Haemorrhagic complications of femoral artery pseudoaneurysms caused by interventional cardiology procedures – a large-population retrospective study

Introduction

For many years femoral access was the preferred approach in interventional cardiology. Due to its large size, which allows for the use of larger catheters, the femoral artery remains the preferred access for many procedures such as transcatheter aortic valve replacement (TAVI) or high-risk percutaneous coronary interventions (PCI). However, during routine coronary angiography and angioplasty, a transradial (TR) approach significantly decreases the chance of cardiac-death, bleeding, and haematoma formation [1, 2] and is widely preferred over the femoral one [2–4]. It is also worth noting that during the procedure, conversion from radial to femoral access may occur, e.g. due to catheterization failure because of anatomical factors, as well as the need to use a specific technique of angioplasty. Moreover, operator preferences and skills also play a major role in the choice of vascular access. A survey conducted in 2016 among 987 academic interventional cardiologists worldwide revealed that 18% of them still preferred mainly the transfemoral (TF) approach [5]. This high percentage is most likely explained by the fact that the TR approach requires more operator experience and a higher learning curve. In other words, the TF approach seems to be an easier and more operator-friendly technique for catheterization and angiography along with shorter duration of procedure and lower radiation exposure [3]. Bleeding and haematomas are the most common complications in both vascular accesses. However, one of the crucial disparities between TF and TR methods of catheterization is the probability of pseudoaneurysm formation, which is extremely rare for radial artery access [6, 7]. It is also worth noting that pseudoaneurysms may also form during other procedures, such as during ablation by inadvertent puncture of the femoral artery [8]. This potentially life-threatening condition requires urgent treatment, especially in the case of large and growing pseudoaneurysms. Its embolization can be performed in several ways including thrombin injection, operative suturing, ultrasound-guided compression, or stenting [9].

Aim of the research

The aim of the presented study is to assess a degree of anaemia and to investigate possible gender differences in patients suffering from femoral artery pseudoaneurysm following interventional cardiology catheterization (mainly due to angiography, but also after electrophysiological procedures and after intra-aortic balloon pumping (IABP)). Most of the patients were successfully treated with an ultrasound-guided thrombin injection. Our research is based on a large population of patients with femoral artery pseudoaneurysm. The data were collected during 16 years, and there has not been similar research investigating that deeply the level of anaemization of patients with post-catheterization pseudoaneurysms. We have taken into account a large group, measuring the level of anaemization and analysing correlating factors, making our study unique.

Material and methods

All data were gathered in the Second Department of Cardiology, Świętokrzyskie Cardiology Centre, Kielce, Poland. In the patients who underwent femoral vascular catheterization in the course of cardiological procedures in the period 2006–2022, control ultrasound examinations were performed in all 21,946 patients. We registered 621 cases of femoral pseudoaneurysms. Most of the participants underwent a procedure of TF approach coronarography. Among them 146 (23.5%) were treated because of ST-segment elevation myocardial infarction (STEMI), 115 (18.5%) because of non-ST-segment elevation myocardial infarction (NSTEMI), 223 (35.9%) because of other acute coronary syndromes (ACS), 35 (5.6%) underwent coronarography for other reasons, 20 (3.2%) had a scheduled angioplasty, 53 (8.5%) because of ablation, and 29 (4.6%) for other reasons. Since 2014, as the frequency of radial access in coronary angiography has increased, we have observed a downward trend in the incidence of femoral artery aneurysms. However, in later years it began to rise slightly with the increase in the frequency of electrophysiological procedures, and since 2016 it has remained at a similar level, i.e. approximately 20 cases per year.
There were 580 (93.4%) cases of femoral artery pseudoaneurysm caused by direct artery catheterization and 38 (6.1%) caused by unintended puncture while obtaining venous access.
Most of the femoral pseudoaneurysms were found immediately after the procedure during the first hospitalization (587 (94.5%)), while 24 (3.9%) cases were diagnosed during another hospitalization, and 10 (1.6%) were out-of-hospital diagnoses.
Out of the total 621 patients with femoral pseudoaneurysm a group of 537 patients had their haemoglobin tested before and after the procedure, and they constitute the population thoroughly analysed in the study. This group of 276 (51.4%) females and 261 (48.6%) males was compared to the control group without femoral pseudoaneurysm comprising 606 patients – 271 females and 335 males. The average age of patients participating in the research was 68 years.
The main method of embolization was an ultrasound-guided thrombin injection. A single thrombin injection was successful in as many as 510 (82.1%) patients, and another 88 (14.2%) required re-injection (Table 1).
The study has been approved by the Bioethics Committee of the Swietokrzyska Chamber of Physicians in Kielce.

Statistical analysis

Continuous data were described by means (with standard deviations) and range (minimum and maximum). Normality of distributions was checked by Shapiro-Wilk test. Categorical data were summarized by frequencies and percentages. Group comparisons were performed using the c2 or Fisher exact test for categorical variables, and the t-test or Mann-Whitney test for normally or non-normally distributed continuous variables, respectively. Haemoglobin difference was defined as haemoglobin after the procedure minus haemoglobin before the procedure. Due to significant non-normality of analysed variables, Spearman rank correlation coefficients were calculated to assess the strength of monotonic association between variables of interest (volume, canal length, time from procedure to diagnosis of pseudoaneurysm, and haemoglobin difference).
Separable logistic regression models were created to examine the relationship between decrease of haemoglobin difference (any decrease, a decrease of more than one unit, a decrease of more than 2 units, and a decrease of more than 3 units) and the patient’s status according to pseudoaneurysm. Odds ratios (OR) and 95% confidence intervals (95% CI) were calculated for these models.
A 2-tailed p-value < 0.05 was considered statistically significant. All statistical analyses were performed using the R software package version 4.0.3.

Results

The occurrence of a pseudoaneurysm statistically significantly increased the chance of any decrease in haemoglobin (HGB) (OR = 3.26, 95% CI: 2.26–4.7, p < 0.0001), as well as a decrease in HGB by more than one unit (OR = 6.34, 95% CI: 4.87–8.26, p < 0.0001), a decrease in HGB by more than 2 units (OR = 7.54, 95% CI: 5.18–10.96, p < 0.0001), or a decrease in HGB by more than 3 units (OR = 20.31, 95% CI: 9.35–44.1, p < 0.0001). Similar changes were observed in both sexes (Table 2).
Among patients who had any decrease in HGB after the procedure, significantly greater decreases on average were observed in the group of patients with a pseudoaneurysm. Similar changes were observed in both women and men (Figure 1).
Men were significantly more likely to be anaemic before the procedure (Table 3).
Among the patients with femoral pseudoaneurysm whose HGB levels dropped the most after percutaneous intervention, STEMI was the most common primary condition (Table 4).
The magnitude of HGB decrease correlates positively with pseudoaneurysm canal volume and length in women (Table 5).
Noteworthy is the higher incidence of aneurysms and anaemia in patients on dual antiplatelet therapy. The main antiplatelet drug, apart from acetylsalicylic acid (ASA), was clopidogrel. Dual antiplatelet treatment based on ASA with clopidogrel was significantly more frequently used among patients with aneurysm comparing with those without: 336 (62.6%) compared to 206 (34.0%), respectively. We did not observe a higher incidence of aneurysms in patients using other antiplatelet and anticoagulant drugs (Table 6).

Discussion

Pseudoaneurysms could be a serious consequence of cardiovascular procedures, clinically presenting usually as painful, throbbing haematoma [10]. When comparing the occurrence of radial and femoral pseudoaneurysm, it was found that femoral pseudoaneurysms are 4 to 8 times more frequent than radial ones [8, 11, 12]. Although the recommended percentage of femoral artery catheterization complicated with pseudoaneurysm is < 0.2%, the widely accepted target occurrence nowadays is < 1% [13]. According to most of the publications available so far, the femoral pseudoaneurysm prevalence is usually between 0.2% and 8% of all patients treated using interventional procedures [14, 15]. In our study, the incidence of femoral artery pseudoaneurysm was 2.8%, which is consistent with literature data. The compilation of data from a large population of patients with femoral artery pseudoaneurysm is a great strength of the presented study.
The data we gathered clearly confirms that the presence of pseudoaneurysm correlates with a higher risk of anaemization. As stated by a significant amount of research, high pseudoaneurysm occurrence is a femoral access characteristic and is rarely present when using transradial intervention [16, 17]. The tendency of bleeding and haemoglobin drop themselves also seem to be strongly connected with transfemoral access catheterization, as was proven by the MATRIX investigators [18]. As was found in our study, patients with pseudoaneurysm who became anaemic after the procedure had lower HGB levels on admission and therefore fell below normal HGB levels more easily. These observations were consistent with literature data, as low red blood cells count (RBC) and HGB levels were found to be predictive factors for pseudoaneurysm formation [18]. This statement could be explained by the impaired rheological properties of blood caused by the reduced interaction of RBC and HGB with the platelets and endothelial cells. It is known that RBCs promote the transport of platelets to the site of vessel wall injury [18, 19]. Therefore, patients with baseline low haemoglobin are more likely to develop an aneurysm and thus become anaemic.
Apart from baseline anaemia, other risk factors for femoral pseudoaneurysm formation are incorrect procedure technique as well as older age, puncture of surrounding veins, hypertension, and artery calcification [13]. Moreover, some research indicates that post-catheterization pseudoaneurysm may develop more often in women [20]. It has also been argued that women have a higher risk of bleeding and anaemia when undergoing PCI compared to men, with female sex being an overall bleeding risk factor [21]. Although in our study femoral pseudoaneurysms were more common in women, the difference was insignificant. We found that men were more likely to be anaemic before the intervention. This observation differs from the results of the study of Ahmed et al., in which occurrence of anaemia was similar in both sexes, and elevated haemoglobin levels in most acute coronary syndromes were connected to male sex [21]. Undoubtedly, the incidence of anaemia is also related to the age of the patients. It has been found that the incidence of anaemia increases with age in men, while it decreases in women [22]. The average age of patients participating in the research was 68 years; however, the range was from 20 to 90 years.
Based on the collected data, we found that STEMI, while not being the most common cause of femoral artery catheterization, was the most common primary condition among the anaemized patients. This result is in cotrast to the research by Bhavanadhar et al., in which anaemia was more common in unstable angina patients than in NSTEMI or STEMI groups [23]. However, the mentioned study included a general population of patients with acute coronary syndromes without considering specific bleeding complications. Regarding our study, the possible explanation of the higher incidence of femoral pseudoaneurysm and anaemia among patients with STEMI is the fact that this population usually demands simultaneous stenting of the infarct-related artery. Hence, therapeutic interventions require sheaths larger in calibre, more frequent device changing, and more time to perform the procedure. All these together increase the risk of both aneurysm and anaemia [8, 24]. Another factor increasing the risk of anaemia in this group of patients is the use of dual antiplatelet therapy at the time of the procedure [24]. On the other hand, there are also studies indicating that the absolute decrease in platelet reactivity after the clopidogrel load is significantly less in anaemic patients compared to patients with normal HGB [25]. The above observation may suggest that patients with anaemia both before and after stenting, e.g. in connection with a pseudoaneurysm, are at risk of recurrent ischaemic events.
Another potentially important finding of our trial is the statement that the magnitude of the decrease in HGB correlates with pseudoaneurysm canal volume and length, especially in women. These are factors that not only increase the risk of anaemia the larger they are, but also have an impact on the choice of pseudoaneurysm treatment. Management includes observation, ultrasound-guided compression, ultrasound-guided thrombin injection, and surgery [15]. Surgical methods are usually necessary in the case of complicated pseudoaneurysms. It is also a second choice after initial failure of other methods. Observation is still used for stable and small pseudoaneurysms that are less than 2 cm [24]. In our study spontaneous resolution was observed in 1.4% of patients. Only fever patients from the study population were treated with ultrasound-guided compression because in our centre the leading method of treating uncomplicated pseudoaneurysms is ultrasound-guided thrombin injection. This is consistent with literature data in which it has been proven that percutaneous thrombin injection is more effective in achieving primary pseudoaneurysm thrombosis and thus shortens the hospitalization time compared to ultrasound-guided compression [12, 19, 26].
The main limitation of our study is that it is a single-centre, retrospective study. Moreover, due to the large number of cases the patient population was heterogeneous. Due to the very good experience of our centre in closing femoral pseudoaneurysms with ultrasound-guided thrombin injection, the vast majority of patients were treated exclusively with this method. We did not observe a higher incidence of aneurysms in patients using antiplatelet drugs other than ASA and clopidogrel and new anticoagulants, although the number of these patients in the studied population was relatively small due to the long period of the study and only recent widespread use of these drugs.

Conclusions

The occurrence of a femoral artery pseudoaneurysm significantly increased the risk of anaemia, which is higher in patients with STEMI and on dual antiplatelet therapy.

Conflict of interest

The authors declare no conflict of interest.

References