Clinical signs of canine pulmonary angiostrongylosis (CPA) are primarily from the respiratory tract and are caused by the presence and migration of adult worms and the first-stage larvae of Angiostrongylus vasorum in the pulmonary arteries and parenchyma. The clinical assessment is often supported by thoracic radiographs. The objective of the study was to investigate the effect of age and worm burden on the development of radiological manifestations of CPA and develop an objective radiological score of pulmonary changes in CPA using an experimental fox model.
Thirty-four foxes were grouped after age (young and adult), infective dose (50 and 200 third-stage larvae) and controls. Ten weeks after inoculation, the radiographic studies were performed. Subsequently, the foxes were euthanized and burdens of adult worms established. Radiographs were assessed blinded by two of the authors following a predefined scoring system. Also, the type of lung pattern and anatomic location in the lung field were recorded. Based on these parameters, the radiologists scored each set of radiographs in one of four categories according to severity. Mean and 95% confidence interval (CI) of worm count was calculated. A Fisher's exact test was performed to determine differences in variables across dose and age groups. Inter- and intra-observer agreement was tested using kappa statistics. Worm count and radiological score was analyzed using linear regression, and one way ANOVA tested the relationship between worm burden and age or dose. The study was conducted under a Danish experimental animal license.
The score for pulmonary changes differed significantly between control and infection groups. Radiological differentiation between high and low-dose groups was possible if data from the accessory and cranial lung only were considered. A linear relationship could be shown between the radiological score versus worm count. The mean worm count of the young, high-dose foxes was significantly higher than in all the other infection groups (p<0001). Inter-observers variation in classifying the pattern of pulmonary changes was poor to moderate (κ-values of 0.12 to 0.53). Despite these inter-observer differences in the pattern of pulmonary changes, the overall radiological score for lung pathology was in agreement with an inter-observer κ-values of 0.62.
In conclusion, an objective, radiological scoring system was developed which may allow reliable, reproducible assessment of the severity of pulmonary changes in an experimental A. vasorum fox model. The anatomical distribution of lesions is more severe and more frequent in the peripheral and caudal areas of the lung, and supports previous data. This is similar to the changes reported in dogs with spontaneous CPA. Finally, the finding of a relationship between age and severity of lesions and number of adult worms recovered seems to indicate an age related immunity in the development of CPA.