Paradoxical Embolism after Declotting of Hemodialysis Fistulae/Grafts in Patients with Patent Foramen Ovale

Abstract

Summary

Background and objectives

The safety of percutaneous endovascular declotting procedures for thrombosed hemodialysis fistulae/grafts is well described in the general population; however, its safety in the presence of a patent foramen ovale (PFO) is not known. The objective of this study is to assess the incidence of symptomatic paradoxical embolic events associated with declotting procedure of thrombosed arteriovenous (AV) graft or fistula in patients with documented PFO.

Design, setting, participants, & measurements

This was a retrospective study in a hospital-based, academic practice. It included 23 patients (10 men; mean age, 65) with PFO and thrombosed hemodialysis graft/fistula who underwent a standardized declotting procedure with 2 mg of Alteplase and balloon thrombectomy. Twenty patients (87%) had AV grafts, and three (13%) had AV fistulae. The PFO shunt was right to left in two (9%), left to right in eight (34%), and bidirectional in ten (44%). The shunt direction was not specified in three patients (13%). The technical success of the declotting procedure and the frequency of clinically manifested paradoxical embolic events in this patient population were calculated.

Results

Fifty declotting procedures were performed on 23 patients with a technical success rate of 96% (48 of 50, 96%). No symptomatic paradoxical embolic events were found in any of the 23 patients with PFO.

Conclusions

Symptomatic paradoxical embolic events after percutaneous endovascular declotting procedures of thrombosed AV grafts and fistulae in patients with documented PFO are rare. This procedure appears to be safe in patients with a PFO.

Introduction

A functioning arteriovenous access is critical for patients with ESRD who receive hemodialysis. These arteriovenous accesses are predominantly arteriovenous fistulae (AVF) or arteriovenous grafts (AVG), collectively referred as AVFG in this manuscript, although a few patients are served with long term hemodialysis catheters (1). Often, the AVFG are complicated by outflow obstruction and thrombosis (2). Current data support the use of percutaneous endovascular techniques to restore patency of the AVFG (3). A percutaneous declotting procedure refers to endovascular thrombectomy of the AVFG using mechanical devices (such as balloons or suction catheters) either with or without concurrent use of thrombolytic agents. During this procedure, the thrombus in the AVFG and outflow veins is often pushed into the central veins. This may result in subclinical pulmonary embolism (PE), although symptomatic PE is occasionally reported (4,5). Many reports have established the safety of this procedure in the general population with no documented cardiac septal defects (4,5). However, the safety of this procedure in patients with documented patent foramen ovale (PFO) is not known. Patients with PFO are at risk of paradoxical embolism, and a few studies have reported occurrence of stroke in these patients, especially after a transient increase in the right heart pressures (6–8). Given the high incidence of subclinical PE during declotting of thrombosed AVFG and consequent elevated right heart pressures, there is real concern of paradoxical embolism in patients with PFO. In this study, we seek to assess the incidence of symptomatic paradoxical embolic events after percutaneous endovascular declotting of thrombosed AVFG in patients with documented PFO.

Materials and Methods

This retrospective study was approved by the institutional review board, and the requirement to obtain the informed consent was waived. The study was Health Insurance Portability and Accountability Act compliant. Patients who underwent a percutaneous endovascular declotting procedure for a thrombosed AVFG between January 2003 and December 2008 and had a documented PFO were included in this study. Our interventional database (HI-IQ software, ConexSys, RI) was interrogated to identify patients who had a declotting procedure during this period. The electronic medical records were reviewed for the presence of PFO. Only patients who had PFO on a contrast-enhanced echocardiography were included in this study.

Patients

During the study period (January 2003 to December 2008), 484 percutaneous endovascular declotting procedures were performed on 238 patients. An echocardiogram was available in 171 patients, and 23 (23 of 171, 13.4%) patients had PFO. These patients constituted our study population. There were 10 men and 13 women with an average age of 65 years. Twenty patients (20 of 23, 87%) had AVG, and three (three of 23, 13%) had AVF. The location of the AVF and AVG in these 23 patients is as follows. Three patients had AVFs: two had (two of 23, 9%) radiocephalic fistulae, and one (one of 23, 4%) had a brachiocepalic fistula. Twenty patients had AVGs: three (three of 23; 13%) had radiocephalic grafts; seven (seven of 23; 30%) had brachiocephalic grafts; three (three of 23, 13%) had brachiobasilic grafts; four (four of 20, 17%) had brachioaxillary grafts; and three patients (three of 23, 13%) had grafts at other locations. A total of 50 declotting procedures were performed in these 23 patients (average of 2.2 procedures/patient) during the study period (Table 1 and Fig. 1).

Table 1.

Clinical characteristics of 23 patients with PFO

Characteristics	n
Mean age (years)	65
Gender
male	10 (43%)
female	13 (57%)
Race
white	15 (65%)
black	5 (23%)
hispanic	1 (4%)
Asian	1 (4%)
others	1 (4%)
Cause of ESRD
diabetes mellitus	6 (26%)
HTN	3 (13%)
diabetes mellitus or HTN	4 (17.5%)
GN	4 (17.5%)
others (pylonephritis, sepsis, RAS)	3 (13%)
unknown	3 (13%)
Type of access
AVG	20
radiocephalic	3 (13%)
brachiocephalic	7 (30%)
brachiobasilic	3 (13%)
brachioaxillary	4 (17%)
other	3 (13%)
AVF	3
radiocephalic	2 (9%)
brachiocephalic	1 (4%)
Total	23

HTN, hypertension; GN, glomerulonephritis; RAS, renal artery stenosis.

Figure 1.

Selection of the study population.

Echocardiography

Echocardiography was retrieved from our electronic medical records, and the report was reviewed. The results of echocardiography were recorded for direction of flow across the PFO. This assessment was made on color Doppler study in 14 (14 of 23, 60%), agitated saline bubble study in seven (seven of 23, 30%), and on both methods in two (two of 23, 10%). The shunt was right to left in two (two of 23, 9%), left to right in eight (eight of 23, 34%), and bidirectional in ten (ten of 23, 44%). The shunt direction was not specified in three (three of 23, 13%) (Table 2).

Table 2.

Shunting patterns of 23 patients with PFO

PFO Shunt Direction	Number of Patients	Percentage (%)
Left to right	8	34
Right to left	2	9
Both directions	10	44
Unknown	3	13
Total	23	100

PFO, patent foramen ovale.

Percutaneous Endovascular Declotting Procedure

In all of the patients, the declotting procedure was performed using a standard technique. After obtaining an informed consent from the patient, the AVG or AVF was accessed at the arterial end with the needle directed toward the outflow with an 18-G angiocath, and 2 mg of alteplase and 2000 units of heparin were injected. After 1 hour, the patient was brought to an angiosuite, and the AVG- or AVF-bearing arm was prepped and draped. The angiocath was exchanged for a 7Fr sheath over a glide wire. A 5Fr Kumpe catheter was advanced in to the subclavian vein, and venography was performed to assess the patency of the central veins. Angioplasty of the venous outflow stenoses, if any, was performed using appropriately sized balloons. Then the thrombus was pushed in to the central veins with angioplasty balloon. Next, the venous end of the AVG/AVF was accessed, and a 6Fr sheath was placed. A 0.035-inch glide wire was inserted and passed through the arterial anastomosis in to the artery. The thrombus plug at the anastomosis was removed using a 5.5Fr Fogarty balloon. Angioplasty of the arterial anastomosis was performed if a stenosis was identified. The declotting procedure was considered technically successful with return of thrill in the graft or fistula. The sheaths were removed, and hemostasis was achieved through purse-string sutures. As per the department protocol for patients with a PFO, a detailed neurologic examination was performed by a board-certified physician for assessment of stroke after every declotting procedure. The presence of any neurologic deficit was recorded and followed with a neurology consult. If there were no neurologic deficits, the patient was discharged from the hospital. All of the patients were informed about the risk of stroke and were requested to go to the hospital emergency room if symptoms of stroke developed after leaving the hospital.

Data Collection

The data were collected from the medical and imaging regards. Paradoxical embolic events such as stroke, myocardial infarction, limb ischemia, visceral ischemia, etc., were recorded if they occurred within 48 hours of the declotting procedure. If a patient had any imaging study (such computed tomography or magnetic resonance imaging) for ischemic symptoms, the results were reviewed, and the findings were recorded.

Statistical Analyses

The technical success of the declotting procedure in this patient population was calculated. This was compared with the technical success of the declotting procedure for all comers in our division (484 procedures in 238 patients during this time). The frequency of paradoxical embolic events in this population was calculated.

Results

As mentioned earlier, 23 patients with PFO underwent 50 declotting procedures during the period of study. Forty-eight (48 of 50, 96%; 95% confidence interval [CI], 91.12% to 99.98%) declotting procedures were successful. Two declotting procedures, one in a patient with an AVF and another in a patient with an AVG, were unsuccessful. In comparison, the technical success of all of the declotting procedures during this period was 97% (471 of 484 procedures in 238 patients; 95% CI, 95.87% to 98.75%).

There were no symptomatic paradoxical embolic events (0%; 95% CI, 0% to 7.1%) after the declotting procedure in any of the patients with PFO. No patients had imaging studies within 48 hours of a declotting procedure. No patients succumbed to death within this period (48-hour mortality, 0%).

Discussion

With the increasing prevalence of ESRD and the consequent rise in hemodialysis, maintenance of a proper vascular access through AVF or AVG has assumed a great importance (9). AVF and AVG are often complicated with recurrent venous stenosis and thrombosis and necessitate frequent interventions through angioplasty and thrombolysis (3,10). Endovascular recanalization procedures are often complicated by bleeding, vein rupture, PE, and downstream arterial embolization (11). The volume of thrombus during declotting procedures is relatively high, although it is still small compared with the large size of the pulmonary arterial bed (5). Subclinical PE during declotting procedures is common, but symptomatic PE is rare (4). One study reported 35% incidence of PE on perfusion scintigraphy after endovascular declotting of thrombosed AVG and AVF, although only 4% were symptomatic (12). Clinical and subclinical PE result in elevation of pulmonary artery and right heart pressures, although such changes rarely lead to cardiopulmonary compromise (13). However, such elevations in right heart pressure may alter the flow direction through a PFO and potentially lead to paradoxical embolism (14).

Given the high incidence of PFO in general population (up to 34%) and its association with stroke in the young (8,15–17), it is important to know the safety of endovascular declotting procedures in patients with a history of untreated PFO. There are only a few case reports of paradoxical emboli secondary to endovascular declotting procedure for thrombosed AVFG (14,18,19). It is hypothesized that such a low incidence of paradoxical embolic events in these patients is secondary to a low incidence of right to left shunt or intermittent nature of the shunt with low volume of blood being transferred (16,20). Similar to the reported studies, our study documented no systemic embolic events during endovascular declotting procedures in patients with echocardiography confirmed PFO. The flow direction across the PFO had no influence in the occurrence of systemic embolic events.

Our study documented the safety of declotting procedures in patients with PFO. As such, patients undergoing declotting procedures may not necessitate a formal echocardiography to rule out the presence of PFO, although a large multi-center study is useful to confirm the preliminary findings of our study. In addition, whether such a recommendation can be applied for surgical thrombectomy needs further clinical studies.

The main limitations of our study are the retrospective nature and the small patient-sample size. Although a prospective trial with a larger number of patients would be highly useful, our study provided follow-up during multiple declotting procedures.

Conclusions

Endovascular declotting procedure of thrombosed arteriovenous grafts and arteriovenous fistulae in patients with PFO is safe and is rarely associated with any symptomatic paradoxical embolic events.

Disclosures

None.