White Blood Cell Scintigraphy for Fracture-Related Infection: Is Semiquantitative Analysis of Equivocal Scans Accurate?

Purpose: White blood cell (WBC) scintigraphy is considered the gold-standard nuclear imaging technique for diagnosing fracture-related infection (FRI). Correct interpretation of WBC scans in FRI is important since a false positive or false negative diagnosis has major consequences for the patient in terms of clinical decision-making. The European Association of Nuclear Medicine (EANM) guideline for correct analysis and interpretation of WBC scans recommends semiquantitative analysis of visually equivocal scans. Therefore, this study aims to assess the diagnostic accuracy of semiquantitative analysis of visually equivocal WBC scans for diagnosing FRI. Methods: A retrospective single-center study was performed in consecutive patients who received WBC scintigraphy in the diagnostic work-up for FRI between February 2012 and January 2017. All the visually equivocal scans were analysed using semiquantitative analysis by comparing leukocyte uptake in the manually selected suspected infection focus with the contralateral bone marrow (L/R ratio). Cut-off points for a ‘positive’ scan result of >0%, >10% and >20% leukocyte increase between the early and late scans were used in separate analyses. The discriminative ability was quantified by calculating the sensitivity, specificity and diagnostic accuracy. Results: In total, 153 WBC scans were eligible for inclusion. After visual assessment of all the scans, 28 visually equivocal scans were included. Dichotomization of the ratios using the cut-off of >0% resulted in a sensitivity of 30%, a specificity of 45% and a diagnostic accuracy of 40%. The >10% cut-off point resulted in a sensitivity of 18%, a specificity of 82% and a diagnostic accuracy of 66%. The >20% cut-off point resulted in a sensitivity of 0%, a specificity of 89% and a diagnostic accuracy of 67%. Conclusion: Semiquantitative analysis of visually equivocal WBC scans is insufficient for correctly diagnosing FRI