Patent Foramen Ovale and High-Altitude Pulmonary Edema FREE

A patent foramen ovale (PFO) is present in approximately 25% of the general population, but a higher prevalence has been reported in disease states known to be associated with pulmonary hypertension such as chronic obstructive pulmonary disease and sleep apnea syndrome.^1- 2 These observations could suggest that under conditions associated with elevated pulmonary vascular and right-sided cardiac pressures, some foramen ovale may reopen, implying a functional rather than an anatomical closing of these foramen. A reopened foramen ovale may contribute to systemic arterial oxygen desaturation from right-to-left intracardiac shunting under any condition that results in a higher right than left atrial pressure.^3- 5

Exaggerated pulmonary hypertension is a hallmark of high-altitude pulmonary edema (HAPE) and plays an important role in its pathogenesis.^6- 8 Moreover, altitude-induced hypoxemia is more pronounced in HAPE-susceptible than HAPE-resistant individuals prior to onset of edema.^9- 10 The underlying mechanisms are incompletely understood. In a preliminary report, Levine et al¹¹ demonstrated right-to-left intracardiac shunting across a PFO in a patient with HAPE and suggested that in persons with PFO a vicious cycle may be operational at high altitude wherein hypoxic pulmonary hypertension initiates right-to-left shunting, which in turn aggravates hypoxemia.

We hypothesized that the frequency of PFO is greater in HAPE-susceptible than in HAPE-resistant individuals and, if so, may contribute to more severe hypoxemia at high altitude. To test this hypothesis, we searched for PFO by transesophageal echocardiography (TEE), estimated pulmonary artery pressure by Doppler echocardiography, and measured arterial oxygen saturation in 16 HAPE-susceptible and 19 HAPE-resistant control participants at low and high altitude.

Thirty-five healthy mountaineers were included in the study. Sixteen participants (10 men and 6 women, mean [SD] age, 43 [12] years) had previously developed at least 1 episode of clinically and radiographically documented HAPE (HAPE-susceptible [HAPE-S] group). The remaining 19 participants (17 men, 2 women, mean [SD] age, 39 [12] years) were known to be resistant to HAPE (repeated alpine-style climbing to peaks above 4000 m without symptoms of pulmonary edema; HAPE-resistant [HAPE-R] group). The study was approved by the institutional review board on human investigation of the University Hospital of Lausanne, and all participants provided written informed consent.

Accompanied by a professional mountain guide, the participants ascended in groups of 2 to 4 from 1130 m to 4559 m within 24 hours. The ascent consisted of a transport by cable car to an altitude of 3200 m; a 1½ hour climb to an altitude of 3611 m, where the participants stayed overnight; and on the next day, a 4- to 5-hour climb to the high-altitude research laboratory at Capanna Regina Margherita (4559 m). The participants then spent 2 days and 2 nights at this hut. Transthoracic echocardiography (TTE) in combination with contrast TEE was performed during the first full day (Table 1) spent at the hut. None of the participants was receiving oxygen or taking any medications before the echocardiographic studies. Arterial oxygen saturation was assessed with a finger pulse oximeter. All measurements were repeated at low altitude (550 m) within 6 months after the high-altitude study.

At high altitude, posteroanterior chest radiographs were obtained for all participants, either at the time when symptoms of HAPE first occurred or before descent. The radiographs were analyzed according to previously described criteria⁸ by a radiologist who was unaware of the participants' clinical history.

Transthoracic echocardiography and TEE were performed with a real-time, phased array sector scanner (model Sonos 5000, Hewlett-Packard, Andover, Mass, and Siemens Acuson Sequoia C256, Mountain View, Calif) with an integrated color Doppler system and a transducer-containing crystal set for imaging (3.5 MHz, 7 MHz for the TEE omniplane probe) and for continuous-wave Doppler recording (1.9 MHz). The recordings were stored on videotape for offline analysis by 2 investigators who were unaware of the participants' clinical history.

Transesophageal echocardiography in combination with injection of 2 mL of echo contrast medium (ad hoc sonicated mixture of 0.2 mL air plus 1.8 mL plasma expander [Physiogel, Pharmacy of the University Hospital, Inselspital, Bern, Switzerland]) into the right antecubital vein and the Valsalva maneuver were used for the search for PFO in at least 2 orthogonal image planes.^4,12- 13 The contrast bolus injection was given at the start of the strain phase of the Valsalva maneuver. The maneuver was considered successful when immediately after the release of the strain phase (lasting 5 to 10 seconds) a leftward deviation of the interatrial septum in the fossa ovalis region related to short right atrial preload and pressure increase was observed. The diagnosis of PFO required the crossing of bubbles from the right to the left atrium (Figure) within 4 heart beats following the release of the strain (see video here).

The size of the PFO was graded semiquantitatively according to the maximum number of bubbles crossing into the left atrium, using a score of 0 to 3: score 1, less than 6 bubbles; score 2, 6 to 20 bubbles; score 3, more than 20 bubbles.^13- 14 At high altitude, TEE could not be performed in 9 participants, so contrast TTE was performed instead (Table 1). At low altitude, 2 participants (Table 1) refused TEE. We did not encounter any complication in performing TEE.

Systolic pulmonary artery pressure was determined by TTE.¹⁵ After tricuspid regurgitation had been localized with Doppler color flow imaging, the peak flow velocity of the transtricuspid jet (v_TR) was measured using continuous-wave Doppler and the pressure gradient between the right ventricle and the right atrium was calculated using the modified Bernoulli equation: Δp_RV-RA = 4 (v_TR)² (Δp, pressure difference; RV, right ventricle; RA, right atrium).² All reported values represent the mean of at least 3 measurements.

Statistical analysis using SAS version 8.2 (SAS Institute Inc, Cary, NC) was performed with paired and unpaired t tests for the comparisons within and between the groups, and the χ² and Fisher exact tests for nonparametric variables (presence or absence of a PFO and HAPE). The McNemar test was used for paired comparisons of nonparametric variables (presence or absence of a PFO at low vs high altitude). For comparisons between HAPE-S participants with large PFOs and HAPE-S participants with small or no PFOs, the Wilcoxon rank-sum test was used. Relations between variables were analyzed by calculating Pearson product-moment correlation coefficients or analysis of variance factorial analysis, as appropriate. Statistical significance was defined at a P value of <.05. Data are expressed as mean (SD).

At low altitude, the systolic transtricuspid pressure gradient and the arterial oxygen saturation were normal and comparable in the 2 groups, whereas the frequency of PFO was 5 times higher in HAPE-S than in HAPE-R participants (56% vs 11%, P = .004, Table 2).

At high altitude, as expected, the systolic transtricuspid pressure gradient was significantly higher in HAPE-S participants than in HAPE-R participants (Table 2). Eight of the 16 HAPE-S participants but none of the 19 HAPE-R participants (P<.001, Table 1, Table 2) developed HAPE. Patent foramen ovale was more than 4 times more frequent (69% vs 16%, P = .001, Table 2), and the arterial oxygen saturation was significantly lower (Table 2) in HAPE-S than in HAPE-R participants. All comparisons made with the t tests remained significant after Bonferroni correction.

At high altitude, in 4 of the 5 HAPE-S participants with large PFOs (Table 1), we observed spontaneous right-to-left shunting across the foramen ovale. Moreover, the 5 HAPE-S participants with large PFOs had more severe arterial hypoxemia (mean [SD] arterial oxygen saturation, 65% [6%] vs 77% [8%], P = .02) and a non−statistically significant higher frequency of HAPE (4/5 vs 4/11, P = .11) than those with no (n = 5) or grade 1 (n = 2) or 2 (n = 4) PFOs, even though the systolic right ventricular to atrial pressure gradient (55 [9] vs 57 [13] mm Hg, P = .68) was similar in the 2 subgroups.

For the entire study population, there existed a significant inverse relationship between arterial oxygen saturation and PFO size (P = .002), as well as oxygen saturation and the systolic right ventricular to atrial pressure gradient (r = −0.57, P<.001). There was no significant relationship between PFO size and the pressure gradient (P = .68).

We found that at both low and high altitude, PFO was 4 to 5 times more frequent in HAPE-S participants than in mountaineers resistant to this condition. At high altitude, the greater frequency of PFO was associated with a lower arterial oxygen saturation and a higher systolic pulmonary artery pressure. Moreover, in the HAPE-S group, the size of the PFO was directly related with arterial hypoxemia, and there existed a non−statistically significant higher occurrence of actual HAPE in this subgroup.

The greater frequency of PFO in HAPE-S participants at low altitude was an unexpected finding and could suggest that, together with exaggerated hypoxic pulmonary hypertension and defective alveolar fluid clearance,¹⁶ PFO may represent an additional constitutional anomaly associated with HAPE susceptibility. Alternatively, in HAPE-S participants, previous episodes of exaggerated pulmonary hypertension during high-altitude exposure may have resulted in the persistent reopening of a previously closed foramen ovale. Consistent with this latter hypothesis, in other clinical conditions associated with pulmonary hypertension, a comparably high prevalence of PFO has been reported.^1- 2

HAPE-susceptible individuals are characterized by exaggerated altitude-induced hypoxemia that has been attributed to relative hypoventilation,¹⁷ an augmented alveolo-arterial oxygen gradient, or intrapulmonary or intracardiac right-to-left shunting.¹⁸ With regard to the latter hypothesis, a small study had suggested that intracardiac shunting across a PFO may exacerbate hypoxemia in HAPE.¹¹ The current observation at high altitude of spontaneous right-to-left shunting in some HAPE-S participants with large PFOs and more pronounced arterial hypoxemia than in those with small or no PFOs suggests that the size, rather than its mere presence, may be clinically relevant in this setting. Observations made in divers with decompression illness and patients with platypnea-orthodeoxia are in line with this hypothesis.^4,13,19

These observations could suggest that right-to-left shunting across a large PFO may play a role in the pathogenesis of HAPE. A report of a 36-year-old man who developed symptoms of HAPE each time he returned to high altitude could be consistent with this hypothesis.²⁰ In this patient, cardiac catheterization revealed an atrial septal defect with left-to-right shunting at rest, but a marked increase in pulmonary artery pressure and inversion of the shunt during a hypoxic challenge. After surgical closure of the atrial septal defect and treatment with a calcium channel blocker, the patient remained free of symptoms on subsequent visits to the same altitude.²⁰ It is important to note, however, that this report does not establish causality between right-to-left shunting and HAPE, and a great deal of additional work is needed before closure of PFO to prevent HAPE may be considered.