The number of percutaneous endovascular procedures has grown rapidly thanks to the technological advances of materials, long term satisfactory clinical results and lower morbidity when compared to traditional surgery. Although the number and the complexity of the procedures are rising, the incidence of local complications has not substantially changed over the years (Tsetis 2010; Kopin et al. 2019).
The arteries of the upper limb (radial, brachial, and rarely axillary) are used as arterial access, especially in cardiologic procedures. The trans-radial approach is associated with a lower incidence of major complications when compared to femoral approach (Rigattieri et al. 2016); the most common major complication is the occlusion of the radial artery, which rarely leads to clinical manifestations, due to the dual arterial supply guaranteed at the hand (Reich et al. 2017). The major complications associated with the brachial approach are thrombosis, PSA and nerve compression; ischemic complications (dissection/thrombosis) are more common at this level than in femoral approach (Romaguera et al. 2012; Johnson et al. 1994).
At the level of lower limbs, the popliteal access is the least common (sometimes used in the recanalization of chronic occlusions of the femoropopliteal arterial axis) (Ortiz et al. 2014); major complications are dissection and PSA (Chan and Common 2004).
The CFA is by far the most used vascular access because it has several advantages over other sites: the large calibre, which allows the introduction of larger catheters and easier cannulation; the femoral head underneath guarantee haemostasis with manual compression (Tsetis 2010; Tonnessen 2011; Keeling et al. 2009; Reich et al. 2017). Due to the highest number of procedures performed via this route, femoral access is subject to more frequent complications than other arterial access (Ortiz et al. 2014).
Bleeding is the most frequent complication; hematoma may present as stable or unstable (uncontrolled bleeding), retroperitoneal haemorrhage or PSA. The incidence of femoral PSA varies from 0.1% to 1.5% after diagnostic angiography and up to 7.7% after interventional procedures and increases with the complexity of procedures, patient age and the presence of bleeding disorders (Graham et al. 1992; Carey et al. 2001; Erol et al. 2015).
SFA or DFA accesses, compared with CFA accesses, are more likely to lead to PSA or AVF formation, due to the smaller size and the lack of bone support against which compress after sheath removal (Morgan and Belli 2003).
The natural history of iatrogenic femoral PSA is uncertain (Graham et al. 1992). In a case series of 147 patients, Thalhammer et al. (2000) reported that 86% had a spontaneous resolution (PSA and AVF) after a mean of 23 days. Although the rupture of the femoral PSA can be a life-threatening condition (Graham et al. 1992), some authors believe sufficient Doppler ultrasound observation alone, especially for asymptomatic patients not receiving anticoagulation therapy with little PSA (diameter ≤ 3 cm) (Johns et al. 1991).
Surgical exploration of a pseudoaneurysm is often challenged by the presence of a haematoma. After the evacuation of the haematoma, another difficulty is the identification of the bleeding site, that may be more than one. Finally, surgical treatment may require simple suture of the defect or a patch angioplasty (surgical repair using a patch). Considering the aforementioned limitations, surgical repair is indicated in the cases of haemodynamically relevant bleeding or shock with a rapidly expanding haematoma, risk of skin necrosis due to pressure by the haematoma and infectious pseudoaneurysm to ensure debridement of infected tissue (Tisi and Callam 2006; Savolainen et al. 2011). In 1991, the ultrasound-guided compression was suggested (Fellmeth et al. 1991); the advantages are its simplicity, low cost and lack of ionizing radiation. However, this procedure has some limitations, related to the pain threshold of the patient and the time necessary to obtain complete closure. Moreover, failure is increased in patients on anticoagulant therapy (Hajarizadeh et al. 1995).
In 1997, Liau et al. (1997) reported the successful use of percutaneous injection of thrombin with ultrasound guidance for the closure of 5 cases of PSA of the CFA. In the series of 15 patients reported by Brophy et al. (2000), all the PSA have been successfully treated with 500–1000 U of bovine thrombin, regardless of the PSA size.
Later, in a series of 54 PSA (divided into simple, 45, and complex, 9) reported by Sheiman et al. (2001), the technical success was 50/54, with the possibility of a second approach in case of failure. In some cases (2–4%) you can proceed to a further ultrasound-guided injection procedure with an overall high success rate. However, in literature, the ultrasound-guided injection technique has a failure rate of 3% -14% (Maleux et al. 2003).
The greatest risk from thrombin injection is distal embolization (which can cause serious complications up to limb loss). Pezzullo et al. (2000) have described distal embolization in one of the 23 patients studied. The placement of the tip of the needle at distance from the neck of the pseudoaneurysm under ultrasound guidance and the slow injection of the drug under ultrasound control minimizes the risk (Maleux et al. 2003; Pezzullo et al. 2000).
Other side effects of thrombin injection include hypotension and bradycardia, bleeding - because of an ‘acquired inhibition of coagulation factor (XI) secondary to immune cross-reactivity of bovine thrombin - and anaphylactic reactions in patients who have had repeated exposure to bovine thrombin (Pope and Johnston 2000). Different forms of thrombin are commercially available, the majority of which are of bovine origin and have been used for many years (Pezzullo et al. 2000; Samal et al. 2001). Because of these risks is currently preferred the human thrombin, which implies a slightly higher cost (Vázquez et al. 2005).
The use of human thrombin would not entail the risks associated with bovine thrombin or any other immunological risk; it must be acknowledged, however, that there may be a small risk of infection, although not yet confirmed (Vázquez et al. 2005).
Quarmby et al. (2002) described the embolization of 10 PSA through the use of autologous thrombin with immediate technical success in 7 cases and the use of the new administration of the same in 3 cases.
Loose and Haslam described the percutaneous technique of blocking flow through percutaneous balloon angioplasty (PTA) inflated for about 15 min to prevent distal embolization during injection under ultrasound guidance in the sac of the PSA. Their method was effective (12/13 cases successfully treated without complications), but is rather expensive and employ ionizing radiation (Loose and Haslam 1998).
The use of stent-graft in the treatment of aortoiliac disease has recently led several authors to employ these devices to exclude aneurysms and peripheral PSA (Xiao et al. 2012; Laganà et al. 2002). In a larger series of stent-graft used in the treatment of pseudoaneurysms and AVF, Waigand et al. (1999) and Thalhammer et al. (2000) reported a technical success rate of 84–88%. However, there are limits to the indication, as the stent-graft may result in the occlusion of the DFA; theoretically, this complication can prevent the use of the site as future access. Hence, a femoral angiography from an oblique projection should be performed to evaluate the suitability of the device insertion, as it allows to check the presence of an adequate landing zone in the CFA respect to the origin of the profunda femoral artery. Besides, hip stress can lead to an increased risk of stent thrombosis (Thalhammer et al. 2000; Xiao et al. 2012).
Retroperitoneal bleedings (RB) are potentially fatal and not easy to diagnose in clinically haemodynamically stable patients; RB may occur spontaneously in the presence of coagulopathy, use of combined antiplatelet drugs (eg. dual antiplatelet therapy), oral anticoagulant therapy with heparin or poorly controlled partial thromboplastin. Generally, the target population includes elderly population. Usually, retroperitoneal hematomas present with abdominal and back pain, less frequently with fatigue, nausea, headache, and dyspnea (Baekgaard et al. 2019; Dolapsakis et al. 2019).
Besides, the RB may be iatrogenic; in particular, they may occur as a complication of arterial or venous catheterization during endovascular procedures. The retroperitoneal bleedings may result from the femoral puncture and are spread either through the Retzius space or directly on the muscles of the pelvic floor without the direct involvement of this space (Terotola et al. 1991).
Usually, AVF are between the SFA, the DFA and the adjacent lateral circumflex femoral vein. The majority of studies reported a high probability of spontaneous closure (Toursarkissian et al. 1997), but Kresowik et al. (1991) reported opposite results. Toursarkissian et al. (1997) reported that 86% of the non-symptomatic low-flow AVF resolves spontaneously, requiring only observation and close Doppler ultrasound monitoring every 2 weeks until 12 weeks. Until recently, the AVF symptomatic were treated surgically (Kresowik et al. 1991; Kelm et al. 2002). More recently, percutaneous treatment with stent-graft showed constant technical success, but with a significant risk of stent occlusion - medium and long-term graft (12% -17%) (Laganà et al. 2002).
Dissection is a frequent but often underdiagnosed complication of arterial catheterization because many cases are asymptomatic (Benjamin et al. 2015). During vascular access, coiling of the guidewire under fluoroscopy or resistance to passage may indicate subintimal passage with resulting retrograde arterial dissection. This complication is more common in the iliac arteries due to the presence of plaques by steno-occlusive atherosclerotic disease and to the tortuosity of these vessels. The operator should immediately recognize the dissection and evaluate the hemodynamic consequences. Generally, the flap of dissection doesn’t cause stenosis and the simple removal of the catheter and guide usually allows a spontaneous resolution, also due to the anterograde flow of blood. If the flap dissection is not felt by the sensitivity of the operator and extends for a longer distance gaining the true lumen, the flap may become occlusive - then counter - with acute ischemia, so that it is necessary an immediate repair by angioplasty or stenting (Tsetis and Belli 2004).
Despite the wide use of vascular closing devices, there are no studies that prove with certainty their superiority in terms of safety compared to manual compression. Only the study of Tavris et al. (2004) demonstrated that vascular complications are lower with the use of VCD. One hundred sixty-six thousand six hundred eighty patients of the National Registry of the American College of Cardiology database, undergoing cardiac catheterization in 2001, have been evaluated; 53,655 were treated with VCD for haemostasis, while in 113,025 was performed the manual compression. The risk of vascular complication was 1.1% in the group with VCD and 1.7% in the group with manual compression, with p < 0.001, statistically significant.
The evolution of metal alloys with flexible and elastic materials and the technological innovation on the porosity of the cover have now led to the use of stents with a higher patency rate and a lower rate of acute thrombosis. All this has led to greater use of devices close to the articular femoral site and the possibility of re-using the access through the stent struts, clearly with catheters and introducers of small arms. A high technical success rate (92.3%) has been achieved, with a low (13.8%) complication rate at 2 years.
Limitations of the study are the lack of long term follow up, except for the cases of stent-graft placement, the retrospectivity of the analysis and the scarcity of data in the literature, necessary to evaluate the congruence and the consistency of the data presented.