Uretero-Arterial Fistula: A Modern Case of an Orphan Disease

Introduction

Uretero-arterial fistula (UAF) is an uncommon source of hematuria, but as a harbinger of mortality, diagnosis can potentially be lifesaving. While rare, there is a general consensus that the incidence is rising. The condition develops when an aberrant connection is made between a ureter and artery, and the presentation varies from microscopic hematuria to frank hemorrhage. UAF is seldom expected, but having a clinical suspicion is paramount for guiding decision-making. There have been some case reports, but data is scarce regarding etiology and management [1, 2, 3, 4]. Thus, we offer our personal experience to help validate current diagnostic and treatment options. Keywords: Uretero-arterial fistula; Cystoscope; Looposcopy; Pyelogram

The Case

A 66-year-old male was diagnosed with stage IV urothelial cell carcinoma of the bladder (pT2N1M0) in 2008. He underwent neoadjuvant chemotherapy with gemcitabine and cisplatin (GC), followed by a radical cystoprostatectomy with ileal loop urinary diversion and bilateral extended pelvic lymph node dissection in 2009. He developed recurrence of disease at the aortic bifurcation, abdominal wall, and ureteral margin in 2010, 2013, and 2014, respectively. This was treated with additional chemotherapy, as well as 3 reoperations (lymph node dissection, resection of the abdominal wall mass, and conduit revision), and adjuvant radiation. He sought the opinion of several major NCCN designated cancer institutions and was enrolled in a clinical trial after a lymph node biopsy was positive for a Src mutation, which led to targeted chemotherapy with bosunitinib. Post-treatment, he developed widespread abdominal and pulmonary recurrences. He was given another 5 cycles of GC, which this time showed regression of metastasis. In October 2015, the patient was admitted to the hospital for a small bowel obstruction and was found to have renal insufficiency with bilateral hydronephrosis. This was due to stricture of the distal ureters which was initially managed with bilateral nephrostomy tubes before being converted to 8.2F single J ureteral stents. Shortly after this procedure, the stents needed to be exchanged due to stent migration and malposition. Following this, he was managed with stent exchanges every 3 months. During his stent exchange in April 2016, wires were passed up the single J stents emanating from the ileal conduit. The right stent was exchanged, followed by the left. Upon removal of the left stent, there was some resistance noted, which was overcome with gentle, steady pressure. Immediately upon removal, profuse bleeding was seen coming from the urostomy. A finger was placed in the conduit and anesthesia was asked to place another IV line, and obtain a type and cross. Unsure of where the bleeding was coming from, Interventional Radiology (IR) was contacted for an immediate intra operative consult. Several minutes later, when the interventional radiologist arrived, the finger was removed and no more blood was seen coming from the stoma. The conduit was irrigated free of the extensive clots that had accumulated. Looposcopy was then performed with a flexible cystoscope showing a small amount of oozing from around the wire at the left ureteral orifice. The concern at this point was for possible erosion of the stent into the parenchyma of the kidney (as evidenced by the resistance upon removal), and thus after discussion it was decided that there was no need for IR intervention. We continued with our procedure, and a 5F open-ended catheter was advanced over the left ureteral wire and a retrograde pyelogram was performed. This showed multiple filling defects within the renal pelvis consistent with blood clots (Figure 1). The new single J stent was placed and our patient was admitted for observation.

Click to enlarge

Figure 1: Retrograde pyelogram at the time of stent exchange. Multiple filling defects are seen within the left renal pelvis representing blood clot. As blood entered the ureter via the fistula, it traveled retrograde and collected in the renal pelvis. Additionally, no contrast was seen entering the fistula or artery either because the fistula had either already clotted off, or the pressure within the fistula was greater than that of our injecting pressure.

Discussion

There have been two studies, dating back to 1997, which looked at UAF retrospectively, both examining the incidence and risk factors for developing UAF, treatment options, and the outcomes with proposed algorithms for disease management [5, 6]. There are two types of UAF: primary and secondary. Primary UAF is less common, and comes from intrinsic pathology in the patient (i.e. arterial aneurysms, atherosclerosis, or arterial-venous malformations). Secondary UAF develops with external risk factors such as a history of pelvic surgery, radiation therapy, or indwelling ureteral stents [1, 7, 8]. The patient in our case had 4 of these conditions. A common element among these secondary causes is the creation of a fixed and more rigid surface for the systolic arterial pulsations to press against, theoretically inducing pressure necrosis and fistula development [7] (Figure 3). About 80 cases of UAF have been reported since 2004 [2, 5, 6, 9] and only a handful of case reports since then. Additionally, for historical reference, there were 3 cases of UAF found in pregnancy, each dating back to the late 1930s [10, 11, 12]. All 3 women had complicated UTIs and died from either hemorrhage or sepsis, and the diagnosis of UAF was made at autopsy.

Click to enlarge

Figure 3: Non-contrast CT of the abdomen and pelvis. a.) Axial view with the abdominal window shows the left ureteral stent passing (arrow) over the left common iliac artery (arrow head). b.) Coronal view in the bone window shows more clearly the stent (arrow) coursing over the iliac artery (arrow head). There is also significant atherosclerosis of the abdominal aorta and iliac arteries. This allows for 2 rigid and relatively fixed environments, which over time sets the stage for fistula formation.

Making the diagnosis can be a major challenge. While hematuria and even bloody efflux seen on cystoscopy should heighten suspicion, this does not make the diagnosis of UAF. Retrograde pyelogram, and CT scan both have low sensitivity for diagnosis, ranging from 45- 50% [7]. Often the fistulous tract can be obscured by a stent on CT scan, plugged by a clot, or the arterial pressure may be enough to overcome the injection pressure of contrast dye on retrograde pyelogram, preventing visualization the fistula. What has become the most sensitive and specific diagnostic test is a provocative angiogram. This involves the manipulation of a ureteral stent just prior to angiogram so that the fistula tract is freshly opened. Tagged RBC scans are not part of the suggested work up because the scan requires active bleeding to make a diagnosis and offers poor anatomic detail. Patients who are bleeding significantly enough to be seen on tagged RBC scan are usually not stable enough for the test. Once the diagnosis is made, prompt treatment is needed, and the options universally require an intervention, either open surgical or endovascular. More specifically, open ligation of the fistula, arterial bypass grafting, and/or nephroureterctomy are the surgical options, while endovascular grafting, or coiling are the minimally invasive techniques. Ureteral stenting is not recommended because at best this will temporize the problem, and potentially exacerbate (i.e. inducing hemorrhage, or enlargement of) the fistula, which is what we saw in our case. There are a multitude of factors to consider when choosing an intervention, such as surgeon abilities, availability of endovascular services, and prior surgical interventions. Due to its minimally invasive approach and ability to perform treatment at the time of diagnosis, endovascular grafting is gaining popularity as the treatment of choice. However, Krambeck et al. views this only as a temporary measure because placing the graft and not addressing the fistulous connection increases the risk of graft infection [7, 6]. Thus, their algorithm would incorporate urinary diversion (i.e. nephrostomy tube) for those patients receiving an endovascular stent. However, only those patients who are poor surgical candidates would receive this stent in the first place, since their model favors primary open surgical repair for those deemed eligible. Our patient ended up having a urinary diversion in concordance with this algorithm.

Closing Thoughts

It can be difficult to anticipate a UAF, and any single physician is unlikely to encounter the disease more than once, if at all, in their career. This is a dangerous combination for providing efficient and successful treatment outcomes. Patients with UAF are often complex both medically and surgically, which adds more layers of difficulty in diagnosis and treatment. Our patient had a significant delay in diagnosis, but was fortunate to have a good outcome. Furthermore, our institution was equipped with the expertise and technology available to eventually provide an accurate diagnosis and curative treatment. This case provides heightened awareness for diagnosing UAF and the current algorithm led to a successful outcome in our patient.