Introduction

Uretero-iliac artery fistula (UIAF) is a recognized but uncommon condition due to multiple factors. Few cases are described in the literature (< 150) [1], but incidence is increased in the last years, since at the beginning of 1990s only 20 cases were described [2]. Diagnosis and management are still a challenge for urologist [2]. Uretero-arterial fistulas can develop with aorta, common iliac artery, external iliac artery and hypogastric artery. Aortic fistulas are usually associated with aneurysms; hypogastric fistulas are very uncommon. Fistulas between ureter and common or external iliac artery are usually associated with a history of pelvic surgery, pelvic irradiation, chronic ureteral stenting and vascular disease [3]. The presentation is usually massive gross hematuria with acute anemia unto to hemorrhagic shock, representing a life-threatening condition. High mortality rate is reported (7–23%) in literature [4]. However, endovascular techniques have improved morbidity and mortality [3, 4]. Gold standard for diagnosis are angiography and computed tomography (CT) scan but, however, these methods have a low sensibility when fistula is too small, and the diagnosis can be misunderstood; in these cases, patients often undergo explorative laparotomy [5]. The aim of this study is to describe four cases of UIAF in three patients treated by classical and endovascular approach.

Materials and methods

This single-center, retrospective observational study was based on data obtained from patients’ medical records from 2013 to 2018 in our institution. We collected four consecutive cases in three patients who were diagnosed with UIAF. UIAF was defined as the ratified presence of an abnormal communication between the ureter and any artery. In all patients, the UIAF was initially evaluated by contrast-enhanced CT angiography. The management strategy was defined individually based on the specific risk profile of each patient. Clinical characteristics, diagnosis, management and follow-up of patients affected by UIAF were recorded in an Excel database and were analyzed retrospectively using descriptive analysis.

Results

All patients were female. Mean age was 66 (57–83). All patients underwent radical hysterectomy and radiation therapy for cervical carcinoma 13 (11–15) years before. Two patients underwent radical cystectomy with Bricker diversion for actinic cystitis. All patients had bilateral ureteral stent for stenosis of ureter (one patient) or ureter-iliac anastomosis stenosis (two patients). According to our database, we do not know exactly how many years the patients had stents. Presentation was the same for all four cases: massive gross hematuria during periodic change of ureteral stent. In our Institution we used silicone stents and we perform the periodical change every 3 months with a fluoroscopy-guided out-patient procedure; the technique used is the placement of a guidewire into the previous stent, the subsequent removal of the stent and the placement of the new stent on the guidewire. In every case, the massive hematuria presented during the removal of the stent after the guidewire placement. Immediate management was the stabilization of vital parameters with infusion of colloidal solutions and adrenergic drugs. Blood transfusions were necessaries in all cases. Hematuria was temporary auto-resolved in all cases. Two patients were submitted to a CT-scan with delayed ureteral phase, finding clots in renal pelvis which, along with hemodinamical instability, raised the suspicion of active bleeding from the renal parenchyma, without identifying an arterial vessel to be embolized. For these reason, a right radical nephrectomy was performed (Fig. 1). After radical nephrectomy, both 2 patients had a massive hemorrhage from the drain in the first post-operative day and angiography finally reported the fistula between distal tract of right ureter, left in place during nephrectomy, and right common iliac artery. One of these patients 3 years after developed a contralateral fistula in the left side, diagnosed with massive hematuria during change of the left ureteral stent, but angiography and CT scan did not show the fistula. The other patient had diagnosis through angiography (Fig. 2) during second episode of massive hematuria because in the first presentation angiography and CT scan resulted negative. In all four cases, after the detection of the UIAF, a vascular procedure was performed, and an endo-prosthesis was put in common iliac artery to solve the fistula; endovascular procedures resulted safe, without complications; observation of patients was continued in hospital for 18 (5–21) days. Ureteral stent was removed in all cases and a nephrostomy tube was placed to drain the kidney. After placement of endoprosthesis, one patient underwent massive deep vein thrombosis despite low molecular weight heparin prophylaxis; an inferior caval vein filter was placed and removed after 3 months of anticoagulant therapy. Follow-up is of 49 (25–66) months and resulted uneventful in all patients (Table 1).

Fig. 1
figure 1

Intra-ureteral contrast medium in areterial phase of CT-Scan

Full size image
Fig. 2
figure 2

Left ureter-iliac artery fistula detected during angiographt

Full size image
Table 1 Clinical characteristics, diagnostic modalities, management strategies, and clinical outcomes in the three cases study patients affected by UIAF
Full size table

Discussion

In our experience, UIAF occurred in all cases in female patients with previous medical history of hysterectomy and radiation therapy for cervical carcinoma. In the literature, 57% of UIAF occur in women at a mean age of 58 years [6]. Two patients underwent also a Bricker diversion for actinic cystitis. All patients had an indwelling stent for stenosis of ureter or ureteral-iliac anastomosis. We know that radiation is a risk factor for stenosis of uretero-iliac anastomosis in Bricker diversion and the ureteral stent is the most adopted treatment. The presence of an indwelling stent and periodical change is a risk factor for inflammation, traumatic lesions, infections and hematuria; the risk is increased in radiated patients [7]. The ureter crossing iliac vessels makes a turn and is narrower than other tracts: it may suffer more traumatism during stent change. Furthermore, radiated tissues are less resistant and softer and are predisposed to more damage. In our cases those factors can have contributed to lesion the ureter forming fistula with the common iliac artery. Maybe also the constancy and the hardness of stent may had contributed to increase the damage. Gross hematuria was episodic. One possible explanation is the increase of blood pressure in some moments, for example due to anxiety for the change of stent, made as outpatient procedure without anesthesia, a possible valve mechanism of fistula can explain passage of blood with arterial pressure from artery to ureter overcoming the flap of valve. Another possible explanation is the presence of stent compressing the fistula and preventing the continuous passage of blood. Those reasons can explain the difficulty to detect fistula with angiography when hematuria is temporary solved; the false-negative result during diagnosis is due to the closing of the flap of valve in that definite moment. In our experience blood refluxed in renal pelvis providing a false positive for kidney hemorrhage and lad to radical nephrectomy in two cases; also, chronic pyelonephritis (probably due to a chronic presence of a stent) increased the false imaging of a blooding kidney. During nephrectomy, ureter was cut proximal to the cross with iliac artery because absence of suspicious of the fistula and without an history of urothelial cancer. For this reason, after the passage in the ureter, blood went in the renal loggia and drained, leading to repeat CT scan and finally diagnose the fistula. The patient with late contralateral fistula and the other patient whit negative first angiography, underwent immediate placement of preventive endoprosthesis relying only on clinical suspicious. Angiography in our experience represents the most valid diagnostic option in emergency because of the possibility to detect immediately the fistula, meanwhile the CT scan can delay diagnosis and treatment. Laparotomy was unnecessary for diagnosis and treatment in our experience, but in many cases reported in the literature it can be an effective method to solve the condition [8]. In cases with achieved hemodynamic control, a multidisciplinary discussion with radiologist, interventional radiologist, vascular surgery is mandatory to settle and solve the problem [9, 10]. Endoprostheis seems to be very effective: with a mean follow-up of 31 months we did not observe any complication. In addition, in the literature is resulted to be the preferred option [11, 12]. Okada et al. reported 36% of recurrence of hematuria in a series of 11 patients underwent placement of endoprosthesis for UIAF, with need for a surgical approach in two patients [13]. The bigger series reported in the literature, by Fox et al. comparing retrospectively endovascular versus surgical approach, did not show a difference between the two methods, reporting the same rate of recurrence of hematuria, complications and mortality with a follow-up of 15 months [14, 15] (Table 2). Open surgical treatment is still described in the literature, especially in patients with first failed an endovascular treatment or enteric contamination, abscess, and local sepsis, but we did not experience this treatment in our experience. One important topic is the management of urological condition. After a placement of endoprosthesis, is advisable to remove indwelling ureteral stent to avoid further complications. Nephrostomy tube [16, 17] is the preferred option. In patients with discussed risk factors a softer and smaller stent or a preventive nephrostomy tube can be the best prevention method. Several limitations of our study should be observed. Our data were retrospective in its design and had a low level of evidence. This is, however, inevitable in studies of highly rare status, resulting in a small sample size. Therefore, future studies on larger cohorts of patients are desirable.

Table 2 Revision of the literature and comparison with our experience
Full size table

Conclusion

Our study identifies different factors that were present in all our patients and might be related to an increased risk of developing UIAFs: concomitant presence of a medical history of previous pelvic malignancy treated with surgery or radiation-therapy, urinary diversion and indwelling ureteral stent in female patients. Gross hematuria during stent change in those patients must be suspected of ureteroarterial fistula and the clinical suspicious is the key factor for immediate diagnosis of this urological emergency. Selective angiography is useful to confirm the diagnosis while subsequent endovascular endoprosthesis represents a safe approach with reduced post-procedural complications.