The effect of volume of interest definition on quantification of lymph node immune response to a monkeypox virus infection assessed by (18)F-FDG-PET

BACKGROUND: 2-deoxy-2-[(18)F]fluoro-D-glucose-positron emission tomography ((18)F-FDG-PET) is applied in the clinic for infection assessment and is under consideration for investigating the inflammatory/immune response in lymphoid tissue in animal models of viral infection. Assessing changes in (18)F-FDG uptake of lymph nodes (LNs), primary lymphoid tissues targeted during viral infection, requires suitable methods for image analysis. Similar to tumor evaluation, reliable quantitation of the LN function via multiple (18)F-FDG-PET sessions will depend how the volume of interest is defined. Volume of interest definition has a direct effect on statistical outcome. The current study objective is to compare for the first time agreement between conventional and modified VOI metrics to determine which method(s) provide(s) reproducible standardized uptake values (SUVs) for (18)F-FDG uptake in the LN of rhesus macaques. METHODS: Multiple (18)F-FDG-PET images of LNs in macaques were acquired prior to and after monkeypox virus intravenous inoculation. We compared five image analysis approaches, SUV(max), SUV(mean), SUV(threshold), modified SUV(threshold), and SUV(fixed volume), to investigate the impact of these approaches on quantification of the changes in LN metabolic activity denoting the immune response during viral infection progression. RESULTS: The lowest data repeatability was observed with SUV(max). The best correspondence was between SUV(fixed volume) and conventional and modified SUV(threshold). A statistically significant difference in the LN (18)F-FDG uptake between surviving and moribund animals was shown using modified SUV(threshold) and SUV(fixed volume) (adjusted p = 0.0037 and p = 0.0001, respectively). CONCLUSIONS: Quantification of the LN (18)F-FDG uptake is highly sensitive to the method applied for PET image analysis. SUV(fixed volume) and modified SUV(threshold) demonstrate better reproducibility for SUV estimates than SUV(max), SUV(mean), and SUV(threshold). SUV(fixed volume) and modified SUV(threshold) are capable of distinguishing between groups with different disease outcomes. Therefore, these methods are the preferred approaches for evaluating the LN function during viral infection by (18)F-FDG-PET. Validation of multiple approaches is necessary to choose a suitable method to monitor changes in LN metabolic activity during progression of viral infection. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s13550-014-0049-z) contains supplementary material, which is available to authorized users.