Evaluation of a 3-hour indium-111 leukocyte image as a surrogate for a technetium-99m nanocolloid marrow scan in the diagnosis of orthopedic infection

OBJECTIVES: This is a retrospective study to evaluate a 3-hour In-111-labeled leukocyte image as a surrogate for a Tc-99m nanocolloid marrow scan in the investigation of suspected orthopedic infection using In-111 leukocyte scintigraphy. METHODS: Images from 51 patients who had received contemporaneous In-111-labeled leukocyte scintigraphy and Tc-99m nanocolloid marrow scintigraphy were reviewed. Initially, the 3-hour and 22-hour In-111-labeled leukocyte images were compared. Sites of abnormal uptake on the 22-hour image were correlated with the 3-hour image and were graded according to the level of concordance or discordance. One week later, the Tc-99m nanocolloid images and 22-hour In-111-labeled leukocyte images of the same patients were compared and graded for concordance or discordance. When discrepancies in grading arose between the observers, a consensus opinion was achieved after additional review of the images a week later. RESULTS: On inspection of the 22-hour In-111 leukocyte images, 93 sites of focal, potentially abnormal leukocyte accumulation were identified. When the grading system was reduced to simply "concordant" or "discordant," there was good agreement between the observers in the majority of cases, with kappa statistics 0.77 for Tc-99m nanocolloid versus 22-hour In-111-labeled leukocyte images and 0.78 for 3-hour versus 22-hour In-111-labeled leukocyte images. Using the comparison of the Tc-99m nanocolloid marrow scan and the 22-hour In-111-labeled leukocyte images to identify concordance or discordance as the "gold standard" for scintigraphic evaluation of suspected orthopedic infection, comparison of the 3-hour In-111-labeled leukocyte images with the 3-hour In-111-labeled leukocyte images gave a sensitivity of 77%, a specificity of 77%, and an accuracy of 77%. CONCLUSIONS: A 3-hour image is helpful using In-111-labeled leukocyte scintigraphy.