- Email: [email protected]
Management of infected aortic grafts: development of less invasive surgery using cryopreserved homografts
Management of Infected Aortic Grafts: Development of Less Invasive Surgery Using Cryopreserved Homografts Paul R. Vogt, MD, and Marko I. Turina, MD Clinic for Cardiovascular Surgery, University Hospital, Zurich, Switzerland
Background. Aortic graft infection is associated with significant mortality and morbidity. Total graft replacement with homografts provides an effective treatment. Partial graft replacement at the site of infection may simplify the surgical procedure. Methods. Between January 1991 and December 1996, homografts were used in 18 patients (mean age, 61612 years; range 41– 85) with thoracic (4/18; 22%) or abdominal (14/18; 78%) aortic graft infection. Sepsis was present in 14 patients (78%); 6 (33%) had various aortic fistulae. Total graft replacement using homografts was performed in 14 (78%), and partial graft replacement at the site of infection in 4 patients (22%).
Results. Hospital mortality was 11%. During the follow-up period of 22615 months (range, 12– 65) there was 1 infection and 1 homograft-related late death after complete homograft replacement, and 1 percutaneous vascular stent placement after partial graft replacement. No other instances of reinfection, suture line rupture or anastomotic aneurysms were observed. Conclusion. Total graft replacement with homografts provides an effective treatment for infected aortic grafts. Partial graft replacement at the site of infection is feasible and safe. (Ann Thorac Surg 1999;67:1986 –9) © 1999 by The Society of Thoracic Surgeons
V
Material and Methods
ascular prosthetic graft infection is a life threatening pathologic entity with a high morbidity and mortality [1, 2]. Above all, the surgical treatment for infected prosthetic grafts of the thoracic or abdominal aorta remains a most demanding technical challenge [3]. Conventional surgical treatment, in situ repair with a new vascular prosthesis or complete removal of the infected aortic graft and reconstruction with an extra anatomical bypass graft, has a substantial operative mortality rate. Despite the improving results [4, 5] of initial surgical treatment, prosthetic aortic graft infection is frequently associated with reinfection, suture line rupture or anastomotic aneurysm, leading to multiple reoperations, distal amputations and prolonged, even life-long antibiotic treatment [6]. Recent findings with cryopreserved heart valve homografts in the treatment of acute infectious endocarditis [7] led us to use cryopreserved homografts for in situ replacement of infected aortic grafts. Their observed clinical resistance to bacterial infections [8] prompted us to limit the surgical procedure, performing only partial graft replacement when the vascular prosthetic infection involved only a small portion of the graft or was limited to the aortic graft anastomosis. This study presents our clinical experience with the treatment of infected aortic grafts, performing either total or partial graft replacement with cryopreserved homografts. Presented at the Aortic Surgery Symposium VI, April 30 –May 1, 1998, New York, NY. Address reprint requests to Dr Vogt, Clinic for Cardiovascular Surgery, University Hospital, Ramistr 100, CH-8091 Zurich, Switzerland; e-mail: [email protected].
© 1999 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
Patients Between January 1991 and December 1996, cryopreserved aortic homografts were used in 18 patients (mean age, 61612 years, range 41– 85) with vascular prosthetic infection of the thoracic (4/18; 22%) or abdominal aorta (14/18; 78%), with or without involvement of the pelvic and groin vessels. Sepsis was present preoperatively in 14 patients (78%), and 6 (33%) had aortobronchial, aortoesophageal or aortoenteric fistulae originating from the infected aortic graft. In 7 patients (39%), as many as 10 previous prosthetic graft infection-related vascular surgical procedures had been done, and as many as 6 different new prosthetic grafts had already been inserted. Preoperatively, blood cultures were positive in 10 patients (55%), and 15 (83%) had received various antibiotic treatments for up to 3 years. Concomitant peripheral arterial vascular occlusion was found in 12 patients (67%), diabetes mellitus in 5 (28%), chronic atrial fibrillation in 3 (17%), coronary artery disease in 13 (72%) and chronic obstructive lung disease in 12 patients (67%). Renal insufficiency was found in 7 patients (39%), 2 were on dialysis, and 2 had had renal transplantation because of end-stage diabetic nephropathy. Emergency surgery was performed in 10 patients (55%), and the remaining 8 (45%) were operated on within 72 hours. Prosthetic vascular grafts of the thoracic aorta were affected in 4 patients (22%). One had infection of an ascending aortic graft previously implanted for an acute type A dissection, and the other had an infected composite graft after repair of annuloaortic ectasia. The other 2 0003-4975/99/$20.00 PII S0003-4975(99)00357-4
Ann Thorac Surg 1999;67:1986 –9
Fig 1. Preoperative computed tomogram of a 46-year-old patient with an aortointestinal fistula from a large juxtarenal and infrarenal aneurysm of the proximal anastomosis of a 7-year-old aorto bifemoral bifurcation graft, rupturing into the retroperitoneal duodenum.
patients had infection of a descending aortic graft: an aortobronchial fistula after aortoplasty with a Dacron patch for coarctation 23 years earlier, and an aortoesophageal fistula originating from the proximal anastomosis of a graft placed for chronic type B dissection 2 months earlier. Abdominal aortic prosthetic grafts were infected in 14 patients (78%). Aortointestinal fistulae were found in 4: 2 patients had had inadvertent intestinal injury during the implantation of an aortobiiliac bifurcation graft years ago and were referred with huge abscesses; the other 2 patients had a proximal aneurysm of their aortobifemoral bifurcation graft rupturing into the retroperitoneal duodenum (Fig 1). One patient presented with persistent bleeding of the sinus tract of an ureterocutaneous fistula draining into the groin along the aortobifemoral prosthesis. Another patient had undergone 10 reoperations for infection of an aortobifemoral bifurcation prosthesis reaching from the infrarenal aorta to the distal part of the superficial femoral artery on both the right and left sides. Two patients were on dialysis because of end-stage diabetic nephropathy, having had renal transplantation, removal of a necrotic renal graft, aortoiliac reconstruction and placement of an extra-anatomic bypass graft with secondary infection; both were taking multiple combinations of several antibiotics, one for 6 months and the other for 2 years.
VOGT AND TURINA LESS INVASIVE SURGERY IN AORTIC INFECTION
1987
infection of the aortic graft was found, involving only a small portion of the graft anastomosis itself. In the patient presenting with an aortoesophageal fistula originating from the proximal anastomosis of a descending thoracic aortic graft, only the proximal third of the descending aortic graft was replaced with a homograft, leaving the distal two-thirds of the aortic prosthesis in place. Distally, the homograft was anastomosed directly to the prosthetic graft, and the anastomosis was sealed with Gentamicin-impregnated fibrin glue. In the second patient, who presented with an aortoduodenal fistula due to the rupture of a proximal anastomotic aneurysm, only the enlarged proximal part of the bifurcation prosthesis containing the aortointestinal fistula was replaced with a descending aortic homograft (Fig 2). Again, the distal anastomosis between the homograft and the remaining bifurcation of the prosthesis was sealed with Gentamicin-impregnated fibrin glue, taking care not to expose the distal part of the bifurcation prosthesis when performing the anastomosis to the homograft. In the remaining 2 patients, extended groin infection was treated by replacing only the distal part of the corresponding leg of the aortobifemoral prosthesis, using a limited retroperitoneal approach. Postoperatively, the patients were subjected to a 2 to 6 week course of antibiotic treatment. With increasing experience, intravenous antibiotic treatment was reduced to a 4-week course, in some cases on an outpatient basis. In patients with partial graft replacement, a 6-week course of intravenous antibiotic treatment was maintained. Patients with Candida albicans, Candida parapsilosis or Aspergillus fumigatus infection had antifungal treatment for 3 months.
Results Early Results The 30-day mortality rate was 5.5%. One patient with an aortointestinal fistula originating from a previously
Homograft Preparation and Surgical Technique Homograft selection and preparation as well as the surgical technique for total in situ replacement of infected vascular prostheses of the thoracic and abdominal aorta including pelvic and groin vessels using cryopreserved aortic or arterial homografts has already been described in detail elsewhere [8, 9]. Partial replacement of an infected aortic prosthesis was performed in 4 patients (22%), in whom only a limited
Fig 2. Same patient as in Figure 1: partial replacement of the straight part of the infected aortobifemoral bifurcation graft with two descending aorta homografts. Note that the distal anastomosis between the homograft and the remaining vascular prosthesis has been secured with aminoglycoside-impregnated fibrin sealant.
1988
VOGT AND TURINA LESS INVASIVE SURGERY IN AORTIC INFECTION
Ann Thorac Surg 1999;67:1986 –9
placed infrarenal bifurcation prosthesis suffered sudden cardiac death 7 days following an uneventful perioperative and early postoperative course. In addition, a 75year-old man suffered an ischemic perforation of his sigmoid colon after implantation of an infrarenal bifurcation graft, and died 6 weeks later from multiorgan failure due to several intraabdominal abscesses. These 2 early deaths represent a hospital mortality rate of 11%. The mean duration of postoperative mechanical ventilation was 3 days (median 1, range 1–20), the mean intensive care stay was 3.6 days (median 2, range 2– 42), whereas the mean hospital stay was 50632 days (median 25, range 21–150). The mean duration of postoperative antibiotic treatment was 43613 days (median 42, range 28 – 84). Reoperations were necessary in 3 patients: in 2 patients, an additional reconstructive procedure had to be performed for associated peripheral vascular occlusive disease. The third patient had a stabilization of an accidental hip fracture 6 weeks after homograft replacement of infrarenal bifurcation grafts, resulting in a total hospital stay of 150 days. All other hospital stays for more than 21 days were due to the need for prolonged intravenous antibiotic treatment. Delayed wound healing, necessitating secondary surgical wound closure, was observed in 3 patients (17%). No other instances of septic or homograft-related complications were observed, even in patients with partial graft replacement.
Late Results There were 2 late deaths (12.5%) during a follow-up of 22615 months (range, 12– 65): an 85-year-old woman with an infrarenal bifurcation homograft died in another hospital from acute upper gastrointestinal tract bleeding. A homograft-duodenal fistula was found, and persistence of Candida albicans infection at the proximal infrarenal suture line was confirmed at autopsy. In this patient, who was treated early in our experience, antifungal treatment was discontinued after 6 weeks due to its severe side effects. The second late death was an 82-year-old man with an infrarenal bifurcation homograft whose prosthetic infection was caused by Aspergillus, Mycobacterium tuberculosis and Staphlyococcus aureus. He died 18 months later from a homograft-duodenal fistula without evidence of persistent infection. There was 1 late reintervention in a patient who had partial replacement of the proximal third of an infected descending aortic prosthesis: 18 months after surgery, a mean pressure gradient of 42 mm Hg at the proximal anastomosis between the homograft and the vascular prosthesis was successfully reduced to 13 mm Hg by percutaneous placement of a stainless steel stent. There were 2 late deaths which were neither infection nor homograft related: a 41-year-old man died 9 months postoperatively from severe diabetes-related complications, and a 59-year-old man died 29 months postoperatively from chronic renal failure years after failed renal transplantation. Postoperative magnetic resonance imaging, computed tomographic, angiographic and transesophageal echocardiographic findings were normal in all surviving pa-
Fig 3. Three dimensional magnetic resonance imaging: normal findings 21 months after partial replacement of an infected aortobifemoral bifurcation graft with two descending aorta homografts.
tients (Fig 3). No instances of reinfection, false aneurysm formation, homograft leakage, suture line rupture or homograft-related reoperations were observed, resulting in a disease-related cumulative survival of 70% at the end of the follow-up period.
Comment The use of cryopreserved arterial homografts provides a safe and effective treatment for infected aortic grafts and compares favorably with current results obtained by conventional surgical techniques [8, 10]. The use of homografts allows a safe in situ replacement of infected vascular prostheses of the thoracic and abdominal aorta, limiting extended debridement at the site of infection and obviating the need for complex extra-anatomic reconstructions with their distinct disadvantages, such as the aortic stump blowout syndrome [11], and subsequent reinfection of the extra-anatomic grafts [4]. In situ replacement of infected aortic grafts may be the only option for certain sites of infection, eg, transverse aortic arch, proximal ascending aorta or upper abdominal aorta, where an extra-anatomic reconstruction may not be possible. In vitro studies demonstrate cryopreserved arterial homografts to be more resistant to infection than vascu-
Ann Thorac Surg 1999;67:1986 –9
lar prostheses [12]: this has been confirmed recently by several clinical studies [8 –10, 13]. Partial graft replacement may be a further step in this concept, and may be indicated when the vascular prosthetic infection is limited, involving only a small portion of the graft or the anastomosis itself, as demonstrated in our patients. This less invasive technique may simplify the operative procedure without increasing the risk of infection or homograft-related reoperations. The local application of antibiotic-impregnated fibrin sealant has reduced the reoperation rate for recurrent endocarditis with the use of a mechanical prosthesis [14]. We used aminoglycoside-impregnated fibrin sealant to cover the anastomosis between the vascular prosthesis and the homograft. Care was taken to limit the exposure of the remaining part of the prosthetic graft, and to seal its whole visible surface with antibiotic-impregnated fibrin glue. Our in vitro studies demonstrated that the local antibiotic concentration achieved by this technique was 30 to 200 times higher than the highest recommended aminoglycoside levels in the blood using standard intravenous treatment, eradicating even aminoglycosideresistant staphylococci. Thus, antibiotic-impregnated fibrin sealant may prove a useful adjunct in the concept of partial replacement of infected aortic grafts. We are aware that this retrospective study represents only a small series of patients. Larger clinical series and a longer follow-up will be necessary to document the feasibility and the safety of this less invasive approach to prosthetic aortic graft infection. The long-term behavior of cryopreserved aortic homografts is unknown, although the low incidence of late homograft-related problems remains encouraging [9], and is superior to the long-term survival rates reported for conventional surgical treatment of major vascular infections [15, 16]. Programmed cryopreservation maintains the basic structure of collagen and elastic fibers [17], so that the mechanical properties of cryopreserved aortic homografts are not different from those of fresh human thoracic aortas [18], even though immunological rejection of cryopreserved allografts has been described [9, 19]. But we believe that cryopreserved aortic homografts will not behave like freshly implanted allografts, which seem to deteriorate uniformly [20]. Since the operative mortality in the surgical treatment of infected aortic grafts [1–3] is of major concern, and this can be decreased by the use of cryopreserved aortic homografts [8 –10, 13], their use is justified even if they do progressively deteriorate: they can help eradicate the infection and allow later implantation of a prosthetic graft without the risk of recurrent infection. Positive experience with their use in the treatment of vascular infections combined with our results in patients with otherwise intractable aortic infections have convinced us that cryopreserved aortic homografts should be the conduits of choice in the management of infected aortic grafts.
VOGT AND TURINA LESS INVASIVE SURGERY IN AORTIC INFECTION
1989
References 1. Yeager RA. Improving results treating aortic prosthetic graft infection: the case for standard excisional therapy. In: Bunt TJ, ed. Vascular Graft Infection. Armonk, New York: Futura Publishing, 1994, pp. 249 –53. 2. Trout HH, Feinberg RL, Kozloff L. Remote bypass for graft infections. In: Bunt TJ, ed. Vascular Graft Infection. Armonk, New York: Futura Publishing, 1994, pp. 267–76. 3. Steed DL. In situ repair of infected aortic grafts. In: Bunt TJ, ed. Vascular Graft Infection. Armonk, New York: Futura Publishing, 1994, pp. 211–22. 4. Yeager RA, Moneta GL, Taylor LM JR, et al. Improving survival and limb salvage in patients with aortic infection. Am J Surg 1990;159:466–9. 5. Bergeron P. Espinoza H, Rudondy P, et al. Secondary aortoduodenal fistulas: value of initial axillofemoral bypass. Ann Vasc Surg 1991;5:4–7. 6. Ricotta JJ, Faggiolo GL, Stella A, et al. Total excision and extra-anatomic bypass for aortic graft infection. Am J Surg 1991;162:145–9. 7. Vogt PR, von Segesser LK, Jenni R, et al. Emergency surgery for acute infective aortic valve endocarditis: performance of cryopreserved homografts and mode of failure. Eur J Cardiothorac Surg 1997;11:53– 61. 8. Vogt PR, von Segesser LK, Goffin Y, et al. Eradication of aortic infection with the use of cryopreserved arterial homografts. Ann Thorac Surg 1996;62:640–5. 9. Goffin YA, Grandmougin D, Wozniak G, et al. Banking and distribution of large cryopreserved arterial homografts in Brussels: assessment of 4 years of activity by the European Homograft Bank (EHB) with references of implantation results in reconstruction of infected infra-renal arterial prostheses and mycotic aneurysms. Vasc Surg 1998;32:19–32. 10. Mestres CA, Mulet J, Pomar JL. Large-caliber cryopreserved arterial allografts in vascular reconstructive operations: early experience. Ann Thorac Surg 1995;60:S105–7. 11. Reilly L, Altman H, Lusby B, et al. Late results following surgical management of vascular graft infection. J Vasc Surg 1984; 1:36– 44. 12. Koskas F, Goeau-Brissonniere O, Nicolas M-H, et al. Arteries from human beings are less infectible by Staphylococcus aureus than polytetrafluoroethylene in an aortic dog model. J Vasc Surg 1996;23:472– 6. 13. Knosalla C, Weng Y, Yankah AC, et al. Using aortic allograft material to treat mycotic aneurysms of the thoracic aorta. Ann Thorac Surg 1996;61:146–52. 14. Watanabe G, Haverich A, Speier R, et al. Surgical treatment of active endocarditis with paravalvular involvement. J Thorac Cardiovasc Surg 1994;107:171–7. 15. O’Hara PJ, Hertger NR, Beven EG, et al. Surgical management of infected abdominal aortic grafts: review of a 25-year experience. J Vasc Surg 1986;3:725–31. 16. Quinones Baldrich WJ, Hernandez JJ, Moore WS. Longterm results following surgical management of aortic graft infection. Arch Surg 1991;126:507–11. 17. Gournier JP, Adham M, Favre JP, et al. Cryopreserved arterial homografts: preliminary study. Ann Vasc Surg 1993; 7:503–11. 18. Courtman DW, Pereira CA, Omar S, et al. Biomechanical and ultrastructural comparison of cryopreservation and a novel cellular extraction of porcine aortic valve leaflets. J Biomed Mater Res 1995;29:1507–16. 19. Neves J, Monteiro C, Santos R, et al. Histologic and genetic assessment of explanted allograft valves. Ann Thorac Surg 1995;60:S141–5. 20. Szilagyi DE, Rodriquez FJ, Smith RF, et al. Late fate of arterial allografts. Observations 6 to 15 years after implantation. Arch Surg 1970;101:721–33.
Background. Aortic graft infection is associated with significant mortality and morbidity. Total graft replacement with homografts provides an effective treatment. Partial graft replacement at the site of infection may simplify the surgical procedure. Methods. Between January 1991 and December 1996, homografts were used in 18 patients (mean age, 61612 years; range 41– 85) with thoracic (4/18; 22%) or abdominal (14/18; 78%) aortic graft infection. Sepsis was present in 14 patients (78%); 6 (33%) had various aortic fistulae. Total graft replacement using homografts was performed in 14 (78%), and partial graft replacement at the site of infection in 4 patients (22%).
Results. Hospital mortality was 11%. During the follow-up period of 22615 months (range, 12– 65) there was 1 infection and 1 homograft-related late death after complete homograft replacement, and 1 percutaneous vascular stent placement after partial graft replacement. No other instances of reinfection, suture line rupture or anastomotic aneurysms were observed. Conclusion. Total graft replacement with homografts provides an effective treatment for infected aortic grafts. Partial graft replacement at the site of infection is feasible and safe. (Ann Thorac Surg 1999;67:1986 –9) © 1999 by The Society of Thoracic Surgeons
V
Material and Methods
ascular prosthetic graft infection is a life threatening pathologic entity with a high morbidity and mortality [1, 2]. Above all, the surgical treatment for infected prosthetic grafts of the thoracic or abdominal aorta remains a most demanding technical challenge [3]. Conventional surgical treatment, in situ repair with a new vascular prosthesis or complete removal of the infected aortic graft and reconstruction with an extra anatomical bypass graft, has a substantial operative mortality rate. Despite the improving results [4, 5] of initial surgical treatment, prosthetic aortic graft infection is frequently associated with reinfection, suture line rupture or anastomotic aneurysm, leading to multiple reoperations, distal amputations and prolonged, even life-long antibiotic treatment [6]. Recent findings with cryopreserved heart valve homografts in the treatment of acute infectious endocarditis [7] led us to use cryopreserved homografts for in situ replacement of infected aortic grafts. Their observed clinical resistance to bacterial infections [8] prompted us to limit the surgical procedure, performing only partial graft replacement when the vascular prosthetic infection involved only a small portion of the graft or was limited to the aortic graft anastomosis. This study presents our clinical experience with the treatment of infected aortic grafts, performing either total or partial graft replacement with cryopreserved homografts. Presented at the Aortic Surgery Symposium VI, April 30 –May 1, 1998, New York, NY. Address reprint requests to Dr Vogt, Clinic for Cardiovascular Surgery, University Hospital, Ramistr 100, CH-8091 Zurich, Switzerland; e-mail: [email protected].
© 1999 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
Patients Between January 1991 and December 1996, cryopreserved aortic homografts were used in 18 patients (mean age, 61612 years, range 41– 85) with vascular prosthetic infection of the thoracic (4/18; 22%) or abdominal aorta (14/18; 78%), with or without involvement of the pelvic and groin vessels. Sepsis was present preoperatively in 14 patients (78%), and 6 (33%) had aortobronchial, aortoesophageal or aortoenteric fistulae originating from the infected aortic graft. In 7 patients (39%), as many as 10 previous prosthetic graft infection-related vascular surgical procedures had been done, and as many as 6 different new prosthetic grafts had already been inserted. Preoperatively, blood cultures were positive in 10 patients (55%), and 15 (83%) had received various antibiotic treatments for up to 3 years. Concomitant peripheral arterial vascular occlusion was found in 12 patients (67%), diabetes mellitus in 5 (28%), chronic atrial fibrillation in 3 (17%), coronary artery disease in 13 (72%) and chronic obstructive lung disease in 12 patients (67%). Renal insufficiency was found in 7 patients (39%), 2 were on dialysis, and 2 had had renal transplantation because of end-stage diabetic nephropathy. Emergency surgery was performed in 10 patients (55%), and the remaining 8 (45%) were operated on within 72 hours. Prosthetic vascular grafts of the thoracic aorta were affected in 4 patients (22%). One had infection of an ascending aortic graft previously implanted for an acute type A dissection, and the other had an infected composite graft after repair of annuloaortic ectasia. The other 2 0003-4975/99/$20.00 PII S0003-4975(99)00357-4
Ann Thorac Surg 1999;67:1986 –9
Fig 1. Preoperative computed tomogram of a 46-year-old patient with an aortointestinal fistula from a large juxtarenal and infrarenal aneurysm of the proximal anastomosis of a 7-year-old aorto bifemoral bifurcation graft, rupturing into the retroperitoneal duodenum.
patients had infection of a descending aortic graft: an aortobronchial fistula after aortoplasty with a Dacron patch for coarctation 23 years earlier, and an aortoesophageal fistula originating from the proximal anastomosis of a graft placed for chronic type B dissection 2 months earlier. Abdominal aortic prosthetic grafts were infected in 14 patients (78%). Aortointestinal fistulae were found in 4: 2 patients had had inadvertent intestinal injury during the implantation of an aortobiiliac bifurcation graft years ago and were referred with huge abscesses; the other 2 patients had a proximal aneurysm of their aortobifemoral bifurcation graft rupturing into the retroperitoneal duodenum (Fig 1). One patient presented with persistent bleeding of the sinus tract of an ureterocutaneous fistula draining into the groin along the aortobifemoral prosthesis. Another patient had undergone 10 reoperations for infection of an aortobifemoral bifurcation prosthesis reaching from the infrarenal aorta to the distal part of the superficial femoral artery on both the right and left sides. Two patients were on dialysis because of end-stage diabetic nephropathy, having had renal transplantation, removal of a necrotic renal graft, aortoiliac reconstruction and placement of an extra-anatomic bypass graft with secondary infection; both were taking multiple combinations of several antibiotics, one for 6 months and the other for 2 years.
VOGT AND TURINA LESS INVASIVE SURGERY IN AORTIC INFECTION
1987
infection of the aortic graft was found, involving only a small portion of the graft anastomosis itself. In the patient presenting with an aortoesophageal fistula originating from the proximal anastomosis of a descending thoracic aortic graft, only the proximal third of the descending aortic graft was replaced with a homograft, leaving the distal two-thirds of the aortic prosthesis in place. Distally, the homograft was anastomosed directly to the prosthetic graft, and the anastomosis was sealed with Gentamicin-impregnated fibrin glue. In the second patient, who presented with an aortoduodenal fistula due to the rupture of a proximal anastomotic aneurysm, only the enlarged proximal part of the bifurcation prosthesis containing the aortointestinal fistula was replaced with a descending aortic homograft (Fig 2). Again, the distal anastomosis between the homograft and the remaining bifurcation of the prosthesis was sealed with Gentamicin-impregnated fibrin glue, taking care not to expose the distal part of the bifurcation prosthesis when performing the anastomosis to the homograft. In the remaining 2 patients, extended groin infection was treated by replacing only the distal part of the corresponding leg of the aortobifemoral prosthesis, using a limited retroperitoneal approach. Postoperatively, the patients were subjected to a 2 to 6 week course of antibiotic treatment. With increasing experience, intravenous antibiotic treatment was reduced to a 4-week course, in some cases on an outpatient basis. In patients with partial graft replacement, a 6-week course of intravenous antibiotic treatment was maintained. Patients with Candida albicans, Candida parapsilosis or Aspergillus fumigatus infection had antifungal treatment for 3 months.
Results Early Results The 30-day mortality rate was 5.5%. One patient with an aortointestinal fistula originating from a previously
Homograft Preparation and Surgical Technique Homograft selection and preparation as well as the surgical technique for total in situ replacement of infected vascular prostheses of the thoracic and abdominal aorta including pelvic and groin vessels using cryopreserved aortic or arterial homografts has already been described in detail elsewhere [8, 9]. Partial replacement of an infected aortic prosthesis was performed in 4 patients (22%), in whom only a limited
Fig 2. Same patient as in Figure 1: partial replacement of the straight part of the infected aortobifemoral bifurcation graft with two descending aorta homografts. Note that the distal anastomosis between the homograft and the remaining vascular prosthesis has been secured with aminoglycoside-impregnated fibrin sealant.
1988
VOGT AND TURINA LESS INVASIVE SURGERY IN AORTIC INFECTION
Ann Thorac Surg 1999;67:1986 –9
placed infrarenal bifurcation prosthesis suffered sudden cardiac death 7 days following an uneventful perioperative and early postoperative course. In addition, a 75year-old man suffered an ischemic perforation of his sigmoid colon after implantation of an infrarenal bifurcation graft, and died 6 weeks later from multiorgan failure due to several intraabdominal abscesses. These 2 early deaths represent a hospital mortality rate of 11%. The mean duration of postoperative mechanical ventilation was 3 days (median 1, range 1–20), the mean intensive care stay was 3.6 days (median 2, range 2– 42), whereas the mean hospital stay was 50632 days (median 25, range 21–150). The mean duration of postoperative antibiotic treatment was 43613 days (median 42, range 28 – 84). Reoperations were necessary in 3 patients: in 2 patients, an additional reconstructive procedure had to be performed for associated peripheral vascular occlusive disease. The third patient had a stabilization of an accidental hip fracture 6 weeks after homograft replacement of infrarenal bifurcation grafts, resulting in a total hospital stay of 150 days. All other hospital stays for more than 21 days were due to the need for prolonged intravenous antibiotic treatment. Delayed wound healing, necessitating secondary surgical wound closure, was observed in 3 patients (17%). No other instances of septic or homograft-related complications were observed, even in patients with partial graft replacement.
Late Results There were 2 late deaths (12.5%) during a follow-up of 22615 months (range, 12– 65): an 85-year-old woman with an infrarenal bifurcation homograft died in another hospital from acute upper gastrointestinal tract bleeding. A homograft-duodenal fistula was found, and persistence of Candida albicans infection at the proximal infrarenal suture line was confirmed at autopsy. In this patient, who was treated early in our experience, antifungal treatment was discontinued after 6 weeks due to its severe side effects. The second late death was an 82-year-old man with an infrarenal bifurcation homograft whose prosthetic infection was caused by Aspergillus, Mycobacterium tuberculosis and Staphlyococcus aureus. He died 18 months later from a homograft-duodenal fistula without evidence of persistent infection. There was 1 late reintervention in a patient who had partial replacement of the proximal third of an infected descending aortic prosthesis: 18 months after surgery, a mean pressure gradient of 42 mm Hg at the proximal anastomosis between the homograft and the vascular prosthesis was successfully reduced to 13 mm Hg by percutaneous placement of a stainless steel stent. There were 2 late deaths which were neither infection nor homograft related: a 41-year-old man died 9 months postoperatively from severe diabetes-related complications, and a 59-year-old man died 29 months postoperatively from chronic renal failure years after failed renal transplantation. Postoperative magnetic resonance imaging, computed tomographic, angiographic and transesophageal echocardiographic findings were normal in all surviving pa-
Fig 3. Three dimensional magnetic resonance imaging: normal findings 21 months after partial replacement of an infected aortobifemoral bifurcation graft with two descending aorta homografts.
tients (Fig 3). No instances of reinfection, false aneurysm formation, homograft leakage, suture line rupture or homograft-related reoperations were observed, resulting in a disease-related cumulative survival of 70% at the end of the follow-up period.
Comment The use of cryopreserved arterial homografts provides a safe and effective treatment for infected aortic grafts and compares favorably with current results obtained by conventional surgical techniques [8, 10]. The use of homografts allows a safe in situ replacement of infected vascular prostheses of the thoracic and abdominal aorta, limiting extended debridement at the site of infection and obviating the need for complex extra-anatomic reconstructions with their distinct disadvantages, such as the aortic stump blowout syndrome [11], and subsequent reinfection of the extra-anatomic grafts [4]. In situ replacement of infected aortic grafts may be the only option for certain sites of infection, eg, transverse aortic arch, proximal ascending aorta or upper abdominal aorta, where an extra-anatomic reconstruction may not be possible. In vitro studies demonstrate cryopreserved arterial homografts to be more resistant to infection than vascu-
Ann Thorac Surg 1999;67:1986 –9
lar prostheses [12]: this has been confirmed recently by several clinical studies [8 –10, 13]. Partial graft replacement may be a further step in this concept, and may be indicated when the vascular prosthetic infection is limited, involving only a small portion of the graft or the anastomosis itself, as demonstrated in our patients. This less invasive technique may simplify the operative procedure without increasing the risk of infection or homograft-related reoperations. The local application of antibiotic-impregnated fibrin sealant has reduced the reoperation rate for recurrent endocarditis with the use of a mechanical prosthesis [14]. We used aminoglycoside-impregnated fibrin sealant to cover the anastomosis between the vascular prosthesis and the homograft. Care was taken to limit the exposure of the remaining part of the prosthetic graft, and to seal its whole visible surface with antibiotic-impregnated fibrin glue. Our in vitro studies demonstrated that the local antibiotic concentration achieved by this technique was 30 to 200 times higher than the highest recommended aminoglycoside levels in the blood using standard intravenous treatment, eradicating even aminoglycosideresistant staphylococci. Thus, antibiotic-impregnated fibrin sealant may prove a useful adjunct in the concept of partial replacement of infected aortic grafts. We are aware that this retrospective study represents only a small series of patients. Larger clinical series and a longer follow-up will be necessary to document the feasibility and the safety of this less invasive approach to prosthetic aortic graft infection. The long-term behavior of cryopreserved aortic homografts is unknown, although the low incidence of late homograft-related problems remains encouraging [9], and is superior to the long-term survival rates reported for conventional surgical treatment of major vascular infections [15, 16]. Programmed cryopreservation maintains the basic structure of collagen and elastic fibers [17], so that the mechanical properties of cryopreserved aortic homografts are not different from those of fresh human thoracic aortas [18], even though immunological rejection of cryopreserved allografts has been described [9, 19]. But we believe that cryopreserved aortic homografts will not behave like freshly implanted allografts, which seem to deteriorate uniformly [20]. Since the operative mortality in the surgical treatment of infected aortic grafts [1–3] is of major concern, and this can be decreased by the use of cryopreserved aortic homografts [8 –10, 13], their use is justified even if they do progressively deteriorate: they can help eradicate the infection and allow later implantation of a prosthetic graft without the risk of recurrent infection. Positive experience with their use in the treatment of vascular infections combined with our results in patients with otherwise intractable aortic infections have convinced us that cryopreserved aortic homografts should be the conduits of choice in the management of infected aortic grafts.
VOGT AND TURINA LESS INVASIVE SURGERY IN AORTIC INFECTION
1989
References 1. Yeager RA. Improving results treating aortic prosthetic graft infection: the case for standard excisional therapy. In: Bunt TJ, ed. Vascular Graft Infection. Armonk, New York: Futura Publishing, 1994, pp. 249 –53. 2. Trout HH, Feinberg RL, Kozloff L. Remote bypass for graft infections. In: Bunt TJ, ed. Vascular Graft Infection. Armonk, New York: Futura Publishing, 1994, pp. 267–76. 3. Steed DL. In situ repair of infected aortic grafts. In: Bunt TJ, ed. Vascular Graft Infection. Armonk, New York: Futura Publishing, 1994, pp. 211–22. 4. Yeager RA, Moneta GL, Taylor LM JR, et al. Improving survival and limb salvage in patients with aortic infection. Am J Surg 1990;159:466–9. 5. Bergeron P. Espinoza H, Rudondy P, et al. Secondary aortoduodenal fistulas: value of initial axillofemoral bypass. Ann Vasc Surg 1991;5:4–7. 6. Ricotta JJ, Faggiolo GL, Stella A, et al. Total excision and extra-anatomic bypass for aortic graft infection. Am J Surg 1991;162:145–9. 7. Vogt PR, von Segesser LK, Jenni R, et al. Emergency surgery for acute infective aortic valve endocarditis: performance of cryopreserved homografts and mode of failure. Eur J Cardiothorac Surg 1997;11:53– 61. 8. Vogt PR, von Segesser LK, Goffin Y, et al. Eradication of aortic infection with the use of cryopreserved arterial homografts. Ann Thorac Surg 1996;62:640–5. 9. Goffin YA, Grandmougin D, Wozniak G, et al. Banking and distribution of large cryopreserved arterial homografts in Brussels: assessment of 4 years of activity by the European Homograft Bank (EHB) with references of implantation results in reconstruction of infected infra-renal arterial prostheses and mycotic aneurysms. Vasc Surg 1998;32:19–32. 10. Mestres CA, Mulet J, Pomar JL. Large-caliber cryopreserved arterial allografts in vascular reconstructive operations: early experience. Ann Thorac Surg 1995;60:S105–7. 11. Reilly L, Altman H, Lusby B, et al. Late results following surgical management of vascular graft infection. J Vasc Surg 1984; 1:36– 44. 12. Koskas F, Goeau-Brissonniere O, Nicolas M-H, et al. Arteries from human beings are less infectible by Staphylococcus aureus than polytetrafluoroethylene in an aortic dog model. J Vasc Surg 1996;23:472– 6. 13. Knosalla C, Weng Y, Yankah AC, et al. Using aortic allograft material to treat mycotic aneurysms of the thoracic aorta. Ann Thorac Surg 1996;61:146–52. 14. Watanabe G, Haverich A, Speier R, et al. Surgical treatment of active endocarditis with paravalvular involvement. J Thorac Cardiovasc Surg 1994;107:171–7. 15. O’Hara PJ, Hertger NR, Beven EG, et al. Surgical management of infected abdominal aortic grafts: review of a 25-year experience. J Vasc Surg 1986;3:725–31. 16. Quinones Baldrich WJ, Hernandez JJ, Moore WS. Longterm results following surgical management of aortic graft infection. Arch Surg 1991;126:507–11. 17. Gournier JP, Adham M, Favre JP, et al. Cryopreserved arterial homografts: preliminary study. Ann Vasc Surg 1993; 7:503–11. 18. Courtman DW, Pereira CA, Omar S, et al. Biomechanical and ultrastructural comparison of cryopreservation and a novel cellular extraction of porcine aortic valve leaflets. J Biomed Mater Res 1995;29:1507–16. 19. Neves J, Monteiro C, Santos R, et al. Histologic and genetic assessment of explanted allograft valves. Ann Thorac Surg 1995;60:S141–5. 20. Szilagyi DE, Rodriquez FJ, Smith RF, et al. Late fate of arterial allografts. Observations 6 to 15 years after implantation. Arch Surg 1970;101:721–33.