New strategies to classify canine pleural effusions and the diagnostic value of acute phase proteins, amylase, and adenosine deaminase in pleural exudates.

BACKGROUND

In dogs, simplified Light's criteria can discriminate transudates from exudates. Other tests used in human medicine are pleural effusion cholesterol (CHOLPE) and butyrylcholinesterase [BChEPE], the pleural effusion/serum ratio of these analytes (CHOLratio and BChEratio), and the serum albumin minus pleural effusion albumin gradient (SEAG).

OBJECTIVES

We aimed to assess the diagnostic accuracies of different biomarkers in dogs with pleural effusion in differentiating exudates from transudates. Secondarily, we evaluated the potential diagnostic utility of pleural effusion acute phase proteins, amylase, and adenosine deaminase in discriminating causes of exudative effusions.

METHODS

Cross-sectional study including 68 client-owned dogs with pleural effusion.

RESULTS

There were 48 exudates (10 septic, 16 neoplastic, 9 hemorrhagic, and 13 classified as other exudates) and 20 transudates. All the variables analyzed, except SEAG, were significantly different between exudates and transudates. Using the cut-off values adopted in human literature, accuracies for CHOLPE, CHOLratio, BChEPE, and BChEratio were between 82.35% and 85.29%; all values were significantly lower compared with the previously published simplified Light's criteria accuracy (i.e., 98%, p < .001 for all comparisons). We found the accuracy of the simplified Light's criteria to be similar to what has been previously reported (95.59%, p = .238). Paraoxonase-1 (PON-1PE) activity and the pleural effusion/serum paraoxonase-1 ratio (PON-1ratio) were significantly lower in exudative neoplastic effusions than in exudative hemorrhagic (p = .004 and p = .001) and septic (p = .004 and p < .001) effusions.

CONCLUSION

Simplified Light's criteria were the best method for discriminating transudates from exudates, and a low PON-1PE activity and PON-1ratio in exudative effusions may suggest an underlying neoplasia.