The Analysis of Risk Factors Associated with Transcatheter Arterial Embolization for Percutaneous Nephrolithotomy

Abstract Background: This study aimed to evaluate the risk factors of transcatheter arterial embolization (TAE) in managing haemorrhage associated with percutaneous nephrolithotomy (PCNL) to improve the surgical effect. Methods: From May 2007 to June 2018, 112 patients (31–60 years) who underwent TAE treatment for haemorrhage after PCNL were retrospectively analyzed. All patient data and embolization details were retrieved from medical records. Univariate analysis was used to identify the risk factors related to clinical outcomes. Results: Technical and clinical success rates were 100% and 93%, respectively. On angiography, we observed injury to the main artery in 1 patient, to secondary branch in 22, to tertiary branch in 58, and to both secondary and tertiary branches in 31. Embolic agents were coils (n =31), gelatin sponge (n=15), and gelatin sponge with coils/microcoils (n =66). Bleeding control failed in 8 patients. Eight patients opted for a second operation, 6 by repeat TAE and 2 by surgery. Bleeding was eventually controlled in all patients. Univariate analysis indicated that extent of transfusion, embolic material used, and injured branches were significantly associated with clinical failure. Conclusions: TAE is effective and safe in treating postoperative bleeding after PCNL. Massive transfusion, embolic material used, and injured branches were related to failure of bleeding control.


Background
With the development of endoscopy, Percutaneous nephrolithotomy (PCNL) has transcended traditional open surgery for renal stone management because it is less injurious and leads to faster recovery [1]. However, postoperative haemorrhage is a lifethreatening illness and remains one of the major causes of death after PCNL [2]. Although conservative approaches such as staying in bed, blood transfusion, and haemostatic drug 3 use can control most haemorrhaging after PCNL, patients with refractory bleeding have been referred to angiography for diagnosis. Moreover, TAE has been regarded as a safe and effective method to treat postoperative bleeding from PCNL, with embolization rates varying from 0.6% to 3.9% [3,4] Previous studies have described risks connected with bleeding after PCNL [5,6]

Angiography and embolisation technique
Transfemoral arteriography was performed under local anaesthesia and performed by five radiologists with 8 to 17 years' experience of interventional radiology work. Firstly, a 5-F vascular sheath was inserted into the right or left common femoral artery. The pigtail catheter was directed to the abdominal aorta for branch angiography to determine the renal artery, accessory renal artery, or other potential bleeding site. A 5-F catheter (Cordis, Miami, FL, USA; Terumo, Tokyo, Japan) was introduced into both internal renal 4 arteries. Superselective aortography of bleeding arteries was conducted by microcatheter (Tracker-18 or Renegade: Boston Scientific, Natick MA, USA; SP or Progreat: Terumo, Tokyo, Japan).
When a bleeding lesion was detected on angiography, embolisation was initiated.
Extravasation of contrast media, arteriovenous fistulae, and pseudoaneurysm were considered active bleeding foci. A variety of embolic materials was used. Gelatin sponge (GS) particles were cut with scissors to about 500-1,000 μm. Coils (stainless steel; Cook, Bloomington, MA, USA) or microcoils (Boston Scientific; Fibered platinum coils, Cook) were used for the feeding arteries, causing rapid extravasation to avoid extensive non-targeted embolisation, or occluding arteriovenous fistulae to avoid pulmonary embolism. After embolisation, angiography was performed through a Cobra diagnostic catheter to identify target vessel occlusion and identify any other potentially bleeding arteries. The end point of embolisation was reached when the contrast agent stopped completely on angiography without further active bleeding.
Patients were monitored by electrocardiography after TAE and remained in bed.
Haemoglobin levels were examined daily until no significant change was noted. Long-term follow-up included renal ultrasonography or computed tomography at 3 and 9 months, along with routine haemoglobin and kidney function tests.

Definitions and statistical analysis
Technical success was defined as the absence of the original bleeding lesion on angiography. Clinical success was defined as stable haemoglobin, absence of clinical symptoms, and radiographic absence of bleeding. The Society of Interventional Radiology divides the postoperative complications into minor or major[7].
Imaging and clinical outcomes of successful and failed TAE were compared with regard to postoperative haemorrhage after PCNL. The differences between the clinical and failed groups were assessed by Fisher's exact test or χ 2 test for categorical variables. Univariate analysis of continuous data was conducted using the Mann-Whitney U test. P <0.05 was considered statistically significant. All statistical analyses were performed by the Statistical Package for the Social Sciences (SPSS, Chicago, IL, USA).

Patients' characteristics
Characteristics of the cohort of 112 patients, including 66 males and 46 females with an average age of 54.3 years (range, 31-60 years), are summarised in Table 1 Thirty-three patients presented with intermittent bleeding, whereby their urine turned red discontinuously with gradual decrease in haemoglobin. Haemodynamic stability was observed in all of these patients. Twenty-two patients presented with slow continuous bleeding, whereby their urine turned reddish-tan without blood clots. Their haemoglobin declined slowly and haemodynamic stability was maintained.

Characteristics of TAE
Bleeding lesions were revealed on angiography in 106 patients, including 56 cases of pseudoaneurysm ( Fig. 1), 33 cases of arteriovenous fistula ( Univariate analysis showed that the injured branch, massive transfusion, and embolic material used were related to clinical failure (Table 1). Clinical success was not associated with age, gender, angiographic findings, number of bleeding arteries, and days spent in 7 hospital.

Discussion
PCNL has now become the first line of treatment for the management of complex renal calculi. However, postoperative bleeding is still a common and serious complication, which is self-limiting in most cases. TAE has been considered as an effective treatment for patients with PCNL who cannot be prevented from bleeding by conservative treatments such as staying in bed, haemostatic drugs, and renal fistula clamping [8,9] In our study, patients who underwent TAE using gelfoam alone had a statistically significantly higher reoperation rate than those with steel coils as embolisation materials.
The incidence of arterial laceration and pseudoaneurysm was high in patients with bleeding after PCNL, and the use of coils/microcoils was more effective than GS particles in achieving devascularisation, especially in the event of coagulation disorders and haemodynamic instability. In addition, in some patients with arteriovenous fistula the use of GS alone may lead to potential pulmonary embolism. Therefore, regardless of angiographic findings on TAE and the size of the haemorrhage, coils/microcoils should be the agents of first choice, followed by gelfoam [14].
With regard to bleeding arteries, in the present study the tertiary branch of the renal artery is the most commonly involved at 70%, compared with the main artery at 4.5%. The clinical success rate of patients with haemorrhage in the main renal artery was significantly lower than that in patients with branch and peripheral vascular haemorrhage.
Patients with injury to the primary artery were more likely to lose more blood than counterparts with collateral vessel injury, and most were already haemodynamically unstable prior to TAE. The reason for main renal artery embolisation may be distal blood reflux leading to continuous postoperative haemorrhage. Moreover, embolisation of the main artery may lead to further loss of renal function, thereby affecting cardiopulmonary function and potentially ending in death. At best, embolisation should occlude both the upstream and downstream vessels of the primary artery injury.
It has been reported that complications of embolisation such as coil migration, loss of renal function, renal artery dissection, and post-embolisation are rare [15]. In this study, minor complications occurred in 11 patients after TAE but with self-limited recovery.
Pneumonia was observed in two patients with arteriovenous fistulae found on angiography, in both of whom gelfoam alone was used. It is thus reasonable to surmise 9 that GS escape to the lungs and cause obstructive pneumonia. After 2 weeks in intensive care, both pneumonia patients were discharged after complete recovery. The clinical success rate in the current study was 93% for the first TAE and 100% for repeat TAE. TAE was thus considered an effective method to resolve post-PCNL severe haemorrhage.
There are several limitations to our study. First, it was retrospective and introduced selection bias. Second, the statistical power of comparing clinical success with clinical failure is low because the latter is a rare event owing to the small sample size (n = 8).
Third, the current study lacks some important variables, such as surgical experience, which may affect the incidence of vascular complications.
Conclusions TAE for haemorrhage after PCNL was effective and safe in this single-centre study.
Massive transfusion, embolic material used, and injured branches were related to clinical failure. TAE should be initiated in patients before haemodynamic instability occurs. This study was not supported by any funding.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors' contributions
HM analyzed and interpreted the patient data regarding the TAE. HM, YC, ZQ, HP and PY designed and performed the operation of TAE. HM was a major contributor in writing the manuscript. Besides, ZQ, HP and YC contributed in collecting the follow-up date afterwards. All authors helped in giving important revising suggestions. All authors read and approved the final manuscript.

Consent for publication
Written informed consent was obtained from the patient for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor of this journal.

Competing interests
The authors declare that they have no competing interests.