The Effect of Negative Pressure Wound Therapy With Antiseptic Instillation on Biofilm Formation in a Porcine Model of Infected Spinal Instrumentation

Objective. This study evaluates the effect of negative pressure wound therapy with antiseptic instillation (NPWTi) in the clearance of infection and biofilm formation in an in vivo model of infected spinal implants compared to traditional treatment modalities. Materials and Methods. Five pigs underwent titanium rod implantation of their spinous processes followed by injection of 1 x 10⁶ CFUs/100µL of methicillin-resistant Staphylococcus aureus through the fascia at each site. At 1 week postoperatively, an experimental arm of 3 pigs received NPWTi, and a control arm of 2 pigs received wet-to-dry dressings. The persistence of local infection in the experimental group was compared to the control group using tissue cultures. Biofilm development on spinal implants was evaluated using scanning electron microscopy. Results. Mean bacterial count showed a statistical difference between the experimental and the control groups (P < .05). Scanning electron microscopy revealed the presence of uniform biofilm formation across the surface of control group instrumentation, whereas the experimental group showed interrupted areas between biofilm formations. Conclusion. The authors concluded that NPWTi is associated with decreased bacterial load and biofilm formation compared to wet-to-dry dressings in an in vivo porcine model of infected spinal instrumentation.

Postoperative infection following placement of spinal instrumentation remains a troublesome complication with potentially devastating consequences.^1-3 Modern treatment paradigms attempt to retain spinal instrumentation whenever possible, however delayed infections or the presence of biofilm-producing pathogens may force physicians to confront a difficult decision: attempt to salvage the infected instrumentation or remove it.^2,4-10 The morbidity resulting from explantation must be weighed against alternate strategies to address infectious complications.

The increasingly recognized role of biofilm formation, especially in implant-associated infections, inspires a search for novel strategies aimed at prevention, disruption, or elimination of biofilms.^1,2,6,11-17 Biofilms form a protected bacterial microenvironment resistant to the innate and humoral immune responses as well as antibiotic penetration. Furthermore, biofilms may breed antibiotic resistance.^6,11 

Negative pressure wound therapy with antiseptic instillation (NPWTi) holds promise as a novel strategy to address biofilm formation in infected wounds. A study evaluating biofilm formation in an ex vivo model of porcine skin found NPWTi (with various antiseptic instillation regimens) significantly reduced bacterial colony forming units (CFUs) compared with NPWT with saline instillation and untreated samples.¹⁸ In this study, instillation treatments included 1% povidone-iodine, 10% povidone-iodine, .05% chlorhexidine gluconate, 0.1% polyhexamethylene biguanide, and 0.2% polydiallyldimethylammonium chloride. Furthermore, this study correlated decreased bacterial load with damaged bacterial cell membranes and disrupted biofilm exopolymeric matrix.¹⁸ In a pilot study¹⁹ measuring clinical outcomes, NPWT with silver nitrate instillation was found to treat complex infected wounds more effectively than standard wet-to-dry dressings. Currently, there is a paucity of in vivo studies assessing the effect of NPWTi on biofilms, and reproducibility of clinical outcomes data are lacking from the literature.

Negative pressure wound therapy utilized following orthopedic procedures has demonstrated decreased infection as well as overall rates of wound complications.^20-23 In addition, NPWT has proven beneficial in the treatment of postoperative spinal infections, though the literature regarding use with infected spinal instrumentation has been inconsistent.^6,24-28 Clinical studies evaluating the use of NPWTi in patients with infected hardware have shown inconsistent clinical outcomes. Although these clinical studies often involve infections with pathogens known to form biofilms, to the best of the investigators’ knowledge there are no studies evaluating the effect of NPWTi on biofilm formation in a controlled, in vivo manner both quantitatively (measuring CFUs) and qualitatively (utilizing scanning electron microscopy [SEM]). 

The purpose of the present study was to evaluate the effect of NPWTi compared with wet-to-dry dressings for the clearance of infection and biofilm formation in an in vivo model of infected spinal instrumentation. The investigators hypothesized that NPWTi would reduce bacterial load and prevent or disrupt biofilm formation. They aimed to demonstrate this by quantifying wound infection with CFU counts and assessing biofilm formation with SEM.

The University of Maryland Institutional Animal Care and Use Committee (IACUC) approved this study. Following approval from IACUC, 5 Yorkshire pigs weighing 25 kg each were obtained for enrollment in a controlled trial. Pigs were chosen due to well-established similarities between human and porcine skin anatomies, wound healing physiologies, and previous experiments defining the porcine spine as a model for human spinal surgery.^29,30 In addition, a porcine model was successfully developed to study NPWT in previously conducted experiments at the authors’ institution.

Pigs were alternatively assigned to the experimental or control arm: 3 pigs to the experimental arm and 2 to the control. Each pig first underwent an acclimation period of 7 to 10 days to an animal jacket designed to secure dressings in place before the initial implantation surgery. In the initial procedure, titanium rods were surgically implanted into thoracic and lumbar spinous processes of each pig through separate incisions to simulate isolated spinal instrumentation sites (Figure 1). Approximately 4 cm of skin was left between sites to adequately support a NPWT dressing and prevent cross-contamination. After instrumentation implantation, the overlying fascia and skin were closed. Just prior to skin closure, 1 x 10⁶ CFUs/100µL of methicillin-resistant Staphylococcus aureus (MRSA) was injected through the fascia at each site. This inoculum was based on previous porcine studies in which dermal and subdermal infections were created as well as studies demonstrating biofilm formation on implanted foreign bodies.^31-35 Closed incisions were dressed with tape, gauze dressings, and a circumferential wrap of self-adherent elastic dressing (Coban Self-Adherent Wrap; 3M Health Care, St. Paul, MN) (Figure 2). On postoperative day 2, the dressings were removed and then monitored daily. A sling was used to restrain animals in a humane manner while the wounds were examined. Postoperatively, animals were monitored for signs of wellbeing including body weight, general physical appearance, vital signs, unprovoked behavior, hyperactivity, hypoactivity, and response to external stimuli. 

Seven days following the initial surgery, all 5 pigs underwent a second surgery in which the incisions were reopened (Figure 3). The experimental arm received NPWT (VAC Ulta; KCI, An Acelity Company, San Antonio, TX) dressings with a wound irrigation solution (Prontosan; B. Braun Medical Inc, Bethlehem, PA) instillation at each incision site. The control arm received traditional wet-to-dry dressings using the wound irrigation solution. This wound irrigant, containing a combination of polyhexanide and betaine, prevents the growth of bacteria by acting as a disinfectant and detergent to help clean the wound. The components of this irrigation have shown improved wound healing times in immature pigs.³⁶ 

All animals were returned to housing designated for swine at the University of Maryland School of Medicine (Baltimore, MD) institutional animal facility. Animals in the experimental group were attached to the NPWT dressing system held in place by a tether apparatus connected to an animal jacket. This apparatus permitted the animal normal range of motion while allowing proper function of the NPWTi dressings (Figure 4). Pigs in the control group received dressing changes twice daily using anesthesia support for sedation as needed. Pigs with NPWTi dressings received twice-daily checks to ensure adequate dressing seal. All animals were euthanized with Telazol (2 mg/kg–8 mg/kg; Zoetis, Parsippany-Troy Hills, NJ) and xylazine (2.2 mg/kg–8.8 mg/kg) intramuscular followed by ≥ 100 mg/kg pentobarbital intravenous overdose 7 days after the second surgery. Following euthanasia, tissue, blood, and instrumentation samples were obtained for culture and additional analysis. 

The persistence of local infection in the experimental group was compared with the control group using tissue cultures. Blood cultures were drawn in all animals to evaluate for hematogenous spread of bacteria. Biofilm development on spinal instrumentation was evaluated using SEM. 

Of the 5 pigs, 3 healthy pigs were assigned to the experimental arm and 2 healthy pigs were assigned to the control arm. Each pig received 6 surgical sites resulting in 18 separate experimental sites and 12 separate control sites. This met the results of an a priori power analysis at a power of 0.8 requiring 7 surgical sites in the experimental group and 10 in the control to meet statistical significance (P = .05). Statistical analysis of culture data was performed using a two-tailed t-test.

All animals remained afebrile throughout the duration of the study, and all blood cultures were negative at the time of necropsy. There were no episodes of instrumentation failure or dressing dislodgment. 

The mean bacterial count in the experimental group was 6647 CFUs/mL and 13 303 CFUs/mL in the control group. This difference was statistically significant (P < .05) (Table). The SEM revealed the presence of uniform biofilm formation across the surface of control group instrumentation (eFigure 5). The experimental group was positive for biofilm formation but with many skip areas with no biofilm (eFigure 6). 

This study compared NPWTi to traditional dressings in a porcine model of infected spinal instrumentation. The investigators hypothesized that NPWTi would reduce bacterial load as well as prevent or disrupt biofilm formation on MRSA-infected instrumentation compared with traditional dressings. Methicillin-resistant S aureus was chosen as the pathogen in this model due to its association with medical device-related infections, ability to form biofilms, and difficulty of eradication.³⁷ Bacterial load was quantified with tissue cultures, and biofilm was assessed with SEM.

The results of this study support the hypothesis that NPWTi decreases bacterial load and reduces biofilm formation. The mean bacterial count in the control was 13 303 CFUs/mL compared with 6647 CFUs/mL in the experimental group (P < .05). Electron micrographs revealed typical biofilm formation in the control group (eFigure 5). Interestingly, biofilm failed to form or was disrupted in several areas in the NPWTi group (eFigure 6). 

The present study demonstrates that NPWTi is a promising therapeutic option that may aid in the prevention or eradication of recurrent infection by reducing bacterial load and interfering with biofilm formation. The reduced bacterial load and incomplete biofilm layer seen in the experimental group is consistent with previous findings¹⁸ evaluating NPWTi in an ex vivo model of biofilm formation on porcine skin. An incomplete biofilm layer on implanted instrumentation is a noteworthy finding because systemic antibiotics and host immune cells may reach sites that would otherwise be inaccessible. This suggests NPWTi may synergize with systemic antibiotic regimens.

Currently, there is no consensus in the literature or clearly codified indications to guide clinicians regarding when to salvage or when to remove infected spinal instrumentation.³⁸ Instrumentation removal risks significant loss of spinal stability, whereas instrumentation retention increases the risk of recurrent infection.³⁸ The porcine spine is an established biomechanical model for human spinal surgery, and the pig is an excellent animal model of human infectious processes.^30,39 Accordingly, this initial animal data are promising and may translate to humans if confirmed by subsequent investigation. 

The limitations of this study design prevent drawing definitive conclusions and further study is needed. The challenges of large animal research and cost constrains restricted the number of animals available for this study. This precluded enrollment of additional study arms to control for confounding variables. The effect of the wound irrigation solution used in this study is unknown due to its use in both the control and experimental groups. In addition, the benefit of NPWTi over standard NPWT was not examined. Ideally, additional study arms would have included standard NPWT, NPWT with saline instillation, saline wet-to-dry dressings, and a true control group devoid of treatment all together. It should also be noted that the biofilm in this model is representative of early infection which may be less stable; thus, it is more susceptible to treatment compared with chronic infection. In future studies, biofilms should be assessed with planimetric analysis. Lastly, it is unclear whether NPWTi impeded biofilm formation or disturbed already-formed biofilms. 

To the best of the investigators’ knowledge, this study is the first in vivo study to demonstrate that NPWTi is associated with decreased bacterial load and biofilm formation compared to wet-to-dry dressings. Based on the results of this study, NPWTi appears to be a superior choice compared with traditional dressings. The results of this study provide a basis for further investigation of this treatment modality. 

From the Division of Plastic Surgery, Anne Arundel Medical Center, Annapolis, MD; Section of Plastic and Reconstructive Surgery, Yale University School of Medicine, New Haven, CT; Division of Plastic and Reconstructive Surgery, University of Maryland School of Medicine, Baltimore, MD; and Department of Plastic and Reconstructive Surgery, The Johns Hopkins Hospital, Baltimore, MD

Address correspondence to:
Devinder P. Singh, MD
Chief & Medical Director
Division of Plastic Surgery
Anne Arundel Medical Center
2001 Medical Parkway
Annapolis, MD 21401
dsingh@aahs.org

Disclosure: Dr. Singh is a consultant for KCI, An Acelity Company (San Antonio, TX). This work was supported by a research grant from Acelity. This device is not US Food and Drug Administration-approved for the indication that is the subject of this manuscript.