Post-procedural pseudoaneurysms: Single-center experience

Introduction

Inadequate closure of an arterial puncture site results in a continued leak of blood in the surrounding tissues forming a hematoma. If a connection remains patent between the hematoma and the artery, a pseudoaneurysm forms. Iatrogenic pseudoaneurysm formation is one of the most common complications of cardiac and peripheral vascular endoluminal procedures. The incidence ranges between 0.2% and 0.5% for diagnostic procedures and may reach as high as 6% for therapeutic procedures.^1–3 Risk factors reported in the literature include: anticoagulation, increased mass index, large sheath size, improper technique (such as unsatisfactory manual compression or maldeployment of a closure device), female gender, chronic renal failure and puncture site access below the femoral bifurcation.^4–8 Duplex ultrasonography is the most common modality used to establish the diagnosis.

Pseudoaneurysms can lead to multiple complications, including pain, distal embolization, compression of adjacent neurovascular structures, skin necrosis and rupture. ⁹ Kent et al. ¹⁰ reported in their study that pseudoaneurysms less than 2 cm are likely to thrombose spontaneously if patients were not anticoagulated.

In the early 1990s, surgical repair was the gold standard for management of iatrogenic pseudoaneurysms. Such surgery, however, can be associated with significant wound morbidity, prolonged hospitalization stays and increased costs. Ultrasound-guided compression, first reported by Fellmeth et al., ¹¹ gained rapid acceptance but is time consuming, associated with patient discomfort, and has significant failure rate. Cope and Zeit ¹³ first described thrombin injection for the management of femoral pseudoaneurysms, but this technique did not gain widespread acceptance until Kang et al.’s ¹⁴ landmark paper. This is currently the preferred treatment with a reported success rate of 91–100%.

The aim of this study was to review the efficacy and safety of thrombin injection of post-procedural pseudoaneurysms (PPPA) at our institution.

Materials and methods

A prospectively maintained vascular laboratory database was queried for all pseudoaneurysms diagnosed between March 2004 and January 2013. Only PPPA were studied. True aneurysms, false aneurysms in dialysis graft or fistula access sites, and pseudoaneurysms referred from outside hospitals were excluded. A chart review was undertaken to record patient demographics and comorbidities, type of procedure causative for to the development of the PPPA, sheath size, arterial access site, use of periprocedural anticoagulation and whether a closure device was used. Follow-up ultrasound studies for recurrence were reviewed.

This study was approved by our Institutional Review Board and conducted in accordance with the Health Insurance Portability and Accountability Act and the prevailing ethical principles governing research. Patient informed consent was waived. There are no competing interests.

Results

Pseudoaneurysms were identified in 204 patients. After excluding pseudoaneurysms in dialysis grafts or fistula access sites and those referred from outside hospitals with incomplete records, we were left with 167 patients. The causative procedure was a diagnostic angiography in 65 patients, percutaneous coronary intervention in 64 patients, peripheral vascular intervention in 25 patients and a vascular access procedure (arterial line or central venous line) in 13 patients (Table 1). Out of the 19,217 diagnostic angiographies and 6598 percutaneous coronary interventions performed during the study period, PPPA was noted in 0.03% and 0.97% of the procedures, respectively (Table 2).

OPEN IN VIEWER

Table 1.

Post-procedural pseudoaneurysms causative procedures.

Procedures	n (%)
Diagnostic angiogram	65 (38.6)
Percutaneous coronary intervention	64 (38.3)
Peripheral vascular intervention	25 (14.7)
Access (arterial and central venous lines)	13 (7.6)

OPEN IN VIEWER

Table 2.

Incidence of post-procedural pseudoaneurysms stratified by causative procedure (2004–2013).

Procedure	Total patients (n)	Number of pseudoaneurysms (n)	Incidence, %
Percutaneous coronary intervention	6598	64	0.97
Diagnostic angiography	19,217	65	0.03

The mean age of affected patients was 66 ± 14.4 years, and 52% (87 patients) were females. The mean diameter of the aneurysm was 2.77 cm ± 1.83 cm with the largest diameter recorded as 17 cm and the smallest as 0.46 cm. Patient comorbidities are noted in Table 3. The majority of our patients (94%) received some sort of periprocedural antithrombotic therapy, which included either one or a combination of antiplatelet agents, warfarin, angiomax, heparin and argatroban. Table 4 summarizes the site of PPPA development.

OPEN IN VIEWER

Table 3.

Patient demographics and clinical characteristics.

Demographics	n (%)
Mean age (years)	66.1 ± 14.4
Females	87 (52%)
Comorbidities
Hypertension	110 (64.7%)
Coronary artery disease	88 (51.8%)
Hyperlipidemia	73 (42.9%)
Diabetes mellitus	44 (25.9%)
Chronic kidney disease	28 (16.5%)
Cerebrovascular accident	21 (12.4%)

OPEN IN VIEWER

Table 4.

Site of origin for post-procedural pseudoaneurysms.

Access site	n (%)
Common femoral artery	133 (78.2)
Superficial femoral artery	26 (15.3)
Other (radial, profunda femoris, and brachial)	8 (6.5)

The vast majority of the 167 patients were treated with thrombin injection (n = 142). A total of 25 patients, however, did not receive thrombin injection. Of these, six patients were treated with ultrasound-guided compression, 10 with surgical repair and 9 were simply observed (Figure 1). On follow-up ultrasound, no recurrence was noted in those patients who underwent ultrasound-guided compression, and all the patients who were observed had spontaneous thrombosis. The mean size of the pseudoaneurysms that were managed with ultrasound-guided compression was 2 cm and with observation was 1.5 cm. Among the 142 patients who received thrombin injection, first attempt success was achieved in 130 patients (93.5%). Of the 12 patients who failed to achieve complete thrombosis, seven were re-injected with successful thrombosis, three underwent surgical repair, one had ultrasound-guided compression and one patient was observed. There was no sonographic evidence of any pseudoaneurysm on follow-up in any of these patients. When patients who had repeat thrombin injection were taken into account, our secondary success rate for thrombin injection was 100%. OPEN IN VIEWER

Demographics, comorbidities, presence or absence of anticoagulation, sheath size, use of a closure device, maximum diameter of aneurysm and access site were all assessed using a univariate analysis for predictors of pseudoaneurysm recurrence. Using forward logistic regression, end-stage renal disease (p = 0.006), larger pseudoaneurysm size (p = 0.013) and superficial femoral artery puncture site (p = 0.031) were found to be significant predictors of failure after first attempt of thrombin injection (Table 5).

OPEN IN VIEWER

Table 5.

Independent predictors for failure of first attempt at thrombin induce thrombosis.

Step	Variable	Adjusted odds ratio (95% CI)	Adjusted p
1	Maximum diameter	2.23 (1.25, 3.96)	0.006
2	End-stage renal disease	1.15 (1.09, 1.78)	0.013
3	Superficial femoral artery	0.20 (0.04, 0.86)	0.031

Complications occurred in six patients (4%). Femoral artery thrombosis occurred in two patients. In one patient the thrombosis was partial – a short tail of partially occluding (<50% of the artery diameter) thrombus originating at the common femoral artery – which resolved with anticoagulation alone. In the second patient, surgical thromboembolectomy was required. Deep venous thrombosis of the femoral vein was noted in two patients and was treated with anticoagulation. Whether this resulted from the procedure itself or from mass effect and compression of the femoral vein by the hematoma/pseudoaneurysm is unclear. The infection rate was 1% (n = 2). Both patients with infection were treated with antibiotics, though one did require surgical evacuation of the infected hematoma.

Discussion

Post-procedure pseudoaneurysms can increase the length of hospital stays and costs. ¹ Risk factors include anticoagulation, obesity, large sheath size and inadequate technique including improper puncture site or unsatisfactory post-procedure compression. Duplex ultrasonography is the most common modality used to establish the diagnosis.¹⁵

The incidence of PPPA has become more frequent with the increased use of anticoagulation, larger sheath size and longer and more complicated endoluminal interventions.^16,17 The most common site for PPPA is the common femoral artery as it is the most frequently accessed site for both diagnostic and interventional procedures. Several treatment strategies for PPPA have been reported including surgery, ultrasound-guided compression and ultrasound-guided thrombin injection. Surgical repair had been the traditional approach to treating of pseudoaneurysm until alternative options were developed. Even though it is successful, surgical repair requires general or regional anesthesia which may be poorly tolerated by patients with poor cardiac reserve and can be associated with significant morbidity. Complications from surgery have been reported in about 20% of patients, and include infection, wound healing problems, prolonged hospital stay, increased cost as well as perioperative death. When Fellmeth et al. ¹¹ introduced ultrasound-guided compression for the treatment of PPPA, the procedure became very popular. It has had a reported success rate of 77–86%. The success rate was reduced to less than 40% in the presence of anticoagulation.^16–20 Even though the overall complication rate is reported to be only 3.6%, this procedure has many disadvantages including patient discomfort despite analgesia, prolonged procedure times (up to 4 h) and a recurrence rate as high as 25–30% in patients on anticoagulation.^12,21,22

Ultrasound-guided thrombin injection was first introduced in 1986, but did not gain popularity until the late 1990s due to complications associated with its initial use.^13,15 Complication rates of 0–3% have been reported in the literature. Table 6 summarizes the largest recent series.^{9,14,18,22–31} These include infection, deep vein thrombosis, arterial thromboembolism and anaphylaxis due to the bovine source of the thrombin injection.^32–34 Thromboembolic complications occur when injection of thrombin is preformed close to the pseudoaneurysm neck at a fast injection rate or in high concentrations; each of these situations can result in a large amount of thrombin reaching the artery. The reported incidence in the literature varies between 0 and 2% in the literature.^12,22,35 We noted two cases of arterial thrombus formation in our series. One resolved with anticoagulation alone, while the other was symptomatic and required surgical thromboembolectomy. We also had two cases of deep vein thrombosis that were managed with anticoagulation for six months. Of the two patients who developed an infection, one underwent surgical evacuation of a large infected hematoma. No allergic reactions were reported in our study.

OPEN IN VIEWER

Table 6.

Summary of recent studies.

	First author	Year	Patient’s number	Primary success, %	Complication (n)
1.	Paulson	2001	114	96	4
2	Mohler	2001	91	98	0
3	Khoury	2002	131	96	3
4	Kang	2000	83	98	1
5	La Perna	2000	70	94	1
6	Chen	2015	121	92	0
7	Schneider	2009	274	97	0
8	Dzijan-Horn	2014	212	79	5
9	Sheiman	2001	54	100	0
10	Demharter	2005	54	98	0
11	Veraldi GF	2005	59	96	0
12	Calton WC Jr	2001	52	94	1
13	Gürel K	2012	55	69	0
14	Mohammad F(current series)		142	93	4
15	Total		1512	93	19 (1.3%)

Thrombin injection has been shown to be well tolerated and is much more effective than ultrasound-guided compression in patients who are anticoagulated.^1,36 Recent studies have shown its success rate to be between 94 and 100%, which is similar to our results.^23,27,37,38 A second, third and fourth injection are sometimes required. ³⁹ In our study, seven patients (4.9%) required a second injection. Patients who were not injected were managed with surgery, ultrasound-guided compression or observation. The reasons for electing surgical treatment in our series included concern for infection, large size and a wide pseudoaneurysm neck.

In his review paper, Kronzon ⁴⁰ discussed the increased incidence of pseudoaneurysms in patients with low puncture sites. In our study, we found that superficial femoral artery access was an independent risk of failure after the first attempt. Other factors associated with failure after the first attempts were end-stage renal disease and large aneurysm size. The largest pseudoaneurysm that was successfully injected in our study was 7.8 cm.

We recommend observing PSA that are less than 2 cm, not expanding and have a small neck; most of these will thrombose spontaneously. Currently, our first line of treatment is thrombin injection. We do not offer ultrasound compression. Our second line of therapy is operative. This is reserved for pseudoaneurysms that have a large neck (greater than 5 mm) which is associated with an infection, skin breakdown or recurrence of 2–3 thrombin injections.

There are several limitations to our study. Patients were investigated with ultrasound only if there was a clinical concern, thus our true incidence of pseudoaneurysm may be higher. Due to the retrospective nature of the study, we were unable to clearly identify the reasons for alternative interventions other than surgery.

Conclusion

Our report includes one of the largest study populations managed with ultrasound-guided thrombin injection for post-procedure pseudoaneurysm. Percutaneous ultrasound-guided thrombin injection is an effective and safe treatment for PPPA. Failure rates are low and associated with large aneurysm size, superficial femoral artery access and end-stage renal disease.