Surgical Closure of Patent Foramen Ovale in Cryptogenic Stroke Patients
Background and Purpose Patent foramen ovale (PFO) is associated with stroke of unknown etiology or cryptogenic stroke. However, optimal treatment to prevent recurrence in cryptogenic stroke patients with PFO is not clearly defined. Since PFO represents a surgically repairable lesion, interest in closing it is high. This report reviews our experience with cryptogenic stroke patients with PFO who underwent surgical PFO closure and were followed for recurrence of neurological events.
Methods We followed 28 cryptogenic stroke patients (17 men, 11 women; mean age, 41±13 years) with transesophageal echocardiograpy–defined PFO who had undergone PFO closure by open thoracatomy. All patients selected for surgery refused, could not take, or failed warfarin therapy. They were followed by physician visits and telephone interviews.
Results There were no surgical complications. With a mean follow-up of 19 months, four patients experienced neurological event recurrence, one stroke, and three transient ischemic attacks. Kaplan-Meier survival analysis demonstrated that the actuarial rate of recurrence was 19.5% (95% confidence limit 2.2-36.8%) at 13 months of follow-up. None of the 17 patients (0%) younger than 45 years suffered a recurrence, whereas four of 11 patients (35%) aged 45 or older experienced a recurrence of neurological event (P<.02). Using a proportional hazards regression model, the increase in relative risk with increasing age was 2.76 per 10 years (95% confidence interval 1.07 to 7.16).
Conclusions Although PFO is easily repairable in patients with cryptogenic stroke, its closure does not consistently prevent recurrence of ischemic events. The recurrence appears to occur more frequently in older cryptogenic stroke patients.
Approximately 40% of ischemic strokes have no clearly definable etiology and are termed cryptogenic strokes.1 PFO, a small communication between the left and right atria, can cause stroke through paradoxical embolization. Recent reports have indicated that patients with cryptogenic stroke have a several-fold higher prevalence of PFO compared with the controls.2 3 4 The recurrent stroke rate for cryptogenic stroke patients with PFO has not been defined, and thus the best therapeutic approach to reduce the recurrence is also undefined. Conventionally, many of the ischemic stroke patients are treated with an antiplatelet agent or anticoagulants.5 6 However, some patients have contraindication to its use, refuse to take oral agents because of potential restrictions on their lifestyle, or experience recurrence of ischemic events while on medical therapy. For these patients, surgical closure of the PFO becomes a potential therapeutic modality. In this paper, we report on our experience with surgical PFO closure for prevention of recurrent cerebrovascular events in patients with cryptogenic stroke and PFO.
Subjects and Methods
This study included 28 stroke patients with stroke subtyping of cryptogenic stroke defined using the previously established criteria.1 They underwent open heart surgery for closure of PFO after a stroke thought to be due possibly to paradoxical embolism. One patient had a history of two previous ischemic strokes, whereas others each experienced one stroke confirmed by CT scan or MRI, considered to be due to an embolism of undefined etiology. Selection criteria for surgical PFO closure included refusal for therapy with warfarin (n=23), contraindication to the use of warfarin (n=3), inability to adequately maintain a therapeutic prothrombin time (n=1), and recurrent stroke on warfarin therapy (n=1). All patients underwent thorough neurological evaluation including carotid duplex Doppler and transcranial Doppler imaging. No patient had history or symptoms suggestive of arrhythmias, and all had normal ECGs. Hematological workup included evaluation for abnormal levels of protein S, protein C, and anticardiolipin antibodies. This was performed in 25 of the 28 patients.
All patients underwent either biplane or multiplane TE (Hewlett-Packard) to evaluate for any other potential causes of cerebral embolization. A saline contrast study for PFO detection was performed by mixing 1 mL of air with 9 mL of normal saline injected through an antecubital vein. Injection was performed with and without the Valsalva maneuver. All patients had PFO diagnosed by the appearance of microbubbles in the left atrium within three cardiac cycles after the opacification of the right atrium (Fig 1⇓). The separation between septum primum and secundum was measured offline using software incorporated in echocardiography equipment. The maximum number of microbubbles appearing in the left atrium within three cardiac cycles from the right atrial opacification was also counted. Atrial septal aneurysm was defined as protrusion of the atrial septum of at least 1 cm into the left or right atrium from its base.7 Significant aortic arch plaque was defined as that measuring more than 4 mm in thickness.8 All echocardiographic images were analyzed by a single experienced observer (S.H.).
Technique of PFO Closure
All patients underwent the procedure by open thoracotomy, with 26 undergoing median sternotomy and one each undergoing the procedure by left thoracotomy and inframammary incision. All patients underwent cardiopulmonary bypass. The presence of all PFOs was confirmed at the time of the procedure. All PFOs were closed by primary anastomoses without the use of patch material. In Fig 2⇓, the PFO before the surgical procedure is demonstrated. The mean cardiopulmonary bypass time was 45±12 minutes, and mean aortic cross clamp time was 13±15 minutes.
All patients were followed on a regular basis by neurologists and cardiologists, and they also were interviewed for any neurological events using a standardized telephone questionnaire. This was administered by an investigator who was unaware of the patients’ clinical history.
All data are expressed as mean±standard deviation for continuous variables and as frequency for categorical variables. Kaplan-Meier estimates of freedom from cerebrovascular events after surgery were constructed, and the influence of each independent variable (age, gender, PFO size, number of microbubbles, presence of ASA, drug therapy after surgery) was assessed with the log-rank test. Estimates of relative risk for each independent variable were obtained using a proportional hazards regression analysis. Data were analyzed using SAS 6.10 (SAS Institute, Cary, NC).
Patients’ clinical characteristics are given in Table 1⇓. Of the 17 men and 11 women, the mean age was 41±13 years at the time of surgery with a range of 17-64 years. Of the 28 patients, none had a history of diabetes mellitus, hypertension, myocardial infarction, or atrial fibrillation. Four patients had a history of cigarette smoking, and one had a history of oral contraceptive use. Three of the 28 patients (11%) had abnormal value in one of the hematological tests for hypercoagulable state; one each had a low level of protein S and protein C, and another patient had an abnormal level of anticardiolipin IgM antibody.
On TE, all had normal ventricular and atrial chamber dimensions. No patient had intracardiac thrombus, vegetation, or spontaneous echo contrast. No patient had significant aortic arch plaque. Although three patients had thin smooth plaque measuring less than 1 mm in thickness, none of the other patients had any plaque. The mean separation of septum primum from the secundum on TE was 3.8±1.9 mm (range 1.0-8.0 mm), and the mean number of microbubbles seen in the left atrium was 20±10. Sixteen of 28 (57%) patients had more than 25 microbubbles in the left atrium on saline contrast study. Thirteen of the 28 patients (46%) had ASA.
No patient experienced any significant complication from surgery. There was no stroke, myocardial infarction, episode of hemodynamic instability, or infectious complication. The mean hospital stay was 7.8±3.8 days with a range of 4-22 days. One patient remained for 22 days since this patient was initially admitted for stroke and underwent surgery during the same admission. All other patients underwent the procedure after an elective admission. Five of the 28 patients (18%) developed postpericardiotomy syndrome. Two were treated with a course of oral steroids, and three were treated with indomethacin. One patient who experienced transient atrial fibrillation in the immediate postoperative period was treated with a short course of a beta-blocker without recurrence.
Cerebrovascular Events After PFO Closure
During 532 patient-months of follow-up (mean, 19 months per patient; range, 1-52 months), 11 patients were placed on aspirin (325 mg/day at the discretion of the referring physicians), and 17 received no medical therapy for stroke prevention. During the follow-up period, four patients experienced recurrent neurological events—one with ischemic stroke and three with transient ischemic attack (TIA). None of the three patients with coagulation abnormalities experienced a recurrence.
The patient with recurrent stroke was the second oldest patient (62 years) in the series who underwent surgical closure for a stroke that occurred during receipt of an adequate level of oral anticoagulation. This patient initially did well after the surgery on no anticoagulant or antiplatelet agents. However, 13 months after the procedure, he presented with changes in vision. The MRI performed at that time revealed new left occipital infarction. The patient was admitted to a hospital and was found also to have pneumonia as well as renal cell carcinoma, which had been undetected previously. He subsequently expired. At autopsy, the site of PFO closure was found to be intact and free of thrombus.
Three patients experienced a TIA. A 45-year-old man, who initially presented with changes in vision associated with a left posterior cerebral artery territory infarction, refused warfarin therapy and subsequently underwent surgery. He was on neither warfarin nor antiplatelet agent after the procedure. Twelve months after the procedure, he presented with similar symptoms, which spontaneously resolved. A TE was performed and demonstrated no separation of septum primum from secundum and fewer than five microbubbles that appeared in the left atrium with the Valsalva maneuver. No intracardiac mass or thrombus was seen.
A 53-year-old woman initially presented with right arm and face numbness and weakness. She was found to have a left parietal infarction. One month later, she experienced nausea, vomiting, and vertigo and was found to have a new cerebellar and calcarine infarction. She refused warfarin therapy, underwent surgical PFO closure, and was discharged on aspirin. Seven months later, she presented with facial numbness and slurred speech. Magnetic resonance imaging showed no new infarctions. A TE was performed, which revealed no intracardiac mass and the PFO to be completely closed without any residual shunt. No intracardiac mass or thrombus was seen.
A 64-year-old patient, the oldest patient in our series, presented with right hemiparesis. He was placed on warfarin; however, 4 months later when prothrombin time decreased to subtherapeutic level, he developed memory and speech difficulty. CT scan showed a left occipital and thalamic subacute infarction consistent with a left posterior cerebral artery territory infarction. He was subsequently placed on warfarin, but 6 months later wanted to stop warfarin therapy and thus underwent surgery. He was discharged on aspirin 325 mg/day. Ten months after the procedure, he presented with an episode of confusion and speech difficulty; however, no new infarction was documented by MRI. He underwent TE at this time and demonstrated no separation of septum primum from secundum and less than five microbubbles appearing in left atrium after Valsalva maneuver. No intracardiac mass or thrombus was seen.
Actuarial Rate of Cerebrovascular Event
In Fig 3⇓, the actuarial rate of recurrent stroke or TIA is shown. The cumulative estimate of stroke or TIA recurrence was 19.5% at 13 months of follow-up (95% confidence interval 2.2-36.8%). Patients with recurrence were significantly older than those without. The mean age of those with recurrence at the time of PFO closure was 56±9 compared with 38±12 for those without recurrence (P=.01). None of 17 patients (0%) younger than 45 years suffered a recurrence, although four of 11 patients (36%) 45 or older experienced a recurrence of a neurological event after PFO closure (P<.02). The risk of recurrence increased with age (P=.01). The increase in relative risk was 1.11 per year, or 2.76 per 10 years (95% confidence limits 1.07-7.16).
None of the other variables considered showed a difference between the group with and without recurrence (Table 2⇓). One of 11 women (9%) experienced recurrence compared with three of 17 men (18%) (P=NS). The mean PFO size and the number of microbubbles were similar between those with and without recurrence (4.8±1.8 versus 3.6±1.9 mm and 20±10 versus 18±9 microbubbles: P=NS). Two of the 13 (15%) patients with ASA experienced recurrence, and two of the 15 patients (13%) without ASA experienced recurrence (P=NS). Two of the 11 patients (18%) on aspirin experienced a recurrent event, compared with two of the 17 (12%) on no medical therapy (P=NS). The risk ratio and 95% confidence interval for each of the variables are presented in Table 2⇓.
Association of Patent Foramen Ovale With Cryptogenic Stroke
According to the Stroke Data Bank,1 a study involving 1273 stroke patients, approximately 40% of ischemic stroke patients have no clearly identifiable etiology (cryptogenic stroke) after extensive work-up. Patent foramen ovale, a hemodynamically insignificant communication between the atria, has been associated with cryptogenic stroke in a variety of studies. This has been demonstrated in the young as well as in the older population with cryptogenic stroke.2 3 4
Demonstration of embolic material reaching the cerebral circulation through PFO has been obtained in several studies using transcranial Doppler after intravenous injection of aerated saline.9 10 11 Furthermore, as a support of PFO as a conduit for paradoxical embolization, several recent case reports document a thrombus lodged in PFO.12 13 Patent foramen ovale is found commonly among the general population, and prevalence figures of 27 and 29% are reported in autopsy series.14 15 Therefore, certain characteristics of PFO may play a role in its becoming a conduit for paradoxical embolization. When the TE characteristics of PFO in those with likely paradoxical embolization are compared with those in the controls, it is noted that the TE-detected conduit size and the degree of shunt are significantly larger in patients with probable paradoxical embolic events.16 17 Upon saline contrast injection, a larger amount of embolic material is also detected with transcranial Doppler in patients with cryptogenic stroke compared with those who have a known cause of stroke,18 consistent with our finding in the current series. Patients in this series had a greater PFO size (3.8 versus 0.57 mm) and amount of shunt (20 versus two microbubbles) than patients with known cause of stroke from our previous study using the same measurement technique.19
Stroke Recurrence in Patients With PFO
Data on 1-year cumulative rate of death or stroke recurrence for various stroke subtypes are available from Northern Manhattan Stroke Study. The rate for all stroke subtypes is 9.4% per year, and 10.0% for cryptogenic stroke.20 These rates do not take into account the patients’ medical therapy, thus there are no data on the efficacy of medical therapy on death or stroke recurrence for patients with cryptogenic stroke. In particular, the recurrence rate of ischemic events in cryptogenic stroke patients with PFO remains undefined.
The recurrence rate may be high. Sharma et al21 report experience in 17 patients with ASA and PFO with 12 months’ follow-up in which they found that although there was no recurrence in patients treated with warfarin, there was 50% recurrence in those treated with aspirin. Comess and colleagues report on 33 patients with PFO with a mean recurrent event rate of 16% per year compared with 7% in the control population.22 Other studies have demonstrated low recurrence. Hanna et al report no neurological event recurrence among 13 medically treated patients with follow-up of up to 41 months.23 More recently, Mas and Zuber24 retrospectively studied 132 patients less than 60 years of age with PFO or ASA who had had a stroke or TIA. Over the mean follow-up of 22.6 months, they report six stroke or TIA with an actuarial risk of 6.4% at 2 years. The low risk may have been due to the patient selection criteria, which included only those younger than 60 years. As a result of conflicting reports, the efficacy of medical therapy for prevention of recurrent neurological events in patients with PFO remains unclear.
Mechanical PFO Closure
Since PFO is easily repairable, the interest in closing them is high. Bridges et al25 report on transcatheter PFO closure in 36 patients with presumed paradoxical embolism with a mean age of 39 years who were followed for a mean duration of 8.4 months. There was no recurrent stroke. More recently, Ende et al26 have reported on the experience of transcatheter closure of PFO in 10 patients with a mean age of 40 years followed for a mean of 32 months. They also report no recurrence of neurological events. It should be noted, however, that mechanical failure of the devices has been reported, and its long-term efficacy and durability remain inconclusively defined.27
Surgical closure has also been attempted with mixed results. Harvey et al report no recurrence in four patients with a mean age of 35 years followed for 7 to 21 months.28 Zhu and colleagues report two patients with recurrence among six patients with a mean age of 35 years with a mean follow-up of 3.9 years.29 Most recently, Devuyst and coworkers have reported on their experience with 30 patients younger than 50 who underwent surgical PFO closure for stroke or TIA.30 No surgical complication is reported, and no recurrence of stroke or TIA is reported after a follow-up period of 2 years. With a similar number of patients and follow-up periods, their experience and ours are in contrast. In our series, four of 28 patients presented with recurrent neurological events at a mean follow-up of 19 months. The recurrence was concentrated in the older population, and other variables such as gender, PFO size, degree of shunt, the presence of ASA and drug therapy did not appear to relate to the recurrence. The significantly higher incidence of recurrence for the older population appears to indicate that they may harbor other undefined causes for their stroke, despite extensive work-up, and that etiology of their stroke may not have been due to paradoxical embolism.
Five patients in our study (18%) experienced postpericardiotomy syndrome requiring medical therapy. Although no significant complication developed as a result of the postpericardiotomy syndrome, its sequel may include constrictive pericarditis with important hemodynamic consequences.31 It is important to note that the surgical complication rate varies, depending on the clinical status of the patients. Based on the series of patients undergoing surgical closure of secundum atrial septal defect, although the mortality for the patients in NYHA classes I and II is 1.6%, it increases to 5% for those in classes III and IV.32 Therefore, older patients with a higher likelihood of having a poor functional capacity will have an increased chance of surgical complication. Combined with the higher recurrence rate seen in our study, the increased surgical risk should be taken into account in considering surgical PFO closure in older cryptogenic stroke patients with PFO.
Limitations in this study exist. These include the small number of patients and relatively short follow-up that are reflected in the wide confidence intervals of the observed risk ratios. In addition, there is no control group of patients receiving medical therapy. In particular, there was no group receiving warfarin. Therefore the efficacy of medical and surgical approaches could not be compared. Routine MRI or CT scan was not performed during the follow-up period. As a result, clinically silent infarctions cannot be ruled out. Aspirin was given after the procedure without a defined criteria, which may have influenced the results. Also, the possibility that the patient with recurrent stroke may have had a hypercoagulable state was not addressed. Finally, not all patients underwent TE after surgery. Therefore, the influence of possible residual shunt on cerebrovascular event was not fully evaluated.
Our experience indicates that even in a highly selected group of cryptogenic stroke patients, surgical PFO closure does not guarantee freedom from recurrent neurological events. Although surgical closure may be effective in younger patients, its efficacy in older patients appears to be poor. We therefore view the surgical PFO closure as a potential but unproven option for the younger cryptogenic stroke patients unable or unwilling to undergo medical therapy. A prospective study with a larger number of patients and a longer follow-up is necessary to better assess the efficacy of medical therapy.
This study was supported in part by National Institute of Neurological Disorders and Stroke grants RO1 NS-32525 and RO1 NS-33248 and a grant from the Mellam Family Foundation. The authors thank Lynette Mendoza and Andrea Caputo for their excellent assistance in the preparation of this manuscript.
- Received June 16, 1997.
- Revision received September 23, 1997.
- Accepted September 23, 1997.
- Copyright © 1997 by American Heart Association
Di Tullio M, Sacco RL, Gopal A, Mohr JP, Homma S. Patent foramen ovale as a risk factor for cryptogenic stroke. Ann Intern Med.. 1992;117:461-465.
Caplan LR. Diagnosis and treatment of ischemic stroke. JAMA. 1991;266;2413-2418.
Sherman DG, Dyken ML, Gent M, Harrison MJG, Hart RG, Mohr JP. Antithrombotic therapy for cerebrovascular disorders: an update. Chest. 1996;104(suppl):444S–456S.
Mügge A, Daniel WG, Angermann C, Spes C, Khandheria BK, Kronzon I, Freedberg RS, Keren A, Dennig K, Engberding R, Sutherland GR, Vered Z, Erbel R, Visser CA, Lindert O, Hausmann D, Wenzlaff P. Atrial septal aneurysm in adult patients. A multicenter study using transthoracic and transesophageal echocardiography. Circulation.. 1995;91:2785-2792.
Teague SM, Sharma MK. Detection of paradoxical cerebral echo contrast embolization by transcranial Doppler ultrasound. Stroke.. 1991;22:740-745.
Chimowitz MI, Nemec JJ, Marwick TH, Lorin RJ, Furlan AJ, Salcedo EE. Transcranial Doppler ultrasound identifies patients with right-to-left cardiac or pulmonary shunts. Neurology.. 1991;41:1902-1904.
Di Tullio M, Sacco RL, Venketasubramanian N, Sherman D, Mohr JP, Homma S. Comparison of diagnostic techniques for the detection of a patent foramen ovale in stroke patients. Stroke.. 1993;24:1020-1024.
Hargreaves M, Maloney D, Gribbin B, Westaby S. Impending paradoxical embolism: a case report and literature review. Eur Heart J.. 1994;15:1284-1285.
Thompson T, Evans W. Paradoxical embolism. Q J Med.. 1930;23:135-150.
Homma S, Di Tullio MR, Sacco RL, Mihalatos D, Li Mandri G, Mohr JP. Characteristics of patent foramen associated with cryptogenic stroke: a biplane transesophageal study. Stroke.. 1994;25:582-586.
Sacco RL, Shi T, Zamanillo MC, Kargman DE. Predictors of mortality and recurrence after hospitalized cerebral infarction in an urban population: the Northern Manhattan Stroke Study. Neurology.. 1994;44:626-634.
Sharma AK, Ofili E, Castello R, Sullivan N, Labovitz AJ. Effect of treatment on recurrent embolic events with atrial septal aneurysm and associated right to left shunting. J Am Soc Echo.. 1991;43:294. Abstract.
Hanna JP, Sun JP, Furlan AJ, Stewart WJ, Sila CA, Tan M. Patent foramen ovale and brain infarct, echocardiographic predictors, recurrence, and prevention. Stroke.. 1994;25:782-786.
Bridges ND, Hellenbrand W, Latson L, Filiano J, Newburger JW, Lock JE. Transcatheter closure of patent foramen ovale after presumed paradoxical embolization. Circulation.. 1992;86:1902-1908.
Rocchini AP. Transcatheter closure of atrial septal defects, past, present and future. Circulation.. 1990;82:1044-1045.
Harvey JR, Teague SM, Anderson JL, Voyles WF, Thadani U. Clinically silent atrial septal defects with evidence for cerebral embolization. Ann Intern Med.. 1986;105:695-697.
Zhu WX, Khandheria BK, Warnes CA, Seward JB, Danielson GK. Closure of patent foramen ovale for cryptogenic stroke in young patients: long-term follow-up. Circulation.. 1992;86:I-147. Abstract.
Devuyst G, Bogousslavsky J, Ruchat P, Jeanrenaud X, Despland P-A, Regli F, Aebischer N, Karpuz HK, Castillo V, Guffi M, Sadeghi H. Prognosis after stroke followed by surgical closure of patent foramen ovale: a prospective follow-up study with brain MRI and simultaneous transesophageal and transcranial Doppler ultrasound. Neurology.. 1996;47:1162-1166.
D’Cruz IA, Overton DH, Pai GM. Pericardial complications of cardiac surgery: emphasis on the diagnostic role of echocardiography. J Cardiol Surg. 1992;7(3):257-268.