Purpose: To determine if tunnel widening, defined as change in maximal tunnel diameter from the time of initial bone tunnel drilling to revision surgery is associated with bacterial deoxyribonucleic acid (DNA) presence and concentration in torn graft tissue from failed anterior cruciate ligament reconstructions (ACLRs).
Methods: Thirty-four consecutive revision ACLRs were included (mean age 27.3 years SD 10.9; median time to failure 4.9 years range 105 days-20 years). Graft selection of the failed reconstruction was 68% autograft, 26% allograft, and 6% autograft/allograft hybrid with a mean drilled tunnel diameter of 8.4 mm SD 0.8. Maximal tunnel diameters prior to revision were measured on pre-operative three-dimensional imaging and compared to drilled tunnel diameters at the time of the previous reconstruction. Tissue biopsies of the failed graft were obtained from tibial, femoral, and intraarticular segments. Sterile water left open to air during revision ACLRs and tissue from primary ACLRs were used as negative controls. Clinical cultures were obtained on all revision ACLRs and PCR with universal bacterial primer on all cases and negative controls. Fluorescence microscopy was used to confirm the presence and location of biofilms in two patients with retrieved torn graft tissue and fixation material. Amount of tunnel widening was compared to bacterial DNA presence as well as bacterial DNA concentration via Welch ANOVA.
Results: Bacterial DNA was present in 29/34 (85%) revision ACLRs, 1/5 (20%) of primary ACLR controls and 0/3 (0%) sterile water controls. Cultures were positive (coagulase negative Staphylococcus sp.) in one case, which also had the greatest degree of tunnel widening. Femoral widening was greater in cases with detectable bacterial DNA (mean widening 2.6 mm SD 3.0) versus without (mean 0.3 mm SD 0.6) (p = 0.003) but was unaffected by bacterial DNA concentration (p = 0.44). Tibial widening was not associated with the presence of bacterial DNA (n.s.); however, higher bacterial DNA concentrations were observed in cases with tibial widening ≥ 3.0 mm (median 2.47 ng bacterial DNA/µg total DNA) versus widening < 3.0 mm (median 0.97 ng bacterial DNA/µg total DNA) (p = 0.046). Tunnel widening was not associated with time to failure, graft selection, or number of prior surgeries (n.s., all comparisons). Fluorescence microscopy confirmed the presence of biofilms on ruptured tendon graft as well as fixation material in 2/2 cases.
Conclusion: Bacterial DNA is commonly encountered on failed ACLR grafts and can form biofilms. Bacterial DNA does not cause clinically apparent infection symptoms but is associated with tunnel widening. Further research is needed to determine whether graft decontamination protocols can reduce graft bacterial colonization rates, ACLR tunnel widening or ACLR failure risk.
Level of evidence: Therapeutic III.
Keywords: ACLR graft failure; ACLR tunnel widening; Bacterial DNA; Bacterial biofilm; Orthopedic biofilm.