Patent Foramen Ovale: a Review of Associated Conditions and the Impact of Physiological Size Edmund K

Journal of the American College of Cardiology Vol. 38, No. 3, 2001 © 2001 by the American College of Cardiology ISSN 0735-1097/01/$20.00 Published by Elsevier Science Inc. PII S0735-1097(01)01427-9 REVIEW ARTICLE Patent Foramen Ovale: A Review of Associated Conditions and the Impact of Physiological Size Edmund K. Kerut, MD, FACC,* William T. Norﬂeet, MD,† Gary D. Plotnick, MD, FACC,‡ Thomas D. Giles, MD, FACC* New Orleans, Louisiana; Houston, Texas; and Baltimore, Maryland

Patent foramen ovale (PFO) is implicated in platypnea-orthodeoxia, stroke and decompression sickness (DCS) in divers and astronauts. However, PFO size in relation to clinical illness is largely unknown since few studies evaluate PFO, either functionally or anatomically. The autopsy incidence of PFO is approximately 27% and 6% for a large defect (0.6 cm to 1.0 cm). A PFO is often associated with atrial septal aneurysm and Chiari network, although these anatomic variations are uncommon. Methodologies for diagnosis and anatomic and functional sizing of a PFO include transthoracic echocardiography (TTE), transesophageal echocardiography (TEE) and transcranial Doppler (TCD), with saline contrast. Saline injection via the right femoral vein appears to have a higher diagnostic yield for PFO than via the right antecubital vein. Saline contrast with TTE using native tissue harmonics or transmitral pulsed wave Doppler have quantitated PFO functional size, while TEE is presently the reference standard. The platypnea-orthodeoxia syndrome is associated with a large resting PFO shunt. Transthoracic echocardiography, TEE and TCD have been used in an attempt to quantitate PFO in patients with cryptogenic stroke. The larger PFOs (approximately Ն4 mm size) or those with signiﬁcant resting shunts appear to be clinically signiﬁcant. Approximately two-thirds of divers with unexplained DCS have a PFO that may be responsible and may be related to PFO size. Limited data are available on the incidence of PFO in high altitude aviators with DCS, but there appears to be a relationship. A large decompression stress is associated with extra vehicular activity (EVA) from spacecraft. After four cases of serious DCS in EVA simulations, a resting PFO was detected by contrast TTE in three cases. Patent foramen ovales vary in both anatomical and functional size, and the clinical impact of a particular PFO in various situations (platypnea-orthodeoxia, thrombo- embolism, DCS in underwater divers, DCS in high-altitude aviators and astronauts) may be different. (J Am Coll Cardiol 2001;38:613–23) © 2001 by the American College of Cardiology

Patent foramen ovale (PFO) has been implicated in several These techniques assess PFO risk based on anatomic and pathologic processes, including paradoxical embolism in functional size. cryptogenic stroke (1–6) and venous-to-arterial gas embolism (AGE) in serious forms of decompression sickness (DCS) (an occupational hazard for underwater divers [7–9] EMBRYOLOGY and high altitude aviators and astronauts [10–13]). En- hanced right-to-left shunting of venous blood through a Patent foramen ovale is a remnant of the fetal circulation. PFO in certain pathologic states (14–16), including the Oxygenated placental blood enters the right atrium (RA) via platypnea-orthodeoxia syndrome (17–24), may cause sys- the inferior vena cava (IVC) and crosses the valve of the temic hypoxemia. Only a few studies have attempted to foramen ovale to enter the systemic arterial system (25). The quantitate the size of a PFO, either functionally or anatom- IVC flow preferentially flows toward the interatrial septum ically. (IAS) and foramen ovale. The crista interveniens directs The availability of noninvasive techniques for assessment superior vena cava flow away from the IAS. Coronary sinus of PFO provides the opportunity for clinicians to diagnose flow is also directed away from the IAS (26). At birth, pathologic states that were heretofore difficult to diagnose. pulmonary vascular resistance and right-sided cardiac pressures drop with a reversal of the RA-to-left atrium (LA) From the *Cardiovascular Research Laboratory, Division of Cardiology, Louisiana pressure gradient. The flap of the foramen ovale (septum State University Health Sciences Center, New Orleans, Louisiana; †National Aero- primum) closes against the atrial septum (septum secun- nautics and Space Administration, Johnson Space Center, Houston, Texas; ‡Division of Cardiology, University of Maryland Hospitals, Baltimore, Maryland. Supported, in dum), with fusion usually occurring within the first two part, by unrestricted funds from the Cardiovascular Research Laboratory, LSU Health years of life. Fusion is incomplete in about 25% of people, Sciences Center, New Orleans, Louisiana, and Toshiba America Medical Systems, resulting in an oblique slit-like defect. Termed a PFO, it Tustin, California. Manuscript received October 3, 2000; revised manuscript received May 4, 2001, functions as a valve-like structure with the “door-jam” on accepted May 21, 2001. the LA side of the atrial septum (25) (Fig. 1). 614 Kerut et al. JACC Vol. 38, No. 3, 2001 Patent Foramen Ovale: Size and Clinical Significance September 2001:613–23

incidence of PFO declined with age, suggesting that ana- Abbreviations and Acronyms tomic closure may occur even in adulthood (28). AGE ϭ arterial gas embolism ASA ϭ atrial septal aneurysm ASSOCIATED ANATOMICAL STRUCTURES ASD ϭ atrial septal defect DCS ϭ decompression sickness Atrial septal aneurysm (ASA) (29,30) (Fig. 2) and Chiari EMU ϭ extravehicular mobility unit network (31) (Fig. 3) are associated with PFO. Atrial septal ϭ EVA extravehicular activity (spacewalk) aneurysm incidence is 1% by an autopsy study (32) and 1.9% IAS ϭ interatrial septum ϭ by transthoracic echocardiography (TTE) (10,803 subjects) ISS International Space Station Ն IVC ϭ inferior vena cava (33), defining ASA as a septal excursion of 10 mm, with LA ϭ left atrium a base diameter of Ն15 mm. Another TTE study using an MRI ϭ magnetic resonance imaging excursion of Ն15 mm found an ASA incidence of 0.22% ϭ NTH native tissue harmonics (34). PAVS ϭ pulmonary arteriovenous shunt ϭ A monoplane transesophageal echocardiography (TEE) PFO patent foramen ovale Ն PV ϭ pulmonary vein study (30) defined ASA excursion 11 mm and base RA ϭ right atrium diameter of 15 mm to compare patients with embolic stroke TCD ϭ transcranial Doppler (n ϭ 133) with controls (n ϭ 277). An ASA was present in ϭ TEE transesophageal echocardiography 15% of the stroke group and 4% of controls. A PFO was TMD ϭ transmitral Doppler TTE ϭ transthoracic echocardiography diagnosed by saline contrast injection in 70% of stroke patients with an ASA and 75% of controls with an ASA. Of the 32 ASAs in this study, six had an associated Chiari INCIDENCE network. Notably, TTE identified only 12/32 (37.5%) of the ASAs. A biplane TEE study (29) found an ASA in Two autopsy studies determined PFO incidence, with the 28/355 (7.9%) of stroke patients and 8/363 (2.2%) of ϭ first (n 1,100) revealing a “probe” patent PFO (0.2 cm to controls. A PFO was found by saline contrast in 56% of 0.5 cm maximum dimension) in 29% and “pencil” patent ASA patients. No other possible source of embolism was ϭ PFO (0.6 cm to 1.0 cm) in 6% (27). The second study (n found in 24/28 (86%) of the stroke patients with an ASA. In 965) recorded a PFO incidence of 27.3%, with PFOs another study, if a mobile IAS or fossa ovalis measured varying in size from 1 mm to 19 mm (mean 4.9 mm). The Ն14 mm, there appeared to be an increased PFO incidence (35). These data indicate that an ASA occurs commonly in patients with unexplained stroke and is more frequently detected by TEE than by TTE. The Chiari network is a remnant of the septum spurium and the right valve of the sinus venosus resulting from incomplete resorption of these structures (36). Lace-like remnants of the original septum may persist and be evident as coarse or fine fibers in the RA. These fibers originate from either the Eustachian or Thebesian valve and have attachments to the upper wall of the RA or IAS. The Eustachian valve may also be relatively mobile and fenes- trated but does not attach to the upper wall of the right atrium or IAS and, therefore, should not be termed a “Chiari network.” The autopsy incidence of the Chiari network has been reported as 2% to 3% (37). A TEE study found a Chiari network in 29/1,436 (2%), whereas TTE identified only 8. A TEE-determined PFO was found in 24/29 (83%) of these, with more “intense” right-to-left shunting noted. An ASA was found by TEE in 33/1,436 (2.3%), with 7/33 having a coexistent Chiari network (31).

DIAGNOSIS OF A PFO Before the advent of echocardiography, diagnosis of a Figure 1. Longitudinal imaging transesophageal echocardiography in the right-to-left shunt was problematic. A PFO without a midupper esophagus. A large patent foramen ovale (PFO) is evident. The arrow illustrates how to measure the PFO width. LA ϭ left atrium; RA ϭ signiﬁcant resting right-to-left shunt is generally not asso- right atrium. ciated with any abnormality of the patient history, physical JACC Vol. 38, No. 3, 2001 Kerut et al. 615 September 2001:613–23 Patent Foramen Ovale: Size and Clinical Signiﬁcance

Figure 2. M-mode (left) and two-dimensional (right) transesophageal echocardiography longitudinal image in the midupper esophagus. The M-mode demonstrates marked interatrial septum (IAS) mobility, with an excursion Ͼ15 mm. By definition this is an atrial septem aneurysm. LA ϭ left atrium; RA ϭ right atrium; SVC ϭ superior vena cava. examination, electrocardiogram or chest radiograph. Heart eration of a dye dilution curve from a dichromatic earpiece catheterization is generally inadequate for diagnosis of PFO, densitometer was reported in patients with cryptogenic unless a right heart catheter crosses the IAS into the LA. stroke. A right-to-left shunt was detected in 24/59 (41%); On occasion, a transvenous pacemaker wire or pulmonary however, it was not compared with any other diagnostic artery catheter may inadvertently cross into the LA via an modality (39). atrial septal defect (ASD) or PFO. Dye dilution methods Echocardiographic techniques have emerged as the prin- are cumbersome and invasive (38). A relatively noninvasive ciple means for diagnosis and assessment of PFO (Table 1). method using peripherally injected indocyanine with gen- In particular, TTE, TEE and transcranial Doppler (TCD) are commonly used. Peripherally injected agitated saline may be utilized with TTE to diagnose a right-to-left shunt, but the yield is rather low compared with TEE. Transcra- nial Doppler has been used for PFO diagnosis by detection of air microemboli in a middle cerebral artery after peripheral injection of agitated saline. With TTE, the Valsalva maneuver may enhance color Doppler or agitated saline contrast detection of a right-to-left shunt, but the acoustic window may be lost during the maneuver. Transmitral Doppler (TMD) of mitral flow using peripheral injection of agitated saline was compared with TEE using both contrast and color Doppler for PFO detection (40). Transesophageal echocardiography detected PFO in 17/44 subjects (39%), with 16 detected by contrast and 1 only by color, the latter having a left-to-right shunt. Transthoracic echocardiography (TTE) was positive in only 12 of 17 positive TEE-detected PFOs, but all 16 contrast positive TEEs were positive by TMD. An additional positive TMD was not detected by TEE. A TMD “bubble score” and automated quantitative power score correlated with PFO size measured by TEE. Transthoracic echocardiography, TTE with native tissue Figure 3. Transesophageal echocardiography in the midupper esophagus at 105°. The arrow points to the mobile Chiari network. IAS ϭ interatrial harmonics (NTH) and TEE were performed with saline septum; LA ϭ left atrium; RA ϭ right atrium; SVC ϭ superior vena cava. contrast and the Valsalva maneuver in a consecutive group 616 Kerut et al. JACC Vol. 38, No. 3, 2001 Patent Foramen Ovale: Size and Clinical Significance September 2001:613–23

Table 1. Evaluation of Patients for Both Anatomical and or the Valsalva maneuver increase the sensitivity of TEE Functional PFO Size PFO detection and that contrast injections via the femoral TTE vein approach are superior to the antecubital route. A Anatomy Chiari network second study comparing the antecubital and femoral vein Interatrial septal mobility/atrial septal aneurysm approach used TEE and TCD in young patients (n ϭ 44) Diameter of fossa ovalis with unexplained neurologic events (Table 3) (43). Al- Color Doppler Evaluate IVC flow direction Right-to-left shunt though the incidence of PFO by antecubital injection was Left-to-right shunt much lower than that generally reported by other investi- Saline contrast (Quantitate bubbles in left heart) gators, contrast injections via the femoral approach had a Resting injection higher PFO detection rate. Valsalva injection A false-negative or false-positive TEE for PFO may TEE Anatomy Chiari network occur. A false-negative TEE may result from inadequate Interatrial septal mobility/atrial septal aneurysm visualization within the esophagus, elevated LA pressures Diameter of fossa ovalis preventing right-to-left passage of contrast (44,45), IVC- Evaluate fossa ovalis for ASD directed flow along the IAS preventing impingement of Measure slit between septum primum and antecubital bubbles against the IAS (42) (Fig. 4) or an septum secundum Resting and Valsalva improperly performed Valsalva maneuver. Color Doppler Evaluate IVC flow direction A TEE false-positive contrast study may occur with a Right-to-left shunt true ASD (Fig. 5) or a pulmonary arteriovenous shunt Left-to-right shunt (PAVS). Contrast that has a delayed appearance (Ͼ3 Measure slit between septum primum and cardiac cycles) within the LA after opacification of the RA septum secundum Resting and Valsalva may result from a PAVS in which contrast may also be Saline contrast (Timing of contrast arrival into LA) observed traversing one or more pulmonary veins (PVs). (Quantitate bubbles in LA) So-called “debris” entering the LA from PVs may result in Resting injection a false-positive TEE (46). Thought to be related to tran- Valsalva injection sient PV stasis, especially during Valsalva maneuver release, TCD Performed in conjunction with TTE and TEE flow was noted to enter the LA. This debris was usually ASD ϭ atrial septal defect; IVC ϭ inferior vena cava; LA ϭ left atrium; PFO ϭ patent foramen ovale; TCD ϭ transcranial Doppler; TEE ϭ transesophageal noted with high gain settings. echocardiography; TTE ϭ transthoracic echocardiography. PFO SIZE AND CLINICAL SIGNIFICANCE of patients referred to an echocardiography laboratory (n ϭ Autopsy correlation. Patients (n ϭ 35) were examined for 109) for TEE (41). Criteria for PFO detection included PFO with both TEE and at subsequent autopsy (47). contrast found in the LA within three cardiac cycles after Patent foramen ovale size was measured with TEE aided by RA opacification, and “significant” shunts were categorized color flow Doppler and by quantitating LA contrast using as Ͼ20 bubbles in the LA. Patent foramen ovale was peripheral venous injections. Nine PFOs found at autopsy identified by TTE in 9/109 subjects (8%), NTH in 21/109 had all been correctly identified ante mortem by color subjects (19%) and TEE in 21/109 subjects (19%). A Doppler TEE. Contrast TEE was negative in one subject “significant” PFO shunt was found in 8 subjects (7%) by described as having a long interatrial channel. Several TTE, 17 (16%) by NTH and 12 (11%) by TEE. Two patients with anatomically confirmed PFO had TEE de- “significant” shunts detected by NTH were not detected by tected left-to-right shunts. All had evidence of high LA TEE. Native tissue harmonics failed to detect four nonsig- pressures. Color flow Doppler slightly overestimated PFO nificant PFOs that were detected by TEE. The advantage of size as determined by autopsy. Left atrium contrast was NTH over TEE was explained by the difficulty patients had graded by the maximum number of microbubbles appearing in performing the Valsalva maneuver due to the presence of in a single frame and subsequently compared with autopsy an endoscope. findings (Table 4). Transesophageal echocardiography is currently consid- The platypnea-orthodeoxia syndrome. The platypnea- ered the reference standard for PFO diagnosis, allowing orthodeoxia syndrome describes both dyspnea (platypnea) direct imaging of the IAS and saline contrast shunting and arterial desaturation in the upright position with im- through a PFO. If color flow Doppler or peripheral saline contrast injection detect right-to-left flow during normal Table 2. respiration, this is termed a resting PFO. Enhanced right- TTE PFO Detection by Site of Injection (42) to-left flow may be noted upon release of the Valsalva Spontaneous maneuver. Respiration Cough Valsalva Total A methodologic study (n ϭ 70) in patients with arterial Arm 1/70 (1%) 4/70 (6%) 9/70 (13%) 9/70 (13%) embolism compared antecubital vein to femoral vein con- Femoral vein 7/70 (10%) 14/70 (20%) 22/70 (31%) 22/70 (33%) trast injection (Table 2) (42). These data suggest that cough PFO ϭ patent foramen ovale; TTE ϭ transthoracic echocardiography. JACC Vol. 38, No. 3, 2001 Kerut et al. 617 September 2001:613–23 Patent Foramen Ovale: Size and Clinical Significance

Table 3. TTE and TCD PFO Detection by Site of Injection (43) Normal Respiration Valsalva Maneuver TEE TCD TEE TCD Arm 5/44 (11.4%) 2/44 (4.5%) 8/44 (18%) 6/44 (13.6%) Femoral vein 17/44 (38.6%) 16/44 (36%) 22/44 (50%) 22/44 (50%)

PFO ϭ patent foramen ovale; TCD ϭ transcranial Doppler; TEE ϭ transesophageal echocardiography; TTE ϭ transthoracic echocardiography. provement in the supine position (orthodeoxia) (48,49). TCD that were not detected by TEE. All six PFOs not The syndrome occurs with an intracardiac or intrapulmo- diagnosed by TCD measured Յ2 mm by TEE, suggesting nary shunt and often with some form of lung disease. The that TCD may only miss small defects. post-pneumonectomy patient (usually the right lung Another study used TCD and monoplane TEE to [18,20]) may develop symptoms many months after the compare 63 controls, 33 patients with stroke that was surgical procedure (17,18,50). A ventilation-perfusion lung explained or “clarified” and 41 patients with cryptogenic scan, obtained for dyspnea and hypoxia, will reveal systemic stroke (52). Shunts were sized by quantitating the number uptake of isotope (21,22,50). of TCD detected spikes or TEE visualized microbubbles Diagnosis of a PFO with platypnea-orthodeoxia has been (Table 5). The TEE score distribution was relatively similar made with a tilt table and saline contrast TTE (17) and among the three groups. By TCD, however, the normal TEE (24). A typical patient history with a significant drop group had a minimal shunt score, and the cryptogenic group in oxygen saturation while in the upright position, along had the largest shunt score, suggesting that TCD may find with a large PFO by TEE found as the only source of a predominately clinically significant shunts. shunt, has been considered diagnostic (22). A TEE study with peripheral contrast injection (18 for Cryptogenic stroke. Efforts to determine anatomical size source of embolism) detected 34 resting PFOs (53). These and functional significance of a PFO have been expended were divided into a small shunt subgroup (n ϭ 18) with Ն3 most often in patients with embolic stroke (Fig. 6). Patients and Ͻ20 LA microbubbles and a large shunt subgroup (n ϭ (n ϭ 49) with stroke were evaluated by TTE, TEE and 16) with 20. All patients were treated with aspirin or TCD (51). Transesophageal echocardiography diagnosed warfarin and were followed for 21 months. Of the large PFO in 19/49 subjects (39%). Transthoracic echocardiog- shunt subgroup (5/16), 31% had an ischemic event, whereas raphy was positive in 9/49 subjects (18%) and TCD in there were none in the small shunt subgroup. 13/49 subjects (27%). No PFOs were detected by TTE and Patients Ͻ60 years of age with unexplained neurologic

Figure 4. Longitudinal transesophageal echocardiography imaging in the midupper esophagus with color Doppler. Blue color is inferior vena cava (IVC) flow directed along the interatrial septum (IAS). Prominent IVC flow may prevent contrast injected in an upper extremity from opacifying the septum along the IAS. 618 Kerut et al. JACC Vol. 38, No. 3, 2001 Patent Foramen Ovale: Size and Clinical Significance September 2001:613–23

Figure 5. Longitudinal transesophageal echocardiography (TEE) imaging in the midupper esophagus. A 32-year-old male presented with an embolic stroke. He was found to have occult thrombi in both calf veins. Transthoracic echocardiography with peripheral saline contrast during normal respiration and Valsalva were negative for a right-to-left shunt. (A) TEE revealed a small restrictive secundum atrial septal defect (ASD) (arrow). Note how this appearance is different from that of a patent foramen ovale. (B) Color Doppler demonstrates a left-to-right shunt with a color mosaic pattern from the shunt in the right atrium (RA). The ASD was subsequently surgically repaired. LA ϭ left atrium; SVC ϭ superior vena cava. events were compared with normal controls by multiplane compared with controls (2%), and 19 of the 21 patients with TEE (54) (Table 6). A PFO was diagnosed using peripheral an ASA had a large PFO (5 Ϯ 2 mm). Peripheral saline contrast injection and by direct visualization of the IAS. contrast shunts were quantitated as minimal (few bubbles in Shunt size was measured during normal respiration and LA), moderate (“cloud” in LA) or severe (intense opacifi- with the Valsalva. Patients with neurologic events had a cation). Although contrast identified patients with PFO, significantly higher PFO incidence and size compared with there was no significant correlation between measured size controls. The large PFO size was found more commonly in and contrast shunt severity. the multiple cerebrovascular accident subgroup. An ASA Underwater diving and DCS. Decompression sickness was more common in the neurologic event group (12%) results from the formation of bubbles in body fluids as JACC Vol. 38, No. 3, 2001 Kerut et al. 619 September 2001:613–23 Patent Foramen Ovale: Size and Clinical Significance

Table 4. TTE PFO Size Comparison by Contrast and of surfacing), “probable” (significant neurologic dysfunction Measured Size (47) but with a normal neurologic examination within 48 h) and Grade (Number of Number of “definite” (significant neurologic dysfunction and an abnor- LA Microbubbles) Patients PFO Autopsy Size mal neurologic examination). A TTE was considered pos- I (1–5) 1 Յ2mm itive if any bubble appeared in the left heart. By TEE, Ն3 2 Ͼ2mmtoϽ10 mm bubbles in the LA were required for an examination to be II (6–25) 2 Ͼ2mmtoϽ10 mm classified as positive, and a study was considered “strongly Ͼ Ն III ( 25) 3 10 mm positive” if Ͼ5 bubbles were seen in a single video frame. LA ϭ left atrium; PFO ϭ patent foramen ovale; TTE ϭ transthoracic echocardio- With normal respirations, no difference in PFO detection graphy. rate was noted. By Valsalva maneuver (Table 7), TCD had ambient pressure is reduced. Arterial gas embolism is usually the best positive and negative predictive value for detection due to air emboli arising from pulmonary barotrauma or of PFO in divers with DCS and was felt to be better than from bubbles in the systemic venous circulation entering the TEE for delineation of a clinically significant shunt. In arterial system. An AGE through an ASD in a scuba diver addition, all controls and divers with a strongly positive was first reported in 1986 (55). Two TTE studies then TEE had a positive TCD, and all with a positive TCD had demonstrated a relationship between PFO and otherwise a strongly positive TEE. unexplained DCS in divers (7,8). In the first study (n ϭ 30), Asymptomatic sport divers (n ϭ 87) were evaluated with 11/30 subjects (37%) had a resting right-to-left shunt (8). bilateral TCD and a sonicated contrast medium injected Of the 30 divers, 18 had type II DCS (i.e., serious disease into an antecubital vein (56). A shunt was considered to be including neurologic or cardiopulmonary dysfunction as of low relevance if Ͻ20 signals occurred after the Valsalva opposed to type I DCS in which only musculoskeletal pain maneuver and high relevance if Ն20 signals were noted. is present), and all 11 divers with a resting shunt were in this Magnetic resonance imaging (MRI) brain scans were also group. None of the other 12 divers, who had type I DCS, performed (Table 8). All MRI lesions were 2 mm to 3 mm had a shunt. in size and located in the distribution of the anterior cerebral In the second study, antecubital injection was performed circulation. Of note, the three divers with many lesions had during normal respiration and “modified” Valsalva (7). The high relevance signals by TCD. incidence of PFO was 23/53 (43%), but further quantifica- A multiplane TEE study with peripheral agitated con- tion, by counting bubbles in the left ventricle, was not trast injection was performed in divers with DCS (n ϭ 37) reproducible from one injection to the next. and control divers (n ϭ 36) (57). Shunts were sized during A TTE, TEE and TCD study compared commercial Valsalva by quantitating microbubbles in the LA (Table 9). divers (n ϭ 26) with control subjects (n ϭ 30) (9). Divers Of the 37 DCS divers, 20 had type II illness and, of these, were categorized as “possible” DCS (transient nonspecific 14/20 (70%) had a grade II PFO. neurologic symptoms and a normal examination within 48 h From these studies, it appears that a PFO is associated

Figure 6. Longitudinal transesophageal echocardiographic imaging in the midupper esophagus. Arrows point to a thrombus wedged through a patent foramen ovale and lodged in both the right atrium (RA) and left atrium (LA). SVC ϭ superior vena cava. 620 Kerut et al. JACC Vol. 38, No. 3, 2001 Patent Foramen Ovale: Size and Clinical Signiﬁcance September 2001:613–23

Table 5. Contrast Comparison of TEE and TCD for PFO Shunt Size (52) (41 ؍ Cryptogenic CVA (n (33 ؍ Clariﬁed CVA (n (63 ؍ Normal (n TEE 27/63 11/33 27/41 I II III I II III I II III

8/27 15/27 4/27 4/11 5/11 2/11 6/27 8/27 13/27 30% 55% 15% 36% 46% 18% 22% 30% 48% TCD 28/63 11/33 22/41 I II III I II III I II III

22/28 5/28 1/28 3/11 5/11 3/11 2/22 6/22 14/22 79% 18% 3% 28% 45% 27% 9% 27% 64%

Transesophageal echocardiography microbubbles or TCD spikes were graded as score I: Ͻ10, score II: 10–50 and score III: Ͼ50. CVA ϭ cerebrovascular accident; PFO ϭ patent foramen ovale; TCD ϭ transcranial Doppler; TEE ϭ transesophageal echocardiography. with unexplained DCS. A resting PFO and one found by serious DCS is reported to be no different from case controls TCD seem to be associated. Recreational divers without a (10,62). However, these data may not be applicable to all clinical history of DCS, but with Ն20 signals detected by types of hypobaric pressure profiles. Specifically, the pres- TCD on a single peripheral injection seemed to have a sure profile used in physiologic training exercises involves a higher likelihood of lesions in the anterior cerebral artery short exposure (about 15 min) to a modest altitude (about territory, as detected by MRI brain scan. 7,600 m) with subjects at rest. A long hypobaric exposure, DCS in high altitude aviators and astronauts. As the during which heavy work is performed, may be distinct from performance of aircraft has evolved over the course of the these physiologic training profiles and may render a PFO last century, both the achieved altitude and rate of climb more important. Presently, the relative risk for altitude DCS have increased. Therefore, hypobaric DCS became a prob- conferred by the presence of a PFO is not well character- lem, particularly evidenced during high-altitude bomber ized. flights in World War II. Several fatal cases occurred in A large decompression stress is associated with extrave- which the pressure profile and postmortem findings were hicular activity (EVA) from spacecraft. During the era of documented in some detail (58), with a PFO noted in 5/19 the International Space Station (ISS) construction and cases. However, no comment was made on the integrity of maintenance, two types of “space suits” were utilized, the the IAS in the reports of several of the remaining autopsies. Extravehicular Mobility Unit (EMU) developed by the U.S. Nevertheless, it is clear that a PFO is not invariably present and the Russian Orlan suit. To maximize space suit flexi- in all serious cases of altitude DCS. bility while maintaining an adequate alveolar oxygen partial Altitude exposures occur yearly in hypobaric chambers pressure, the suits contain nearly pure oxygen at subatmo- utilized by the armed forces of several nations to familiarize spheric pressures. The absolute pressure within the EMU is aircrew with the physiologic phenomena associated with 29.6 kPa, while in the Orlan suit it is 38.6 kPa. Under most hypoxia and rapid decompression. Altitude DCS is an conditions, the cabin atmosphere within all spacecraft uti- uncommon but persistent problem in these exposures; an lized in this program (Space Shuttle, proposed “lifeboat” overall incidence of about 0.1% is experienced by the U.S. spacecraft, the Soyuz spacecraft and ISS itself) approximates Armed Forces, with about 10% of these cases manifesting sea-level conditions with a composition of 21% oxygen, neurologic or other serious symptoms (59–61). balance nitrogen and trace gases, at an absolute pressure of The contribution of a PFO to these cases of DCS is not 101 kPa. Consequently, when a crew member transitions clear, but the prevalence of PFO in U.S. Navy aviators with from a spacecraft cabin into a space suit for EVA, there is a substantial decompression. If this decompression were per- Table 6. Transesophageal Echocardiography Measured PFO formed without specific preparation, serious DCS would be Size in Otherwise Unexplained TIA, Single CVA and Multiple CVAs (54) virtually inevitable. To reduce risk of DCS during EVA, Table 7. TTE, TEE and TCD in Controls and Divers With (121 ؍ CNS Event (n DCS (9) Control TIA 1 CVA >2 CVAs Control Probable/Definite All Divers (20 ؍ n) (46 ؍ n) (55 ؍ n) (123 ؍ n) (26 ؍ n) (15 ؍ n) (30 ؍ n) ϩPFO 38 (31%) 33 (60%) 37 (80%) 17 (85%) PFO size (mm) 2 Ϯ 13Ϯ 14Ϯ 25Ϯ 2 TTE 5 (17%) 3 (20%) 8 (31%) PFO 1–4 mm 28 (51%) 24 (52%) 4 (20%) TEE 14 (47%) 9 (60%) 15 (58%) Ͼ4 mm 5 (9%) 13 (28%) 13 (65%) TCD 7 (23%) 7 (47%) 13 (50%)

CNS ϭ central nervous system; CVA ϭ cerebrovascular accident; PFO ϭ patent DCS ϭ decompression sickness; TCD ϭ transcranial Doppler; TEE ϭ transesoph- foramen ovale; TIA ϭ transient ischemic attack. ageal echocardiography; TTE ϭ transthoracic echocardiography. JACC Vol. 38, No. 3, 2001 Kerut et al. 621 September 2001:613–23 Patent Foramen Ovale: Size and Clinical Signiﬁcance

Table 8. TCD Detected Shunts Compared With MRI Brain during normal respiration and the Valsalva maneuver, with Lesions in Asymptomatic Sports Divers (56) attention to proper timing during the Valsalva maneuver. It TCD Shunt Number of Divers With appears that femoral venous injection of contrast, as com- (Microbubbles) Divers MRI Lesions pared with antecubital injection, increases the incidence of None 62 7 (1 lesion in 7 divers) right-to-left shunt detection, especially during TTE studies. Low (Ͻ20) 12 1 (1 lesion in 1 diver) During a TEE study, observation of agitated contrast High (Ն20) 13 3 (39 lesions in 3 divers) adjacent to the RA-side of the IAS after antecubital contrast MRI ϭ magnetic resonance imaging; TCD ϭ transcranial Doppler. injection may be helpful in that the antecubital injection probably is accurate. several protocols are utilized that include reductions in cabin Individuals who have PFO can be subgrouped according pressure or periods of oxygen breathing before decompres- to many PFO characteristics (large vs. small PFO, provoked sion. vs. unprovoked trans-septal gas passage, associated anatom- Human terrestrial hypobaric chamber studies indicate ical structures). Which methodologic techniques for detec- that significant gas phase evolution takes place in body tion of those characteristics that are clinically important in tissues with all decompression protocols. In particular, assessing risk of embolic stroke, diver’s DCS and altitude monitoring of the pulmonary artery with Doppler ultra- DCS is not yet entirely clear. The size of PFO probably sonography reveals that heavy burdens of circulating gaseous matters in determining risk, as does the ease with which emboli are present in from 6% to 39% of subjects, depend- trans-septal passage of emboli occurs in relation to intratho- ing on the protocol used (63–65) (J. Conkin, personal racic pressure excursions. Patent foramen ovales are very communication, 2000). In these studies, serious DCS has common, and not all PFOs are the same. The challenge that been encountered, including unusual forms such as massive remains is to determine which PFOs and clinical contexts cutaneous “marbling,” orthostatic instability and severe ce- confer an increased risk of significant disease. rebral dysfunction ([66], J. Conkin, personal communication, 2000). A PFO screening was performed after four cases of Reprint requests and correspondence: Dr. Thomas D. Giles, serious DCS resulting from EVA-like simulations. A rest- Louisiana State University Health Sciences Center, 1542 Tulane Avenue, Room 331E, New Orleans, Louisiana 70112-2822. ing PFO was detected by contrast TTE in three (J. E-mail: [email protected]. Waligora, personal communication, 2000). Conclusions. Patent foramen ovale has been implicated in platypnea-orthodeoxia, embolic stroke and paradoxical gas embolism in DCS. Few studies have attempted to quanti- REFERENCES tate PFO size. Current literature suggests that, when 1. Webster MW, Chancellor AM, Smith HJ, et al. Patent foramen ovale evaluating a patient for source of embolism, the size of any in young stroke patients. Lancet 1988;2:11–2. detected PFO should be considered since the risk of 2. Jones HR, Jr, Caplan LR, Come PC, et al. Cerebral emboli of trans-septal passage of emboli seems to be greater with paradoxical origin. Ann Neurol 1983;13:314–9. 3. Lechat P, Mas JL, Lascault G, et al. Prevalence of patent foramen larger lesions. ovale in patients with stroke. N Engl J Med 1988;318:1148–52. During TTE and TEE examinations for PFO, one 4. Ranoux D, Cohen A, Cabanes L, et al. Patent foramen ovale: is stroke should evaluate for the presence of an ASA and a Chiari due to paradoxical embolism? Stroke 1993;24:31–4. 5. Hausmann D, Mugge A, Becht I, Daniel WG. Diagnosis of patent network because of an association with significant PFOs. foramen ovale by transesophageal echocardiography and association M-mode of the IAS helps facilitate measurement of septal with cerebral and peripheral embolic events. Am J Cardiol 1992;70: excursion. 668–72. 6. Kasper W, Geibel A, Tiede N, Just H. Patent foramen ovale in Color flow Doppler by TTE and TEE should be per- patients with hemodynamically significant pulmonary embolism. Lan- formed to evaluate IVC flow direction into the RA, exam- cet 1992;340:561–4. ining in particular if IVC flow runs along and, therefore, 7. Wilmshurst PT, Byrne JC, Webb-Peploe MM. Relation between interatrial shunts and decompression sickness in divers. Lancet 1989; potentially across, the IAS. Color flow Doppler should be 2:1302–6. used to look for both right-to-left and left-to-right shunting 8. Moon RE, Camporesi EM, Kisslo JA. Patent foramen ovale and during normal respiration and after the Valsalva maneuver. decompression sickness in divers. Lancet 1989;1:513–4. 9. Kerut EK, Truax WD, Borreson TE, et al. Detection of right to left Agitated saline contrast injection should be performed shunts in decompression sickness in divers. Am J Cardiol 1997;79: 377–8. Table 9. TEE Detected Shunts and Shunt Size in Divers With 10. Gallagher DL, Hopkins EW, Clark JB, Hawley TA. U.S. Navy and Without DCS (57) experience with type II decompression sickness and the association with patent foramen ovale. Aviat Space Environ Med 1996;67:712. Grade II PFO ؉ > 11. Bendrick GA, Ainscough MJ, Pilmanis AA, Bisson RU. Prevalence of Divers PFO ( 20 Microbubbles) decompression sickness among U-2 pilots. Aviat Space Environ Med DCS 22/37 (59.5%) 19/37 (51.3%) 1996;67:199–206. 12. Conkin J, Bedahl SR, Van Liew HD. A computerized databank of Controls 13/36 (36.1%) 9/36 (25.0%) decompression sickness incidence in altitude chambers. Aviat Space DCS ϭ decompression sickness; PFO ϭ patent foramen ovale; TEE ϭ transesopha- Environ Med 1992;63:819–24. geal echocardiography. 13. Pilmanis AA, Meissner FW, Olson RM. Left ventricular gas emboli in 622 Kerut et al. JACC Vol. 38, No. 3, 2001 Patent Foramen Ovale: Size and Clinical Significance September 2001:613–23

six cases of altitude-induced decompression sickness. Aviat Space for detection of patent foramen ovale. Acta Neurol Scand 1998;97: Environ Med 1996:67;1092–6. 231–6. 14. Thompson RC, Finck SJ, Leventhal JP, Safford RE. Right-to-left 40. Kerr AJ, Buck T, Chia K, et al. Transmitral Doppler: a new shunt across a patent foramen ovale caused by cardiac tamponade: transthoracic contrast method for patent foramen ovale detection and diagnosis by transesophageal echocardiography. Mayo Clin Proc quantification. J Am Coll Cardiol 2000;36:1959–66. 1991;66:391–4. 41. Camp GV, Franken P, Melis P, et al. Comparison of transthoracic 15. Bansal RC, Marsa RJ, Holland D, Beehler C, Gold PM. Severe echocardiography with second harmonic imaging with transesophageal hypoxemia due to shunting through a patent foramen ovale: a echocardiography in the detection of right to left shunts. Am J Cardiol correctable complication of right ventricular infarction. J Am Coll 2000;86:1284–7. Cardiol 1985;5:188–92. 42. Gin KG, Huckell VF, Pollick C. Femoral vein delivery of contrast 16. Begin R, Gervais L, Bureau MA. Patent foramen ovale and hypoxemia medium enhances transthoracic echocardiographic detection of patent in chronic obstructive pulmonary disease. Eur J Respir Dis 1981;62: foramen ovale. J Am Coll Cardiol 1993;22:1994–2000. 373–5. 43. Hamann GF, Schatzer-Klotz D, Frohlig G, et al. Femoral injection of 17. Seward JB, Hayes DL, Smith HC, et al. Platypnea-orthodeoxia: echo contrast medium may increase the sensitivity of testing for a clinical profile, diagnostic workup, management, and report of seven patent foramen ovale. Neurology 1998;50:1423–8. cases. Mayo Clin Proc 1984;59:221–31. 44. Siostrzonek P, Lang W, Zangeneh M, et al. Significance of left-sided 18. Smeenk FW, Postmus PE. Interatrial right-to-left shunting develop- heart disease for the detection of patent foramen ovale by transesoph- ing after pulmonary resection in the absence of elevated right-sided ageal contrast echocardiography. J Am Coll Cardiol 1992;19:1192–6. heart pressures: review of the literature. Chest 1993;103:528–31. 45. Movsowitz HD, Movsowitz C, Jacobs LE, Kotler MN. Negative 19. Godart F, Rey C, Prat A, et al. Atrial right-to-left shunting causing air-contrast test does not exclude the presence of patent foramen ovale severe hypoxaemia despite normal right-sided pressures: report of 11 by transesophageal echocardiography. Am Heart J 1993;126:1031–2. consecutive cases corrected by percutaneous closure. Eur Heart J 46. Camp GV, Cosyns B, Vandenbossche J. Non-smoke spontaneous 2000;21:483–9. contrast in left atrium intensified by respiratory manoeuvres: a new 20. Bakris NC, Siddiqi AJ, Fraser CD, Jr, Mehta AC. Right-to-left transesophageal echocardiographic observation. Br Heart J 1994;72: interatrial shunt after pneumonectomy. Ann Thorac Surg 1997;63: 446–51. 198–201. 47. Schneider B, Zienkiewicz T, Jansen V, et al. Diagnosis of patent 21. Murray KD, Kalanges LK, Weiland JE, et al. Platypnea-orthodeoxia: foramen ovale by transesophageal echocardiography and correlation an unusual indication for surgical closure of a patent foramen ovale. with autopsy findings. Am J Cardiol 1996;77:1202–9. J Card Surg 1991;6:62–7. 48. Altman M, Robin ED. Platypnea (diffuse zone I phenomenon?). 22. Dear WE, Chen P, Barasch E, et al. Sixty-eight-year-old woman with N Engl J Med 1969;281:1347–8. intermittent hypoxemia. Circulation 1995;91:2284–9. 49. Robin ED, Laman D, Horn BR, Theodore J. Platypnea related to 23. Davidson A, Chandrasekaran K, Guida L, et al. Enhancement of orthodeoxia caused by true vascular lung shunts. N Engl J Med hypoxemia by atrial shunting in cystic fibrosis. Chest 1990;98:543–5. 1976;294:941–3. 24. Herregods MC, Timmermans C, Frans E, et al. Diagnostic value of 50. Durand E, Bussy E, Gaillard JF. Lung scintigraphy in postpneumo- transesophageal echocardiography in platypnea. J Am Soc Echocar- nectomy dyspnea due to a right-to-left shunt. J Nucl Med 1997;38: diogr 1993;6:624–7. 1812–5. 25. Moore KL. The Developing Human: Clinically Oriented Embryol- 51. Di Tullio M, Sacco RL, Venketasubramanian N, et al. Comparison of ogy. 2nd ed. Philadelphia, PA: W.B. Saunders, 1977. diagnostic techniques for the detection of a patent foramen ovale in 26. Heymann MA, Creasy RK, Rudolph AM. Quantitation of blood flow stroke patients. Stroke 1993;24:1020–4. patterns in the foetal lamb in utero. In: Proceedings of the Sir Joseph 52. Job FP, Ringelstein EB, Grafen Y, et al. Comparison of transcranial Barcroft Centenary Symposium: Foetal and Neonatal Physiology. contrast Doppler sonography and transesophageal contrast echocardi- Cambridge, MA: Cambridge University Press, 1973:129–35. ography for the detection of patent foramen ovale in young stroke 27. Thompson T, Evans W. Paradoxical embolism. Q J Med 1930;23: patients. Am J Cardiol 1994;74:381–4. 135–52. 53. Stone DA, Godard J, Corretti MC, et al. Patent foramen ovale: 28. Hagen PT, Scholz DG, Edwards WD. Incidence and size of patent association between the degree of shunt by contrast transesophageal foramen ovale during the first 10 decades of life: an autopsy study of echocardiography and the risk of future ischemic neurologic events. 965 normal hearts. Mayo Clin Proc 1984;59:17–20. Am Heart J 1998;131:158–61. 29. Agmon Y, Khandheria BD, Meissner I, et al. Frequency of atrial septal 54. Schuchlenz HW, Weihs W, Horner S, Quehenberger F. The associ- aneurysms in patients with cerebral ischemic events. Circulation ation between the diameter of a patent foramen ovale and the risk of 1999;99:1942–4. embolic cerebrovascular events. Am J Med 2000;109:456–62. 30. Pearson AC, Nagelhout D, Castello R, et al. Atrial septal aneurysm 55. Wilmshurst PT, Ellis BG, Jenkins BS. Paradoxical air embolism in a and stroke: a transesophageal echocardiographic study. J Am Coll scuba diver with an atrial septal defect. BMJ 1986;293:1277. Cardiol 1991;18:1223–9. 56. Knauth M, Rica S, Pohimann S, et al. Cohort study of multiple brain 31. Schneider B, Hofmann T, Justen MH, Meinertz T. Chiari’s network: lesions in sport divers: role of a patent foramen ovale. BMJ 1997:314; normal anatomic variant or risk factor for arterial embolic events? J Am 701–5. Coll Cardiol 1995;26:203–10. 57. Germonpre P, Dendale P, Unger P, Balestra C. Patent foramen ovale 32. Silver MD, Dorsey JS. Aneurysms of the septum primum in adults. and decompression sickness in sports divers. J Appl Physiol 1998;84: Arch Pathol Lab Med 1978;102:62–5. 1622–6. 33. Olivares-Reyes A, Chan S, Lazar EJ, et al. Atrial septal aneurysm: a 58. Dixon JP. Death from altitude-induced decompression sickness: major new classification in two hundred five adults. J Am Soc Echocardiogr pathophysiologic factors. In: Pilmanis AA, editor. Proceedings of the 1997;10:644–56. 1990 Hypobaric Decompression Sickness Workshop. Brooks Air 34. Hanley PC, Tajik AJ, Hynes JK, et al. Diagnosis and classification of Force Base, TX: Air Force Systems Command, 1992:97–105. atrial septal aneurysm by two-dimensional echocardiography: report of 59. Bason R, Yacavone D. Decompression sickness: U.S. Navy altitude 80 consecutive cases. J Am Coll Cardiol 1985;6:1370–82. chamber experience 1 October 1981 to 30 September 1988. Aviat 35. Louie EK, Konstadt SN, Tao TLK, Scanlon PJ. Transesophageal Space Environ Med 1991;62:1180–4. echocardiographic diagnosis of patent foramen ovale in adults without 60. Weien RW. Altitude decompression sickness: the US Army experi- prior stroke. Circulation 1991;84:II–451. ence. In: Pilmanis AA, editor. Proceedings of the 1990 Hypobaric 36. Corliss CE. Patten’s Human Embryology: Elements of Clinical Decompression Sickness Workshop. AL-SR-1992-0005. Brooks Air Development. New York, NY: McGraw-Hill, 1976. Force, TX: Air Force Systems Command, 1992:379–83. 37. Chiari H. Uber netzbildungen im rechten vorhof des herzens Beitr 61. Baumgartner N, Weien RW. Decompression sickness due to USAF Pathol Anat 1897;22:1–10. altitude chamber exposure (1985 to 1987). In: Pilmanis AA, editor. 38. Cheng TO. Paradoxical embolism: a diagnostic challenge and its Proceedings of the 1990 Hypobaric Decompression Sickness Work- detection during life. Circulation 1976;53:565–8. shop. AL-SR-1992-0005. Brooks Air Force Base, TX: Air Force 39. Karttunen V, Ventila M, Hillbom M, et al. Dye dilution and oximetry Systems Command, 1992:363–376. JACC Vol. 38, No. 3, 2001 Kerut et al. 623 September 2001:613–23 Patent Foramen Ovale: Size and Clinical Significance

62. Clark JB, Hayes GB. Patent foramen ovale and type II altitude phase formation in simulated microgravity. NASA technical memo- decompression sickness (abstr). Aviat Space Environ Med 1991;62: randum 104762. Houston, TX: National Aeronautics and Space 445. Administration, 1993. 63. Waligora JM, Horrigan D, Jr, Conkin J, Hadley AT, III. Verification 65. Waligora JM, Horrigan DJ, Jr, Conkin J. The effect of extended O2 of an altitude decompression sickness prevention protocol for shuttle prebreathing on altitude decompression sickness and venous gas operation utilizing a 10.2 psi pressure stage. NASA technical memo- bubbles. Aviat Space Environ Med 1987;58:A110–12. randum 58259. Houston, TX: National Aeronautics and Space Ad- 66. Powell MR, Norfleet WT, Kumar KV, Butler BD. Patent foramen ministration, 1984. ovale and hypobaric decompression. Aviat Space Environ Med 1995; 64. Powell MR, Waligora JM, Norfleet W, Kumar KV. Project Argo-gas 66:273–5.