Fungus infection in immunocompromised rabbits: correlation of thin-section CT findings and histopathology

Objective: To compare the thin-section CT findings of pulmonary candidiasis, aspergillosis and cryptococcosis with histopathology in immunocompromised rabbits and improve the diagnostic accuracy of fungus infection. Methods: Healthy New Zealand white rabbits were used for immunocompromised animal models. Thin-section CT scan was performed before and 2, 4, 6, 8, 10, 12, 14 d after inoculation. The pattern and distribution of the pulmonary abnormalities were retrospectively assessed by two thoracic radiologists and compared with histopathology. The granulocyte count was compared before and after administration of immunosuppressive agents. The paired t test, chi square test and the Fisher's exact test were used for the statistics. Results: Fourteen rabbits had candidiasis, 16 rabbits had eryptococcosis, 15 rabbits had aspergillosis. The granulocyte counts before and after administration of immunosuppressive agents were (2.91±0.92) and (0.35±0.19) x 109/L respectively in candidiasis group, there was a significant difference (t=12.484, P<0.05); ( 2.51±0.82) and (0.76±0.71) x 109/L in aspergillosis group, there was a significant difference (t=5.792, P<0.05); (2.10±0.65) and (0.48±0.22) x 109/L in cryptococcosis group, there was a significant difference (t=8.199, P<0.05). The onset time of infections on CT were not significantly different in three groups (P>0.05). Ground glass opacity (GGO) and consolidation were the two most common findings in immunocompromised rabbits with three fungus infections, areas of GGO was correlated with the congestion, hemorrhage, inflammatory cell infiltration and interstitial hyperplasia in pathology. Consolidation was correlated with the severe congestion, hemorrhage, inflammatory cell infiltration, interstitial hyperplasia, necrosis and vascular embolism in pathology. Conclusion: GGO and consolidation are the two most common findings of fungus infections in immunocompromised animal models and thin-section CT findings can reflect the pathological changes. (authors)