Ventricular Assist Device-Specific Infections

Abstract

Ventricular assist device (VAD)-specific infections, in particular, driveline infections, are a concerning complication of VAD implantation that often results in significant morbidity and even mortality. The presence of a percutaneous driveline at the skin exit-site and in the subcutaneous tunnel allows biofilm formation and migration by many bacterial and fungal pathogens. Biofilm formation is an important microbial strategy, providing a shield against antimicrobial treatment and human immune responses; biofilm migration facilitates the extension of infection to deeper tissues such as the pump pocket and the bloodstream. Despite the introduction of multiple preventative strategies, driveline infections still occur with a high prevalence of ~10-20% per year and their treatment outcomes are frequently unsatisfactory. Clinical diagnosis, prevention and management of driveline infections are being targeted to specific microbial pathogens grown as biofilms at the driveline exit-site or in the driveline tunnel. The purpose of this review is to improve the understanding of VAD-specific infections, from basic "bench" knowledge to clinical "bedside" experience, with a specific focus on the role of biofilms in driveline infections.

Keywords: biofilms; driveline infections; driveline tunnel; epidemiology; exit-site; prevention; treatment; ventricular assist device.

Figure 1

Ventricular assist device (VAD)-specific and VAD-related infections. Microorganisms cause infections (text in red) at different anatomic sites where VAD components are placed (text in black) via three routes (text in blue). Microorganisms form biofilms on the driveline at the skin exit-site, causing superficial driveline infections. Microbial dispersal from established biofilms often results in bloodstream infections. Migration of biofilms along the driveline tissue tunnel leads to ascending tunnel infection or pump pocket infection (route 1). Intraoperative microbial contamination of VAD components (route 2), or hematogenous seeding from other infection sites (route 3) can also lead to pump pocket infection, pump/cannula infections, or other VAD-related infections.

Figure 2

Driveline exit-site infection. (A) Uninfected control. The patient received a HeartMate II VAD system (Abbott, Plymouth, MN, USA). (B) A typical driveline exit-site infection that demonstrates tissue destruction and granulation tissue. A HeartWare HVAD system (Medtronic, Minneapolis, MN, USA) was used in the patient.

Figure 3

In vitro biofilm formation on different components of a VAD driveline [9]. (A) Biofilms formed by S. aureus ATCC 25923 on the smooth section of the driveline; (B) Biofilm formation of P. aeruginosa PAO1 on the smooth section of the driveline; (C) Three-dimensional structure of the driveline velour; (D) Biofilm formation of P. aeruginosa PAO1 on the velour section. Drip flow biofilm reactor assay was used to mimic the clinical environment where an implanted driveline and invading microorganisms might encounter. Reprinted from Reference [9], Copyright (2020), with permission from Elsevier.

Figure 4

Clinical biofilms observed on an infected VAD driveline [8]. A driveline was explanted from a VAD patient with clinically diagnosed S. aureus driveline infection. The driveline was sectioned into small pieces, washed to remove planktonic cells, structurally fixed with glutaraldehyde, and imaged with scanning electron microscopy. (A) S. aureus monolayer biofilms formed on the smooth section of the driveline around the exit-site; (B) Microcolony-alike biofilms formed at the tissue-driveline interface (upper panels) or on the velour fibers (lower panels). Reprinted from Reference [8], Copyright (2020), with permission from Elsevier.

Figure 5

The presence of microgaps in the driveline tissue tunnel. (A) Micro-CT image of an explanted driveline shows numerous microgaps in the sectioned velour; (B) Scanning electron microscopy (SEM) of the velour (top-bottom view) shows microgaps [9]; (C) SEM of the explanted driveline (cross-section view) also shows numerous microgaps within the velour, suggesting insufficient tissue integration [9]. Reprinted from Reference [9], Copyright (2020), with permission from Elsevier.