When investigating any case of suspected VAD infection, prompt investigation is required, and testing as outlined below should be pursued (◘ Table 49.2) [2]. VAD-specific infections include infections that are specific to patients with VADs, are related to the device hardware, and do not occur in non-VAD patients, e.g., pump and cannula infections, pocket infections, and percutaneous driveline infections [3–6]. VAD-related infections refer to those that can also occur in patients who do not have VADs; however, in patients with VADs, there may be unique considerations with respect to making the correct diagnosis or determining the cause-effect relationship, e.g., mediastinitis and IE. Non-VAD infections are essentially not affected by the presence of the VAD and are unlikely related to the VAD presence but are included to encourage comprehensive and comparable data recording of all infections in this patient population to facilitate multicenter review.

VAD-specific infections may be of the hardware itself or the body surfaces that contain them and include pump, cannula, and anastomotic infections, pocket infections, and percutaneous driveline or tunnel infections. Accurate VAD-specific infections required new definitions to be constructed to reflect the specifics of such infection to enable study of the potential sources or risk factors for these infections. Guidelines on the diagnosis of prosthetic joint infection (PJI) [7], IE [8], cardiovascular device infections [5, 9], intra-abdominal infections [10], catheter-related bloodstream (CRBS) [11], and skin and soft tissue infections [12] have provided the basis on which the definitions were constructed. These infections share many features of VAD-specific infections as they are often difficult to diagnose conclusively and difficult to treat due to the presence of biofilms on prosthetic surfaces that markedly reduce the likelihood of successful treatment with anti-infectives alone.

The first group of VAD-specific infections is hardware related (e.g., pump and/or cannula infections); see ◘ Fig. 49.1. The “pump” or “VAD” refers to that part of the device that is involved in the propulsion of blood and includes both continuous-flow (cf) and/or pulsatile (intracorporeal and paracorporeal)-flow devices, though the majority of implants today are CF pumps, >90% [1]. The term “inflow cannula” refers to that part of the device connecting the ventricle to the pump device. The term “outflow cannula” refers to that part of the device connecting the pump device to the patient’s cardiovascular system. Suture lines refer to the surgical anastomoses between pump and patients’ cardiovascular system. These generic terms have been chosen to allow as many VAD devices (including LVADs and RVADs) as possible to be incorporated into this definition framework.

A patient must have at least one of the microbiological, histopathological, radiological, or clinical criteria to achieve a firm diagnosis as outlined in ◘ Tables 49.3 and 49.4. The retrieval of a pathogen or an indistinguishable organism from more than one site is critical for validating the microbiological criterion.

The term “pocket” in these definitions is used to describe infections that occur in the body space or pocket that holds the pump inside the body of the patient. Classically the pocket may be newly created within the abdominal wall or within proximity to the pericardium and the diaphragm. The most recent devices use natural body cavities and are placed entirely within the left ventricle or within the pericardial sack (◘ Tables 49.5 and 49.6). Pocket infections in those devices still requiring a surgical pocket may be diagnosed without removing the VAD at the time of surgery if samples from the inner surface of the pocket and the exterior surface of the VAD are taken (◘ Table 49.2, ◘ Fig. 49.1). Cardiothoracic surgeons, cardiologists, and, in specialized centers, interventional radiologists working closely with microbiology teams may be able to aspirate diagnostic fluid surrounding devices by imaging guidance.

Percutaneous driveline infections (PDIs) are important but challenging to define. Objectively they lie between existing standards for tunneled central lines and implantable intraports [5, 9, 11]. It is difficult to strike a balance between fully comprehensive definitions and definitions that are practical and useful for clinicians. Consequently, PDI definitions have been adapted from CDC/NHSN surveillance definitions of healthcare-associated infection and classified as superficial or deep, based on the depth of the infection [13], and subclassified into proven, probable, and possible superficial and deep infections. Each subclassification is described under the following four categories, general appearance, microbiology, surgical/histology, or clinical criteria (◘ Table 49.7), and should allow for detailed analysis of the etiology and risk factors for both superficial and deep PDI. This is considered the most useful way to define driveline infections as management of driveline infections typically depends upon the depth of the infection, which likely correlates with the source of the infection (◘ Table 49.7) [12]. All percutaneous drivelines should be surgically examined at time of removal or replacement, and where infection involves both superficial and deep incision will be classified as a deep infection. PDI are the most commonly occurring infections in VAD patients and may reflect the presence of a deeper infection of the pocket space or pump and/or cannula. This infection may be the result of local trauma at the exit site during device implantation, which may act as a cutaneous source of infection at a later date [14]. Sonography and CT angiography can reveal cuffs of fluid around the drivelines, cannula, and pump. Indium-labeled WBC scanning may also be helpful but, as yet, has not been validated for diagnosing these infections. The intraoperative exploration of the percutaneous driveline exit site (PDL-ES) at explantation or revision is required to satisfy these definitions, making them more useful for epidemiologic study than for clinical diagnosis. More recently different types of clinical grading systems as illustrated in ◘ Table 49.9 are used in combination with this ISHLT classification to aid diagnosis and management of infection.

VAD-related infections include IE, bloodstream infections (BSIs), mediastinitis, and sternal wound infection and are outlined in ◘ Table 49.8. Imaging has a particular role in revealing new inflammatory change in the mediastinum, and newer cardiac CT can show large valve vegetations and cannula insertion infections. It has been reported that CT may have a role in sternal wound infection characterization though we would mainly use it today to define the extent of deep-seated infection or collection and occasionally to guide tissue sampling by core biopsy for culture if swabs have not yielded a specific diagnosis [15, 16].

Diagnosing VAD-related BSI in the presence of a CVC may be particularly difficult. The technique of “differential time to positivity” (DTP) as a method of determining which infections are due to the VAD and which are due to the CVC is recommended, consistent with recent IDSA guidelines [11]. This method uses a 2 h time to positivity differential to determine the source of infection when a CVC is present. This method, though not 100% accurate, may implicate the CVC as the source of the bacteremia if the CVC blood cultures become positive >2 h before peripheral blood cultures become positive. Efforts can be made to avoid secondary seeding of the infection to the VAD by prompt removal of the CVC. If the DTP of CVC and peripheral blood cultures are less than 2 h, it is possible the VAD is the source of infection. Other causes such as non-VAD infection should also be considered, as the VAD may not always be the source of the BSI [17 ]. Mycotic aneurysms have also been reported in patients using VADs and associated with persistent or relapsing BSI [18]. Mycotic aneurysms may be visceral or intracerebral (usually presenting as an intracranial hemorrhage).

Non-VAD infections are essentially “independent” or not directly related to the presence of the VAD but infections occurring in a “cardiac sick” population of immunocompromised hosts with underlying comorbidities such as diabetes, prolonged hospitalization, multiple drug regimens, and renal impairment. The purpose of including non-VAD infection is to provide a comprehensive overview of all infections in this “cardiac sick” population (◘ Table 49.9).

Patients with VAD infections may present in a variety of ways making definitive diagnosis difficult. Patients often present with non-specific symptoms such as lethargy, fatigue, fever, or anorexia as well as a wide spectrum of ailments ranging from minor erythema at the PDL-ES to severe sepsis and clinical shock. All clinicians must be alert to the possibility of infection in VAD patients and should be educated regarding the clinical symptoms and signs, which ensure early detection and guide the most efficient diagnostic algorithm.

The initial evaluation should include a careful history and review of symptoms. Physical examination, review of the VAD function, surgical wounds, and PDL-ES are essential as early detection and treatment of a localized process may prevent progression to more serious VAD infections. It can also help to direct the clinician to non-VAD infections that may be present such as a UTI or CDI. Laboratory studies including a full blood count [19–21], serial erythrocyte sedimentation rate (ESR) [19], or C-reactive protein (CRP) are recommended in all patients [19–21, 23]. If there is pus visible at the PDL-ES, then an aspirate of this pus should be sent for bacterial and fungal cultures. Routine surveillance cultures of exit sites may be considered as colonization often precedes infection and can serve as valuable information for subsequent infection. Initial imaging should include a standard chest radiograph; an echocardiogram will be needed if there is suspicion of native valve infective endocarditis (IE) or concomitant cardiac implantable electrical device infective endocarditis (CDIE). At least three sets [24] of blood cultures should be obtained and at least two sets taken from a peripheral site at times consistent with modified Duke’s criteria [8] (outlined in ◘ Table 49.3) before commencing anti-infectives and where a CVC is present one set from the line concurrent with one of peripheral blood cultures [11]. Difficulty in obtaining blood samples from children and concerns about drawing large volumes may result in lower volumes of blood being submitted for culture and may reduce the negative predictive value of the culture [11]. When clinical, laboratory, and microbiology culture data point to a particular VAD or non-VAD infection, imaging can play a role in supporting such suspicions or directing tissue samples. Further when infection source eludes standard evaluation, imaging can have a role in primary diagnosis.

Sonography is a useful tool to visualize fluid around percutaneous and tunneled drivelines and in pump pockets and can be used to direct tissue samples or lavage. HIDA scanning was the gold standard for diagnosing cystic duct obstruction and remains a first-line test in many centers. Due to ease of access, widespread practice and rapid diagnosis and ultrasound have become the first-line study for infections such as cholecystitis [10]. Various CT protocols beyond the scope of this chapter can be used to evaluate the lungs, pleura, and mediastinum as well as other organ structures. They are clearly of value in investigating suspected infection in a patient using a VAD [15, 16, 25]. Since MRI is largely precluded, CT and digital subtraction angiography are the tests of choice for mycotic pseudoaneurysm assessment and treatment. The ability of modern scanners to provide whole-body assessment is very helpful. We have also found it of value to assess cannulae, thrombi, and vegetations. Sternal wound infections have been assessed by CT, or bone/indium-labeled leukocyte [26] scan with specific protocols may have a role in characterization of infection but in the surgically “damaged” sternum may have limited value [15].

In selected patients, the VAD may need to be removed due to uncontrolled infection or for technical reasons. When this happens, the VAD should always be sent to the laboratory for processing. Sterile aspirates or sterile syringe aspirates (from surgery) should be taken for Gram stain, KOH and Calcofluor white stain, bacterial and fungal cultures, and broad-range PCR at the time of explantation from the internal and external aspect of the inflow cannula and from the internal and external aspect of the outflow cannula when a VAD is removed. A small volume of sterile water (<5 ml) should be instilled into the explanted VAD and then aspirated and sent for bacterial and fungal culture. Defining the optimal method of culture of VADs is beyond the scope of these guidelines; however in the future, it would be beneficial to devise a standardized culture process for VADs so that the microbiology laboratory practice can be standardized across all centers (e.g., VAD sonication or even scraping of the biofilm) [26]. In particular, the use of broth cultures for the retrieval of organisms (currently used for explanted heart valves) including broad-range PCR should be considered where possible. Cardiothoracic surgeons, cardiologists, and, in specialized centers, interventional radiologists working closely with microbiology teams may be able to aspirate diagnostic fluid surrounding devices by imaging guidance [10]. Risk of introducing infection into a sterile fluid collection using this technique should be considered and performance of such procedures must have direct oversight for specimen handling by those involved in the infection management.

Any purulence present in the pocket area should also be sent for Gram stain, KOH, and Calcofluor white stain, bacterial and fungal culture, broad-range PCR, and a further two swabs processed in the same way taken from the external surface of the VAD, anterior and posterior. Finally at least two samples of tissue from the pocket area and insertion site of the cannulas into the heart should be sent for histology and tissue stains for bacteria and for microbiology, Gram stain, KOH and Calcofluor white stain, bacterial and fungal cultures, and broad-range PCR.

It may also be necessary to send additional samples to the microbiology laboratory if non-VAD infections are suspected (e.g., urine, stool for C. difficile toxins A and B, sputum, and wound swabs). The investigation of suspected VAD infections should be done in consultation with an infectious disease physician or clinical microbiologist, a cardiologist, and a cardiothoracic surgeon to optimize both the diagnosis and management of the potential infection. The anti-infective regimen must be carefully chosen since prolonged, even lifelong, therapy may be required (◘ Table 49.10).