Antibiotic-loaded bone graft for reduction of surgical site infection in spinal fusion

Background context: Infections remain a leading complication associated with spinal arthrodesis, regardless of the use of prophylactic antibiotics and improved surgical techniques, with incidence of infection as high as 8.2%. Infection prolongs antibiotic usage, increases hospital time, and inevitably inflates overall treatment costs. Local antibiotics, such as vancomycin, have been used in combination with fusion materials over the past decade to decrease infection risk. An ideal graft material would serve a dual role: encouraging vertebral fusion while reducing the incidence of infection.

Purpose: The objective of this study was to thoroughly evaluate the use of a vancomycin-loaded demineralized bone matrix (vDBM) for fusion capability while reducing the incidence of surgical site infection.

Study design: Antimicrobial efficacy and spinal fusion were evaluated using a preclinical rabbit model of posterolateral fusion.

Materials and methods: Vancomycin-loaded demineralized bone matrix was prepared and evaluated for in vitro release kinetics and bacterial inhibition. In vivo antibacterial efficacy and fusion capability were performed using a model of posterolateral fusion in a rabbit. First, 10 New Zealand white rabbits underwent a bilateral posterolateral fusion procedure, were inoculated with Staphylococcus aureus, and were treated with either demineralized bone matrix (DBM) or vDBM. Fourteen days after the procedure, the animals were anesthetized and euthanized, and the transverse process was harvested and enumerated for bacterial quantification. Concurrently, 21 New Zealand white rabbits underwent the same procedure and were euthanized 8 weeks after surgery and were evaluated for fusion by manual palpation and radiographic scoring. In addition, two groups of six animals received the DBM or vDBM material as described, but the graft was combined with equal volumes of milled harvest iliac crest bone graft (ICBG). Eight weeks after surgery, these animals were euthanized and also evaluated for fusion by manual palpation and radiographic scoring.

Results: Vancomycin continued to be released from the vDBM over the course of 6 days while maintaining sufficient eluate concentrations to maintain a zone of inhibition similar or larger than a vancomycin control. In vivo, vDBM significantly reduced the amount of bacteria within the fusion site compared with DBM, with a 4-log decrease in bacterial bioburden. The use of vDBM, however,showed a decrease in the fusion rate compared with DBM when used in a sterile wound. In a S. aureus-contaminated wound, both the DBM and the vDBM showed decreased fusion rates.Considering DBM materials were most commonly used as autograft extenders, additional animals received either DBM plus ICBG in a sterile wound or vDBM plus ICBG in a contaminated wound. Both groups had similar fusion rates and similar fusion volumes after 8 weeks in vivo.

Conclusions: Whereas vDBM reduced the overall bioburden within a contaminated surgical site of posterolateral fusion, the addition of the vancomycin to the DBM reduced the fusion capability of the DBM graft. The addition of ICBG to vDBM restored the fusion capability of the graft while reducing the overall infection.