details of the patient demographics are listed in Table 1. There were no significant differences between ASD and PFO groups regarding the demographic data. History of previous transient ischemic attack (TIA) or stroke in the PFO group was significantly higher than in the ASD group (P=.001). 

Implantation of the device was technically successful in all patients (PFO, n = 14; ASD, n = 16). There were no periprocedural major or minor complications. In all patients, DW-MRI was performed before (E1) and 72 hours after (E2) the transcatheter closure procedure. In preprocedural DW-MRI scans, a total of 18 patients showed hyperintense white-matter lesions, 11 showed lacunar defects, 6 had chronic infarcts, and 7 had signs of brain atrophy. No patient revealed acute ischemic lesions at the preprocedural DW-MRI. Old territorial infarcts were significantly more common in PFO patients compared to ASD patients (P<.05) (Table 2).

After transcatheter closure, 1 PFO patient showed an acute ischemic lesion on DW-MRI scan (Figure 1). The lesion was located in the right parietal cortex and had a volume of <1 cm3. That patient’s neurological exam revealed no abnormality. This patient had a fluoroscopy time of 9 minutes and total procedure time of 20 minutes, which is not statistically different from the rest of the patients.

Individual characteristics, detailed MRI findings, procedure details, and NSE levels in consecutive PFO-ASD closure patients are shown in Table 3.

Considering all patients in both groups, NSE levels did not significantly change between 0 and 12 hours, between 12 and 24 hours, or between 0 and 24 hours (Independent samples t-test: P=.559, P=.649, and P=.312, respectively). Table 4 summarizes NSE levels in both groups.

When PFO and ASD groups are evaluated separately, NSE levels were not significantly different at 0 hours, 12 hours, or 24 hours (P>.05) (Table 5). 

There was no statistically significant change in NSE levels between 0 and 12 hours, between 12 and 24 hours, or between 0 and 24 hours (P=.429, P=.156, and P=.310, respectively, in PFO patients; P=.120, P=.426, and P=.559, respectively, in ASD patients)  (Figure 2).

The mean fluoroscopy time was 11.0 ± 4.2 minutes in PFO patients and 9.6 ± 3.5 minutes in ASD patients; there was no significant difference between the groups (P=.334). There was also no significant difference between the two groups regarding total procedure times (22.9 ± 5.7 minutes in the PFO group and 22.0 ± 4.1 minutes in the ASD group; P=.609).

Fluoroscopy time, which is the time period when actual manipulation of the PFO or ASD takes place and thus might be a risk factor for cerebral microembolic events, was tested against NSE levels for any correlation. There was no statistically significant correlation between fluoroscopy time and changes in NSE levels between 12 and 24 hours or between 0 and 24 hours. However, fluoroscopy time and changes in NSE levels between 0 and 12 hours was, albeit slightly, positively correlated (Pearson’s coefficient, 0.398; P=.029).


To the best of our knowledge, this is the first study investigating the periprocedural cerebral embolism risk with DW-MRI combined with clinical and serologic parameters of brain injury during transcatheter closure of ASD and PFO. There are few studies in the literature that report the peri-interventional cerebral embolism risk and neurological complications during this procedure. In our study, we detected 1 acute lesion on DW-MRI scan (3.3%). This was a PFO patient; her neurologic exam revealed no abnormality and plasma NSE levels did not change significantly (10.6 ng/mL at baseline, 9.8 ng/mL at 12 hours, 8 ng/mL at 24 hours).

NSE is an isoenzyme of the glycolytic enzyme enolase (2-phospho-D-glycerate hydrolase), and has been shown to be highly specific for neuronal tissue that is released into the cerebrospinal fluid and the cerebral and systemic circulation after neuronal damage. In healthy individuals, NSE is negligibly present in the peripheral blood. Various investigators have found an increase in serum NSE levels after neuronal damage associated with intracerebral hemorrhage, ischemic stroke, and brain injury, suggesting that NSE is a sensitive quantitative marker of parenchymal brain injury.2-5 One study suggested that after a cardiac arrest with at least 5 minutes of cardiopulmonary resuscitation, NSE levels in the first 48 hours may facilitate the prediction of survival to hospital discharge.6 Another study concluded that NSE could be used as a simple and reliable prognostic factor for predicting postoperative brain dysfunction after cardiac surgery.7 However, in our study, NSE did not show any significant changes during the immediate postprocedure period. NSE levels did not predict the clinically silent embolic event that was demonstrated by DW-MRI in the aforementioned PFO patient either. Having only 1 ischemic event, however, is a limitation for any conclusions, implying that the plasma NSE levels offer no additional benefit for monitoring ischemic events in ASD and PFO patients after transcatheter closure procedures.

In a recent study, Gharem et al investigated the incidence of clinically silent peri-interventional cerebral embolic events and neurological impairment after transfemoral  aortic valve implantation (TAVI) with DW-MRI. They reported 3 new neurological findings in 22 patients (10%) and 75 new embolic cerebral lesions (72.7%)  on DW-MRI after TAVI. Their study showed no relation between acute cerebral lesions and plasma concentrations of NSE either.8 Although there was a high rate of silent peri-interventional embolism risk after TAVI, the incidence of persistent neurological impairment was reported to be as low as 3.6%.

Five prospective studies with DW-MRI have detected new silent ischemic lesions after cardiac catheterization, with an incidence ranging between 5% to 22%.9 After cardiac surgery, especially valve replacement surgery, DW-MRI detected new focal brain lesions in 47% of asymptomatic patients.10

Ferrari et al monitored middle cerebral artery for high-intensity transient signals (HITS) during transcatheter closure of PFO and ASD in order to investigate the incidence of cerebral embolism with or without neurological deficit. They detected HITS in 33 of 35 patients (96%) with a median rate of 8 HITS and reported that highest rates were observed during the times when the septum was crossed with a guidewire and the LA disc was deployed. Despite this high rate, there were no clinically apparent neurological deficits.11

Dorenbeck et al reported new microembolic lesions on the DW-MRI in 3 of 35 symptomatic PFO patients after transcatheter closure (8.6%). Clinically, 2 of these 3 lesions were silent and detected solely by MRI. The third patient suffered from a transient right-sided hemihypesthesia for 12 hours, meaning the rate of clinically apparent temporary lesion was 2.8%. The mentioned study did not evaluate the utility of neurochemical markers. Three different occluder devices were used and 2 silent DW-MRI lesions were reported after implantation of a CardioSeal occluder system, and 1 microembolic infarct with a Helex occluder system. They detected no acute ischemic defects after closure with the Amplatzer septal occluder system. Although different devices were used for PFO closure, as the authors have mentioned, this was a small study and comparison between these different devices needed larger numbers.12

Only one brand of closure device was used in order to increase the statistical power of this study. We used the newest commercially available nitinol meshwire occluder devices at the time of the study (Occlutech Figulla PFO and ASD Occluder N). The major feature of these devices is the absence of LA clamp, thus theoretically minimizing any chance for trauma and clot formation on the LA disc. The PFO occluder device’s LA disc consists of a single layer without a hub, thereby minimizing the amount of material on the LA side in contrast to other occluder systems.13,14

A previous study concluded that balloon sizing, although reported as the gold standard for estimation of defect diameter and device size, caused a tendency to oversize the ASD and this might cause the operator to select a larger device than necessary. It was also associated with increased fluoroscopy time and total procedure time.15 In another study, Wang et al evaluated the safety and feasibility of transcatheter closure of ASD without balloon sizing and found no significant difference in success rates between balloon sizing and TEE guided closure.16 In this study, we measured maximal diameters of the defects on intraprocedural TEE images and did not use balloon sizing; this may have shortened our procedure and fluoroscopy times, which is believed to be linked to thrombus formation and peri-interventional embolism risk. 

DW-MRI examination of patients with vascular dementia implied that small, clinically silent lesions may contribute to cognitive deterioration.17 We did not detect clinically apparent cerebral emboli in any patient; however, silent embolic events may also cause deficits with more subtle manifestations, such as cognitive and memory dysfunctions, which are not assessable by simple neurological examinations. Although cerebral MRI examination is not indicated in asymptomatic patients during routine practice, DW-MRI may be helpful in follow-up of patients undergoing potentially high-risk angiographic procedures. Long-term observations are needed to determine if these silent ischemic events culminate in any form of clinically apparent cerebral dysfunction.

The optimal therapy for prevention of recurrent stroke in PFO patients with cryptogenic stroke has not been adequately defined. The choice between medical therapy and percutaneous closure caused intense debate over the past years.18 In two recent multicenter prospective trials, Meier et al and Carroll et al concluded that closure of a PFO for secondary prevention of cryptogenic embolism did not result in a significant reduction in the risk of  recurrent embolisms or death as compared with medical therapy.19,20


Although the representative value of our study is limited by the rather small number of investigated patients, it may be concluded that the incidence of clinically apparent or silent peri-interventional cerebral embolic lesions after transcatheter closure of ASD and PFO is low with the new-generation Occlutech ASD and PFO occluder devices. In long-term follow-up, the importance of peri-interventional silent embolic lesions may be investigated with appropriate tests, such as DW-MRI and detailed neurocognitive examinations.


  1. King TD, Thompson SL, Steiner C, Mills NL. Secundum atrial septal defect. Nonoperative closure during cardiac catheterization. JAMA. 1976;235(23):2506-2509.
  2. Schoerkhuber W, Kittler H, Sterz F, et al. Time course of serum neuron-specific enolase. A predictor of neurological outcome in patients resuscitated from cardiac arrest. Stroke. 1999;30(8):1598-1603.
  3. Schaarschmidt H, Prange HW, Reiber H. Neuron-specific enolase concentrations in blood as a prognostic parameter in cerebrovascular diseases. Stroke. 1994;25(3):558-565.
  4. Mabe H, Suzuki S, Mase M, et al. Serum neuron-specific enolase levels after subarachnoid hemorrhage. Surg Neurol. 1991;36(3):170-174.
  5. Cunningham RT, Watt M, Winder J, et al. Serum neuron-specific enolase as an indicator of stroke volume. Eur J Clin Invest. 1996;26(4):298-303.
  6. Auer J, Berent R, Weber T, et al. Ability of neuron-specific enolase to predict survival to hospital discharge after successful cardiopulmonary resuscitation. Can J Emerge Med. 2006;8(1):13-18.
  7. Konstantinou EA, Venetsanou K, Mitsos AP, et al. Neuron-specific enolase: a valuable prognostic factor of central nervous system dysfunction following cardiac surgery. Brit J Anaes Rec Nursing. 2008;9:22-28.
  8. Ghanem A, Mueller A, Claas P, et al. Risk and fate of cerebral embolism after transfemoral aortic valve implantation: a prospective pilot study with diffusion-weighted magnetic resonance imaging. J Am Coll Cardiol. 2010;55(14):1427-1432.
  9. Jurga J, Nyman J, Tornvall P, et al. Cerebral microembolism during coronary angiography: randomized comparison between femoral and radial arterial access. Stroke. 2011;42(5):1475- 1477.
  10. Knipp SC, Matatko N, Schlamann M, et al. Small ischemic brain lesions after cardiac valve replacement detected by diffusion-weighted magnetic resonance imaging study: relation to neurocognitive function. Eur J Cardiothorac Surg. 2005;28(1):88-96.
  11. Ferrari J, Baumgartner H, Tentschert S, et al. Cerebral microembolism during transcatheter closure of patent foramen ovale. J Neurol. 2004;251(7):825-829.
  12. Dorenbeck U, Simon B, Skowasch D, et al. Cerebral embolism with interventional closure of symptomatic patent foramen ovale: an MRI-based study using diffusion-weighted imaging. Eur J Neurol. 2007;14(4):451-454.
  13. Krizanic F, Sievert H, Pfeiffer D, et al. Clinical evaluation of novel occluder device for percutaneous transcatheter closure of patent foramen ovale. Clin Res Cardiol. 2008;97(12):872-877.
  14. Krizanic F, Sievert H, Pfeiffer D, et al. The Occlutech Figulla PFO and ASD occluder: a new nitinol wire mesh device for closure of atrial septal defects. J Invasive Cardiol. 2010;22(4):182-187.
  15. Queck SC, Wu WX, Chan KY, et al Transcatheter closure of atrial septal defects — is balloon sizing still necessary? Ann Acad Med Singapore. 2010;39(5):390-393.
  16. Wang JK, Tsai SK, Lin SM, et al. Transcatheter closure of atrial septal defect without balloon sizing. Catheter Cardiovasc Interv. 2008;71(2):214-221.
  17. Choi SH, Na DL, Chung CS, et al. Diffusion-weighted MRI in vascular dementia. Neurology. 2000;54(1):83-89.
  18. O’Gara PT, Messe SR, Tuzcu EM, et al. Percutaneous device closure of patent foramen ovale for secondary stroke prevention: a call for completion of randomized clinical trials: a science advisory from the American Heart Association/American Stroke Association and the American College of Cardiology Foundation. Circulation. 2009;119(20):2743-2747.
  19. Meier B, Kalesan B, Mattle HP, et al. Percutaneous closure of patent foramen ovale in cryptogenic embolism. N Engl J Med. 2013;368(12):1083-1091.
  20. Carroll JD, Saver JL, Thaler DE, et al. Closure of patent foramen ovale versus medical therapy after cryptogenic stroke. N Engl J Med. 2013;368(12):1092-1100.

From the 1Department of Cardiology, Bakirkoy Sadi Konuk Education and Research Hospital, Istanbul, Turkey, 2Department of Radiology, Marmara University Faculty of Medicine, Istanbul, Turkey, and 3Department of Radiology, Bakirkoy Sadi Konuk Education and Research Hospital, Istanbul, Turkey.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.

Manuscript submitted February 11, 2013, provisional acceptance given March 28, 2013, final version accepted June 24, 2013.

Address for correspondence: Dr Hande Oktay Tureli, Kardiyoloji Klinigi, Bakırkoy Sadi Konuk EA Hastanesi, Tevfik Saglam Cad. 11, Bakirkoy, Istanbul, Turkey. Email: handetureli@yahoo.com