Ab H. Boontje, M.D., Ph.D., Groningen, Holland

Transcription

1 Aneurysm formation in human umbilical vein grafts used as arterial substitutes Ab H. Boontje, M.D., Ph.D., Groningen, Holland A series of 257 human umbilical vein grafts for femoropopliteal bypass in 203 patients, inserted between 1978 and 1984, is presented. The 6-year cumulative patency rate is 74% for above-knee and 44% for below-knee cases. Late complications, such as formation of aneurysms in the human umbilical vein graft, occurred in nine grafts, 2V2 to 6 years after implantation. This corresponds to 3.5% of the total number of 257 human umbilical vein grafts. Three of the grafts developed a second aneurysm at a later time, making a total of 12 aneurysms. The aneurysms were evident as a painful, pulsating mass. All patients were operated on, primarily by resection and interposition grafting. All aneurysms were saccular and false. Macroscopic and microscopic examination revealed that all aneurysms except four originated from a gap in the umbilical vein wall and the Dacron mesh. The adjacent graft wall had a completely normal architecture without biodegradation. The cause of the origin of the gap in the wall remains obscure. Four aneurysms were anastomotic and were located at the site of the suture line, placed at the factory, joining the two components of a composite human umbilical vein graft; these anastomotic anettrysms were caused by breaking of the Prolene suture. (J VASC SURG 1985; 2:524-9.) The success of any arterial substitute depends on a high long-term patency rate and a minimal incidence of complications. This applies especially to biologic grafts, such as the human umbilical vein graft (, or the so-called Dardik Biograft, Meadox Medicals, Inc., Oakland, N.J.). In the past this vascular conduit has given promising results as femoropopliteal bypass. However, the long-term behavior and the incidence of biodegradation and of aneurysm formation of the graft are not known. In this context we report the results of 257 femoropopliteal bypass procedures with an observcd since Wc describe specifically 12 cases of aneurysm formation in nine s and the distinct morphologic types of these aneurysms. PATIENTS AND METHODS From January 1978 to January 1984, 257 s were implanted as a femoropopliteal bypass graft in the control of atherosclerotic occlusive disease. The series of 257 operations with s comprises 203 patients; 54 had a bilateral bypass graft. Indications for operation were a long-lasting and disabling short-distance claudication in 190 cases From the Department of Vascular Surgery, University Hospital. Reprint requests: Ab H. Boontje, M.D., Ph.D., Professor of Vascular Surgery, Department of Vascular Surgery, University Hospital, 9713 EZ Groningen, Holland. (74%) and limb-threatening ischemia (rest pain or gangrene) in 67 cases (26%). The proximal anastomosis was to the common femoral artery in the majority of patients, to the inguinal end of a previously inserted proximal bypass in 24 cases (9%), and to the aorta in one patient. The distal anastomosis was above the knee in 159 grafts (62%) and below the knee (to the distal popliteal artery or to one of the lower leg arteries) in 98 grafts (38%). Pertinent data on handling of the, on operative techniques, and on classification of the out- " flow tract have been described in previous publications. 1,2 Patency of the grafts was determined by good palpable pulsatility and by follow-up angiography. During the entire follow-up period, from January 1978 up to the present, aneurysms formed in nine s. In three patients a second aneurysm developed in another part of the same after repair of the first aneurysm. Therefore, there was a total of 12 aneurysms in nine s. One patient had an ancurysm in the in the other leg at a later time. One patient had a bilateral aneurysm simultaneously. The aneurysms were detected as a painful pulsating mass at regular check-ups in the outpatient department. All aneurysms were operated on after translumbar aortography was performed. 524

2 Volume 2 Number 4 July 1985 Aneurysm formation in human umbilical vein grafis % nl ,~,... '""Q,. w n3 e.~,,49... a.. L ~ 74% AK (n=159) ne~ "~.~ "'"~"... e.. "'~"~e 8v n6,... ~ 64% AK+BK (n=257) "g "~ ""'"~- 44%T BK (n=98) 20 1 month years Fig. 1. Cumulative patency rates of s. Table I. Aneurysms in s Patient Sex Age Graft diameter Right (R) (mm); Time Duration of or left (L) distal Graft interval pulsating leg anastomosis function (yr) mass ~ Location Operation Result A1 F 67 L 6, BK Open 4 B1 M 65 R 6/5, BK Open 3 C1 M 70 L 6/5/4, BK Open 2 DI M 70 R 6, BK Occlusion 6 D2 M 70 L 6, BK Occlusion 6 E1 M 65 L 6/5, BK Open 4 F M 66 L 6, AK Open 6 A2 F 69 L 6, BK Open 5i/2 C2 M 71 L 6/5/4, BK Open 21/2 G M 76 R 6/5, BK Open 41/2 E2 M 66 R 6, AK Open 41/2 B2 M 67 R 6/5, BK Occlusion 5 *Duration of 2 weeks and V2 year are approximate figures. 1/2 yr 1/2 yr Groin Interposition Good Groin Dacron Interposition Good Dacron No. of aneurysms; morphologic type 2; defect in 1; anastomotic (6/5 mm ) 1; anastomotic (5/4 mm ) 1; defect in 1; defect in 1; 1; 1/2 yr Upper leg Interposition Occlusion 1; l/2 yr 1; 1; 1/2 yr 3; Upper leg Jump graft Occlusion 3; 1/2 yr Groin Dacron Good 1; patch defect in anastomotic (6/5 mm ) defect in defect in anastomotic (6/5 mm ) defect in defect in defect in RESULTS The results of the series of 257 femoropopliteal bypass grafts with an have been analyzed according to the life-table method (Fig. 1). No patient was lost to follow-up. The cumulative patency rate for the entire group of 257 s was 64% at 6 years postoperatively. This corresponds to a cumulative patency rate of 74% after 6 years for patients with the distal anastomosis above the knee and to 44% for patients with the distal anastomosis below the knee. However, successful thrombectomies were not entered again in the life tables as patent grafts. During the follow-up period seven patients presented with 12 aneurysms in nine s (Table I). In all cases there was a painful and growing pulsating mass, which had been evident for about 2 weeks in six cases and for halfa year in six other cases. Eight aneurysms were located in the thigh above the

3 526 Boontje Journal of VASCULAR SURGERY Fig. 2. Angiogram showing composite with aneurysm (patient B1). knee and four in the groin. Six of the aneurysms developed in 6 mm s, (two to the proximal and four to the distal popliteal artery), and six in a composite graft to the distal popliteal artery. All but three s functioned well. Three grafts had an asymptomatic occlusion previously for a number of years, but an aneurysm developed in the groin. The time interval between the original insertion of the and the operative correction of the aneurysm varied from 21/2 to 6 years. Only two of the patients had other Signs of aneurysmal disease, particularly dilatation of the abdominal aorta. All aneurysms were operated on after preceding translumbar aortography (Fig. 2). Resection and interposition grafting with a new short was the method of correction in eight cases. In one aboveknee case a jump graft () to the distal popliteal artery was inserted after resection of the aneurysm. A partial resection and correction with a Dacron tube prosthesis or patch was done in three cases in which the aneurysm was located in the groin. Complete replacement of the was never performed. Surgical intervention was successful in 10 cases. Twice the graft occluded after correction, but this event was not followed by amputation of the leg. Examination of the removed aneurysms showed that they were of the saccular and false type, contalning thrombus. In two cases the pulsating mass consisted in fact of two and in two other cases of three distinct aneurysms. Thus the total number of aneurysms was actually 18. Microscopic study of the removed specimens revealed interesting findings. The wall consisted of fibrous tissue without elements of the umbilical vein wall. The false aneurysm originated from a defect in the umbilical vein wall and the surrounding Dacron mesh and had no relationship with any anastomosis in 14 of the 18 cases (Fig. 3). Likewise, the three cases of aneurysm in the groin were not anastomotic aneurysms, since the original proximal anastomosis with Ti-cron (Davis and Geck) sutures between the common femoral artery and the was intact. There existed a defect in the "roof" or "cobra mouth" of the end-to-side anastomosis, consisting of umbilical vein wall and mesh, which caused a false aneurysm. There were four cases of anastomotic aneurysms. These four aneurysms were located at the factory-made anastomosis between both ends of a composite. Examination of the aneurysm revealed a Prolene suture, visible on both sides of the entrance to the false aneurysm, protruding from the, tissues. The Prolene suture had been broken, which caused a dehiscence of the anastomosis (Fig. 4). The patient with a mm composite aortotibial graft had an anastomotic aneurysm at the 5-4 mm anastomosis after 2 years and one at the 6-5 mm anastomosis after 21/2 years, both occurring after the fracture of the Prolene suture material, which was clearly visible. DISCUSSION Since the clinical introduction of the in 1975, there has been a rather extensive clinical experience with this graft material as a vascular conduit for femoropopliteal bypass. Our long-term patency rates for above-knee and below-knee procedures, 74% and 44%, respectively, after 6 years, are comparable to those obtained by Dardik et al? Although

4 Volume 2 Number 4 July 1985 Aneurysm formation in human umbilical veingrafls 527 Fig. 3. False saccular aneurysm of after resection (patient A1). Note gap m umbilical vein wall and Dacron mesh. Fig. 4. Removed and opened specimen (patient B1). At entrance to false aneurysm broken Prolene suture is visible on both sides. these results are quite promising, biodegradation and formation o f aneurysms are well-known complications o f biologic graft material. Biodegradation o f H U V G s is the subject o f a separate study. In current experience the incidence o f aneurysm formation in H U V G s seems to be rather low. In our series o f 257 H U V G s in 203 patients with a followup period o f 6 years, one or more aneurysms de- veloped in nine H U V G s (3.5%). Giordano and Keshishian 4 described three patients without giving the total number o f bypass grafts they inserted. Dardik et al. s described only seven aneurysms in a total o f 756 H U V G s (0.9%). As stated earlier, the 12 cases o f aneurysmal lesions comprised in fact 18 distinct aneurysmal sacs. Eleven o f these aneurysms in the body o f the H U V G

5 528 Boontje Journal of VASCULAR SURGERY Big. 5. Specimen with three false aneurysms (patient G). Entrance to left one has defect in wall of graft and mesh. In entrance of two others, there is no umbilical vein wall but only remnants of Dacron mesh with small holes. were caused by a gap in the wall of the graft and in the Dacron mesh. The morphologic type of the aneurysms varied slightly. In the larger aneurysms there was a rim of normal media of the and mesh with its collagenous fibrous layer around the entrance of the false aneurysm (Fig. 3). Sometimes the medial layer was a little thinner than in the original (a small piece of which, for later comparison, was kept in stock after the primary operation), but the architecture itself was quite normal. Remarkable was the fact that the umbilical vein wall showed no signs of biodegradation. Smaller aneurysms also had a wall of fibrous tissue, without elements of the vein wall, but remnants of the Dacron mesh could be found, either as stretched fibers, split up with rather large spaces in between, or with holes, after rupture of the fibers. In such cases the mesh was the inner layer of the lumen (Fig. 5). Aneurysm formation also does occur in occluded s, not in the body of the graft, but at the level of the proximal anastomosis. In our three cases the occlusion of the existed for 6 months, 1, and 4 years. The cause of the origin of a circumscribed gap in the wall of the remains obscure. It seems reasonable to suppose that the defect in the vein wall comes first and that the breaking of the mesh is secondary, But the reason for a defect in the, in the presence of normal architecture of the remaining adjacent graft, is still unknown. In our cases there was no relationship with biodegradation, char- acterized as loss of cellular details. According to Dardiket al., 3 the glutaraldehyde tanning protects the graft from biodegradation. Diffuse and irregular changes in the wall of the graft characteristic of biodegradation will probably cause occlusion of the graft rather than aneurysm formation. (This will be the subject of a separate study.) Furthermore, true aneurysms caused by fusiform dilatation of the and of the mesh, as described by Dardlk et al.,s were not encountered in our series. The four anastomotic aneurysms of the presutured anastomosis of a composite had a logical cause, namely fracture of the Prolene suture. Although polypropylene (Prolene) is used widely in vascular surgical procedures, it has been known for some time that Prolene can break after long-term implantation, 6 as was the case with silk in earlier experience. For this reason Starr et al.6 recommend the use of braided Dacron sutures for prosthetic graft material. Concerning the best surgical treatment of aneurysms in the, resection of the aneurysm and interposition grafting seem to be logical. It was our procedure of choice with all aneurysms in the body of the graft. In favor of this type of treatment is the fact that the operation is simple and fast. But it must be said that in three of the nine s in which aneurysms formed, a second aneurysm developed within 6 months to 2 years in another part of the same graft. Because complete replacement of the en-

6 Volume 2 Number 4 July 1985 Aneurysm formation in human umbilical vein grafts 529 tire is much more complicated and timeconsuming, segmental replacement of the aneurysm remains our preferred method of treatment. In conclusion, an gives reasonable results as a femoropopliteal bypass, having good long-term patency rates, but late complications, such as formation of aneurysms, may occur. The possibility of the development of false aneurysms must be kept in mind. Not only the routine periodic assessment of the function of the is necessary, but special attention must be paid to the early detection of this complication. REFERENCES 1. Boontje AH. Glutaraldehyde-tanned human umbilical cord veins for femoropopliteal bypass. Vase Surg 1981; 15: Boontje AH. The Biograft for femoropopliteal bypass. J Cardiovasc Surg 1983; 24: Dardik H, Baler RE, Meenaghan M, Natiella J, Weinberg S, Turner R, Sussman B, Kahn M, Ibrahim IM. Morphologic and biophysical assessment of long-term human umbilical cord vein implants used as vascular conduits. Surg Gynecol Obstet 1982; 154: Giordano JM, Keshishian JM. Aneurysm formation in human umbilical vein grafts. Surgery 1982; 91: Dardik H, lbrahim IM, Sussman B, Kahn M, Sanchez M, Klausner S, Baier RE, Meyer AE, Dardik II. Biodegradation and aneurysm formation in umbilical vein grafts. Ann Surg 1984; 199: Starr DS, Weatherford SC, Lawrie GM, Morris Jr GC. Suture material as a factor in the occurrence of anastomotic false aneurysms. Arch Surg 1979; 114:412-5.