[Aorto-enteric fistulas].

The aorto-enteric fistula (AEF) is a direct communication between aorta and intestinal lumen. There are primary and secondary forms. Primary AEFs are usually due to erosion of an aortic aneurysm (AAA) into the intestine, while secondary forms are caused by reconstructive procedures on the abdominal aorta. The incidence of primary AEF ranges from 0.1 to 0.8%, and secondary from 0.4% to 2.4% [2-4]. The mortality rate after surgical treatment of secondary AEFs is from 14% to 70% [5]. Therefore, they are of great medical importance. The aim of this paper is the presentation of 9 new cases. Over a 33-year period (1966-1999) a retrospective analysis of patients' records identified 9 patients with AEFs. All were males with average age of 66.62 (51-70) years. In Tables 1 and 2 are presented data on our cases. Of the total number of 9 patients, there were 4 primary and 5 secondary AEFs. All primary fistulas were caused by AAA rupture. Secondary AEFs developed after aortic abdominal surgery in the period between one and seven years after the operation. In 7 cases fistula involved the duodenum, in one the sigmoid and in one the transversal colon. The dominant manifestation of fistulas was gastrointestinal bleeding: melaena--8 (89%); haematemesis and melaena--2 (22%); proctorrhagia--1 (11%). In cases of primary AEFs gastrointestinal bleeding was followed by low back pain and haemorrhagic shok, while in cases of secondary AEFs by sepsis (fever, increased leucocytes count, sedimentation). In two cases the final diagnosis was established by gastrography and colonoscopy, while in two patients Duplex ultrasonographic examination suspected AEF. In all other cases the diagnosis was established intraoperatively (Figure 1). After aneurysmal resection in cases of primary AEFs, revascularization of the lower limbs was performed with extra-anatomic axillo-bifemoral bypass graft (one case) and with "in situ" graft placement (three cases) (Figure 2). The duodenal defect was closed transversally with standard two layers suture techniques in two patients without fistula excision, and in two cases after fistulas excision. In one case associated gastero-entero and entero-entero anastomosis was performed. In all cases with secondary AEFs, after removing of the previously implanted aortic graft, the aorta was closed just below the renal arteries root, and wrapped with a vascularized pedicle of omentum, to separate it from the bowel and the contained area. The duodenal defect was closed after fistulas excision using two layers transversal suture technique in two cases, and in one patient with large fistula a partial duodenectomy and Roux's procedure were necessary. In two patients in whom AEFs involved the transversal and sigmoid colon colostoma was performed. In three cases an extra-anatomic axillo-bifemoral bypass graft was performed for lower limbs revascularization, and in one patient bypass from the ascendent aorta to the femoral artery, using retroperitoneal route was carried out. In one patient the revascularization of the lower limbs was not done because of intraoperative death of the patient. Seven of our patients died during the first 15 postoperative days. One died during the operation after massive acute myocardial infarction. In other six cases the mortality causes were: MOFS-3 cases, and secondary enteric fistula-3 cases. Two of our patients survived. One has been followed-up for 15 years, and his axillo-bifemoral bypass is patent. The other with bypass from the ascendent aorta to the femoral artery died 7 years after the operation, also with patent graft. More details are given in Table 3. Sir Astley Cooper was the first who described primary AEFs caused by AAA rupture in 1817 [6], and Brock in 1953, first described secondary AEF developed 6 months after aortic homograft implantation [8]. In 1957, Haberer successfully treated primary AEF by suture of the duodenal defect and aneurysmorrhaphy [9]. In our country Stojanovitsh and Vujadinovitsh in 1966, first treated primary AEF [16]. Their patient died due to MOFS. However, in 1984 and 1985, Lotina successfully treated two patients with secondary AEFs [11] (Figure 3, Sheme 1). The authors also analyzed literature data on the aetiology, pathogenesis, clinical manifestations, diagnosis and treatment of AEFs. In conclusion, the authors suggest: 1. "Omega" extra-anatomic bypass from supraceliac artery trough retroperitonely to femoral arteries; 2. "In situ" replacement of the abdominal aorta using cadaveric homografts; 3. Intraoperative control of bleeding with endoluminal balloon occlusive aortic catheter.

Davidovic LB ，Spasic DS ，Lotina SI ，Kostic DM ，Cinara IS ，Svetkovic SD ，Djordic PM ... -  《Srpski Arhiv za Celokupno Lekarstvo》

[Surgery of abdominal aorta with horseshoe kidney].

Seventy one surgical procedures on abdominal aorta in patients with horseshoe kidney have been reported in literature until 1980. Bergan reviewed 30 operations of abdominal aortic aneurysms (AAA) in these patients until 1974. Of them 3 AAA were ruptured. Gutowitz noticed 57 surgically treated AAA in patients with horseshoe kidney until 1984. Over the period from 1991 to 1996 thirty nine new cases were reported , including 2 ruptured AAA. The surgery of the abdominal aorta in patients with horseshoe kidney is associated with the following major problems: -reservation of anomalous (aberrant) renal arteries; reservation of the kidney excretory system; approach to the abdominal aorta (especially in patients with AAA) and graft placement The aim of the paper is the presentation of 5 new patients operated for abdominal aorta with horseshoe kidney. Over the last 12 years (January 1, 1984 to December 31, 1996) at the Centre of Vascular Surgery of the Institute of Cardiovascular Diseases of the Clinical Centre of Serbia, 5 patients with horseshoe kidney underwent surgery of the abdominal aorta. There were 4 male and one female patients whose average age was 57.8 years (50-70). Patient 1. A 50-year-old male patient was admitted to the hospital for disabling claudication discomforts (Fontan stadium IlI) and with significantly decreased Ankle-Brachial indexes (ABI). The translumbal aortography showed aorto-iliac occlusive disease and horseshoe kidney with two normal and one anomalous renal artery originating from infrarenal aorta (Crawford type II). Intravenous pyelography and retrograde urography showed two separated ureters. The aorto-bifemoral (AFF) bypass with Dacron graft was done with end-to-end (TT) proximal anastomosis just under the anomalous renal artery. The graft was placed behind the isthmus. During a 12-year follow-up renal failure, renovascular hypertension and graft occlusion were not observed. Patient 2. A 53-year-old male patient was admitted to the hospital for symptomatic AAA. Two years before admission the patient underwent coronary artery bypass grafting. The Duplex scan ultrasonography and translumbal aortography showed an infrarenal AAA, aneurysm of the right iliac artery and horseshoe kidney with two normal and one anomalous renal artery originating from the left iliac artery (Crawford type III). Intravenous pyelography and retrograde urography showed two separated ureters. After partial aneurysmectomy, the flow was restaured using bifurcated Dacron graft placed behind the isthmus. The right limb of the bifurcated graft was anastomosed with the common femoral artery and the left limb with left iliac artery just above the origin of the anomalous renal artery. The first day after operation thrombosis of the left common femoral artery with leg ischaemia was observed. (That artery was cannulated for ECC during coronary artery bypass grafting 2 years ago). The revascularisation of the left leg was done with femoro-femoral cross over bypass. During a 11-year follow-up period, the graft was patent and renal failure or revascular hypertension were not observed. Patient 3. A 66-year-old male patient was admitted to the hospital for rest pain (Fontan stadium III) and significantly decreased ABI. The patient had diabetes mellitus and myocardial infarction two months before admission. Translumbar aortography showed an aorto-iliac occlusive disease associated with horseshoe kidney with 5 anomalous renal arteries. (Crawford type III). Due to high risk, the axillo-bifemoral (AxFF) extra-anatomic bypass graft was performed. Five years after the operation the patient died due to new myocardial infarction. During the follow-up period the graft was patent and there were no signs of renal failure and renovascular hypertension. Patient 4. A 50-year old male patient was admitted to the hospital for high asymptomatic AAA. The diagnosis was established by Duplex scan and translumbal aortography. The large infrarenal AAA (transverse diameter 7 cm) associated with horseshoe kidney with two normal renal arteries (Crawford type I) were found. Intravenous pyelography and retrograde urogrpahy showed two separated ureters. After partial aneurysmectomy the tubular Dacron graft was placed behind the isthmus. During a 15-month follow-up the graft was patent and there were no signs of renal failure and renovascular hypertension. Patient 5. A 70-year-old female patient was admitted to the hospital for large asymptomatic AAA. The Duplex ultrasonography, CT scan, NMR and translumbal aortography showed an infrarenal AAA, aneurysms of the both common iliac arteries, aneurysm of the left hypergastric artery and horseshoe kidney with two normal and two anomalous renal arteries. One of the anomalous renal arteries originated from AAA, and the other from the left common iliac artery (Crawford type II). Intravenous pyelography and retrograde urography showed two separated ureters. After partial aneurysmectomy the flow was restaured using bifurcated Dacron graft placed behind the isthmus. The right limb of the graft was anastomosed (TT) with bifurcation of the common iliac artery and the left limb with the distal part of the common iliac artery (end-to-side) just above the origin of the second anomalous renal artery. The first anomalous renal artery that originated from AAA was removed from the aneurysm wall and anastomosed with graft using Carrel patch technique. During a 9-month follow-up the graft was patent and there were no signs of renovascular hypertension and renal failure. The horseshoe kidney is a rare anomaly of the urinary system. The incidence of this anomaly is from 1:1600 to 1:400 In 95% of cases the kidneys are connected with the lower poles, while in 5% with the upper poles In most cases, the isthmus structure is parenchimatous structure, and rarely it consists of the connective tissue. Usually the isthmus is located in front of the abdominal aorta and inferior vena cava, and very rarely behind them In two thirds of patients anomalous vascularization is present There are two classifications of anomalous vascularization: Papin's and Crawford's. According to Papin's classification, based on the number of renal arteries, there are three types of horseshoe kidney vascularization: Papin I (20%): There are two normal renal arteries only. (One of our 5 patients); Papin II (66%): There are 3-5 renal arteries. (Four of our 5 patients); Papin III (14%): There are more than 5 renal arteries. The Crawford's classification based on the origin of renal arteries, is of greater surgical importance than Papin's. According to it there are also three types of vascularization: Crawford I: There are two renal arteries with normal origin. (One of our 5 patients); Crawford II: Besides two normal, there are 1-3 anomalous renal arteries originating from the infrarenal aorta or iliac arteries (Three of our 5 patients); Crawford III: All renal arteries have an anomalous origin. (One of our 5 patients). The patients with horseshoe kidney can also have two separated, or one connected excretory urinary systems. All our 5 patients had two separated ureters. There is no specific clinical manifestation of the horseshoe kidney. Urinary infection or calculosis are very frequent as are in other urinary anomalies. The diagnosis of horseshoe kidney is established by Dupplex ultrasonography, CT scan, NMR, radionuclide scintigraphy and angiography. Very often the diagnosis is established occasionally during the examination of aneurysmal and occlusive diseases of the abdominal aorta. In cases of AAA or AIO associated with horseshoe kidney preoperative vascularization and condition of the excretory system should be established. Besides standard translumbar aortography selective renovasography is often neccessary. In some cases the intraoperative angiography or arterial identification, with metallic probe must be done. All renal arteries are "terminal" without significant anastomosis on the side of the kidney. Therefor its preservation is neccessary. There are three ways. The first is the location of anastomosis (3 of our patients). The second is an AxFF bypass, but only in patients with AIO (One of our patients and in the third reimplantation of the renal artery using Carrel patch technique was performed (One of our patients). The Isthmus of the kidney aggravates aortic preparation especially in patients with AAA. Sometimes isthmectomy is neccessary. In such cases there is danger of urinary fistula. Therefor many authors suggest the left extraperitoneal approach to abdominal aorta. In our patients, the transperitoneal approach was used, isthmectomy was not neccessary and graft was placed behind the isthmus. The operation of the abdominal aorta in patients with horseshoe kidney can be difficult due to anomalous renal arteries, anomalous excretory urinary system and is Ehmus. In these patients a more precise preoperative diagnosis is neccessary.

Lotina SL ，Davidović LB ，Kostić DM ，Velimirović DV ，Petrović PLj ，Perisić-Savić MV ，KovacevićN S ... -  《Srpski Arhiv za Celokupno Lekarstvo》

Surgical treatment of aorto-enteric fistulas.

Socrate AM ，Rosati L ，Locati P 《MINERVA CARDIOANGIOLOGICA》

[Factors determining late patency of aortobifemoral bypass graft].

Most of the patients with aortoiliac occlusive diseases have a multilevel localization of atherosclerotic diseases. In patients with aortoiliac occlusive diseases, the femoro-popliteal segment is involved in 28 to 66% of cases. These patients are usually old persons with many risk factors. Therefore, simultaneous proximal and distal reconstruction is often associated with a higher morbidity and mortality rates. In contrast, can proximal reconstruction help only patients with multilevel occlusive diseases? The aim of this paper is: definition of factors determining late patency rate of aortobifemoral bypass graft in patients with multilevel occlusive diseases; definition of factors determining clinical effects after aortobifemoral bypass procedures. This prospective study included 283 aortobifemoral reconstructions due to aortoiliac occlusive diseases operated between January 1st, 1984 and December 31st, 1992 at the Institute of Cardiovascular Diseases of the Serbian Clinical Centre in Belgrade. Bifurcated polytetrafluorethylene (PTFE) grafts were used in 136 patients, and standard nonimpregnated knitted Dacron grafts in 147 paetients. There were 25 (8.8%) female and 258 (91.2%) male patients, average age 56.88 years. Ninety one (32.2%) patients had claudication discomfort (Fonten stadium II), 91 (32.2%) disabling claudication discomfort (Fonten stadium IIb), 45 (15.9%) rest pain (Fonten stadium III), and 56 (19.8%) gangrene (Fonten stadium IV). In 45 (15.9%) patients previous vascular procedures were performed. Prior to operation Doppler ultrasonography and translumbar aortography were done. Isolated aortoiliac occlusive diseases with intact femoro-popliteal segment (Type I) were found in 83 (29.3%) patients; combined aorto-iliac and diseases of superficial femoral artery (Type II) in 170 (60%) patients; and combined aorto-iliac and femoro-popliteal diseases (Type III) in 30 (10.7%) individuals. Transperitoneal approach to abdominal aorta and standard inguinal approach to femoral arteries, were used. In 154 (54.4%) patients proximal anastomosis had an end to side (TL), while in 129 (45.6%) end to end (TT) form. In 152 (26.88%) patients distal anastomosis was found on the common femoral artery (AFC), while in 414 (73.2%) on the deep femoral artery (APF). In 7 patients the aorto-femoro-popliteal "jumping" bypass was performed, and in 29 subjects the simultaneous sequential femoro-popliteal bypass graft (Figures 1, 2, 3, 4a and 4b). The patients were followed-up over a period from one, six and twelve months after reconstruction, and later once a year, using physical examination and Doppler ultrasonography. In patients with suspected graft occlusion, anastomotic stenosis, pseudoaneurysms, progression of distal diseases, Duplex ultrasonography and angiography were also used, and leukoscintigraphy in patients with suspected graft infection. Statistical analysis was performed by Long Rank and Student's t-test. Inhospital mortality rate was 11 (7%). Simultaneous distal reconstructions significantly increased the mortality rate (p< 0.01). The follow-up period was from 2 months to 9.5 years (mean 3.6 years). Configuration of proximal anastomosis showed no significant influence on graft patency (p>0.05) (Graphs 1, 2, 3). Location of distal anastomosis at the deep femoral artery contributed to statistically significant increase in graft patency (p < 0.01) (Graphs 4, 5, 6). Simultaneous distal bypass showed statistically significant increase in graft patency (p < 0.01), but also significant increase in inhopsital mortality rate (p < 0.01) (Graphs 7, 8, 9). The type of occlusive diseases had no statistically significant influence on graft patency (p > 0.05) (Graphs 10, 11, 12). Six (2.1%) early unilateral limb occlusions were observed. The reasons for early graft occlusions were: stenosis of distal anastomosis in 3 patients and pure run off in 3 subjects. In 5 patients urgent reoperations (limb thrombectomy and profundoplasty or femoro-popliteal bypass graft above the knee) were performed with complete recovery of patients. However, in one patient an above the knee amputation had to be done. During the follow-up period 14 (5.2%) late graft occlusions were recorded: 11 unilateral limb and 3 bilateral graft occlusions. The reasons for late graft occlusion were: distal progression of atherosclerotic diseases, distal anastomotic stenosis, proximal progression of atherosclerotic diseases and anastomotic neointimal hyperplasy. All patients with late graft occlusion underwent successful redo-operations. Next late redo-procedures had to be done: three new aorto-bifemoral bypass grafts (patients with bilateral occlusion), two limb thrombectomies, 6 limb thrombectomies with profundoplasty and 3 femoro-femoral "cross-over" bypass grafts. Configuration of proximal anastomosis and type of occlusive disease showed no statistically significant influence on the number of early and late graft occlusions (p > 0.05). Location of distal anastomosis at the deep femoral artery and simultaneous distal bypass, statistically significantly decreased the number of early and late graft occlusions (p < 0.05). "Small aorta syndrome" statistically significantly increased the number of late graft occlusions. Eleven distal anastomotic pseudoaneurysms were noted. In 8 patients pseudoaneurysms were infected and in 3 noninfected. In all patients new redo-operations were carried out. Graft infection was recorded in 5 (1.7%) patients. One (0.3%) secondary aortoduodenal fistula was found. During the follow-up period new disabling claudication discomforts were found in 46 patients. The causes were distal anastomotic stenosis in 30 patients and progression of distal arterial diseases in 16 subjects. Of the total number of 30 patients with distal anastomotic stenosis 14 were reoperated (profundoplasty) and 16 patients refused a new operation. Also, 16 patients with progression of distal atherosclerotic diseases were reoperated. The operation was a kind of femoropopliteal or crural bypass grafts. During the follow-up period 97 patients were asymptomatic, 128 showed significant improvement, 29 had disabling claudications, and 111 had amputations. Distal anastomosis at the deep femoral artery and patent superficial femoral artery, statistically significantly influenced the clinical course after operation (p 0.01), while configuration of proximal anastomosis and simultaneous distal bypass had no significant effects (p < 0.05). (1) Only location of distal anastomosis has a statistically significant influence on the patency of aorto-bifemoral bypass graft. (2) The location of distal anastomosis and type of occlusive disease have a statistically significant influence on the clinical effect of the operation. (3) The simultaneous distal bypass had no influence on the late patency of aortobifemoral bypass graft and on the number of asymptomatic patients. Also, it increased inhospital mortality rate.

Davidović LB ，Lotina SI ，Kostić DM ，Cinara II ，Cvetković SD ，Stojanov PL ，Velimirović DB ，Marković MM ，Pejkić SA ，Vukotić AM ... -  《Srpski Arhiv za Celokupno Lekarstvo》

[Aorto-enteric and paraprosthetic fistulas. Apropos of 5 cases].

Glock Y ，Bouissou E ，Tasrini J ，Fourtanier G ，Fournier D ，Puel P ... -  《journal de chirurgie》