Perspective: options in managing aortic graft infections

Abstract

Thoracic aortic graft infections are infrequent, but are associated with high mortality and morbidity risk. To prevent these life-threatening infections, one must be aware of pathogenesis. When it occurs, a multidisciplinary (surgeon, radiologist, microbiologist, nuclear specialist, infectiologist, anesthesiologist, intensive care specialist) cascade must be initiated. A fast and accurate diagnosis using diagnostic criteria is vital. The appropriate treatment consists of a combination of antibiotics and surgery. Whether or not the vascular prosthesis is preserved depends on a case-by-case basis (tailor-made). Several graft preservation strategies have proven their effectiveness.

Introduction

Thoracic aortic graft infections (TAGI) are rare. The incidence of this severe complication following open surgery is approximately 3% [1], with a mortality of 25 to 75% [2–4]. Stent grafting in thoracic endovascular aortic repair (TEVAR) has a lower risk of infection, with an incidence of less than 1% and a mortality of 25% [5].

In our center (General Hospital Saint-John, Bruges, Belgium), we could only report one case of TAGI after thoraco-abdominal aorta replacement in a total of 299 patients (0.33%) who underwent open thoracic aortic replacement (root, ascending, arch, descending, thoraco-abdominal) between January 2010 and the end of June 2017.

Prevention of infection is of paramount importance. It is essential that the complete perioperative healthcare staff have full knowledge of the risks and do every effort to prevent this severe complication. However, when this infrequent but serious infection occurs, only a rapid and correct diagnosis followed by effective multidisciplinary treatment can be lifesaving.

Pathogenesis

Vascular graft infections can be divided into early (within 4 months postoperatively) and late (after 4 months postoperatively). Thoracic aortic surgery is clean surgery. Almost all cases of early aortic graft infections are caused by surgical site infections (SSI), by contamination and direct seeding of the graft during implantation. Resident flora of the skin, falling bacteria, and breaks in sterile technique are the main causes of SSI. Extension from a contiguous infected site or hematogenous seeding from a remote site is a rare cause of TAGI after open procedure.

In cases of stent graft infection, the number caused by surgical site infection (specifically groin infection) is much smaller. More possible sources seem to be any other infectious complication in the perioperative period (most common urinary tract infection) and placement of stent grafts in an infected area bed. One third of the source of the stent graft infections are apparently not procedure related. Any later occurring serious or chronic infection can be a spreading source. The mean time between the implantation and diagnosis of infection in stent graft is 18 months (0.2–158) [6].

Host-related factors also play an important role: diabetes mellitus, obesity, renal failure, liver disease, malnutrition, use of immuno-modulating medication [7]. Besides breaks in surgical technique, other procedure-related risk factors are grafts placed in the groin, prolonged operative time, emergency procedure, re-operative surgery, prolonged preoperative hospitalization, and inadequate perioperative antibiotic prophylaxis [8]. Pure wound complications as necrosis, long draining hematoma, or lymphocele increase the risk of developing a vascular graft infection.

In approximately 75% of all cases, gram-positive bacteria, especially Staphylococcus aureus, are the prevalent pathogens [9]. Staphylococcus aureus and coagulase-negative Staphylococci have the ability to produce an adherent biofilm on prosthetic surfaces. A biofilm is an extracellular matrix containing infective microbes. It is a hiding place for microorganisms against the host immune response and it prevents the penetration of antibiotics [10].

Methicillin-resistant Staphylococcus aureus (MRSA) and Pseudomonas aeruginosa have high virulence properties causing severe infections with higher rates of morbidity and mortality, compared to low virulence organisms such as coagulase-negative Staphylococcus, Corynebacterium, and Propionibacterium species. The former mentioned are usually associated with early onset graft infections, the latter with late onset infections [11].

Prevention

Open replacement of native aorta with a prosthetic vascular graft is considered as a highly invasive surgical procedure. In comparison to other types of surgery, closer attention needs to be paid to prevent graft infection.

Recently (May 2017), the Centers of Disease Control and Prevention published a new version of the Guidelines for the Prevention of Surgical Infection [12]. They provided a new and updated list of evidence-based (strong category) recommendations:

Before surgery, patients should shower or bathe (full body) with soap (antimicrobial or nonantimicrobial) or an antiseptic agent on at least the night before the operative day. Antimicrobial prophylaxis should be administered only when indicated based on published clinical practice guidelines and timed such that a bactericidal concentration of the agents is established in the serum and tissues when the incision is made. Skin preparation in the operating room should be performed using an alcohol-based agent unless contraindicated. For clean and clean-contaminated procedures, additional prophylactic antimicrobial agent doses should not be administered after the surgical incision is closed in the operating room, even in the presence of a drain. Topical antimicrobial agents should not be applied to the surgical incision. During surgery, glycemic control should be implemented using blood glucose target levels less than 200 mg/dL, and normothermia should be maintained in all patients. Increased fraction of inspired oxygen should be administered during surgery and after extubation in the immediate postoperative period for patients with normal pulmonary function undergoing general anesthesia with endotracheal intubation. Transfusion of blood products should not be withheld from surgical patients as a means to prevent SSI.

To reduce a rare but severe complication as aortic graft infection, also “weaker” and non-harmful recommendations must be implemented in daily practice. Although it is not evidence-based proven that demonstrated bacterial count reduction by measures as wearing high-level barrier masks, complete head coverings, operating room air treatment and ventilation, and reducing room traffic can reduce the prevalence of surgical site infections [13], we strongly believe that these details are essential. Avoiding contact between prosthetic graft and skin by using antiseptic (povidone-iodine) soaked wound protection coverage and plastic adhesive drapes with antimicrobial properties is not proven effective, but is standard in our practice. The same applies for gentle tissue handling with careful wound opening (without “tactic of the scorched earth”), covering of prosthesis with native aorta (if possible), and draining of all cavities and meticulous (with reduced incidence of hematoma and lymphocele formation) wound closure with antimicrobial-coated absorbable suture. Even eradication of preoperative nasal colonization with MRSA in prevention of SSI is controversial [14]. In elective cases, we always eradicate with harmless local mupirocin therapy, whereas in urgent cases, the combination antimicrobial prophylaxis with cefazolin and vancomycin is used. Preoperatively, we also search for (cone beam computed tomography (cone beam CT) scanning) and cure paranasal sinus infections, dental, and periodontal disease. In sum, every curable infective disease in the preoperative setting is a contraindication for elective aortic surgery.

Several study groups promote the use of coated vascular grafts in the prevention of infection: silver-coated vascular prostheses [15], rifampicin-bonded gelatin-sealed grafts [16].

Late onset prosthetic graft infections can occur 20 years or even more postoperatively. These are unlikely caused by SSI, but tend to have parallel causes of infective endocarditis. For that reason, every aortic graft patient must receive antibiotic prophylaxis when undergoing invasive diagnostic or therapeutic procedure. After placement of aortic grafts, future infections should be addressed radically.

Diagnosis and diagnostic criteria

An accurate diagnosis is important in acute graft infections (AGI). A false-positive diagnosis can lead to unnecessary surgery. Failure to diagnose is associated with high-risk morbidity.

Until recently (2016), well-defined diagnostic criteria for the diagnosis of AGI were lacking, in contrast to the longer existing evidence-based guidelines for managing infections of other implanted prostheses (heart valves, joints).

It was the Management of Aortic Graft Infection Collaboration (MAGIC) that gave us a useful definition of AGI in routine clinical practice [17]. They specified and gave “weight” to a combination of clinical (surgical), radiological, and laboratory findings. These three diagnostic categories (clinical/surgical, radiological, laboratory) were within each category ranked as “major” or “minor.”

The summary of the results is shown in Fig. 1 [17].

Fig. 1.

Summary of results

AGI can be defined as follows:

• Suspicion of AGI in a patient: any isolated major criterion or minor criteria from two of the three categories

• Diagnosis of AGI: a single major criterion, plus any other minor or major criterion from another category

Clinical/surgical criteria

The specificity of clinical manifestations divides the criteria into “minor” or “major.” Localized clinical features can be caused by postoperative wound inflammation or superficial wound infection without AGI. Fever ≥ 38 °C lacks specificity and is therefore withheld as a minor criterion, only in the clinical absence of another focus of infection.

Other systemic manifestations as anorexia, malaise, and weight loss are common in patients with AGI. They lack specificity to include as a criterion. Ten percent of infected TEVAR stent grafts are asymptomatic [6].

Major clinical/surgical criteria are more specific, with direct involvement of the aortic graft. Pus (demonstrated pus cells by direct microscopy) around graft or in an aneurysm sac at surgery is a strong criterion. Also the circumstances in which there is direct communication between the aortic graft and a nonsterile site are “major” arguments for AGI. Almost half of the reported infected stent grafts had a fistula, most common aorto-esophageal. [6].

Radiological criteria

CT is considered as the standard imaging modality. It can provide the following findings: presence of perigraft fluid (with density measurement), presence of perigraft gas, loss of normal tissue planes in the perigraft and mediastinal structures with an increased amount of soft tissue between the graft and the surrounding sac and pseudoaneurysm formation. Gas around the vascular graft and fluid (hematoma) around the aorta are respectively absorbed within 1 and 7 weeks after surgery. The presence of gas after 4–7 weeks indicates AGI [18]. Persistence of perigraft fluid after 3 months is highly suspicious [19]. In this situation, CT-guided aspiration of perigraft fluid for microbiological analysis is recommended [20]. In this stage of diagnosis, the placement of draining catheters should be avoided because of the risk of introducing infection. Signs of secondary infection involvement of adjacent structures by contiguous spread need the presence of another major criterion. The formation of a pseudoaneurysm is a well-known feature of AGI, but it can also be a consequence of a non-infective dehiscence of a suture line. A late endoleak in TEVAR cases can possibly be counted as a minor criterion [6].

Magnetic resonance imaging (MRI) has so far not been as extensively studied as CT in diagnosis of AGI. It provides better soft tissue resolution that may help to distinguish inflammatory fluid from hematoma.

Fluorodeoxyglucose positron emission tomography (FDG-PET)/CT shows promise in the diagnosis of AGI. Multiple reports claim that PET/CT is a reliable technique as represented by 93% sensitivity, 91% specificity, 91% positive predictive value, and 96% negative predictive value [21]. The interpretation is essential for the accuracy. Synthetic graft material can provoke chronic aseptic low-grade inflammation for several years after surgery with associated diffuse tracer uptake. The knowledge of the specific pattern uptake (focal, irregular) correlated with infection versus diffuse (linear, circular) correlated with aseptic inflammation, together with the findings on the CT component, is necessary to obtain an accurate diagnosis [22]. This technique has also potential for monitoring response to treatment.

To detect late onset AGI 99mTc-hexamethylpropyleneamineoxime (99mTc-HMPAO)-leukocyte, single-photon emission computed tomography could be more useful than conventional radiology [23].

Laboratory criteria

Identification of the causative pathogen is highly important. However, in a third of the patients with AGI, no positive tissue or blood cultures are obtained. The main reason is prior empiric antibiotic use. Positive microbiology from intraoperative specimens (perigraft fluid, pus, tissue biopsies) and prosthetic graft (surface) is a major diagnostic criterion. The same applies to the positive culture of microorganisms from radiological-guided, percutaneous aspirate of perigraft pus/fluid. An aseptic puncture technique minimizing contamination is necessary. When the culture by standard microbiology is negative, other high-sensitive molecular techniques like broad range bacterial polymerase chain reaction (BRB PCR) to identify microbial DNA and matrix-assisted laser desorption/ionization time of flight (MALDI-TOF) when available have diagnostic value [24]. There are no studies on sonication of vascular grafts. Sonication is a culturing method in which sound waves are used to release microorganisms from their enveloping biofilm. In orthopedic prosthetic infections, culturing of this sonication fluid has been shown to be equally specific but more sensitive compared to conventional culturing techniques [25].

The identification of high-virulence pathogens is always regarded as clinical significant. Contrarily, the interpretation of isolation of possible contaminant, low-virulence bacteria (Staphylococcus epidermidis, Propionibacterium species, Corynebacterium), is more difficult. In these cases, the MAGIC consensus proposes that exactly the same microorganism must be present in at least two samples (blood cultures and/or intraoperative specimen) to have a major criterion.

Repeated positive blood cultures with the same microorganism are a classic feature of endovascular infection. Half the cases of AGI have negative blood cultures. Positive blood cultures give no direct site specification. Therefore, a positive blood culture is withheld as a minor criterion only in the absence of an apparent infective source. Anyway, the clinician must be suspicious when any bacteremia, irrespective of origin, occurs within the first postoperative month. The reason is the higher likelihood of hematogenous seeding to the prosthetic surface before the presumed endothelial covering.

Inflammatory blood markers (e.g., C-reactive protein, fibrinogen, white blood cells, lymphocyte CD4/CD8, tumor necrosis factor-alpha, procalcitonin) are classified as minor criteria. They lack specificity and are mostly elevated in the immediate postoperative period. But the diagnosis of AGI is highly unlikely with normal inflammatory blood markers. Presepsin, a new biomarker, is more specific for infection and sepsis, with low values in patients with systemic inflammatory response or invasive trauma without infection [26].

Treatment

The appropriate treatment of the patient with an AGI is the combination of medical and surgical intervention. Nutritional management and strict blood glucose control in diabetic patients must also be applied to improve immune functions.

Antibiotic treatment

Before starting antibiotic therapy, great effort should be made to collect as many as possible specimens for microbiologic laboratory (bacterial, fungal, and mycobacterial stains and cultures): blood, wound, ultrasonic- or radiologic-guided aspiration of perigraft fluids, intraoperative perigraft fluid, and tissue. Additionally, serological testing can be useful. Empiric broad-spectrum parenteral antibiotics, which include coverage of virulent resistant gram-negative and gram-positive organisms including MRSA, are initiated, such as vancomycin and piperacillin/tazobactam. Antibiotics are tailored after the identification of the causative pathogen. Close collaboration with the microbiologist and the infectiologist is recommended. There are no randomized data of the total duration of antimicrobial therapy. It depends on clinical response and type of surgical therapy. Most experts agree that a minimum of 4–6 weeks of parenteral antibiotics is necessary, counted from the moment of the surgical debridement of the infected prosthesis and the surrounding tissues. If cultures are negative, it is recommended that broad-spectrum antibiotics are continued for at least 4 weeks. When infection control is achieved after 4 weeks, the therapy is generally switched to oral antibiotics. The duration of this oral therapy is controversial. Some of these patients receive antibiotics during several years or even lifelong, particularly in case of partial or complete graft preservation strategy.

Surgical treatment

The poor results of non-operative management alone in AGI justify the use of aggressive surgical treatment [27]. There are no evidence-based practice guidelines to address this problem, but there is expert agreement to individualize treatment in different clinical situations.

In ideal treatment of infection of a vascular prosthesis, surgical treatment should consist of extensive removal of infected and necrotic tissues, debridement, and complete removal of infected graft followed by vascular graft reconstruction. In non-thoracic AGI, it can be performed by an extra-anatomic bypass. Anatomic limitations preclude the use of this surgical modality in thoracic AGI.

There are several outcome reports of thoracic aortic graft replacement with prosthetic grafts and homografts after complete infected graft removal, with omental filling procedures after re-thoracotomy and debridement [28]. The mortality rates were high, especially in cases of more extent and complex reconstruction, like the aortic root. Despite these poor results, more aggressive treatment remains probably the first choice in case of severe sepsis, anastomotic disruption, aortic fistula, and involvement of virulent organisms such as Pseudomonas or MRSA.

Arguments in favor of preservation or partial excision of the (partial infected) graft were published. In one step, four main procedures were performed: (1) chest wall, mediastinal, and perigraft tissue debridement; (2) intraoperative mediastinal antibiotic irrigation; (3) locoregional tissue transposition to cover the graft and to fill dead space, preferable an omental flap; (4) sternal fixation and complete wound closure [29]. Also, a two-step approach with repeated cleaning and packing of the mediastinum with 10% povidone-iodine every 8 h during 48 h, prior to omental filling and chest closure, gave favorable results [30]. Not only omentum, but also muscular flaps (latissimus dorsi, rectus abdominis, pectoralis major) are used for wrapping and filling. When infection reaches aorta-graft suture lines with fragile anastomotic tissues, (partial) removal and replacement is needed to prevent anastomotic rupture or pseudo-aneurysm formation.

The usefulness of vacuum-assisted closure (VAC) therapy in refractory postoperative refractory mediastinitis has been demonstrated as a less invasive treatment, also after prosthetic vascular prosthesis in the absence of pseudoaneurysm or intra-graft vegetation [31]. It can be installed after the debridement and is changed two or three times a week. Even direct contact with a prosthetic graft seems to be safe when using a two-sponge technique (non-adherent polyvinyl alcohol foam sponge with very small pores covers the anastomosis) [32]. After sterilization (two consecutive negative cultures), omental or muscular filling and wound closure follow.

To lower the risk of reinfection of (partial) replaced aortic graft after (partial) removal of prosthetic graft, a variety of conduits are used: homografts, xenografts, xeno-pericardial grafts, prosthetic grafts impregnated with silver or antibiotics such as rifampicin [33].

To treat or to prevent AGI, companies have developed prostheses with a drug delivery system and with special coatings (liposomal encapsulated antibiotics, polycationic peptides, quinupristin-dalfopristin). In addition, bio-soluble beads impregnated with vancomycin and tobramycin placed around an infected graft were used. A wide range of bactericidal coatings such as graphene [34] are still being studied.

There are limited data regarding outcomes of the less occurring stent graft infections. These cases need, besides parenteral antibiotics, similar aggressive surgical treatment. Medical therapy is insufficient because of the high rate of fistulization. Explantation of the infected endograft is the best approach for TEVAR cases, even if the mortality and reinfection ratio is high. The profile of these patients is different, some of them were not suitable for an open procedure due to increased comorbidity. Endograft explantation requires thoracotomy, more proximal and distal aortic clamping to ensure adequate landing zones, frequently cardiopulmonary bypass (complete or left-left, with or without deep hypothermic circulatory arrest and selective anterograde cerebral perfusion) and coincident esophageal or airway repair (with intercostal or omental flap) [35].

Conclusions

The use of synthetic aortic grafts and stent grafts is inevitable in treatment of aortic pathology. Despite precautionary actions, these rare mortal infections occur. The importance of prevention cannot be emphasized enough. The recently published MAGIC consensus about the diagnostic criteria is a big step forward to achieve an accurate diagnosis. Only a multidisciplinary treatment can assure a better prognosis. Surgical therapy (in situ replacement or graft-preserving technique) must be individualized according to the severity, the extent, the location of infection, and the general condition of the patient.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical approval

In the paper there were no results of new (self-done) research involving human participants and/or animals published. So there is no need for informed consent.