The annual incidence of TAE procedures performed in the three participating tertiary European vascular centres for the management of severe haemorrhage in patients with uncorrected bleeding diathesis, was surprisingly high (12.7%; approximately one out of ten patients who underwent TAE in 2019). The results of this retrospective, real-life, study, demonstrated that TAE in the specific cases was safe and effective and could increase the therapeutic window in carefully selected patients. Specifically, technical success was 100% and only one case of rebleeding/repeat TAE was recorded that was successfully treated with further TAE. The clinical safety and effectiveness of TAE has been widely reported in the literature, establishing its role in bleeding control (Spiliopoulos et al. 2019; Hur et al. 2014; Melloul et al. 2015; Padia et al. 2020). However, uncorrected coagulopathy related to anticoagulation, haemophilia, cirrhosis or severe trauma is a relative contraindication for TAE, as it has been previously reported as a significant negative predictor for recurrent bleeding, in-hospital mortality and procedure-related complication rates, up to 3.5 times more likely to result in clinical failure (Hur et al. 2014; Valek and Husty 2013; Angle et al. 2010; Loffroy et al. 2010; Lopera 2010). Loffroy et al., have reported that patients with impaired coagulation demonstrate three times higher risk for rebleeding following successful TAE and ten times higher bleeding-related risk of death compared to patients with normal coagulation profile. Nevertheless, the authors highlighted the fact that these patients should not be refused emergency TAE while efforts to correct coagulation before, during, and after embolization, are of most importance for definite clinical success (Loffroy et al. 2010). This study reports similar technical success and rebleeding rates between patients with severe bleeding diathesis and those with normal coagulation profile at the time of embolization. Moreover, no procedure-related complications were noted, while the 30-day survival rate was considered extremely satisfactory, given the advanced age and multiple comorbidities noted in the specific cohort.
The authors speculate that the above-mentioned outcomes could be attributed to several factors, among which are the long-standing local experience and expertise in the multi-disciplinary management of haemorrhage, that includes optimal periprocedural coagulation improvement, as well as specific considerations regarding the timing and technique of TAE. Notably, efforts to improve the coagulation profile prior, during and after embolization were made in all patients and in 58% of the cases, abnormal INR, aPTT and PLT values were normalized within 24 h, a fact that certainly contributed to the minimal rebleeding rate and overall clinical success. The decision to timely proceed to emergency embolization in patients with uncorrected bleeding diathesis was balanced between the risk of technical failure/procedure-related complications and patient survival, which was severely compromised by ongoing bleeding causing hemodynamic instability, leaving no time for further improvement of the coagulation parameters. Such emergency multidisciplinary clinical decisions are currently supported by international experts’ consensus (Patel et al. 2012). Of note, time to embolization has been correlated with better clinical outcomes and waiting for a long time period in order to normalize clotting time is not a good option in cases of large-volume bleeding causing hemodynamic instability (Chakraverty et al. 2012).
In this study, the choice of embolic material was heterogeneous and highly individualized, a fact that should be expected in a retrospective analysis. It is well-known that correct choice of the embolization agents is essential. However, in patients with impaired coagulation, the choice of embolic agents is even more crucial and should be based on unique particularities. The use of metallic coils has been correlated with ineffective haemostasis, as their mechanism of action is based on the organism’s intrinsic ability to form clot. This is less likely to occur when the coagulation cascade has been impaired. In this series, blood flow through the coils has been noted even 5 min after coil deployment in a patient with factor VIII deficiency and an undefinable aPTT value (Fig. 1). The same applies for gelfoam as its mechanism of action also involves mechanical obstruction, blood flow reduction and therefore thrombus formation. However, dense coil packing or a combination of dense coiling with gelfoam could provoke complete arterial blockage and successful hemostasis (Fig. 1). Solely dense coil embolization was successfully performed in four cases, without any bleeding recurrence. Although, not commonly preferred, successful coil-only embolization has been previously reported for patients with significantly prolonged INR (> 3) (Ramaswamy 2014). Choosing to use only coils, is usually based on the safety provided by coiling in order to avoid non-targeted embolization and severe ischemic complications, such as in cases in which selective catheterization is not possible or for lesions close to arterial origins and bifurcations (Fig. 1), where glue or particles could easily migrate. It is important to understand that correct dense coil packing can be efficient even in cases of deranged clotting, especially if prompt normalization of the coagulation profile is awaited shortly after TAE. Microparticles, also act as blood-flow blockers, but their mechanism of action also involves an immediate inflammatory reaction of the arterial wall, focal angionecrosis and eventual vessel fibrosis and theoretically should be less influenced by deranged coagulation. In this study, microparticles were successfully used but only in combination with microcoils, and generally various combination of embolic agents was used in most cases (in order to enhance the occlusive effect, as previously described (Aina et al. 2001). On the other hand, n-butyl cyanoacrylate glue and ethylene vinyl alcohol copolymer (Onyx®) have been proposed as the most effective embolic agents in patients with bleeding disorders, as polymerization (glue) and solidification (Onyx) abruptly occludes the target vessel independently from the coagulation process (Shi et al. 2017; Toyoda et al. 1996; Né et al. 2018).
In the single case of rebleeding, gel foam was used a spontaneous bleeding originating from a DFA branch, in a patient with acquired factor VIII deficit and hemodynamic instability. Re-embolization using a combination of microcoils and gel foam was successfully performed, and bleeding control was achieved following concomitant factor VIII administration, which was not available at the time of initial embolization. Despite bleeding relapse, the first TAE procedure, contributed in stabilising the patient hemodynamically. The second hemorrhagic event occurred after 12 h. Subsequent second successful TAE in combination with factor VIII administration resulted in successful haemostasis.
Another important issue is access-site haemostasis and related complications. Manual compression in patients with prolonged coagulation time can be extremely long-lasting, tedious and ineffective. A closure device was used in all patients and no access-site complications were noted. In two cases the sheath was removed the next day and in one case after approximately 72 h. Choosing not to remove the sheath immediately after the procedure is always a choice if abnormal coagulation parameters are expected to normalize or significantly improve over time. Smaller 4Fr or 5Fr sheaths would be an option in these high bleeding risk cohort, but hemostasis should be performed using manual compression, as available closure devices use 6Fr sheaths. The authors preferred the use of a closure device over manual compression for smaller access diameters in the specific cohort, although this is not supported by evidence.
Limitations of this study include the retrospective design which could generate a selection bias, while some cases and data could have been missed. Moreover, cases that were discussed by the MDT for possible TAE, but finally underwent endoscopy or surgery, could not be identified. Additionally, this was not designed to be a comparative study with a control group and comparative data of rebleeding rate with the group of patients demonstrating normal coagulation profile could be confounded by significant heterogeneity. Notably, additional conservative treatments given to patients (for example cryoprecipitate) may have increased procedural and clinical success, but study design did not allow for the assessment of the contribution of such factors to the study outcome measures. Finally, the relatively small number of patients included, limits the validity of outcomes.