Longitudinal time-lapse in vivo micro-CT reveals differential patterns of peri-implant bone changes after subclinical bacterial infection in a rat model

Subclinical infection associated with orthopedic devices can be challenging to diagnose. The goal of this study was to evaluate longitudinal, microcomputed tomography (microCT) imaging in a rat model of subclinical orthopedic device-related infection caused by Staphylococcus epidermidis and four different Cutibacterium (previously Propionibacterium) acnes strains, and compare outcomes with non-inoculated and historical S. aureus-inoculated controls. Sterile screws or screws colonized with bacteria were placed in the tibia of 38 adult Wistar rats [n = 6 sterile screws; n = 6 S. epidermidis-colonized screws; n = 26 C. acnes-colonized screws (covering all three main subspecies)]. Regular microCT scans were taken over 28 days and processed for quantitative time-lapse imaging with dynamic histomorphometry. At euthanasia, tissues were processed for semiquantitative histopathology or quantitative bacteriology. All rats receiving sterile screws were culture-negative at euthanasia and displayed progressive bony encapsulation of the screw. All rats inoculated with S. epidermidis-colonized screws were culture-positive and displayed minor changes in peri-implant bone, characteristic of subclinical infection. Five of the 17 rats in the C. acnes inoculated group were culture positive at euthanasia and displayed bone changes at the interface of the screw and bone, but not deeper in the peri-implant bone. Dynamic histomorphometry revealed significant differences in osseointegration, bone remodeling and periosteal reactions between groups that were not measurable by visual observation of still microCT images. Our study illustrates the added value of merging 3D microCT data from subsequent timepoints and producing inherently richer 4D data for the detection and characterization of subclinical orthopedic infections, whilst also reducing animal use.