A Paraprosthetic-Enteric Fistula Associated with a Duodenal Tumor

Case reports

A Paraprosthetic-Enteric Fistula Associated with a Duodenal Tumor R o b e r t D a c h s , M D , R o n a l d E. C l e m e n t , M D , B r u c e D z i u r a , a n d J o e l B e r m a n , M D , Springfield, Massachusetts

MD,

An unusual etiology and presentation of a paraprosthetic-enteric fistula is reported involving a duodenal tumor. Review of the possible mechanisms of graft-enteric fistula formation and diagnostic evaluation is presented. Initial insertion of an extraanatomic bypass followed by graft excision was performed in the present case and is encouraged in hemodynamically stable patients. Additionally, resection of the eroded duodenum with end-to-end anastomosis was performed in order to restore intestinal continuity. (Ann Vasc Surg 1990;4:65-68) KEY WORDS:

Paraprosthetic-enteric fistula; duodenal tumor; graft-enteric fistula.

bility. Additionally, he noted intermittent passage of melanotic stools during the preceding four weeks. He denied any gastrointestinal disturbances or episodes of fever or chills. Physical examination at the time of admission revealed a thin, elderly, white man with stable vital signs. The abdomen was soft and non-tender with active bowel sounds and no masses. Rectal examination revealed occult blood. Admission laboratory data showed a hemoglobin level of 6.6 and hematocrit of 21.3%, with a white blood celt count of 12,600. CT scan of the abdomen displayed several small loculations of gas within an abnormal soft tissue density that surrounded the aortic graft (Fig. 1). Endoscopy revealed a considerable amount of fresh blood within the duodenum. No duodenal ulcer or prosthetic graft was visualized. After stabilization, the patient was taken to the operating room with the preoperative diagnosis of an aortoduodenal fistula. Initially a right axillobifemoral bypass was performed with an 8 mm P T F E graft. All incisions were then closed prior to laparotomy. Using a retroperitoneal approach, a thick, fibrotic inflammatory reaction surrounding the body of the aortic graft was noted. Careful dissection revealed a 6 cm hemorrhagic erosion of the posterior wall of the third portion of the duodenum in direct contact with the body of the aortic graft. Soilage of the graft was

A o r t o e n t e r i c fistula is an u n c o m m o n y e t d e v a s tating c o m p l i c a t i o n o f a o r t i c s u r g e r y . W h e t h e r t h e s e fistulae a r e the r e s u l t o f p r i m a r y graft infection o r m e c h a n i c a l e r o s i o n o f t h e b o w e l r e m a i n s c o n t r o v e r s i a l . M a l i g n a n c y h a s o n l y r a r e l y b e e n ass o c i a t e d w i t h p r i m a r y a o r t o e n t e r i c fistulae [ I - 3 ] a n d n e v e r w i t h g r a f t - e n t e r i c fistulae. T h i s r e p o r t d o c u m e n t s the first c a s e o f a s e c o n d a r y a o r t o d u o d e n a l fistula as a r e s u l t o f an infiltrating t u m o r .

CASE REPORT A 77-year-old man was referred for surgical evaluation at Baystate Medical Center in June 1987, having previously undergone resection of a ruptured abdominal aortic aneurysm in March 1980. An ultra-lightweight, Dacron aortobiiliac graft had been used for initial reconstruction. He subsequently did well until January 1987 when he noted a 20 pound weight loss, anorexia, and easy fatiga-

From the Departments of Surgery and Pathology, Baystate Medical Center, Springfield, Massachusetts. Reprint requests: Joel Berman, MD, Department of Surgeo', Baystate Medical Center, Springfield, Massachusetts 01199. 65

PARAPROSTHET1C ENTERIC FISTULA

66

ANNALS OF VASCULAR SURGERY

Fig. 1. CT scan of abdomen displaying several small Ioculations of gas within abnormal soft tissue density surrounding aortic graft.

noted. There was no evidence of pseudoaneurysm formation, disruption of the suture line, or breakdown of the graft material. The graft was excised and the aortic stump and common iliac arteries oversewn. A 13 cm segment of duodenum containing the erosion was resected and an end-to-end anastomosis performed. After a lengthy postoperative course, the patient was discharged home. Histopathologic examination of the duodenum revealed a malignant, poorly differentiated spindle and epithelioid cell neoplasm associated with the hemorrhagic ulceration (Fig. 2). Immunoperoxidase stains suggested an epithelial origin of the tumor, and a final diagnosis of poorly differentiated adenocarcinoma was made.

DISCUSSION Since Brock reported the first case of a graftenteric fistula in 1953 [4], two distinct types of secondary aortoenteric fistulae have been described. First, direct communication between the lumen of the bowel and a disrupted aortic suture line results in a true aortoenteric fistula. In contrast, erosion of the bowel wall next to the vascular prosthesis without involvement of the suture line is a well-recognized entity termed paraprosthetic-enteric fistula by Elliott, Smith, and Szilagyi [5], which was apparent in our patient. In 1983, Bunt reviewed 256 cases of aortoenteric fistula and 38 cases of paraprosthetic fistula since Brock's initial report [6]. In neither that review, nor subsequent reports, has an aortoenteric or paraprosthetic-enteric fistula been associated with malignancy. Thus, the present report documents the first case of a secondary aortoduodenal fistula associated with a tumor. Opinion remains divided as to whether primary infection of the prosthesis or mechanical erosion of the bowel wall by a noncompliant graft is the initiating event in the development of these fistulae.

Fig. 2. Histopathologic examination of duodenum revealing malignant, poorly differentiated spindle and epithelioid cell neoplasm associated with hemorrhagic ulceration.

Animal studies have supported each of these hypotheses [7,8]. As for the development of paraprosthetic fistulae, it is apparent that the bowel wall is eroded by the mechanical pulsations of the aortic prosthesis. In the present case, a highly anaplastic adenocarcinoma infiltrated the third portion of the duodenum and fixed this portion of bowel to the aortic graft. We hypothesize that mechanical forces transmitted by the graft caused eventual erosion of the duodenum since the erosion involved only that portion of bowel overlying the graft. While aortoenteric fistulae classically present as a "herald" bleed followed by a significant gastrointestinal hemorrhage, paraprosthetic-enteric fistulae are more frequently associated with signs of sepsis. O'Mara and Imbembo reviewed 21 cases of paraprosthetic fistulas and noted that the majority of the patients presented with signs of sepsis [9]. Other authors have also stressed the septic presentation of these patients [10-14]. Yet, as the present case demonstrates, occult gastrointestinal bleeding may

VOLUME 4 NO 1 - 1990

PARAPROSTHETIC ENTERIC FISTULA

be the only clue to the presence of a graft-enteric fistula. Thus, whether sepsis is present or not, all patients with a previous aortic replacement that have evidence of gastrointestinal hemorrhage require prompt investigation to rule out graft-enteric fistula formation. Diagnostic evaluation utilizing computed tomography and gastroduodenoscopy resulted in an accurate preoperative diagnosis in the present case. The CT finding of an inflammatory response containing small air pockets surrounding the aortic prosthesis (see Fig. 1) was first described by Haaga and associates [15] in four patients with infected aortic grafts. Subsequent reports have described the CT scan's ability to identify aortoenteric fistule [I 6-19], yet occasional false negative studies have been noted [20]. In contrast, aortography and barium contrast studies are unreliable in diagnosing this condition [21], Gallium, labelled leukocyte, and 99m-Tc pertechnetate scanning have separately been reported to be helpful diagnostic tools [22-24]. Upper gastrointestinal endoscopy has been shown to be useful in a number of case reports [25-31]. While visualization of the aortic prosthesis has been reported, the ability of the endoscope to visualize the entire upper gastrointestinal tract and eliminate other potential causes of hemorrhage is invaluable in making the diagnosis of aortoenteric fistula. However, while an armentarium of diagnostic studies can be employed, an accurate preoperative diagnosis is not always established. A recent study revealed that of 33 cases of graft-enteric fistula, only 10 diagnoses were established preoperatively [32]. Hence, while CT scanning and endoscopy may be helpful, the suspicion of an aortic or paraprosthetic-enteric fistula demands operative intervention. While occasional reports suggest alternative forms of management, graft excision with extraanatomical bypass still remains the treatment of choice for graft-enteric fistula. Bunt's review of the mortality rates associated with a variety of management options demonstrates the superiority of graft excision and remote bypass [6]. Alternatively, replacement of the infected aortic graft with arterial or venous autografts has recently been suggested [33,34]. While early reports of this procedure appear promising, further experience and follow-up is required. In contrast, Thomas and Baird in the British literature have suggested a "conservative" approach (closure of the duodenum, repair of the anastomotic defect, and interposition of omentum) in those cases where the aortic prosthesis is "noninfected" [35]. Despite their results with four patients, this form of management must be viewed skeptically because all grafts in contact with bowel lumen must be considered potentially infected. Thus, until further experience supports this ther-

67

apy, graft excision and extraanatomic bypass should be performed in cases of graft-enteric fistula. While debate over the sequence of operative management still exists, recent studies have advocated initial insertion of the remote bypass followed by removal of the infected aortic graft [36,37]. While secondary infection of the remote bypass due to bacteremia at the time of removal of the infected aortic graft has been described [38], no significant difference in the rate of new graft infection has been associated with the reversal of the traditional operative sequence [37]. The incidence of this complication can be reduced with the use of high systemic levels of appropriate broad spectrum antibiotics and meticulous attention to protecting these wounds from contamination during removal of the infected aortic prosthesis. Further, a functioning remote bypass at the time of removal of the aortic prosthesis obviates the complications associated with distal ischemia. Thus, in the hemodynamically stable patient in whom the diagnosis of an aortoenteric or paraprosthetic fistula is reasonably certain, initial insertion of a remote bypass is encouraged. Last, while closure of the enteric defect is all that is often needed, the extensive destruction of the bowel wall in the current case required resection of the eroded duodenum. An end-to-end anastomosis was performed to restore intestinal continuity. Although histological examination revealed tumor invading the duodenum in the present case, surgical repair of the fistula with removal of the infected prosthesis and extraanatomical bypass remains the primary objective whether or not a tumor is present in cases of aortoenteric or paraprosthetic-enteric fistulae.

REFERENCES 1. GHAPHERY AD, GUPTA RK, CURRIE RA. Carcinoma of the head of the pancreas with aortoduodenal fistula, Am J Surg 1966:111:580-583. 2. GUPTA RK, ROGERS KE. A unique case of aortoduodenal fistula following carcinoma of cervix. Am J Obstet Gynecol 1978;131:110-113. 3. POSTOLOFF AV, CANNON WM. Genesis of aortic perforation secondary to carcinoma of the esophagus. Arch Pathol 1946;41:533-539. 4. BROCK RC. Aortic homografting: a report of six successful cases. Guys Hosp Rep 1953:102:204. 5. ELLIOTT JB, SMITH RF, SZtLAGYI DE. Aortoenteric and paraprosthetic-enteric fistulas. Arch Surg 1974;108:47% 490. 6, BUNT TJ. Synthetic vascular graft infections. II. Graftenteric erosions and graft-enteric fistulas. Surgeo' 1983; 94:1-9. 7. BUSUTTIL RW, REES W, BAKER JD, WILSON SE. Pathogenesis of aortoduodenal fistula: experimental and clinical correlates. Surget3' t979;85:1-13. 8. IKONONOPOULOS DC, SPANOS PK, LAZARIDES DP. Pathogenesis of aortoenteric fistula: an experimental study. Int Angiol 1986;5:33-37.

68

PARAPROSTHET1C ENTERIC FISTULA

9. O'MARA C, IMBEMBO AL. Paraprosthetic-enteric fistula. Surgery 1977;81:556-566. 10. MARTIN A, COPEMAN PWM. Aorto-jejunal fistula from rupture of Teflon graft, with septic emboli in the skin. Br Med J 1967;2:155-156. 11. GORDON SL, NICHOLAS GG, CARTER SL, MIRA AJ. Aorto-enteric fistula presenting as multicentric osteomyelitis. C/in Orthop 1978;131:255-258. 12. DALINKA MK, REGINATO AJ, BERKOWlTZ HD, et al. Hypertrophic steoarthropathy as indication of aortic graft infection and aortoenteric fistula. Arch Surg 1982;117:13551359. 13. BECKER RM, B L U N D E L L PE. Infected aortic bifurcation grafts: experience with fourteen patients. Surgery. 1976; 80:544-549. 14. CRIADO FJ, CLASSEN JN, WILSON TH. Secondary aorto-enteric fistulas: prosthetic and paraprosthetic. Am Surg 1981 ;47:313-321. 15. HAAGA JR, BALDWIN N, REICH NE, et al. CT detection of infected synthetic grafts: preliminary report of a new sign. A JR 1978;131:317-320. 16. BROWN OW, STANSON AW, PAIROLERO PC, HOLLIER LH. Computerized tomography following abdominal aortic surgery. Surgery. 1982;91:716-722. 17. MARK A, MOSS AA, LUSBY R, KAISER J. CT evaluation of complications of abdominal aortic surgery. Radiology 1982;145:409-414. 18. KUKORA JS, RUSHTON FW, CRANSTON PE. New computed tomographic signs of aortoenteric fistula. Arch Surg 1984;119:1073-1075. 19. ACKROYD J, WILLIAMS TG, THOMAS ML, BURNAND KG. The diagnosis of aortic graft-enteric fistulae by computed tomography. Br J Surg t985;72:72-73. 20. FREMANIS IE~ KOZAK B, TAYLOR LM, PORTER JM. Failure of CT scanning to diagnose aortic graft infection. J Vasc Surg 1987;6:531-532. 21. BERNHARD VM. Management of graft infections following abdominal aortic aneurysm replacement. World J Surg 1980; 4:679-688. 22. CAUSEY DA, FAJMAN WA, PERDUE GD, et al. 67-GA scintigraphy in postoperative synthetic graft infections. A JR 1980;134:1041-1045. 23. STEVICK CA, FAWCETT HD. Aortoiliac-graft infection: detection by leukocyte scan. Arch Surg 1981; 116:939-942.

mmm

ANNALS OF VASCULAR SURGERY

24. MAKHIJA M, SCHULTZ S. Imaging of an actively bleeding aortoduodenal fistula. C/in N u c / M e d 1985;10:372-373. 25. MIR-MADJLESSI SH, SULLIVAN BH, FARMER RG, BEVEN EG. Endoscopic diagnosis of aortoduodenal fistula. Gastrointest Endosc 1973;19:187-188. 26. BUNT TJ, DOERHOFF CR. Endoscopic visualization of an intraluminal dacron graft: definitive diagnosis of aortoduodenal fistula. South Med J 1984;77:86--87. 27. SKIBBA RM, GREENBERGER N J, HARDIN CA. Paraprosthetic-enteric fistula: role of preoperative endoscopy. Dig Dis 1975;20:1081-1086. 28. BAKER MS, FISHER JH, VAN DER REIS L, BAKER BH. The endoscopic diagnosis of aortoduodenal fistula. Arch Surg 1976;111:304. 29. BRADY PG. Aortoduodenal fistula: role of endoscopy in diagnosis. Am J Gastroentero/ 1978;69:705-707. 30. BRAND EJ, SIVAK MV, SULLIVAN BH. Aortoduodenal fistula: endoscopic diagnosis. Dig Dis Sci 1979;24:940-944. 31. PUPPALA AR, MUNASWAMY M, DOSHI AM, STEINHEBER FU. Endoscopic diagnosis of aortoduodenal fistula: complication of abdominal aortic bypass grafts. Am J Gastroenterol 1980;73:414---417. 32. REILLY LM, ALTMAN H, LUSBY RJ, et al. Late results following surgical management of vascular graft infection. J Vasc Surg 1984;1:36-44. 33. EHRENFELD MK, WILBUR BG, OLCOTT CN, STONEY RJ. Autogenous tissue reconstruction in the management of infected prosthetic grafts. Surgery 1979;85:8292. 34. LARENTZEN JE, NIELSEN OM, Aortobifemoral bypass with autogenous saphenous vein in treatment of paninfected aortic bifurcation graft. J Vasc Surg 1986;3:666---668. 35. THOMAS WEG, BAIRD RN. Secondary aorto-enteric fistulae: towards a more conservative approach. Br J Surg 1986;73:875-878. 36. TROUT HH, KOZLOFF L, GIORDANO JM. Priority of revascularization in patients with graft enteric fistulas, infected arterial prosthesis. Ann Surg 1984;199:669-683. 37. REILLY LM, STONEY RJ, GOLDSTONE J, EHRENFELD WK. Improved management of aortic graft infection: the influence of operation sequence and staging. J Vasc Surg 1987;5:421-431. 38. YASHER J J, WEYMAN AK, BURNARD RJ, YASHER J. Survival and limb salvage in patients with infected arterial prostheses. Am J Surg 1978;135:499-504.