March 2009.indd SQU MED J, APRIL 2009, VOL. 9, ISS. 1, PP. 89-94, EPUB 16TH MAR 2009 SUBMITTED - 31ST AUG 08 REVISION REQ. 18TH NOV 08, REVISION RECD. 6TH DEC 08, ACCEPTED - 7TH DEC 08 POST-TRAUMATIC ABDOMINAL AORTIC aneurysms are relatively uncommon. We re-port a patient with post-traumatic pseudo-an- eurysm of the abdominal aorta, which presented with life-threatening complications due to its communica- tion with the portal venous system. This patient is im- portant to report because of the unusual problems en- countered during his management, especially during attempts at percutaneous embolisation of the fistula. C A S E R E P O R T A 35-year-old man was stabbed in the right upper quadrant of the abdomen. An exploratory laparotomy revealed a tear in the left lobe of the liver that was su- tured, and a retroperitoneal haematoma in the upper abdomen which was left undisturbed. His postopera- tive course was uneventful and he was discharged on the tenth postoperative day. Five days later, he had a small amount of haematemesis which stopped sponta- neously. After that, he had recurrent episodes of hae- matemesis once or twice a week. None of these was associated with hypotension, nor did he require hospi- talisation or blood transfusions. Two months after the initial injury, he presented at the Department of Gas- trointestinal Surgery and Liver Transplantation of the All India Institute of Medical Sciences with a history Traumatic Aorto-Mesenteric-Portal Fistula: Percutaneous management Case Report *Rajeev Jain,1 Girish K Pande,2 Peush Sahni,2 Dev N Dwivedi2 C A S E R E P O R T اجللد طريق عن العالج : يّ حِ الرَضْ ٌّ البَابِيّ يّ املَساريقِ رِي بْهَ الناسور األَ حالة تقرير دويفدي ديف ، ساهني باند ، بوش جيريش ، راجيف جني مع دموي وصدمة يعاني من تقيؤ ــنة 35 س عمره للعلوم الطبية رجل الهند كل معهد الى جاء البطن في جرح قاطع ــهرين من ش امللخص: بعد البطني األبهر في دَمٍ كاذِبَة مُّ أَ بني يربط ناسورا كبيرا ــفت كش البطنية األوعية وتصوير للبطن ــب َوسَ احملُ يٌّ عِ طَ َقْ ويرٌ امل التَصْ . البابِيّ مِ الدَّ طِ غْ ضَ فَرْطُ الناسور الستئصال عملية نهائية إجراء ثم ومن الدموي ، مت تثبيت الضغط اجللد حيث طريق ــور عن الناس اغلق . لْوِيّ العُ يُّ ــاريقِ َس امل الوَريدُ من فرع و واصالح العيب األبهري. الهند. ، حالة ، تقرير التدخل عمليات ، األبهر تصوير ، أم الدم الكلمات: األبهر ، مفتاح 1Department of Radiology, College of Medicine & Health Sciences, Sultan Qaboos University, Muscat, Sultanate of Oman; 2Department of Gastroin- testinal Surgery and Liver Transplantation All India Institute of Medical Sciences, Ansari Nagar, New Delhi, India *To whom correspondence should be addressed. Email: rajeev.squ.edu.om ABSTRACT Two months after a stab injury to the abdomen, a 35-year-old male presented at the All India Institute of Medical Sciences with haematemesis, shock and portal hypertension. Computed tomography of the abdomen and abdominal angiography revealed a large fistulous communication between an abdominal aortic pseudoaneurysm and a branch of the superior mesenteric vein. The fistula was occluded percutaneously, which allowed the patient to stabilise haemodynamically and, finally, to undergo a definitive surgical excision of the pseudoaneurysm and repair of the aortic defect. Keywords: Aorta; Trauma; Aneurysm; Aortography; Interventional procedures; Case report; India. 90 R A J E E V J A I N , G I R I S H K PA N D E , P E U S H S A H N I A N D D E V N D W I V E D I 2 of progressively increasing abdominal distension for one week, decreased urine output for three days and a large bout of haematemesis just prior to admission. At the time of examination he was anxious and anaemic. He was tachycardiac (140/minute), hypoten- sive (systolic 70mmHg), and tachypnoeic (36/minute). There were tense ascites and an upper abdominal bruit. All his peripheral pulses were palpable and equal in volume. A diagnostic paracentesis revealed straw-col- oured fluid, and the nasogastric aspirate showed fresh blood. His haemoglobin was 7 g/dl, blood urea 241 mg/dl, serum creatinine 3.5 mg/dl, serum sodium 129 mEq/ dl and serum potassium 6.6 mEq/dl. The serum bi- lirubin was 0.7 mg/dl, alkaline phosphatase 313 U/L, serum aspartate aminotransferase/serum alanine ami- notransferase (AST/ALT) 97/108 U/L [Table 1], and the prothrombin time was markedly prolonged (more than 60 seconds; control of 12 seconds). Blood gas analysis showed metabolic acidosis. He was resuscitated with intravenous crystalloids and blood transfusions. Endoscopy showed congested and oozing gastric mucosa, but no gastro-oesophageal varices. A large volume paracentesis was done to im- prove the patient’s respiratory status. Despite this, he remained oliguric, acidotic, tachypnoeic and haemo- dynamically unstable. The ascites re-accumulated rapidly (within hours). Administration of fresh frozen plasma improved his prothromin time to within 5 sec- onds of the control value. A contrast enhanced computed tomography scan showed massive ascites. There was a dilated, tortuous vascular channel anterior to the aorta at the level of the origin of the renal arteries. This structure (pseu- doaneurysm) was in direct communication with a branch of the superior mesenteric vein (SMV). The draining superior mesenteric and portal veins were di- lated (2.5cms diameter) and showed marked enhance- ment with contrast [Figure 1]. The liver was markedly hyperdense compared to the spleen, which showed lit- tle enhancement. The kidneys showed poor enhance- ment and the aorta below the site of abnormality had a smaller lumen, indicating a shunt from the aorta into the portal venous system, resulting in ‘functional’ por- tal hypertension. An abdominal angiogram confirmed the presence Figure 1: Contrast enhanced scan of the abdo- men at the level of the renal vessels shows the following structures. [1] - aorta, [2] - a large, vascular channel (pseudoaneurysm) just anterior to the aorta, [3] - the draining branch of the SMV, [4] - dilated SMV, and [5] - retrograde flow into splenic vein. Note the presence of gross ascites and the differential enhancement of the liver and spleen. The punctate high-density structures in the head of the pancreas represent dilated venules secondary to high-pressure ret- rograde flow in the mesenteric-portal venous system Days after embolisation 0 1 2 5 12 15 Urea (mg/dl) 212 209 201 117 12 12 Serum creatinine (mg/dl) 2.3 3.2 3.4 1.7 0.6 0.6 Serum bilirubin (mg/dl) 0.8 0.8 0.7 0.7 0.7 0.7 Serum alkaline phosphatase (u/L) 332 330 412 504 653 290 Aspartate aminotransferase (u/L) 481 5730 2090 147 52 36 Alanine aminotransferase (u/L) 436 2080 1295 365 106 23 Table 1: Liver and renal function tests after embolisation 91 TR AU M AT I C A O R T O - M E S E N T E R I C - P O R TA L F I S T U L A : P E R C U TA N E O U S M A N A G E M E N T of an 8mm wide communication between the aorta (at the level of the origins of the renal arteries) and a large pseudoaneurysm. The latter, in turn, was drain- ing through a 1cm wide defect into a dilated (1.5cm) branch of the SMV [Figures 2a and 2b]. As an emergency measure, it was decided to attempt an occlusion of the aortic opening of the pseudoaneu- rysm, with the possibility of encouraging thrombosis of the abnormal vascular channel. An 8mm diameter balloon catheter was placed across the aortic end of the pseudoaneurysm, and the balloon gently inflated until a waist was identified, to block the aortic opening temporarily [Figure 3], and with the hope of inducing thrombosis of the pseudoaneurysm. Over the next 6-8 hours the patient became normotensive, his urine out- put increased, he did not require further paracentesis and his respiratory parameters improved. A follow-up angiogram after 12 hours revealed a patent fistulous tract. An embolisation of the fis- tula was attempted using large diameter coils. Two 12mm diameter 3.5-turn platinum coils were placed at the junction of the branch vein and the SMV. This was followed by release of multiple steel coils into the pseudoaneurysm [Figure 4a]. This resulted in cessa- tion of the blood flow through the fistula into the SMV [Figure 4b]. The abdominal bruit disappeared and the patient had a rapid clinical recovery with resolution of the ascites. The serum biochemistry also showed a gradual but sustained improvement [Table 1]. As the ascites disappeared, a small, ill-defined, non-expansile lump became palpable in the epigastric region. Ten days later, the abdominal bruit reappeared. Figure 2: Abdominal aortogram, catheter tip at L1 vertebral level - (a) lateral, and (b) anteroposte- rior projections. Angiogram demonstrates a large fistulous communication between the aorta and a branch of the SMV, at the L2 vertebral level. Most of the injected contrast is diverted into the portal venous system, with hardly any flow into the renal arteries. The superior mesenteric and portal veins are grossly dilated. Numeric labelling in the figures indicates the same structures as in Fig. 1. ([6] - portal vein) 92 R A J E E V J A I N , G I R I S H K PA N D E , P E U S H S A H N I A N D D E V N D W I V E D I 2 Angiography demonstrated recurrence of a small amount of flow across the fistula and into the SMV. Release of more steel coils into the pseudoaneurysm resulted in cessation of flow into the SMV, but a small amount of flow persisted through the aortic opening into the pseudoaneurysm. Therefore, the patient was offered a surgical repair. He, however, refused surgery at that time due to social reasons. He was discharged from the hospital with normal renal and liver function tests. He was readmitted 3 weeks later and had remained well during this period. The abdominal lump and bruit were still present but there was no ascites. The haemo- gram, renal and liver function tests were within nor- mal limits. An ultrasound Doppler study confirmed the persistence of blood flow across the aortic open- ing into the pseudoaneurysm. However, the superior mesenteric and portal venous blood flow was normal with no evidence of arterial pulsations. The patient underwent surgical exploration through a midline laparotomy. A 4 x 5cm size retroperitoneal mass was palpated. The mass was densely adherent to the SMV. The abnormal communication was success- fully excised with repair of the aorta and the SMV. Postoperatively, the patient had an uneventful re- covery. At follow-up clinical examinations 2 and 4 months after the surgical procedure, his renal and liver functions were normal, and Doppler studies on both occasions did not demonstrate any recurrence of the fistula. D I S C U S S I O N A number of patients with aortic injury exsanguinate prior to reaching a hospital. The mortality has been reported to be up to 70% even among those who sur- vive long enough to undergo surgery. These injuries are often associated with other visceral and vascular injuries of the abdomen that may obscure the diagno- sis. In some patients, a fistulous communication may occur between the aorta and a major vein, either as a result of trauma or spontaneous rupture of a pseu- doaneurysm. This may control the aortic bleeding and might lead to the diagnosis being missed at the initial laparotomy.1 Such patients usually survive the initial injury and may present weeks to months later, with manifestations of portal hypertension or congestive heart failure depending on the site of communication - the portal system or the vena cava.2 We came across only four previous reports in the literature of traumatic aorto-portal fistulae. In one patient, a preoperative angiogram demonstrated the abnormality, which was corrected surgically.3 Two patients presented with signs of portal hypertension, 14 and 18 months after the initial laparotomy. The aorto-portal fistulae in these patients were managed surgically,4 and with angiographic embolisation,5 re- spectively. The fourth patient had presented one year after the initial laparotomy, with a two months’ his- tory of melaena, intermittent haematemesis and signs of right heart failure, and was treated surgically.6 In our patient, the aortic injury was overlooked at the initial laparotomy. The recurrent gastrointesti- nal bleeding and the gradually progressive abdominal distension preceding his presentation at our institu- tion suggested a progressive increase in the blood flow across the fistula leading to severe ‘functional portal Figure 3: Angiographic balloon dilatation cath- eter inflated across the aortic opening. The site and size of the opening are identified by the waist of the balloon (arrow). The distal part of the balloon, tip and coiled guidewire are across the aortic opening within the pseudoaneurysm 93 TR AU M AT I C A O R T O - M E S E N T E R I C - P O R TA L F I S T U L A : P E R C U TA N E O U S M A N A G E M E N T hypertension’, and prerenal azotaemia compounded by hypovolaemia due to multiple episodes of haemate- mesis. The gastrointestinal bleeding was the result of marked congestion of the gastric mucosa - a picture akin to severe portal hypertensive gastropathy. How- ever, the absence of varices suggests that there was a rapid increase in the blood flow across the fistula over a few days prior to his presentation to us. Angiographic management of the aorto-mesenter- ic-portal fistula presented an interesting challenge. The ideal technique for occlusion of a similar fistula in a more peripheral location would have been place- ment of a covered metallic stent graft. However, in this case, the site of the fistula at the level of renal arter- ies, and in close proximity to the origin of the superior mesenteric artery, precluded the use of a covered stent graft. The next option could have been a detachable balloon, but a balloon large enough to occlude the 4cm diameter pseudoaneurysm was not available. The presence of rapidly accumulating large volume ascites prevented pre-embolisation percutaneous transhepat- ic occlusion of the junction of the branch vein with the SMV. Therefore, to tide over the immediate crisis, we inflated a balloon catheter across the aortic opening of the fistula. This was successful in occluding the open- ing, and the patient’s condition stabilised. Subsequently, the pseudoaneurysm was success- fully embolised with multiple metallic coils and the patient improved rapidly. However, the aortic end of the pseudoaneurysm could not be fully obliterated due to the rapid flow and its large size. This resulted in persistent blood flow into the pseudoaneurysm (but not into the mesenteric vein) with reappearance of a Figure 4a: Lateral projection shows large diam- eter platinum coils occluding the junction of the branch vein with the SMV (arrow), and multiple steel coils within the pseudoaneurysms (arrow- heads) Figure 4b: Abdominal aortogram performed after embolisation of the fistula with steel coils, showed good flow in the distal aorta, with filling up of the major branches. The pseudoaneurysm (arrows) showed minimal filling from the aortic opening. There was no flow of contrast into the portal venous system. Good flow of contrast is now shown in both renal arteries (arrowheads) 94 R A J E E V J A I N , G I R I S H K PA N D E , P E U S H S A H N I A N D D E V N D W I V E D I 2 bruit necessitating surgical repair. Though we did not attempt this, an injection of thrombin solution has been used successfully to effect complete thrombosis of pseudoaneurysms.7 In retrospect, two alternative techniques could have been employed for occlusion of the fistula. The first was the placement of an atrial septal defect oc- cluder across the aortic opening; the potential draw- backs of this technique would have been the possibili- ty of occluding the renal artery origins, and prevention of any subsequent surgical repair of the aortic defect. Theoretically, it should also have been possible to per- form a transjugular transhepatic approach into the SMV (similar to a transjugular intrahepatic porta- systemic stent shunt [TIPSS]) followed by placement of a large diameter covered stent across the opening of the branch vein into the SMV, prior to transaortic embolisation of the pseudoaneurysm; however, we did not have the expertise for performing TIPSS at our in- stitution. Also, this approach would have resulted in a permanently dilated (2.5cm diameter) segment of the SMV (due to the implanted metallic stent), with the potential for thrombosis and infection. After successful embolisation, the serum transami- nases increased [Table 1] and then gradually normal- ised. This probably happened due to a relative ischae- mia secondary to the disruption of arterial blood flow through the portal vein. The subsequent normalisation suggested that the hepatic arterial flow had returned to normal volume. Unlike the previous case of aorto-portal fistula treated successfully by embolisation,5 in our patient angiographic embolisation did occlude the fistula, but could not block the large aortic opening of the pseudoaneurysm. However, it helped to stabilize the patient, restore normal liver function and allowed a delayed elective surgical procedure. To conclude, post-traumatic aorto-portal fistula is a rare complication. This patient presented with sec- ondary functional portal hypertension, and shock due to the steal phenomenon because of the large diameter, high-flow aorto-mesenteric-portal fistula. Unconven- tional percutaneous techniques had to be employed to occlude the fistula, which subsequently permitted elective surgery. C O N C L U S I O N This report describes the rare condition of post-trau- matic aorto-mesenteric-portal fistula with functional portal hypertension and its emergent radiological management as a bridge to definitive surgical therapy. The techniques involved in the successful radiologi- cal management were dictated by the location of the large fistulous communication in proximity to origins of important vascular structures, which could not be occluded, hence the difficulty encountered in using standard occlusion devices. This in turn prompted the use of unconventional equipment and procedures in an emergent setting. A C K N OW L E D G M E N T S This work was performed at the Departments of Ra- diodiagnosis and Gastrointestinal Surgery and Liver Transplantation at the All India Institute of Medical Sciences, Ansari Nagar, New Delhi, India. R E F E R E N C E S 1. Robertson G. Combined stab wounds of the aorta and vena cava of the abdomen. Arch Surg 1967; 95:12-15. 2. Saunders MS, Riberi A, Massullo EA. Delayed traumat- ic superior mesenteric arteriovenous fistula after a stab wound: case report. J Trauma 1992; 32:101- 6. 3. Buscalgia LC, Blaisdell FW, Lim RC Jr. Penetrating ab- dominal vascular injuries. Arch Surg 1969; 99:764-9. 4. Little JM, Sheldon DM, Mills FH. Traumatic aorto-por- tal vein fistula: Repair using intraperitoneal hypother- mia. Am Surg 1968; 34:350-3. 5. Raabe R, Lawrence PF, Luers PR, Miller FJ. Radiograph- ic and clinical findings in unusual abdominal aortic an- eurysms. Cardiovasc Intervent Radiol 1986; 9:176-81. 6. Epstein BM, Bocchiola FC, Andrews JC, Bester L. Case report: Traumatic arterio-venous fistula involving the portal system. Clin Radiol 1987; 38:91-3. 7. Krüger K, Zähringer M, Söhngen F-D, Gossmann A, Schulte O, Feldmann C, et al. Femoral pseudoaneu- rysms: management with percutaneous thrombin in- jections—success rates and effects on systemic coagula- tion. Radiology 2003; 226:452