Extended combined Tc-99m-white blood cell and bone imaging improves the diagnostic accuracy in the detection of hip replacement infections

Although the diagnosis of hip prosthesis infection is clinically important, X-ray studies, blood chemistry and synovial fluid aspiration may be unreli able for this purpose. The aim of this study was to evaluate whether extend ing the time for technetium-99m labelled leucocyte imaging to 24 h post inj ection improves the accuracy of diagnosis of hip replacement infections. We studied 64 symptomatic patients with hip prostheses. The presence of infec tions was verified by intraoperative bacterial cultures, and infection was excluded either by negative operative findings or by follow-up for at least 1 year. Leucocyte imaging was done with Tc-99m-hexamethylpropylene amine o xime (HMPAO)-labelled leucocytes at 2-4 h (routine images) and at 24 h (lat e images) after the injection of the leucocytes. In addition, bone imaging was carried out with Tc-99m-hydroxydiphosphonate (HDP) at the arterial, sof t tissue and metabolic phases. A standardised method was used to compare le ucocyte images with bone metabolic images. In this material, there were six confirmed infections. All the bone imaging methods had a sensitivity of 10 0% in detecting prosthesis infections whereas the specificity varied from o nly 2% to 82%. Routine leucocyte imaging was less sensitive (50% vs 83%) an d less specific (90% vs 100%) than late leucocyte imaging. All tests had a high negative predictive value for excluding infection (95%-100%). However, both bone (10%-38%) and routine leucocyte imaging (33%) showed a poor posi tive predictive value (PPV), whereas late leucocyte imaging had a PPV of 10 0% and a diagnostic accuracy of 98%. We conclude that late leucocyte imagin g improves the specificity of diagnosis of infected hip prostheses. This ty pe of imaging procedure should be combined with three-phase bone scintigrap hy in studies of patients with painful joint replacement.