Scintigraphy of infected total hip arthroplasty (THA): A canine model

Differentiating low-grade sepsis from aseptic loosening of an orthopedic prosthesis is difficult. This study was designed to compare the ability of Tc-99m-HMDP, Ga-67, and In-111 leukocytes (WC) to differentiate low-grade sepsis from aseptic THA component loosening in a canine model. A canine THA was implanted in 14 dogs. Six dogs were given infected femoral components by injecting 10/sup 5/ colony-forming units of Staphylococcus aureus into the femoral canal 6y0 to 90 seconds prior to cementing. Four dogs had an aseptic loose femoral component, and four dogs had an aseptic tight femoral component (control). At six months all dogs were evaluated with X-ray, lab scintigraphy, and tissue quantitation of each tracer. Diagnosis was confirmed by histology and quantitative microbiology. White blood cell counts and differentials were normal in all dogs, and in only one out of six infected dogs was the sedimentation rate abnormal. X-rays were interpreted as possible infection in five dogs and probable infection in only one dog. In-111 WBC scans were more accurate than sequential Tc-Ga scans (sensitivity 94% vs 61%, specificity 86% vs 71% accuracy 90% vs 67%). Quantitative counting of gamma camera data and tissue samples demonstrated significantly (P < .01) higher accumulation of In-111 WBC about the infected than the loose or control component. No significant difference was demonstrated between the loose and septic components with TC-HMDP or Ga. These results correlate well and confirm our clinical data that In-111 WBC scanning is accurate and useful in the workup of the painful orthopedic prosthesis