Selective preservation of infected prosthetic arterial grafts. Analysis of a 20-year experience with 120 extracavitary-infected grafts

Intra-arterial device infections

As an increasing number and variety of prosthetic devices are used in cardiovascular medicine, novel infectious complications have been described. Infection of intra-arterial devices, including arterial closure devices, prosthetic carotid patches, coronary artery stents and endovascular stents, and stent-grafts, is now being reported. Prosthetic vascular graft infection is an older, more common, and better-characterized entity, but recent developments in the surgical management of these infections have prompted a re-examination of the syndrome. Staphylococcal species account for most intra-arterial device infections, and often, morbidity and mortality rates are high. An update on intra-arterial device infections is warranted.

Intra-abdominal aortic graft infection: complete or partial graft preservation in patients at very high risk

BACKGROUND

Total graft excision with in situ or extra-anatomic revascularization is considered mandatory to treat infection involving the body of aortic grafts. We present a series of nine patients with this complication and such severe comorbid medical illnesses or markedly hostile abdomens that traditional treatments were precluded. In these patients selective complete or partial graft preservation was used.

METHODS

Over the past 20 years we have treated nine infected infrarenal aortic prosthetic grafts with complete or partial graft preservation, because excision of the graft body was not feasible. In all nine patients infection of the main body of the aortic graft was documented at computed tomography or surgery. Essential adjuncts included percutaneous or operative drain placement into retroperitoneal abscess cavities and along the graft, with instillation of antibiotics three times daily, repeated debridement of infected groin wounds, and intravenous antibiotic therapy for at least 6 weeks.

RESULTS

One patient with purulent groin drainage treated with complete graft preservation died of sepsis. One patient with groin infection treated with complete graft preservation initially did well, but ultimately required total graft excision 5 months later, after clinical improvement. In four patients complete graft preservation was successful; two patients required excision of an occluded infected limb of the graft; and one patient underwent subtotal graft excision, leaving a graft remnant on the aorta, and axillopopliteal bypass. In summary, seven of nine patients survived hospitalization after complete or partial graft preservation; amputation was avoided in all but one patient; and no recurrent infection developed over mean follow-up of 7.6 years (range, 2-15 years).

CONCLUSIONS

Although contrary to conventional concepts, partial or complete graft preservation combined with aggressive drainage and groin wound debridement is an acceptable option for treatment of infection involving an entire aortic graft in selected patients with prohibitive risks for total graft excision. This treatment may be compatible with long-term survival and protracted absence of signs or symptoms of infection.

Use of a local fasciocutaneous flap for treatment of exposed vascular grafts to the dorsalis pedis artery

Exposed or infected peripheral vascular grafts pose a significant challenge to the vascular surgeon. Although graft removal and extraanatomic bypass is feasible in selected circumstances, this procedure is generally not applicable for bypass to the pedal vessels. Preservation of patent grafts is almost always required for limb salvage. We present a case report of an exposed vein graft to the dorsalis pedis artery. We conclude that a local fasciocutaneous flap is an excellent treatment option, and describe the procedure in detail.

Salvage of infected expander prostheses in breast reconstruction

Periprosthetic infection is a devastating complication following breast reconstruction with prostheses. Traditional surgical principles dictate removal of the prosthesis to control infection. Although successful salvage of prostheses in the presence of periprosthetic infections has been reported in the plastic and other surgical literature, salvage procedures remain seldom practiced. Reports in the plastic surgery literature have been limited to implant salvage following cosmetic breast augmentation and subcutaneous mastectomy with implants. Salvage of saline-filled expander prostheses used in breast reconstruction following mastectomy for cancer has not been previously reported. The authors review their experience with implant salvage in patients with periprosthetic infections following breast reconstruction for a 6-year period. Fourteen patients (13 with saline-filled expander prostheses and one with silicone prosthesis) underwent implant salvage. Salvage of the breast reconstruction was successful in nine patients. Staphylococcus aureus infection was associated with poorer salvage rate (p = 0.023). Previous radiotherapy to the chest wall did not affect the salvage outcome (p = 0.50). In selected patients, immediate salvage of a breast reconstruction in the presence of prosthesis-related infection remains an alternative to implant removal followed by delayed reconstruction.

Risk factors associated with infection of lower extremity revascularization: analysis of 365 procedures performed at a teaching hospital

Infection of arterial reconstructions is associated with high rates of mortality and limb loss despite optimal treatment. Lower extremity revascularization procedures performed at a teaching hospital were reviewed to identify risk factors associated with wound infection. Medical records, postoperative infection surveillance forms, and a computerized vascular registry for lower extremity revascularizations involving a common femoral or more distal artery during a 3-year period were reviewed. There were 335 bypass operations (184 femoral-distal, 36 popliteal-distal, 17 aortofemoral, 13 femorofemoral, 11 axillofemoral, 74 graft revisions) and 30 other vascular procedures (arterial thrombectomy or endarterectomy). Factors analyzed included age, gender, diabetes mellitus, dialysis dependence, malnutrition, obesity, ipsilateral foot ulcer or gangrene, separate admissions within the month preceding surgery, length of hospital stay before surgery, length of operation, wound hematoma requiring reoperation, vein or prosthetic grafts, or redo surgery. Risk factors commonly thought to increase wound infection following lower extremity revascularizations, such as diabetes, obesity, renal failure, redo surgery, and prosthetic grafts, did not predict this complication in this series. Given the correlation of operative time with infection, efforts to minimize operative time by "double-teaming" staff participation in teaching cases may decrease infection rates, although this is speculative. Vascular services should institute strategies to ensure that appropriate prophylactic antibiotics are administered in a timely fashion before lower extremity revascularizations.

Prosthetic patch remnants to treat infected arterial grafts

PURPOSE

Our previous experience with the traditional management of infected prosthetic arterial grafts, which included graft excision and vein patch repair of the involved artery, was complicated by a high incidence of vein patch rupture. This study assessed the treatment of infected prosthetic grafts with subtotal graft excision and oversewing of small graft remnants.

METHODS

During the last 20 years, we treated 53 wounds involving 45 infected prosthetic grafts in 42 patients by means of subtotal graft excision and oversewing of a residual 2- to 3-mm graft remnant (patch) at an intact arterial anastomosis. This technique was selectively used to maintain patency of small-diameter arteries (41 common femoral, five deep femoral, three axillary, two iliac, and two popliteal), which were critical for limb salvage or amputation healing. This strategy avoided difficult dissection of the underlying artery in scarred wounds and obviated the placement of a new patch in an infected field. Graft remnants were polytetrafluoroethylene in 51 cases and Dacron in two cases. Of the 45 grafts, 31 were occluded and 14 were patent. All infected tissue was widely debrided, wet-to-dry dressing changes were performed three times daily, and appropriate intravenous antibiotics were administered for at least 1 week. Secondary bypass grafting procedures were performed as needed to achieve limb salvage. The follow-up period in surviving patients averaged 32 months (range, 1 to 218 months).

RESULTS

No complications were directly attributable to prosthetic patch remnants in 92% of cases (49 of 53 cases). Six of 42 patients (14%) died during hospitalization (three of cardiac complications and three of sepsis with multiple organ failure). Two infected pseudoaneurysms developed 8 and 34 months after surgery, and two wounds failed to heal. Sixteen secondary bypass grafting procedures were necessary to achieve limb salvage. Patch oversewing led to limb salvage without the need for secondary revascularization in 26 other cases and to the successful healing of 10 amputated limbs when secondary revascularization was not possible.

CONCLUSION

Prosthetic patch remnants are a useful adjunct that simplify management of infected prosthetic grafts, are associated with a low incidence of wound complications, and help maintain patency of essential collaterals to achieve limb salvage or heal an amputation.

Management of Patients with Prosthetic Vascular Graft Infection

Management of patients with infected prosthetic vascular grafts is one of the most difficult challenges faced by the vascular surgeon. Patients often present with nonspecific symptoms, but delay in treatment can lead to life-threatening sepsis and/or hemorrhage. Fortunately, prosthetic vascular graft infection is uncommon, with the incidence varying between 1 and 6 per cent, depending on the location of the graft. Initially, the potentially infected vascular graft should be imaged using either CT or magnetic resonance imaging, with radionuclide studies being reserved for those instances in which imaging studies do not confirm or exclude the diagnosis of infection. Current treatments for prosthetic vascular graft infection include attempted graft preservation, graft removal with in situ graft replacement (using autogenous or new prosthetic grafts), and graft removal with extra-anatomic bypass. Morbidity and mortality associated with treatment, likelihood of long-term limb salvage, and likelihood of persistent or recurrent infection vary among these types of treatment. Therefore, in an individual patient with a prosthetic vascular graft infection, many things must be considered to appropriately determine the treatment most likely to achieve eradication of the infection and long-term limb salvage with the lowest risk. Regardless, with appropriate application of the techniques currently available for treatment of prosthetic vascular graft infection, long-term elimination of infection and limb preservation can be achieved in the great majority of patients with this grave problem.

Bacterial components inhibit fibroblast proliferation in vitro

Perigraft fluid from Staphylococcus epidermidis infected grafts in a mouse model significantly inhibits fibroblast proliferation (60-98% at 7 and 28 days), compared with perigraft fluid from sterile grafts. The fibroblast inhibitor was trypsin-heat resistant and dependent primarily upon the bacteria, not the host proinflammatory mediators or the vascular graft biomaterial. We tested the inhibitory properties of S. epidermidis strains RP62A (slime producer) and RP62NA (nonslime producer) and Staphylococcus aureus strain 502a, using an in vitro tritiated thymidine murine fibroblast (ATCC CCL-12) proliferation assay. Whole killed bacteria, disrupted bacteria (live and killed), bacterial supernatants, and purified cell wall products (peptidoglycan, teichoic acid, and lipoteichoic acid from disrupted bacteria) were studied. Significant fibroblast inhibition occurred for all three bacterial strains with disrupted bacteria (live or killed) and cell free bacteria derived supernatants. The fibroblast inhibitor from disrupted slime producing S. epidermidis was trypsin-heat resistant. The fibroblast inhibitor from disrupted S. aureus and supernatants for all three bacterial strains at 1 x 10(7) were trypsin-heat sensitive. Fibroblast inhibition was not dependent upon bacterial viability and not mediated by bacterial cell wall products. In conclusion, components of slime and nonslime producing S. epidermidis and S. aureus inhibit fibroblast proliferation.

Comparison of healing in fresh and preserved arterial allografts in the dog

The use of aortic allografts for the management of vascular prosthetic infections has recently been reintroduced. Impressive results have been obtained; however, the possibility of late degeneration remains a major concern. The healing behavior of aortic allografts, either fresh or preserved, in antibiotic-supplemented nutrient medium at 4 degrees C for 1 week and used as thoracic aorta substitutes in dogs was investigated after 6 months of implantation. Four dogs received a fresh aortic allograft from four different donors, and four dogs received a preserved allograft from two different donors. Autografts in two dogs were performed as controls. The in vivo investigation was conducted to describe (1) the histological characteristics of the arterial wall, (2) the macroscopic and thrombogenic aspect of the luminal surface, (3) the integrity of the endothelial lining by scanning electron microscopy, and (4) its biochemical function by prostacyclin (PGI2) and thromboxane A2 (TXA2) secretion. Immune-mediated reactions directed toward the grafts were measured by sequential screening of donor-specific serum antibody development. All donor-recipient pairs of dogs were major histocompatibility complex (MHC)-incompatible according to a mixed lymphocyte reaction (MLR) assay. From the results of this study we concluded that although preserved arterial allografts exhibited similar surface characteristics as those of fresh allografts in terms of re-endothelialization and long-term graft function, an elicited immune response, a degenerative process in the media, and a hyperplasic reaction in the intima could not be prevented using this method of preservation.

Bacterial products primarily mediate fibroblast inhibition in biomaterial infection

PURPOSE

The stimulation of fibroblast growth is essential for the normal healing and tissue integration of biomaterials. The local elevation of proinflammatory mediators in infected perigraft fluid (PGF) may inhibit this growth. We sought to determine whether infected PGF inhibited fibroblast growth, and, if so, whether this was primarily dependent on the biomaterial, bacteria, or host.

METHODS

In vivo Dacron or expandable polytetra-fluoroethylene (ePTFE) grafts, sterile or colonized with slime-producing (RP-62A, viable or formalin-killed) or nonslime-producing (RP-62NA) Staphylococcus epidermidis (1 x 10(7) CFU/cm2), were implanted in Swiss Webster mice, and the PGF was harvested at 7 and 28 days. Antibodies to tumor necrosis factor alpha, interleukin 1 alpha, interferon gamma (7 micrograms/day), and indomethacin (50 micrograms/day) were administered by microinfusion pumps for 7 days and the PGF was harvested. Inhibition of the proinflammatory mediators was confirmed by enzyme-linked immunosorbant assay. The nontreated, heat-treated, or trypsin-digested in vivo PGF was incubated with an in vitro [3H]thymidine murine fibroblast (ATCC CCL-12) proliferation assay.

RESULTS

Fibroblast inhibition was significant at 7 and 28 days with infected PGF incubation compared with sterile and was not dependent on bacterial slime production or viability. Dacron sterile PGF did not significantly inhibit fibroblasts compared with control, whereas sterile ePTFE stimulated (P < 0.05) fibroblasts. Treatment of the PGF with proinflammatory cytokines, heat, and trypsin failed to reverse fibroblast inhibition in the infected state.

CONCLUSION

Biomaterial infection is associated with fibroblast inhibition that is dependent primarily on bacterial products and not the host or biomaterial. Conservative intervention strategies for graft infection need to address the problem of poor healing as well as bacterial clearance.

Salvage of femoropedal bypass graft complicated by interval gangrene and vein graft blowout using a flow-through radial forearm fasciocutaneous free flap

We report the case of a 71-year-old man who had interval gangrene of his calf with subsequent vein graft blowout 3 months after undergoing a femoral-to-dorsalis pedis saphenous vein bypass grafting procedure. To provide wound coverage, restore vascular continuity, and preserve functional ambulation, a flow-through radial forearm fasciocutaneous free flap was interposed between cut ends of the bypass graft. Venous drainage of the flap was from the cephalic vein to the popliteal vein. At 1 month after the operation, the patient had complete wound healing and began to ambulate. At 11 months an asymptomatic high-grade stenosis in the distal radial artery segment of the reconstruction was successfully treated with percutaneous angioplasty. After 22 months of follow-up there have been no further complications, and the patient continues to have full, functional ambulation. The radial forearm flow-through free flap allows single-stage restoration of bypass graft continuity and coverage of extensive, complex tissue defects. This technique represents a novel approach to this difficult problem and provides a viable alternative to major limb amputation.

Alternative approach for management of infected aortic grafts

PURPOSE

Prosthetic infection after aortic reconstructive surgery historically has been treated with extraanatomical bypass, graft excision, and aortic stump closure, but at the cost of substantial mortality and amputation rates. Alternatives to this strategy include in situ prosthetic replacement in the infected area, as well as autogenous reconstructions. Inherent to all of these procedures, however, is either the creation of an aortic stump, which carries a significant risk of subsequent blowout, or the placement of a bypass conduit in the infected field, thereby maintaining the potential for subsequent infectious complications. To avoid such problems, we have used retroperitoneal in-line aortic bypass with polytetrafluoroethylene through dean tissue planes.

METHODS

Since 1987 we have treated 16 graft infections in this manner. The surgical approach consisted of obtaining retroperitoneal proximal aortic control outside of the infected field (above or below the renal arteries), followed by infrarenal division and oversewing of the distal aorta. A polytetrafluoroethylene bifurcated graft was then sewn to the proximal aorta and tunnelled through the psoas sheath laterally to the profunda femoris artery on the ipsilateral side and via the space of Retzius to the contralateral appropriate femoral vessel, so as to avoid any contact with the infected areas. After the closure of the wounds, a plastic barrier was placed over all incisions and the patient was placed supine. The old infected graft was removed transperitoneally. Extensive cultures were taken at various sites to demonstrate no cross-contamination.

RESULTS

All patients were followed-up clinically and with tagged white cell scans at 6-month intervals. There were no immediate postoperative deaths and no amputations. One patient had a myocardial infarction and died at 5 months, and a second patient died at 2 months. Of the remaining 14 patients, none had recurrent sepsis and all have had negative Indium-labeled white cell scans in follow-up. Eleven (78%) are still alive, with a mean follow-up of 32 months (range, 20 to 106 months).

CONCLUSIONS

In-line aortic bypass for treatment of aortic graft infections yields excellent results and has become our treatment of choice in dealing with this difficult problem.

Management and outcome of infrapopliteal arterial graft infections with distal graft involvement

PURPOSE

The purpose of this study was to determine the outcome of patients with infrapopliteal artery graft infections (InfraPopGIs) who presented with graft infection distal to the popliteal artery.

PATIENTS AND METHODS

Between July 1, 1979 and June 30, 1994, 27 patients presented with infrapopliteal artery graft infections (18 polytetrafluoroethylene [PTFE], 9 autologous vein). The infection involved the anastomosis in 22 cases (8 anterior tibial, 8 posterior tibial, 4 peroneal, 2 dorsalis pedis arteries) and was localized to the body of the graft in 5 cases (4 calf, 1 ankle). All bypasses were originally performed for limb salvage. Twelve patients with patent grafts and intact anastomoses were managed by complete graft preservation. Fifteen patients presented with occluded grafts (10), anastomotic hemorrhage (4), or systemic sepsis (1) and were treated by total or subtotal graft excision.

RESULTS

The hospital mortality rate was 19% (5 of 27) and the amputation rate in survivors was 27% (6 of 22). These results were compared with a mortality rate of 13% (15 of 114; P > 0.05) and a limb loss rate of 10% (10 of 99)(P = 0.05) in 114 patients during this period who presented with infection proximal to the tibial arteries. Of 6 survivors with graft infections who required amputations, 5 lacked a suitable outflow artery for a secondary bypass and 1 developed progressive gangrene despite a patent secondary bypass. Among the other 16 survivors, 7 (44%) limbs remained viable without requiring a secondary bypass, 6 (37%) limbs were salvaged with successful preservation of patent grafts, and 3 (19%) required secondary bypasses to prevent limb loss.

CONCLUSIONS

Patients presenting with infrapopliteal artery graft infections have higher amputation rates than patients with more proximal infected peripheral grafts. Selective graft preservation and selective revascularization when outflow arteries are available are essential adjuncts to minimize high rates of limb loss associated in patients with graft infections.

The thrombosed prosthetic graft is a risk for infection of an adjacent graft

BACKGROUND

A bland thrombosed graft may be more susceptible to the future risk of infection than a patent graft. Once infected, that graft can threaten other patent grafts. Therefore, the purpose of the following study was to assess the role a thrombosed graft might play in infection of contiguous patent bypasses.

METHODS

From 1990, a retrospective review was performed using the operative and medical records of cases in which a prosthetic graft infection was identified arising in association with an adjacent thrombosed graft.

RESULTS

A total of 22 cases of prosthetic arterial bypass infection were treated at our institution from January 1990 through September 1995. Of these, 7 (32%) were identified by the operative report as arising in a thrombosed prosthetic graft and spreading to an attached or adjacent patent prosthetic graft. All patients had multiple bypasses prior to infection, mean 5.4 +/- .75 (range 3 to 8). All thrombosed infected grafts were infrainguinal polytetrafluoroethylene (PTFE) for limb salvage: 6 femoralpopliteal and 1 femorotibial. Mean interval time between placement of the primarily infected graft and removal was 14.6 +/- 6.7 months (range 1 to 53). The secondarily infected patent bypasses were inflow procedures to the same limb in 6 cases: 1 aortofemoral, 2 ileofemoral, 2 axillofemoral, and 1 femoral femoral graft. The thrombosed infrainguinal bypass was directly attached to the secondarily infected bypass in 5 cases and near but not attached in 1 case. One secondarily infected prosthetic graft was a femoraldistal bypass placed adjacent to the thrombosed graft. Four patients had above-knee amputations with a clinically bland graft divided at the time of amputation. In these 4 patients and 2 additional cases, wet gangrene or infection was present in the distal extremity prior to the development of prosthetic graft infection. At the point that infection became clinically apparent, the thrombosed graft was removed in all cases and the secondarily infected graft was removed in 4 of 7 cases. Overall mortality was 57%.

CONCLUSIONS

A thrombosed prosthetic graft near a patent prosthetic bypass may become secondarily infected and threaten the patent graft. We recommend total removal of any thrombosed prosthetic graft in proximity to a patent prosthetic bypass when the risk of infection is high or at the time of subsequent amputation for gangrene.

Differences in early versus late extracavitary arterial graft infections

PURPOSE

The purpose of this report was to determine differences in presentation, bacteriology, management, and outcome of early (EGIs) versus late extracavitary arterial graft infections (LGIs).

METHODS

Between July 1, 1979, and June 30, 1994, we treated 141 patients with infected extracavitary arterial grafts (112 prosthetic, 29 vein) with selective partial or complete graft preservation.

RESULTS

A total of 99 (70%) EGIs (< 2 months) and 42 (30%) LGIs (4 to 96 months) were involved. The hospital mortality rate was 14% (20 of 141), and the amputation rate in survivors was 13% (16 of 121). No significant difference in mortality (16% [16 of 99] vs 10% (4 of 42]) or limb loss (16% [13 of 83] vs 8% [3 of 38]) was seen between EGIs and LGIs, respectively (p > 0.05). Patients with EGIs were as likely to have a disrupted anastomosis (17% [17 of 99] vs 21% [9 of 42]) or systemic sepsis (4% [4 of 99] vs 4% [2 of 42]) as patients with LGIs, respectively (p > 0.05). Patients with EGIs were more likely to have patent, intact grafts and to be treated by complete graft preservation (61% [61 of 99] vs 26% [11 of 42]) (p = 0.0001). In comparison, patients with LGIs were more likely to have occluded grafts and to require subtotal graft excision (48% [20 of 42] vs 18% [18 of 99]) (p = 0.0001). Surviving patients with EGIs treated by complete graft preservation were more likely to have successful healing of their wounds after long-term follow-up (average 3 years) than patients with LGIs (79% [41 of 52] vs 40% [4 of 10], respectively) (p = 0.03). The pathogens cultured from wounds of EGIs versus LGIs were pure gram-positive bacteria in 49 (49%) versus 19 (46%), pure gram-negatives in 18 (18%) versus 11 (26%), and both types in 33 (33%) versus 12 (28%) (p > 0.05).

CONCLUSION

Complete graft preservation can be attempted more frequently and is more likely to be successful in EGIs than in LGIs. No difference in bacteriology was seen between the two groups. Graft-preserving treatment can be successful but should only be cautiously attempted in patients with late extracavitary arterial graft infections.

Recommendations for initial antibiotic treatment of extracavitary arterial graft infections

BACKGROUND

Initial antibiotic treatment of extracavitary arterial graft infections is usually empiric or based on Gram's stain findings. Increasing virulence of bacteria causing extracavitary arterial graft infections may render previous choices of antibiotics obsolete. The purposes of this study were to correlate Gram's stain findings of gram-positive bacteria and gram-negative bacteria with wound cultures and provide a microbiologic basis for appropriate initial antibiotic therapy.

METHODS

Between July 1, 1979 and June 30, 1994, specimens obtained on the day of admission from purulent wounds involving 113 extracavitary arterial graft infections were retrospectively reviewed for Gram's stain and culture and sensitivity results.

RESULTS

Gram's stain findings correlated with final cultures on only 28 of 113 cases (25%), including 20 of 48 pure gram-positive, 2 of 24 pure gram-negative, and 6 of 41 mixed bacterial cultures. Staphylococcus aureus was the most common gram-positive bacteria cultured (43 isolates) and Pseudomonas species was the most common gram-negative bacteria (25 isolates). Bacteria were sensitive to a first-generation cephalosporin in only 32% (36 of 113) of infections. A combination of vancomycin and either ticarcillin-clavulanic acid or ceftazidime, which have minimal toxicity and provide excellent coverage against staphylococci, Pseudomonas, and other gram-negative bacteria, would have covered 96% (109) and 95% (107) of cultured organisms, respectively.

CONCLUSIONS

Regardless of Gram's stain findings, current recommendations for initial treatment of extracavitary arterial graft infections should include vancomycin and ceftazidime or ticarcillin-clavulanic acid until final culture and sensitivity results dictate the use of more selective antibiotics.